Insights into the effect of ketylimine, aldimine, and vinylene group attachment and regiosubstitution on the fluorescence deactivation of fluorene

Article abstract of DOI:10.1139/V10-089

The spectroscopic and electrochemical properties of a 9-substituted fluorene ketylimine (3) were investigated and compared with those of its vinylene analogue (4) to determine the origins of the quenched fluorescence of these compounds. The predominate mo

Full text of DOI:10.1139/V10-089

173
Insights into the effect of ketylimine, aldimine,
and vinylene group attachment and
regiosubstitution on the fluorescence deactivation
of fluorene
´
Stephane Dufresne, Thomas Skalski, and W.G. Skene
Abstract: The spectroscopic and electrochemical properties of a 9-substituted fluorene ketylimine (3) were investigated
and compared with those of its vinylene analogue (4) to determine the origins of the quenched fluorescence of these com-
pounds. The predominate mode of singlet excited state deactivation of the heteroatomic fluorene was found to be internal
conversion involving bond rotation. Meanwhile, its carbon counterpart was found to undergo deactivation preferentially by
intersystem crossing to form its triplet, which was confirmed by laser flash photolysis. Both 3 and 4 quenched the fluores-
cence of fluorene with diffusion-controlled rate constants, implying that the singlet excited states of 3 and 4 are also
quenched by intramolecular photoinduced electron transfer (PET). This deactivation mode was found to be exergonically
favorable (–90 kJ/mol for 3 and –81 kJ/mol for 4) according to the Rehm–Weller equation. The position of the heteroa-
tomic bond on the fluorene moiety was further found to influence the singlet excited state deactivation pathway. The 2-
substituted regioisomer decayed predominately by intramolecular PET and its fluorescence can be restored by acid proto-
nation. Conversely, the PET mechanism is a minor deactivation mode for the 9-substituted fluorene derivative and its fluo-
rescence can be enhanced by suppressing bond rotational modes, possible at low temperature and potentially in thin films.
Key words: fluorenyl ketylimines, intramolecular photoinduced electron transfer, Stern–Volmer quenching, azomethines.
´
´
´
´ ´ ´ ´
´
´
´
´ ´
Resume : Afin de determiner les origines de la fluorescence degeneree de ces composes, on a etudie les proprietes spec-
´
´
`
´
troscopiques et electrochimiques des cetylimines de fluorenes portant un substituant en position 9 (3) et on les a comparees
`
`
´
´
´
´
a celles de son analogue vinylene (4). On a trouve que le mode dominant de la desactivation de l’etat singulet excite du
`
´ ´
´
fluorene heteroatomique se produit par une conversion interne impliquant une rotation de liaison. On a trouve par ailleurs
´ ´
´
´
`
`
que sa contrepartie carbonee subit une desactivation preferentielle par un croisement intersysteme conduisant a la forma-
´
´
´
´
tion de son triplet, ce qui est confirme par la photolyse laser eclair. Les composes 3 et 4 desactivent tous la fluorescence
ˆ ´
`
´
´
´
du fluorene avec des constantes de vitesse controlees par la diffusion impliquant que l’etat singulet excite des composes 3
´
´
´
´
´
´
et 4 est aussi desactive par un transfert d’electron photoinduit (TEP) intramoleculaire. On a trouve que, suivant l’equation
´
de Rehm–Weller, ce mode de desactivation est favorable d’un point de vue exorgonique (–90 kJ/mol pour 3 et –81 kJ/mol
´ ´
´
`
pour 4). On a trouve aussi que la position de la liaison heteroatomique dans la portion fluorene influence aussi la voie de
´
´
´
´
´
`
´
desactivation de l’etat singulet excite. On a aussi observe que le regioisomere portant un substituant en position 2 se desac-
´
´
ˆ
´
tive d’une fac¸on predominante par un TEP intramoleculaire et que sa fluorescence peut etre restauree par protonation
´
´
´
´
`
acide. Inversement, le mecanisme de TEP n’est qu’un mode mineur de desactivation pour les derives du fluorenes portant
ˆ
´
un substituant en position 9 pour lesquels la fluorescence peut etre augmentee par une suppression des modes de rotation
`
´
des liaisons possibles a basse temperature et possiblement dans des films minces.
