Spirobifluorene Regioisomerism: A Structure–Property Relationship Study

Article abstract of DOI:10.1002/chem.201700570

The present works report the first structure–property relationship study of a key class of organic semiconductors, that is, the four spirobifluorene positional isomers possessing a para-, meta- or ortho-linkage. The remarkable and surprising impact of the ring bridging and of the linkages on the electronic properties of the regioisomers has been particularly highlighted and rationalised. The impact of the ring bridging on the photophysical properties has been stressed with notably the different influence of the linkages and the bridge on the singlet and triplet excited states. The first member of a new family of spirobifluorenes substituted in the 1-position, which presents better performance in blue phosphorescent OLEDs than those of its regioisomers, is reported. These features highlight not only the great potential of 1-substituted spirobifluorenes, but also the remarkable impact of regioisomerism on electronic properties.

Full text of DOI:10.1002/chem.201700570

DOI: 10.1002/chem.201700570
Full Paper
&
Organic Semiconductors |Hot Paper|
Spirobifluorene Regioisomerism: A Structure–Property
Relationship Study
Lambert Sicard,[a] Cassandre Quinton,[a] Jean-David Peltier,[a] Denis Tondelier,[b]
Bernard Geffroy,[b, c] Urelle Biapo,[b] Rꢀmi Mꢀtivier,[d] Olivier Jeannin,[a] Joꢁlle Rault-Berthelot,[a]
and Cyril Poriel*[a]
Abstract: The present works report the first structure–prop-
erty relationship study of a key class of organic semiconduc-
tors, that is, the four spirobifluorene positional isomers pos-
sessing a para-, meta- or ortho-linkage. The remarkable and
surprising impact of the ring bridging and of the linkages on
the electronic properties of the regioisomers has been par-
ticularly highlighted and rationalised. The impact of the ring
bridging on the photophysical properties has been stressed
with notably the different influence of the linkages and the
bridge on the singlet and triplet excited states. The first
member of a new family of spirobifluorenes substituted in
the 1-position, which presents better performance in blue
phosphorescent OLEDs than those of its regioisomers, is re-
ported. These features highlight not only the great potential
of 1-substituted spirobifluorenes, but also the remarkable
impact of regioisomerism on electronic properties.
Introduction
class of OSCs. Indeed, since the discovery of the “spiro con-
cept”, the 9,9’-spirobifluorene (SBF) has become a central mo-
lecular scaffold in organic electronics.[5–9] The 2-substituted
SBFs are, in this context, the most developed class of SBF-
based polymers and oligomers. The para-linkage between the
pendant substituent in the 2-position and the constituted
phenyl ring of the fluorene ensures a good delocalisation of p-
electrons, essential to develop efficient fluorophores. However,
in recent years, the growing necessity to design efficient host
materials for blue phosphorescent organic light-emitting
diodes (PhOLEDs)[10,11] has led to a huge demand of new gen-
erations of SBF-based materials with wide energy gap (ca.
4 eV) and hence a restricted p-conjugation. Indeed, in order to
obtain a high triplet energy (ET), a key feature in the design of
host materials for blue PhOLEDs, which are still the weakest
link of this technology, it is mandatory to restrict the p-elec-
trons delocalisation within the OSC. This p-conjugation disrup-
tion has been successfully investigated with ortho-linked SBFs
(substitution in the 4-position)[12–16] and meta-linked SBFs (sub-
stitution in the 3-position),[2,17–20] leading to high performance
blue PhOLEDs. However, despite the recent very high efficiency
devices obtained by Jiang et al.[17] and by our groups,[15] only
few examples of 3- and 4-substituted SBFs have been de-
scribed to date and more importantly no rational structure–
property relationship studies have been reported on SBF re-
gioisomerism. Such studies are nevertheless the foundation of
materials design for electronics. Thus, the combination of
steric hindrance (ortho-position) and electronic decoupling
(meta-position) found in 1-substituted SBF, has never been
studied and could nevertheless be the best way to obtain high
ET OSCs based on the SBF scaffold. Hence, we report herein
not only the first example of a 1-substituted SBF, namely 1-
phenyl-SBF 1 (Scheme 1), but also a detailed study describing
Regioisomerism, also called positional isomerism, is an impor-
tant concept in organic chemistry that can have remarkable
consequences on the properties of molecules.[1] Indeed,
a simple structural modification can drastically influence the
electronic and physical properties of an organic semi-conduc-
tor (OSC), which in turn strongly modifies the performance and
stability of the corresponding electronic device. Although very
promising, this concept remains barely used in optoelectronics.
Recently, our groups have reported the use of regioisomerism
to finely tune the singlet and triplet energies of dihydroinde-
nofluorenes, leading to highly efficient optoelectronic devi-
ces.[1,2] Similarly, Haley and co-workers have shown the impact
of the regioisomerism on the electronic properties of a promis-
ing family of antiaromatic indenofluorene derivatives.[3,4] Spiro-
configured compounds constitute one of the most important
[a] L. Sicard, Dr. C. Quinton, J.-D. Peltier, Dr. O. Jeannin, Dr. J. Rault-Berthelot,
Dr. C. Poriel
UMR CNRS 6226- Universitꢀ Rennes 1
35042 Rennes (France)
E-mail: cyril.poriel@univ-rennes1.fr
[b] Dr. D. Tondelier, B. Geffroy, U. Biapo
LPICM, CNRS, Ecole Polytechnique, Universitꢀ Paris Saclay
91128, Palaiseau (France)
[c] B. Geffroy
LICSEN, NIMBE, CEA, CNRS, Universitꢀ Paris-Saclay
CEA Saclay 91191 Gif-sur-Yvette Cedex (France)
[d] Dr. R. Mꢀtivier
UMR CNRS 8531-PPSM, ENS Cachan
94235 Cachan (France)
Supporting information and the ORCID identification number(s) for the au-
thor(s) of this article can be found under:
http://dx.doi.org/10.1002/chem.201700570.