´
´
´
´
´
´
Mots-cles : cetylimines de fluorenyles, transfert d’electron photoindut (TEP) intramoleculaire, desactivation de Stern–
´
Volmer, azomethines.
´
[Traduit par la Redaction]
which make them suitable for organic electronics such as
field-effect transistors, light-emitting diodes, and photovol-
taics, to name but a few.1–4 In particular, fluorene-based ma-
terials have been extensively studied because of their
Introduction
Conjugated materials have received much attention as a
result of their photophysical and electrochemical properties,
Received 30 March 2010. Accepted 8 June 2010. Published on the NRC Research Press Web site at canjchem.nrc.ca on 13 January 2011.
This article is part of a Special Issue dedicated to Professor J. C. Scaiano.
1
2
´
´
S. Dufresne, T. Skalski, and W.G. Skene. Centre for Self-Assembled Chemical Structures, Departement de Chimie, Universite de
´
Montreal, CP 6128, Centre-ville, Montreal, QC H3C 3J7, Canada.
1
´
´
´
Present address: Universite d’Artois, Unite de Catalyse et de Chimie du Solide, UMR CNRS n8 8181, Faculte Jean Perrin, rue Jean
Souvraz, S.P. 18, 62307, Lens CEDEX, France.
2Corresponding author (e-mail: w.skene@umontreal.ca).
Can. J. Chem. 89: 173–180 (2011)
doi:10.1139/V10-089
Published by NRC Research Press

174
Can. J. Chem. Vol. 89, 2011
inherent high fluorescence, which makes them ideal for use
in emitting devices.5 Fluorene and its derivatives can addi-
tionally withstand the harsh oxidative and reductive environ-
ment of a device, further making them ideal materials for
use in emitting devices.6–8 Despite these advantages, the flu-
orescence quantum yield and color emitted by fluorene are
highly sensitive to its environment and its substitution. This
is evidenced by keto defects in the 9-position of fluorene,
resulting in both decreased emission and undesired green
emission.6,9,10
Chart 1.
Although the fluorenone defect (1) (Chart 1) in polyfluor-
enes is undesired because it reduces device efficiency and
results in color contamination, the ketone group is nonethe-
less interesting because it affords the means to introduce
functionality into the conjugated fluorene polymer and to
subsequently modulate the polymer’s properties.7,11–13 Sim-
ple condensation of amines with fluorene’s ketone leads to
ketylimines. The properties of the polyfluorenes such as sol-
ubility, electronic effects, color, and emission yields, to
name but a few, can potentially be tailored by judicious
choice of the amine used in preparing the ketylimine.
Despite the potential ease of preparing fluorene ketyli-
mines such as 3 by simple condensation methods, few ex-
amples of these compounds are known.14–16 Interestingly,
no modern reports have described the successful synthesis
of 3 or its derivatives, while only classic literature reports
the preparation of some derivatives using harsh and ex-
tremely stringent reaction conditions.17,18 The limited num-
ber of reports detailing the preparation of 3 and its
derivatives, together with the absence of reported character-
ization, prompted us to prepare 3 to investigate the influence
of the heteroatomic bond on the photophysical properties of
fluorene. We were further incited to investigate the photo-
physical properties of both 3 and 4 by the lack of previous
photophysical studies and by a desire to understand the ori-
gins of their unknown quenched fluorescence. The knowl-
edge gained from investigating these model compounds is
of importance not only for understanding the influence of
the homo- and heteroatomic bonds on the photophysical
properties, but also for designing and preparing future gener-
ations of fluorenyl compounds with tailored properties for
specific applications. We additionally examined 3 because
it is complementary to our previous aldimine studies of flu-
orene derivatives such as 5 and 6 and would provide vital
structure–property information as well as information about
the regiosubstitution effect on the photophysical proper-
ties.19–24 The preparation and photophysical studies of 3 are
therefore reported herein to understand the effect of the
N=C group and its placement on the fluorenyl excited state
properties.
Spectroscopic measurements
The absorption measurements were done on a Cary 500
spectrometer, while the fluorescence studies were performed
on an Edinburgh Instruments FLS-920 fluorimeter after dea-
erating the samples thoroughly with nitrogen for 20 min.
Fluorescence absolute quantum yields were measured at
10–5 mol/L by exciting the compounds of study at the corre-
sponding maximum absorption in spectroscopic grade di-
chloromethane at room temperature in an integrating
sphere.25 The DE values were calculated from the intercept
of normalized absorption and fluorescence spectra of the
corresponding compounds. The spectroscopic band gaps
(Eg) were calculated from the tangent from the absorption
onset.
Electrochemical measurements
Cyclic voltammetric measurements were performed on a
Bioanalytical Systems EC epsilon potentiostat. Compounds
were dissolved in anhydrous and deaerated dichloromethane
at 10–4 mol/L with NBu4PF6 at sufficient concentration to
achieve highly conductive solutions. A platinum electrode
was used as the auxiliary and the working electrode, while
a saturated Ag/AgCl electrode was used as the reference
electrode. Ferrocene was added at the end of the analyses to
serve as an internal reference. The electrochemical band
gaps (Eg) were calculated from the measured oxidation and
reduction potentials according to Epa + Epc + 0.4.
Experimental procedures
General procedures
Laser flash photolysis
All reagents, including 1, 2, and 4, were commercially
available from Sigma-Aldrich and were used as received un-
less otherwise stated. Anhydrous and deaerated solvents
were obtained with a Glass Contour solvent purification sys-
Laser flash photolysis experiments were done on a Luz-
chem mini-LFP system exciting at 355 nm with the third
harmonic of a Nd-YAG laser. The solutions were prepared
with matched absorbances between 0.3 and 0.5 at the excita-
tion wavelength and deaerated for 20 min with nitrogen.
Highest occupied molecular orbital (HOMO) and lowest
1
tem. H NMR and 13C NMR spectra were recorded on a
Bruker 400 MHz spectrometer with the appropriate deuter-
ated solvents.
Published by NRC Research Press

Dufresne et al.
175
unoccupied molecular orbital (LUMO) energy levels were
calculated semiempirically using density functional theory
(DFT) calculation methods available in Spartan 06 (Wave-
function, Inc.) with the 6-31g* basis set. The bond angles,
distances, torsions, and other parameters were experimen-
tally derived from the X-ray data from the corresponding
structures.
N=C bond with a limited degree of conjugation on the fluo-
rophores. However, the desired product could not be ob-
tained despite the various reaction conditions employed. Its
instability is most likely a result of its extreme sensitivity to
hydrolysis and it spontaneously decomposes under ambient
conditions. Nonetheless, the unprecedented and complete
characterization by standard methods confirmed the forma-
tion of 3 in high purity and sufficient amounts for spectro-
scopic measurements, albeit in low yield.
N-(9H-Fluoren-9-ylidene)benzenamine (3)
Aniline (153 mg, 1.64 mmol) was dissolved with DABCO
(500 mg, 4.45 mmol) in anhydrous toluene at 0 8C. This was
followed by the slow addition of 1.0 mol/L TiCl4 solution in
toluene (2.77 mL, 2.77 mmol). Compound 1 (100 mg,
0.55 mmol) was added and then refluxed for 3–4 h. The sol-
vent was removed and the product was isolated as a yellow
solid after purification by recrystallization in 50:50 dichloro-
methane/hexanes or by flash chromatography on alumina
with neat hexanes with 1% triethylamine increased to hex-
X-ray crystallographic data
Unlike aldimines, ketylimines lack the characteristic
imine peak, allowing for unequivocal product identification
1
by H NMR. The lack of characteristic proton resonance at
8.5 ppm makes product identification challenging. Undeni-
able evidence for correct product formation is, however,
possible by X-ray diffraction. In addition to this, the relation
of the phenyl unit to the fluorene moiety can be determined
from the X-ray diffraction data, allowing for comparison of
the structures with analogous aldimines such as 6.
1
anes/ethyl acetate (95:5) (57 mg, 41%); mp 92–94 8C. H
NMR (acetone-d6) d: 7.85 (d, 1H, J = 7.6 Hz), 7.77 (m,
2H), 7.52 (m, 1H), 7.43 (m, 4H), 7.21 (m, 1H), 6.98 (m,
2H), 6.55 (d, 1H, J = 8.0 Hz). 13C NMR (acetone-d6) d:
164.0, 153.9, 145.6, 143.6, 139.4, 139.3, 133.9, 132.7,
131.2, 130.7, 130.3, 130.0, 129.5, 128.5, 127.1, 125.7,
124.8, 122.4, 121.8, 119.7.