Chem. Eur. J. 2017, 23, 1 – 10
1
ꢂ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
age,[22] Scheme 1. Thanks to a comparison with the structurally
fluorene analogues, this work provides a foundation on the
impact of SBF regioisomerism on the electronic properties.
This study notably shows the surprising consequences of the
ring bridging on the p-conjugation and the different influence
of the linkages and the bridge on the singlet and triplet excit-
ed states. This approach has allowed the design of the first
member (1) of a new family of OSCs, which possesses a very
high ET (2.86 eV), one of the highest reported for SBF-based
materials. As a first electronic application, 1 has been used as
a host in blue PhOLEDs with higher performance than those
using its regioisomers, highlighting the great potential of this
family of OSCs.
Scheme 1. SBF and the four positional isomers of phenyl substituted SBFs.
the impact of SBF regioisomerism on the electronic properties
and device efficiency.
Results and Discussions
This work reports the first structure–property relationship
study of the four SBF positional isomers (substituted with
a phenyl ring) namely 2-phenyl-SBF 2 possessing a para-link-
age,[21] 1-phenyl-SBF 1 and 3-phenyl-SBF 3 both possessing
a meta-linkage and 4-phenyl-SBF 4 possessing an ortho-link-
The synthesis of 2 and 4 has been previously described in liter-
ature,[21,22] that of 1 and 3 is described in Scheme 2. The syn-
thesis of 1 (Scheme 2, top) starts with the synthesis of the key
fluorenone c1, substituted in the 1-position with an iodine
Scheme 2. Synthetic routes to 1 and 3.
&
&
Chem. Eur. J. 2017, 23, 1 – 10
www.chemeurj.org
2
ꢂ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
atom. First, the Suzuki–Myaura cross-coupling between 3-bro-
moaniline and [2-(ethoxycarbonyl)phenyl]boronic acid in the
presence of [Pd(dppf)Cl2] (dppf=1,1’-bis(diphenylphosphino)-
ferrocene) as catalyst and potassium carbonate as base pro-
vides the corresponding biphenyl a1 bearing a carboxylate
and an amine function (yield 41%). The electrophilic intramo-
lecular cyclisation of the ester a1 was then performed in meth-
anesulfonic acid at high temperature (1308C), providing the
fluorenone b1 substituted in the 1-position (43% yield). We
note that the reaction also provides the fluorenone substituted
in the 3-position. The regioselectivity has not been studied in
detail herein, but preliminary works seem to show that the
temperature and the nature of the solvent have an important
effect, as previously reported in literature for similar aromatic
electrophilic substitutions.[23,24] The substitution of the amine
by an iodine atom is then performed through a Sandmeyer re-
action on fluorenone b1 providing the 1-iodofluorenone c1
with 75% yield. We note that compound c1 reported herein
widens the scope of substituted fluorenone isomers as key
building blocks for the synthesis of spiro derivatives for organ-
ic electronics.[14,25–27] Pd-catalysed cross-coupling ([Pd(dba)2]
(dba=dibenzylideneacetone), PCy3, KF, DMF, 1308C) between
c1 and phenylboronic acid then gives the 1-phenylfluorenone
d1 with a high yield of 85%. Thus, despite a sterically hindered
environment, incorporation of pendant substituents in the 1-
position through Pd-catalysed cross-coupling is an interesting
strategy to obtain 1-substituted fluorenone derivatives. Finally,
the synthesis of 1 was then carried out through a classical
two-step procedure. The lithium–iodine exchange of 2-iodobi-
phenyl with n-butyllithium at low temperature, followed by ad-
dition of fluorenone d1 afforded the corresponding fluorenol
(not isolated) and further involved in an intramolecular cyclisa-
tion reaction (AcOH/HCl) to provide the expected 1-phenyl spi-
robifluorene 1 with an overall yield of 38% over the two steps.
This synthetic strategy is straightforward and provides to the
best of our knowledge the first example of a 1-substituted SBF
for organic electronics.
Figure 1. ORTEP drawing of 1–4 (ellipsoid probability at 50% level).
encumbered phenyl/fluorene para-linkage,[29] maximises the
conjugation between the two fragments. In 3, the meta-link-
age leads to an even smaller dihedral angle of 34.28 between
the fluorene and its attached phenyl ring (Figure 1, bottom
right). In 4, the presence of the pendant phenyl ring in ortho-
position of the biphenyl linkage leads to an impressive en-
hancement of the dihedral angle, recorded at 51.28 (Figure 1,
bottom left).[21] This structural feature, assigned to the steric in-
teraction between the hydrogen atoms in the ortho-position of
the pendant phenyl ring and that of the fluorenyl core is at
the origin of the partial p-conjugation breaking of 4-substitut-
ed SBFs.[22] In 1, this dihedral angle is larger, reaching 75.48
(Figure 1, top left) for one molecule and 66.78 for the other
(2 molecules are indeed present in the asymmetric unit, see
Supporting Information). This structural particularity is the con-
sequence of the substitution in ortho-position of the spiro
carbon, which leads to a sterically hindered environment due
to the presence of the cofacial fluorene. Indeed, very short Cꢀ
C distances are measured between carbon atoms of the non-
substituted fluorene and those of the pendant phenyl ring
(3.25 and 3.31 ꢃ for one molecule and 3.20, 3.29 and 3.33 ꢃ for
the other molecule, see Supporting Information). These CꢀC
distances are shorter than the sum of their van der Waals radii
(3.4 ꢃ) and translate strong interactions between the two cofa-
cial fragments, as confirmed by the electrostatic potential sur-
face obtained by molecular modelling (see Supporting Infor-
mation). Thus, in the four SBF isomers, the position of the
phenyl ring leads to different steric hindrances with the substi-
tuted or the non-substituted fluorene. This structural feature
will be one of the key parameters involved in the electronic
properties (see below). The electronic properties of com-
pounds 1–4 are given in Table 1.