Compound 3 was crystallized by slow evaporation from a
saturated solution of acetone to give small yellow plates.
The resolved structure (Fig. 1, left panel) confirms that 3
was formed and crystallizes in an orthorhombic unit as per
Table 1. Three different molecules of 3 were found within
the crystal lattice for a total of 24 molecules per lattice. De-
spite the large number of different molecules per lattice,
only small variances were found between them. As seen in
the right panel of Fig. 1, depicting the crystal structure
shown along the b-axis, the phenyl moiety is not coplanar
with the fluorene moiety. In fact, the planes described by
the two aromatic moieties are twisted by 708, 808, and 858
for each of the three distinct molecules isolated per lattice.
The twisting between the two planes is required to minimize
the steric hindrance between the ortho protons of the phenyl
and fluorene units. This is evident from the calculated
HOMO and LUMO energy levels seen in Fig. 2 using the
X-ray crystallographic data for the optimized geometry. It
should be noted that the observed twisting between the two
planes is greater than that for its aldimine regioisomer (6),
whose two terminal fluorenes are twisted by 268 and 658 rel-
ative to the central fluorene.24 Although the mean plane an-
gles differ for 3 and its regioisomer (6), the corresponding
bond distances and angles for the two analogues are identi-
cal within experimental error. The large twisting angle be-
tween the aromatic units nonetheless implies not only that
the 9-position of the fluorene is sterically hindered, but that
3 has a limited degree of conjugation extending only
through the fluorene=N portion. The latter is supported spec-
troscopically (vide infra). The severe steric hindrance and
the large twisting of the fluorenyl and phenyl mean plane
angles observed from the crystallographic data for 3 prevent
any higher ordered structure arising from intermolecular in-
teractions. In fact, the three-dimensional crystal network of
3 is governed by only weak face-to-p interactions. This is
in contrast to 6, which has a high degree of 3D ordering ow-
ing to many intermolecular p-stacking and C–H–p interac-
tions.
Crystal structure determination
Diffraction data for 3 were collected on a Bruker FR591
diffractometer using graphite-monochromatized Cu Ka radi-
˚
ation (1.54178 A). The structures were solved by direct
¨
methods (SHELXS97, University of Gottingen). All non-
hydrogen atoms were refined based on Fobs2 (SHELXS97,
¨
University of Gottingen), while hydrogen atoms were re-
fined on calculated positions with fixed isotropic U, using
riding model techniques.
Synthesis
Given the limited experimental protocols for the prepara-
tion of 3 and the lack of its detailed characterization, we set
out to prepare this compound. We first attempted to prepare
3 via our previously successful methods for aldimines by
simple condensation of 2 with aniline in absolute ethanol
and in neat aniline.26–28 Although some desired product was
detected by both TLC and crude NMR, purification by
standard methods yielded insufficient material for character-
ization and photophysical studies. The low yield is most
likely a result of reduced reactivity because of steric hin-
drance of the ketone, while the product formed was most
likely extremely hydrolytically sensitive and subsequently
decomposed with standard chromatographic preparative
methods.14,17,29,30 The latter is supported by the absence of
isolated products after silica gel chromatography. Stringent
reaction protocols using TiCl4 and an excess of DABCO
were used for activating the hindered ketone and preventing
imine hydrolysis, respectively. Product isolation was possi-
ble by column chromatography over activated basic alu-
mina. Alternatively, the desired product could also be
obtained by recrystallization of the crude reaction mixture
from a 50:50 dichloromethane/hexanes mixture. We also at-
tempted to prepare an aliphatic derivative of 3 using butyl-
amine to understand the photodeactivation modes of the
Published by NRC Research Press

176
Can. J. Chem. Vol. 89, 2011
Fig. 1. ORTEP X-ray structure of 3 with the ellipsoids shown at 50% probability along the a-axis (left) and b-axis (right).
Table 1. Details of crystal structure determination for 3.
tion and fluorescence spectra of 3 and 4 that the heterocon-
jugated fluorene is more conjugated than its all-carbon
counterpart. This is further supported by the shorter N=C
Compound 3
C19H13N
Formula
˚
bond distance (1.279 A) compared with the corresponding
Mw (g/mol); F(000)
Crystal color and form
Crystal size (mm)
T (K); Dcalcd. (g/cm3)
Crystal system
Space group
255.30; 3216
Yellow plate
0.15 Â 0.13 Â 0.05
150(2); 1.247
Orthorhombic
Pbca
˚
C=C bond distance (1.340 A) for 4, derived from the X-ray
crystallographic data. The shorter bond distance implies a
greater sp2 character to the bond and hence a higher degree
of conjugation, which is further corroborated by the band
gaps (Eg).
Both 3 and 4 possess substantially lower fluorescence
quantum yields than 1 in dichloromethane, as seen in Ta-
ble 2. More specifically, 4 does not fluoresce, while 3 only
weakly fluoresces at room temperature. To better understand
the origins of the differences between the two analogues and
their quenched fluorescence relative to 1, the fluorescence at
77 K was measured. All modes of rotational deactivation are
suppressed at this temperature and an observed fluorescence
increase would denote internal conversion deactivation by
bond rotation. The fluorescence yield of 3 increased to 0.6
at 77 K, while no change was observed for 4 at this low
temperature. Although an accurate ratio between the emis-
sion at 77 K and that at room temperature cannot be deter-
mined owing to the weak room temperature fluorescence, it
can nonetheless be confirmed that the major mode of singlet
excited state deactivation of 3 is by nonradiative means in-
volving bond rotation, most likely the N–phenyl bond.
Meanwhile, the singlet excited state deactivation of 4 occurs
by other means. It can be concluded from the temperature
fluorescence measurements that deactivation of excited fluo-
rene is highly influenced by its substituents. This is evident
from the temperature-dependent fluorescence of the re-
gioisomers 3 and 4, where two different deactivation modes
occur despite the vinylene and imine bonds being isoelec-
tronic. The different effect of the homoaryl and heteroconju-
gated bonds on the fluorene properties is further evidenced
from the calculated HOMO and LUMO levels for 3 and 4.
As seen in Fig. 2, the HOMO is localized exclusively over
the phenyl–N moiety and the LUMO is distributed across
the fluorenyl moiety for 3. This is in contrast to 4, whose
HOMO is distributed across the fluorenyl moiety while the
Unit cell
˚
a (A)
9.6721(2)
˚
b (A)
16.4370(3)
51.3027(10)
90.000
90.000
90.000
8156.1(3); 24
1.76–72.26; 0.996
8022/7112; 0.041
0.556
Semiempirical
0.0571; 0.1549
0.0629; 0.1580
1.297
˚
c (A)
a (8)
b (8)
g (8)
3
˚
V (A ); Z
q range (8); completeness
Reflections: collected/independent; Rint
m (mm–1)
Abs. corr.
R1(F); wR(F2) [I > 2s(I)]
R1(F); wR(F2) (all data)
GoF (F2)
–3
˚
Max. residual e density (e A )
–0.426
Spectroscopy
Given the limited reports examining the fluorescence of 3
and its analogous vinylene derivative (4),31–33 we examined
the photophysical properties of these two compounds. Such
studies are required to understand the effect and the place-
ment of the heteroconjugated bond on the fluorescence and
the origin of the weak fluorescence of these compounds.
This is possible by comparing the spectroscopic properties
of 3–5. As seen in Table 2, the three compounds exhibit ap-
proximately the same absorption maxima and differ only in
their long wavelength absorption. The same trend of spectral
shifts as a function of structures is also observed for their
fluorescence spectra. It can be concluded from the absorp-
Published by NRC Research Press

Dufresne et al.
177
Fig. 2. HOMO (left) and LUMO (right) energy levels of 3 calculated using the DFT 6-31g* basis set using the X-ray crystallographic data
for the optimized geometry.
Table 2. Various spectroscopic data.
Fluorescence quantum yielda
Absorbance
(nm)b,c
258 (377)
261
258, 298 (388)
259 (328)
361
Fluorescence
(nm)b
DE
(eV)
Eg spec
(eV)
CH2Cl2 +
TFA
0
0.45
0.03
0
Compound
1
2
3
CH2Cl2
0.02
0.50
0.04
0
77 K / RTd
7
4
15
0
—
505
302
478
402
445
2.9
4.4
2.8
3.4
2.9
3.6
3.9
2.6
3.1
2.9
4
5e
0.40
—
aAbsolute fluorescence quantum yield measured using an integrating sphere.
bMeasured in CH2Cl2.
cValues in parentheses denote the most bathochromically shifted absorption maximum.
dFluorescence quantum yield ratio measured at 77 K in a 4:1 methanol/ethanol matrix and at room temperature (RT) in methanol.