The synthesis of 3 (Scheme 2, bottom) follows a similar strat-
egy involving the synthesis of the 3-bromofluorenone c3 as
key fragment. Thus, (2-aminophenyl)(4-bromophenyl)metha-
none, possessing an amine function in a-position of the
ketone, was first diazotised by conventional means and then
reacted in situ to form a carbon–carbon bond leading to the
corresponding fluorenone backbone.[28] 3-Bromofluorenone c3
is thus obtained with 55% yield. Suzuki–Myaura cross-coupling
between c3 and phenylboronoic acid ([Pd(dppf)Cl2], K2CO3,
DMF, 1508C) leads to the formation of 3-phenylfluorenone d3
with a high yield of 89%. Finally, following the same sequence
described above, the 3-phenyl-substituted SBF 3 is obtained
with a very high yield of 90% (2 steps).
The structural arrangement of compunds 1–4 obtained by
X-ray diffraction are depicted in Figure 1. The most important
structural feature is the relative position of the pendant phenyl
ring with respect to the fluorene. Compound 2 presents a dihe-
dral angle between the mean plane of the pendant phenyl
ring and that of its attached phenyl ring of the fluorene of
37.48 (Figure 1, top right). This angle, characteristic of a non-
The UV/Vis absorption spectra of the SBF regioisomers are
presented Figure 2, top left and time-dependant density func-
tional theory (TD-DFT, Figure 3) calculations have been per-
formed using the B3LYP functional and the extended 6-311+
G(d,p) basis set on the optimised geometry of S0 (B3LYP/6-
31G(d). The unsubstituted SBF exhibits two characteristic
Chem. Eur. J. 2017, 23, 1 – 10
www.chemeurj.org
3
ꢂ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
Table 1. Electronic data of 1–4 and SBF.
labs [nm] (e [ꢄ104 Lmolꢀ1 cmꢀ1])[a]
1
2
3
4
SBF
298 (0.98)
309 (1.66)
313, 323
0.61
296 (2.44) 308 (2.31)
297 (0.92) 310 (1.49)
297 (1.07)
309 (1.49)
359
297 (0.72)
308 (1.45)
310, 323
0.40
4.60
0.87
319 (1.60)
334, 350
0.87
316 (0.64)
332, 343
0.74
lem[a] [nm]
QY[a]
0.42
4.20
1.00
1.40
tf [ns][a]
5.16
1.22
0.72
1.56
5.60
0.83
5.74
1.29
0.45
kr (ꢄ108) [sꢀ1
]
knr (ꢄ108) [sꢀ1
]
1.30
HOMO [eV]
electrochemical[b]
calcd[d]
ꢀ5.94
ꢀ6.00
ꢀ5.86
ꢀ5.90
ꢀ5.94
ꢀ5.97
ꢀ5.95
ꢀ6.02
ꢀ5.95
ꢀ6.03
LUMO [eV]
DE [eV]
ET [eV]
electrochemical[b]
calcd[d]
ꢀ1.73
ꢀ1.32
ꢀ1.99
ꢀ1.55
ꢀ1.77
ꢀ1.34
ꢀ1.87
ꢀ1.38
ꢀ1.74
ꢀ1.30
optical[c]
electrochemical[b]
3.95
4.21
3.70
3.87
3.78
4.17
3.82
4.08
3.97
4.21
optical[e]
calcd[f]
2.86
2.65
2.56
2.38
2.83
2.64
2.78
2.57
2.88
2.67
tp [s][e]
5.8
3.3
5.4
4.7
5.3
[a] In cyclohexane. [b] From CVs. [c] From UV/Vis spectra. [d] From DFT calculations. [e] In 2-Me-THF. [f] From TD-DFT calculations.
Figure 2. Top: Absorption in cyclohexane of 1–4 and SBF (left) and 1F–4F and F (right). Bottom: Emission at room temperature (cyclohexane, left) and at 77 K
(2-Me-THF, right) of 1–4 and SBF.
bands at 297 and 308 nm corresponding to p-p* transitions
(See TD-DFT calculations in SI).[22] The four phenyl-substituted
SBF isomers all display these two thin bands (Figure 2) at a sim-
ilar wavelength. In addition to these bands, 2 displays an extra
and large band centred at 319 nm. This band clearly signs an
extension of the p-conjugation from fluorene in SBF to
phenyl-fluorene in 2 and is assigned to an HOMO!LUMO
transition (with both orbitals centred on the phenyl-fluorene
&
&
Chem. Eur. J. 2017, 23, 1 – 10
www.chemeurj.org
4
ꢂ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Figure 3. Representation of the energy levels and the main molecular orbitals involved in the electronic transitions of 1 (top left), 2 (top right), 3 (bottom left)
1/2
and 4 (bottom right) obtained by TD-DFT B3LYP/6–311+G(d,p), shown with an isovalue of 0.04 [e bohrꢀ3
] .
fragment), possessing a high oscillator strength (f=0.59,
Figure 3, top right). This extension of conjugation is due to the
combination of two parameters: the para-linkage and the
small dihedral angle (Figure 1) formed between the phenyl
and the fluorene. Instead of this large band at 319 nm, the
spectrum of 4 only presents a weak tail between 309 and
325 nm assigned to an HOMO!LUMO transition with a weak
oscillator strength (f=0.11, Figure 3, bottom right). The pres-
ence of this tail reflects a certain degree of p-conjugation be-
tween the fluorene moiety and the phenyl ring, induced by
the ortho-linkage. There is however a strong p-conjugation dis-
ruption in 4 due to the large angle formed between the fluo-
rene and the phenyl in C4. Indeed, the calculated electronic
density accompanying this first electronic excitation shows
that this pendant phenyl ring is involved with a weak contribu-
tion (see Supporting Information). The meta-linkage found in
both 1 and 3 leads to very different results. Indeed, in 3, we
note the presence of a large band at 316 nm, very similar to
that observed for 2, and assigned in the light of TD-DFT to
a transition possessing two major contributions, HOMO!