The change in refractive index is minimal for the two temperatures, and therefore no temperature-dependent fluorescence correction
is required.
eRef. 21.
Fig. 3. HOMO (left) and LUMO (right) energy levels of 4 calculated using the DFT 6-31g* basis set using the X-ray crystallographic data
for the optimized geometry.
LUMO covers the cyclopentyl, imino, and phenyl moieties
(Fig. 3).
triplet state. As seen in Fig. 4, a strong transient signal is
apparent at 450 nm for the excitation of 4 at 355 nm by la-
ser flash photolysis. The quenching of this transient with
1,3-cyclohexadiene and the unimolecular decay confirm the
triplet nature of the transient. The similar transient inten-
sities observed for optically matched samples of xanthone
and 4 suggest that the triplet quantum yield of 4 is similar
to that of xanthone (FTT = 1).34–37 Conversely, no transient
signal was detected for 3, implying the absence of triplet
The nonexistent room temperature fluorescence and the
absence of increased fluorescence at a reduced temperature
of 4 would suggest the presence of deactivation modes other
than internal conversion. A possible deactivation mode
would be intersystem crossing to form the triplet state. Both
3 and 4 were therefore further investigated by laser flash
photolysis to spectroscopically confirm the formation of a
Published by NRC Research Press

178
Can. J. Chem. Vol. 89, 2011
Fig. 4. Transient absorption spectra of xanthone (black) and 4 (red)
measured at 0.12 ms after the 355 nm laser pulse in deaerated di-
chloromethane. Inset: monoexponential decay of transient 4 moni-
tored at 450 nm.
Fig. 5. Stern–Volmer fluorescence quenching of 2 with 0 (black),
20 (red), 40 (green), 60 (blue), and 80 (orange) mmol of 3 in di-
chloromethane. Inset: change in fluorescence of 2 with the addition
of 3, monitored at 318 nm.
formation and hence no deactivation by intersystem crossing
for the heteroatomic compound. The time-resolved and
steady-state spectroscopic methods confirm that 3 preferen-
tially deactivates by internal conversion, while the singlet
excited state of 4 is predominately deactivated by intersys-
tem crossing.
vated by both 3 and 4. The results are consistent with pre-
vious quenching studies of 1 with aliphatic imines that were
deactivated by PET. The similarity of the quenching rate
constants of 3 and 4 with their fluorophore-free quenching
counterparts, concomitant with the favorable energetics
(vide infra), implies that the singlet excited state deactiva-
tion occurs by intramolecular PET.19–21
Although the temperature-dependent fluorescence con-
firms that the singlet excited state of 3 deactivates by inter-
nal conversion, the quantum yield of this process cannot be
accurately measured. This is a result of the large error asso-
ciated with comparing the intense fluorescence signal at
77 K with the extremely weak room temperature signal.
Therefore, internal deactivation modes other than rotational
internal conversion energy dissipation, such as photoinduced
electron transfer (PET), cannot be dismissed. Stern–Volmer
quenching studies of 3 and 4 were subsequently undertaken
to determine whether other modes of singlet excited state
deactivation occurred.38 Owing to the quenched fluorescence
of 3 and 4, quantitative and accurate quenching studies were
not possible using these two compounds as fluorophores.
The quenching measurements were therefore done with fluo-
rophore 2 because of its inherent high fluorescence yield. As
seen in Fig. 5, 3 quenches the fluorescence of 2 with a bi-
The diffusion-controlled kq quenching of 2 by both 3 and
4 implies deactivation by PET. The energetics of PET were
empirically calculated according to the Rehm–Weller equation
to corroborate the fluorescence quenching data: DG ꢀ = Epa
eT
(fluorophore) – Epc (quencher) – DE0,0 – l.39 The equation
takes into account the fluorophore’s oxidation potential
(Epa), the quencher’s reduction potential by accepting the
transferred electron from the fluorophore (Epc), and the sol-
vent reorganization energy (l). The energy gap (DE0,0) be-
tween the ground and excited singlet states of 2 must also
be taken into account, since PET involves electron transfer
from the excited state. The required value is calculated
from the intercept of the normalized plot of the absorption
and fluorescence spectra of the fluorophores (see the Sup-
plementary data). The Epa and Epc values (Table 3) were
measured by cyclic voltammetry and correspond to the oxi-
dation and reduction potentials of 2 and quenchers (3 and 4),
respectively. The combined spectroscopic and electrochemi-
cal data (Tables 2 and 3) were used to calculate the ener-
getics of PET. The exergonic values of –90 and –84 kJ/mol
calculated for the quenching of 1 by 3 and 4, respectively,
support the Stern–Volmer quenching data and imply that de-
activation of the singlet excited states of 3 and 4 also occurs
by intramolecular PET. This deactivation mode is consistent
with the 2-fluorenyl aldimine regioisomers that also deacti-
vate by intramolecular PET.19–21 Interestingly, the PET de-
activation mode of 2-substituted regioisomers can be
suppressed by protonating the azomethine, resulting in fluo-
rescence yields near unity.19–21 However, the fluorescence
yield of 3 does not increase upon protonation (Table 2).