LUMO and HOMO!LUMO+2 (f=0.15, Figure 3, bottom-left).
This band at 316 nm translates a clear extension of the p-con-
jugation with nevertheless a molar absorption coefficient
2.5 times lower than that observed for the para-isomer 2 (2:
104 Lmolꢀ1 cmꢀ1). However, the presence of this band appears
to be very surprising in the light of literature. Indeed, the ab-
sorption spectrum of the meta-terphenyl (analogue of 1 and 3
but without the bridge) possesses a lmax at 246 nm, strongly
blue shifted compared to that of its regioisomer, the para-ter-
phenyl (analogue of 2), lmax =277 nm.[2] This p-conjugation dis-
ruption observed for meta-terphenyl finds its origin in the
shape and distribution of the molecular orbitals involved
(small contributions are indeed found on meta-carbon
atoms).[30] Indeed, it is admitted that there is a better delocali-
sation of p-electrons following the para/ortho/meta-sequence
and numbers of studies have tried to elucidate the origin of
the restricted p-conjugation between para-, ortho- and meta-
substituted oligophenylenes.[30–36] In our case, it is clear that 3
displays a different behaviour compared to its building block
meta-terphenyl as it presents a relatively intense degree of
conjugation between the phenyl and the fluorene. Thus, the
“linkage” effect cannot explain by itself this feature and other
parameters should be invoked. We believe that the p-conjuga-
tion extension of 3 finds its origin in the electron donating
effect of the bridge, herein the spiro carbon. Indeed, the spiro
carbon may increase the electron density in its para-position
that is in C3 allowing the p-conjugation extension.[17]
To unravel this “ring bridging” effect, we have chemically
modified the bridge through the synthesis of the related fluo-
e319nm =1.6ꢄ104 Lmolꢀ1 cmꢀ1
;
3:
e316nm =0.64ꢄ
Chem. Eur. J. 2017, 23, 1 – 10
www.chemeurj.org
5
ꢂ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
rene analogues (see Supporting Information), namely 1-phenyl-
fluorene (1F), 2-phenylfluorene (2F), 3-phenylfluorene (3F), 4-
phenylfluorene (4F), possessing a methylene bridge instead of
the spirofluorene bridge. As observed for 3, the absorption
spectrum of the meta-isomer 3F surprisingly displays a large
band, lmax =308 nm, at an almost identical wavelength to that
not discussed here). Indeed, 2 and 3 possess an almost identi-
cal emission spectrum, which are the most red-shifted in the
series due to their extended conjugation (lmax =334 and
332 nm, respectively). This result, although in full accordance
with that described above for the absorption, appears again
very surprising as the meta-linkage of 3 should strongly restrict
the p-conjugation compared to the para-linkage of 2.[22,25] This
is again the consequence of the spiro bridge in para-position
of the phenyl ring. Both fluorophores 2 and 3 also display
a high quantum yield (87 and 74%, respectively), indicating
a weak non-radiative pathways from S1 to S0. Thus, from a spec-
tral shape point of view, para- and meta-linkages are noticea-
bly almost indistinguishable. Important differences can be nev-
ertheless found in the activation/deactivation processes.
Indeed, the fluorescence decay curve of 2 provides a single
lifetime of 1.56 ns, which is noticeably shorter than that of 3
(5.74 ns). The radiative rate constant (kr) of 2 is calculated to
be 5.6ꢄ108 sꢀ1, which is about four times that of 3 (1.29ꢄ
108 sꢀ1). This feature is in good agreement with the oscillator
strength difference observed for the first electronic transitions
(f=0.59 for 2 and 0.15 for 3). However, the non-radiative rate
constant (knr) of 2, (knr =0.83ꢄ108 sꢀ1) is twice that of 3 (knr =
0.45ꢄ108 sꢀ1); hence this feature shows that vibrational deacti-
vation pathways are more favourable for the former than for
the latter despite their identical environment. Thus, despite 2
and 3 possessing a similar quantum yield and spectrum shape,
they nevertheless present very different radiative and non-radi-
ative constants, highlighting the importance of the linkages on
the photophysical processes. Remarkably, 1 displays a blue-
shifted emission spectrum (lmax =313 nm) compared to that of
3 despite the identical meta-linkages of both molecules. The
spectrum of 1 is even almost identical to that of its building
block SBF, completely erasing the effect of the pendant phenyl
ring on the p-conjugation pathway at the excited state. The
quantum yield of 1 is nevertheless higher than that of SBF
(0.61 and 0.40, respectively), indicating that 1-substituted SBFs
are very efficient near-UV emitters. The higher quantum yield
of 1 compared to that of SBF is due to a combination of
a higher kr (1.22ꢄ108 sꢀ1 for 1 and 0.87ꢄ108 sꢀ1 for SBF) and
a smaller knr (0.72ꢄ108 sꢀ1 for 1 and 1.30ꢄ108 sꢀ1 for SBF). If
one compares the two meta-substituted isomers 1 and 3, it is
worth noting that the decrease of quantum yield in case of
1 mainly results in more efficient internal conversion processes
(knr of 3 smaller than knr of 1) and not in a much lower elec-
tronic transition moment (identical kr for 1 and 3). Thus, the
absorption and emission spectra at room temperature follow
the same surprising trend. This is not the case at 77 K.
observed for the para-isomer 2F, lmax =310 nm (3F: e308nm
=
0.60ꢄ104 Lmolꢀ1 cmꢀ1 e310nm =2.68ꢄ104 Lmolꢀ1 cmꢀ1,
;
2F:
Figure 2, top right). This large band can be assigned to the p-
conjugation extension and the clear difference observed be-
tween the absorption spectra of the two meta-isomers, 3F and
1F, sheds light on the key role played by the bridge. Thus,
meta-isomers 3F and 3 display a similar behaviour showing
a surprising conjugation extension, less intense than the para-
analogues 2 and 2F, but more intense than their ortho-isomers
4F and 4, hence confirming the impact of the bridge on the
optical properties. Thus, the rigidification of the meta-terphen-
yl core by one bridge cancels, at least partially, the effect of
the linkages on the conjugation length, which appears as an
interesting way to tune the electronic properties of bridged
oligophenylenes.
The other meta-linked SBF 1 displays a very different absorp-
tion spectrum almost identical to that of SBF with a main thin
band at 309 nm and no trace of extended conjugation at
higher wavelengths. TD-DFT of 1 reveals for this band two
main transitions (HOMO!LUMO and HOMO!LUMO+1,
Figure 3, top left) all involving only the fluorene fragment with
no electronic density found on the phenyl unit (see also the
calculated electron density changes in Supporting Informa-
tion). This feature highlights a strong similitude with the transi-
tions observed for SBF. Thus, the complete p-conjugation
breaking of 1 finds its origin not only in the meta-linkage,
which cannot completely break the conjugation as exposed
above for 3, but also in the very large dihedral angle measured
between the phenyl unit and the fluorene. This large angle is
caused by the presence of the spiroconjugated fluorene,
which strongly restricts the rotation of the phenyl ring. Remov-
ing this bulky spirofluorene, such as in the fluorene analogue
1F mentioned above, confirms its importance as a long tail in
the absorption of 1F is observed (Figure 2, top right), reflecting
a clear conjugation between the phenyl and the fluorene
moiety (see the calculated electron density changes in Sup-
porting Information). It is noteworthy that meta-isomer 1F and
ortho-isomer 4F possess an almost identical absorption spec-
trum (1F: e302nm =0.55ꢄ104 Lmolꢀ1 cmꢀ1
; 4F: e300nm =0.58ꢄ
104 Lmolꢀ1 cmꢀ1, Figure 2, top right), confirming that the
bridge rigidification cancels the effect of the linkages on the
conjugation length. Hence, there is a better delocalisation fol-
lowing the 2-/3-/4-/1-substitution sequence, which translates
into an opening of the optical gap DEopt from 2 (3.70 eV) to
1 (3.95 eV). Thus, the most efficient conjugation is found for 2
and 3, which do not present any steric congestion highlighting
its importance whatever the linkages involved.
At 77 K, the emission spectra of 1–4 present a well-resolved
phosphorescence contribution, with a first band centred at
431 nm for 1, 483 nm for 2, 438 nm for 3 and 445 nm for 4
(Figure 2, bottom right). The corresponding ET of 1–4 were
thus, estimated to be about 2.86, 2.56, 2.83 and 2.78 eV, re-
spectively. Due to the p-conjugation disruption, the meta-sub-
stituted terphenyl core of 1 and 3 leads to an increase of the
ET compared to the para-substituted terphenyl core of 2 and
to a lesser extent to that of the ortho-substituted terphenyl
core of 4. The para-, ortho- and meta-terphenyl analogues of
The properties of 1–4 in their excited state confirm the key
steric and electronic roles of the spiro bridge; their fluores-
cence spectra (Figure 2, bottom left) show the same trend as
that observed above (except 4, which is a very particular case
&
&
Chem. Eur. J. 2017, 23, 1 – 10
www.chemeurj.org
6
ꢂ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
1–4, without any bridge, follow the same trend with ET of 2.55,
2.67 and 2.82 eV respectively.[2,37] Two important features need
to be stressed:
impact of the nature of the linkages on the phosphorescence
lifetimes.
The cyclic voltammetry (CV) of 1–4 allows to determine their
HOMO energies[38] at ꢀ5.94, ꢀ5.86, ꢀ5.94 and ꢀ5.95 eV, re-
spectively (Figure 5, right). Thus, despite their phenyl substitu-
tion, 1, 3 and 4 possess the same HOMO energy as that of SBF
1) The ET of 1 (2.86 eV) is almost identical to that of SBF (ET =
2.88 eV) confirming that the pendant phenyl has no influ-
ence on the T1 state (Figure 4 bottom), which is a key point
for further use as host in blue PhOLED (see below),
Figure 5. Cyclic voltammetry (CV) of 1–4 and SBF recorded in DMF/Bu4NPF6
0.1m (reduction, left) and in dichloromethane/Bu4NPF6 0.2m (oxidation,
right). Sweep-rate: 100 mVsꢀ1. Platinum working electrode. All CVs are nor-
malised at the peak potential of the reduction (left) or oxidation (right) pro-
cess.