This is most likely a result of the steric hindrance at the
molecular quenching rate constant (kq) of 3.3 Â 1011
L
mol–1 s–1, from the measured fluorescence lifetime of 2
(to = 10 ns). Similarly, 4 also quenches 2 with kq = 4.0 Â
1011 L mol–1 s–1. Higher quenching ratios of fo/f would be
beneficial to derive accurate kq values. However, these were
not possible because of spectroscopic screening effects of
the quenchers at higher concentrations leading to deviations
from linearity of the Stern–Volmer plots. The rate constants
measured are slighter greater than the diffusion-controlled
limit in dichloromethane (kq = 2 Â 1010 L mol–1 s–1).38 This
is a result of overlapping absorbances between the fluoro-
phore and the quencher at the excitation wavelength making
it difficult to make accurate quenching measurements.
Nonetheless, the calculated diffusion-controlled kq confirm
that the fluorene singlet excited state is efficiently deacti-
Published by NRC Research Press

Dufresne et al.
Table 3. Electochemical data recorded in anhy-
179
the fluorescence quenching mechanism. The future genera-
tion of materials for a given emitting or sensing application
can therefore be prepared with the knowledge gained from
the discovered fluorescence quenching mechanism.
drous and deaerated dichloromethane with a satu-
rated Ag/Ag+ reference electrode.
Compound
Epa (eV)
0.86
0.91
1.50
1.73
Epc (eV)
–1.03
–1.31
–1.40
–1.85
Eg (eV)
2.3
2.6
3.3
4.0
1
2
3
4
Supplementary data
Supplementary data for this article are available on the
journal Web site (canjchem.nrc.ca). CCDC 781760 contains
the crystallographic data for this manuscript. These data can
be obtained, free of charge, via www.ccdc.cam.ac.uk/conts/
retrieving.html (or from the Cambridge Crystallographic
Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK;
fax +44 1223 336033; or deposit@ccdc.cam.ac.uk).
fluorenyl 9-position and as such, the imine cannot be effi-
ciently protonated. Nonetheless, the spectroscopic and elec-
trochemical data support the conclusion that the singlet
excited states of 3 and 4 deactivate by intramolecular PET.
While this degenerate manifold is also efficiently deacti-
vated by rotational internal conversion in the case of 3, 4 is
efficiently deactivated predominately by intersystem cross-
ing to form its triplet. The net outcome of these different de-
activation processes is suppressed fluorescence for both 3
and 4.
Acknowledgements
The Natural Sciences and Engineering Research Council
of Canada (NSERC) is thanked for a Discovery Grant, Stra-
tegic Research Grant, and Research Tools and Instruments
grant allowing this work to be performed, and the Canada
Foundation for Innovation (CFI) is thanked for additional
equipment funding. W. G. S. also thanks the Royal Society
of Chemistry for a JWT Jones Travelling Fellowship, thus
allowing this manuscript to be completed. S. D. and T. S.
Conclusion
The origin of the previously unknown fluorescence sup-
pression for 9-substituted fluorenes has been determined.