(ꢀ5.95 eV). This is in accordance with the electronic distribu-
tion of their HOMO (Figure 4 middle), which does not (or
weakly for 3) present any density on the pendant phenyl ring
due to the phenyl/fluorene steric hindrance for 1 and 4 and to
the meta-conjugation for 3. The HOMO of 2 (ꢀ5.86 eV), which
presents an electron density delocalised through the phenyl/
fluorene fragment, is obviously the highest in the series. In the
fluorene series, a different sequence is detected. Indeed, the
HOMO levels of 1F-4F are found at ꢀ5.94, ꢀ5.72, ꢀ5.89 and
ꢀ5.93 eV, respectively (see CVs and molecular modelling in
Supporting Information). In contrast to the SBF series, the
HOMO of 3F is significantly higher than that of 1F and 4F. This
is in accordance with the electronic density detected on the
pendant phenyl ring of 3F and highlights that the steric hin-
drance found in 1F and 4F is more efficient than the nature of
the linkage to restrain the electronic density within the fluo-
rene core. This important difference observed between the F
and SBF series indicates that the bridge (methylene vs. spiro-
fluorene) has a considerable impact on the HOMO distribution
and energy levels. The cathodic explorations (Figure 5, left)
have again revealed a different behaviour in each series. In the
SBF series, the LUMO energy of 1 (ꢀ1.73 eV) is almost identical
to that of SBF (ꢀ1.74 eV), indicating, as for the HOMO, that
the phenyl ring in the 1-position does not influence the LUMO
energy (Figure 4, top). However, 3 and 4 display a different be-
haviour with a deeper LUMO calculated with a deeper
ꢀ1.77 eV and ꢀ1.87 eV, respectively, translating the non-negli-
gible influence of the phenyl unit on the LUMO energy. This is
particularly pronounced for 4, which presents a significant con-
tribution of the phenyl ring in the LUMO distribution, which is
not the case for its HOMO level. Thus, the phenyl ring has a dif-
ferent influence on the benzenoidal HOMO/quinoidal LUMO
distribution depending of the regioisomer involved. This signif-
icantly different influence of the phenyl ring over HOMO/
LUMO levels is also observed in the fluorene series and is as-
Figure 4. Frontier molecular orbitals (top: LUMO, middle: HOMO) and SDD
triplet (bottom) with isovalues of 0.04 and 0.004 respectively.
2) The emission from T1 state follows a classical para/ortho/
meta-sequence (ET increases as follows 2/4/3/1), being
hence different to that from S1 state.
Thus, in contrast to our observations in absorption and fluo-
rescence, the nature of the linkage fully drives the ET values.
Indeed, the triplet exciton of 3 is localised along the substitut-
ed fluorene with no contribution of the pendant phenyl
(Figure 4, bottom), maintaining a high ET of 2.83 eV. In addition,
the delocalisation of the triplet exciton of 4 presents a signifi-
cant contribution of the pendant phenyl, which in turn de-
creases the ET to 2.78 eV. This feature is very different to that
observed for HOMO and LUMO distribution (see below) and in-
dicates the peculiar behaviour of these isomers. Thus, the sin-
glet and triplet energies follow different trends with a remark-
able different contribution of the pendant phenyl group. The
bridge seems to have a strong impact in the singlet-state
energy, whereas the triplet-state energy is fully driven by the
nature of the linkages. Finally, radiative deactivation of the trip-
let state of SBF and 1–4 has been measured and appears to
be very slow under these experimental conditions: the phos-
phorescence decay was measured and the lifetime of the T1
state of 1 and 3 was found to be 5.8 s and 5.4 s, respectively.
Thus, the meta-linkages of 1 and 3 lead to the longer lifetimes
very similar to that measured for SBF. The para-linkages of 2
have an important influence on the photophysical properties
of the T1 state, since its lifetime (3.3 s) was found to be shorter
than its meta- and ortho-isomers (4.7 s for 4), highlighting the
Chem. Eur. J. 2017, 23, 1 – 10
www.chemeurj.org
7
ꢂ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
signed to the ring bridging. The LUMO levels of 1F-4F were re-
corded at ꢀ1.77, ꢀ1.92, ꢀ1.71 and ꢀ1.82 eV, respectively. We
particularly note that there is a strong contribution of the
phenyl ring in the LUMO of 1F (see Supporting Information),
which strongly decreases its energy compared to that of F.
This is a significant difference with the LUMO of 1, which only
implies the fluorene, highlighting the key steric role of the
spiro bridge.
Experimental Section
Details on material and methods, electrochemical properties, struc-
tural properties, photophysical properties, molecular modelling,
device fabrication and characterisation, and copies of NMR spectra
can be found in the Supporting Information. CCDC 1495896 (1)
and 1495849 (3) contain the supplementary crystallographic data.
These data can be obtained free of charge by The Cambridge Crys-
tallographic Data Centre.
In order to finally explore the potential of the 1-substituted
SBF family in electronics and more generally the impact of re-
gioisomerism on device performance, 1, 3 and 4 have been
used as host in blue PhOLEDs containing very low amount of
FIrpic (5%, see structure in Supporting Information). Indeed, 2
cannot be used (ET =2.56 eV), since in order to ensure efficient
energy transfers within the emissive layer,[39] the host should
possess a higher ET than that of the blue phosphor FIrpic (ET =
2.62 eV).[40] The device using 3 has a current efficiency (CE) of
12.8 cdAꢀ1, a power efficiency (PE) of 4.36 lmWꢀ1 and an exter-
nal quantum efficient (EQE) of 4.7% (at 10 mAcmꢀ2). The
ortho-substituted 4 displays slightly higher performance, with
an EQE of 5.5%, a CE of 14.3 cdAꢀ1 and a PE of 5.8 lmWꢀ1. The
performance of the meta-substituted isomer 1 appears to be
the highest in the series with an EQE of 5.9% and correspond-
ing CE and PE of 15.9 cdAꢀ1 and 5.9 lmWꢀ1, respectively. As
the device architecture is identical, the different performance,
despite being small, can only be imputed to the efficiency of
the host and more precisely to the position of the phenyl ring
within the molecular structure (see PhOLED data of 1, 3 and 4
in Supporting Information). In the light of these performances,
the 1-substituted SBF scaffold appears promising to host blue
phosphors. All the electroluminescent spectra (see Supporting
Information) exclusively show the emission of FIrpic, indicating
an efficient energy-transfer cascade. Thus, this first electronic
application of a 1-substituted SBF shows a good confinement
of the excitons within the emitting layer, which is a crucial
point for optoelectronics. It should be stressed that these blue
PhOLEDs performances remain modest compared to the best
recently reported for pure hydrocarbons.[17] However, with
a more accurate design (incorporation of donor and/or accept-
or units),[15,41] device performance could be easily increased
and this new family of 1-substituted SBF can play an interest-
ing role in the future.
Acknowledgements
The authors thank the CDIFX (Rennes) for X-ray data collection,
the CRMPO and P. Jꢀhan (Rennes) for mass spectrometry,
GENCI (France) for allocation of computing time at CINES
(Montpellier) under project c2016085032. C.P. and C.Q. wish to
highly thank Prof J. Cornil (Mons) for his precious help in theo-
retical calculations. We wish to thank the ANR “Men In Blue”
(n814-CE05-0024) for financial support, for a post-doctoral
grant (C.Q.) and for a studentship (L.S.). The Region Bretagne
and the ADEME (Dr Bruno Lafitte) are also warmly acknowl-
edged for a studentship (J.D.P.).
Keywords: organic light-emitting diodes · phosphorescence ·
regioisomerism
conjugation
·
ring bridging
·
spirobifluorene
·
p-
[1] M. Romain, D. Tondelier, J.-C. Vanel, B. Geffroy, O. Jeannin, J. Rault-Ber-
thelot, R. Mꢀtivier, C. Poriel, Angew. Chem. Int. Ed. 2013, 52, 14147–
14151; Angew. Chem. 2013, 125, 14397–14401.
[2] M. Romain, S. Thiery, A. Shirinskaya, C. Declairieux, D. Tondelier, B. Geff-
roy, O. Jeannin, J. Rault-Berthelot, R. Mꢀtivier, C. Poriel, Angew. Chem.
Int. Ed. 2015, 54, 1176–1180; Angew. Chem. 2015, 127, 1192–1196.
[3] A. G. Fix, D. T. Chase, M. M. Haley in Topics in Current Chemistry (Eds.:
J. S. Siegel, Y.-T. Wu), Springer, Heidelberg, 2014, pp. 159–195.
[4] A. G. Fix, P. E. Deal, C. L. Vonnegut, B. D. Rose, L. N. Zakharov, M. M.
Haley, Org. Lett. 2013, 15, 1362–1365.
[5] T. P. I. Saragi, T. Spehr, A. Siebert, T. Fuhrmann-Lieker, J. Salbeck, Chem.
Rev. 2007, 107, 1011–1065.
[6] L.-H. Xie, J. Liang, J. Song, C.-R. Yin, W. Huang, Curr. Org. Chem. 2010,
14, 2169–2195.
[7] X.-F. Wu, W.-F. Fu, Z. Xu, M. Shi, F. Liu, H.-Z. Chen, J.-H. Wan, T. P. Russell,
Adv. Funct. Mater. 2015, 25, 5954–5966.
[8] J. Yi, Y. Wang, Q. Luo, Y. Lin, H. Tan, H. Wang, C.-Q. Ma, Chem. Commun.
2016, 52, 1649–1652.
[9] N. J. Jeon, H. G. Lee, Y. C. Kim, J. Seo, J. H. Noh, J. Lee, S. I. Seok, J. Am.
Chem. Soc. 2014, 136, 7837–7840.
[10] K. S. Yook, J. Y. Lee, Adv. Mater. 2012, 24, 3169–3190.
[11] K. S. Yook, J. Y. Lee, Adv. Mater. 2014, 26, 4218–4233.
[12] C. Fan, Y. Chen, P. Gan, C. Yang, C. Zhong, J. Qin, D. Ma, Org. Lett. 2010,
12, 5648–5651.
[13] Z. Jiang, H. Yao, Z. Zhang, C. Yang, Z. Liu, Y. Tao, J. Qin, D. Ma, Org. Lett.
2009, 11, 2607–2610.
[14] S. Thiery, D. Tondelier, C. Declairieux, B. Geffroy, O. Jeannin, R. Mꢀtivier,
J. Rault-Berthelot, C. Poriel, J. Phys. Chem. C 2015, 119, 5790–5805.
[15] S. Thiery, D. Tondelier, B. Geffroy, E. Jacques, M. Robin, R. Mꢀtivier, O.
Jeannin, J. Rault-Berthelot, C. Poriel, Org. Lett. 2015, 17, 4682–4685.
[16] C. Quinton, S. Thiery, O. Jeannin, D. Tondelier, B. Geffroy, E. Jacques, J.
Rault-Berthelot, C. Poriel, ACS Appl. Mater. Interfaces. 2017, 9, 6194–
6206.
Conclusion
To summarise, this work reports the first rational study on SBF
regioisomerism, highlighting the chief influence of the bridge
and of the linkages on the electronic properties. Of particular
interest, the impact of the ring bridging on the photophysical
properties has been evidenced, notably with different influen-
ces of the linkages and the bridge on the singlet and triplet ex-
cited states. The first member of a new family of SBFs substi-
tuted in the 1-position is reported; it possesses a very high
triplet energy and presents better performance in blue
PhOLEDs than those of its regioisomers. These features high-
light not only the potential of 1-substituted SBFs, but also the
remarkable impact of regioisomerism on electronic properties.
[17] L.-S. Cui, Y.-M. Xie, Y.-K. Wang, C. Zhong, Y.-L. Deng, X.-Y. Liu, Z.-Q. Jiang,
L.-S. Liao, Adv. Mater. 2015, 27, 4213–4217.
[18] M.-M. Xue, Y.-M. Xie, L.-S. Cui, X.-Y. Liu, X.-D. Yuan, Y.-X. Li, Z.-Q. Jiang, L.-
S. Liao, Chem. Eur. J. 2016, 22, 916–924.
&
&
Chem. Eur. J. 2017, 23, 1 – 10
www.chemeurj.org
8
ꢂ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
[19] Y. Liu, L.-S. Cui, X.-B. Shi, Q. Li, Z.-Q. Jiang, L.-S. Liao, J. Mater. Chem. C
2014, 2, 8736–8744.
[31] S. Karabunarliev, M. Baumgarten, N. Tyutyulkov, K. Mꢆllen, J. Phys. Chem.
1994, 98, 11892–11901.
[20] C. Poriel, R. Mꢀtivier, J. Rault-aBerthelot, D. Thirion, F. Barriꢅre, O. Jean-
nin, Chem. Commun. 2011, 47, 11703–11705.
[32] S. Y. Hong, D. Y. Kim, C. Y. Kim, R. Hoffmann, Macromolecules 2001, 34,
6474–6481.
[21] S. Thiery, C. Declairieux, D. Tondelier, G. Seo, B. Geffroy, O. Jeannin, R.
Mꢀtivier, J. Rault-Berthelot, C. Poriel, Tetrahedron 2014, 70, 6337–6351.
[22] S. Thiery, D. Tondelier, C. Declairieux, G. Seo, B. Geffroy, O. Jeannin, J.
Rault-Berthelot, R. Mꢀtivier, C. Poriel, J. Mater. Chem. C 2014, 2, 4156–
4166.
[33] N. Fomina, S. E. Bradforth, T. E. Hogen-Esch, Macromolecules 2009, 42,
6440–6447.
[34] C. S. Hartley, Acc. Chem. Res. 2016, 49, 646–654.
[35] J. He, J. L. Crase, S. H. Wadumethrige, K. Thakur, L. Dai, S. Zou, R. Ra-
thore, C. S. Hartley, J. Am. Chem. Soc. 2010, 132, 13848–13857.
[36] H.-H. Huang, C. Prabhakar, K.-C. Tang, P.-T. Chou, G.-J. Huang, J.-S. Yang,
J. Am. Chem. Soc. 2011, 133, 8028–8039.
[23] C. Poriel, J. Rault-Berthelot, D. Thirion, F. Barriꢅre, L. Vignau, Chem. Eur.
J. 2011, 17, 14031–14046.
[24] C. Poriel, F. Barriꢅre, D. Thirion, J. Rault-Berthelot, Chem. Eur. J. 2009, 15,
13304–13307.
[37] H. Yersin, Highly Efficient OLEDs with Phosphorescent Materials, Wiley-
VCH, Weinheim, 2007.
[25] M. Romain, M. Chevrier, S. Bebiche, T. Mohammed-Brahim, J. Rault-Ber-
thelot, E. Jacques, C. Poriel, J. Mater. Chem. C 2015, 3, 5742–5753.
[26] D. Thirion, C. Poriel, J. Rault-Berthelot, F. Barriꢅre, O. Jeannin, Chem. Eur.
J. 2010, 16, 13646–13658.
[38] A. P. Kulkarni, C. J. Tonzola, A. Babel, S. A. Jenekhe, Chem. Mater. 2004,
16, 4556–4573.
[39] M. A. Baldo, D. F. O’Brien, Y. You, A. Shoustikov, S. Sibley, M. E. Thomp-
son, S. R. Forrest, Nature 1998, 395, 151–154.
[27] M. Romain, C. Quinton, D. Tondelier, B. Geffroy, O. Jeannin, J. Rault-Ber-
thelot, C. Poriel, J. Mater. Chem. C 2016, 4, 1692–1703.
[28] L.-S. Cui, S.-C. Dong, Y. Liu, M.-F. Xu, Q. Li, Z.-Q. Jiang, L.-S. Liao, Org.
Electron. 2013, 14, 1924–1930.
[40] E. Baranoff, B. F. E. Curchod, Dalton Trans. 2015, 44, 8318–8329.
[41] C. Poriel, J. Rault-Berthelot, S. Thiery, C. Quinton, O. Jeannin, U. Biapo, B.
Geffroy, D. Tondelier, Chem. Eur. J. 2016, 22, 17930–17935.
[29] D. Thirion, C. Poriel, F. Barriꢅre, R. Mꢀtivier, O. Jeannin, J. Rault-Berthelot,
Org. Lett. 2009, 11, 4794–4797.
Manuscript received: February 7, 2017
[30] P. Guiglion, M. A. Zwijnenburg, Phys. Chem. Chem. Phys. 2015, 17,
17854–17863.
Final Article published: && &&, 0000
Chem. Eur. J. 2017, 23, 1 – 10
www.chemeurj.org
9
ꢂ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
FULL PAPER
The first structure–property relation-
ship study of a key class of organic
semiconductors, the four spirobifluor-
ene positional isomers, is reported. The
impact of the ring bridging on the pho-
tophysical properties is stressed and the
first member of a new family of spirobi-
fluorenes substituted in position 1 is de-
scribed, presenting better performance
in blue phosphorescent OLEDs than
those of its regioisomers.
&
Organic Semiconductors
L. Sicard, C. Quinton, J.-D. Peltier,
D. Tondelier, B. Geffroy, U. Biapo,
R. Mꢀtivier, O. Jeannin, J. Rault-Berthelot,
C. Poriel*
&& – &&
Spirobifluorene Regioisomerism: A
Structure–Property Relationship Study
&
&
Chem. Eur. J. 2017, 23, 1 – 10
www.chemeurj.org
10
ꢂ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!