Deactivation of the fluorophore excited state occurs by com-
bined efficient rotational internal conversion and intramolec-
ular photoinduced electron transfer for the heteroatomic
ketylimine 3. Although the singlet excited state of its carbon
analogue (4) also deactivates by PET, the major mode of de-
activation is by intersystem crossing resulting in efficient
triplet formation. While the triplet manifold was spectro-
scopically confirmed by laser flash photolysis, evidence for
the intramolecular PET mechanism was supported by exam-
ining the fluorescence rate constants for quenching of fluo-
rene by both the ketylimine and vinylene fluorene
derivatives. The PET mechanism was further supported by
calculating the energetics according to the Rehm–Weller
equation. Not only was the PET mechanism found to be ex-
ergonically favorable with these two quenchers, but deacti-
vation of the singlet excited state occurred with diffusion-
controlled rate constants. Unlike its 2-substituted fluorenyl
regioisomers, the fluorescence of 3 cannot be restored by
suppressing the PET mechanism by protonation. Nonethe-
less, the fluorescence quenching of 3 occurs predominately
by rotational internal conversion, while that of its 2-substi-
tuted regioisomer occurs predominately by PET. The fluo-
rescence of 3 can therefore be dramatically increased by
reducing bond rotation deactivation modes, which is possi-
ble at low temperature or potentially in the solid state (e.g.,
by spin coating as thin films). The fluorescence enhance-
ment possible with 9-heteroatomic fluorenyl imines makes
these compounds beneficial and better suited for emitting
devices than both their vinylene and 2-substituted re-
gioisomer analogues, which waste their desired singlet ex-
cited state energy by uncontrollable means. Although the
N=C bond is understood to be isoelectronic to the C=C
bond, it was found that deactivation of the singlet excited
state occurs by different modes depending on the type of un-
saturated bond. The location of the N=C moiety on the fluo-
rene was also found to play an important role in determining
´ ´
thank NSERC and Le conseil general du Nord Pas de Calais
and Le conseil municipal de Vaulx Vraucourt for graduate
scholarships, respectively. T. S. also thanks the Direction
des Relations Internationales of the Universite de Montreal.
´
´
References
´
(1) Bredas, J.-L.; Norton, J. E.; Cornil, J.; Coropceanu, V. Acc.
Chem. Res. 2009, 42 (11), 1691. doi:10.1021/ar900099h.
PMID:19653630.
(2) Bredas, J.-L.; Durrant, J. R. Acc. Chem. Res. 2009, 42 (11),
1689. doi:10.1021/ar900238j. PMID:19916562.
(3) Kamtekar, K. T.; Monkman, A. P.; Bryce, M. R. Adv. Mater.
2010, 22 (5), 572. doi:10.1002/adma.200902148. PMID:
20217752.
(4) Lee, K.; Nair, P. R.; Scott, A.; Alam, M. A.; Janes, D. B. J.
Appl. Phys. 2009, 105 (10), 102046. doi:10.1063/1.3116630.
(5) Grimsdale, A. C.; Chan, K. L.; Martin, R. E.; Jokisz, P. G.;
Holmes, A. B. Chem. Rev. 2009, 109 (3), 897. doi:10.1021/
cr000013v. PMID:19228015.
(6) Chan, K. L.; Sims, M.; Pascu, S. I.; Ariu, M.; Holmes, A.
B.; Bradley, D. D. C. Adv. Funct. Mater. 2009, 19 (13),
2147. doi:10.1002/adfm.200990055.
(7) Jo, J.; Vak, D.; Noh, Y.-Y.; Kim, S.-S.; Lim, B.; Kim, D.-Y.
J. Mater. Chem. 2008, 18 (6), 654. doi:10.1039/b714138f.
(8) Rathnayake, H. P.; Cirpan, A.; Delen, Z.; Lahti, P. M.; Kar-
asz, F. E. Adv. Funct. Mater. 2007, 17 (1), 115. doi:10.1002/
adfm.200600089.
(9) Montilla, F.; Mallavia, R. Adv. Funct. Mater. 2007, 17 (1),
71. doi:10.1002/adfm.200600141.
(10) Romaner, L.; Pogantsch, A.; Scandiucci de Freitas, P.;
Scherf, U.; Gaal, M.; Zojer, E.; List, E. J. W. Adv. Funct.
Mater. 2003, 13 (8), 597. doi:10.1002/adfm.200304360.
(11) Pijper, T. C.; Pijper, D.; Pollard, M. M.; Dumur, F.; Davey,
S. G.; Meetsma, A.; Feringa, B. L. J. Org. Chem. 2010, 75
(3), 825. doi:10.1021/jo902348u. PMID:20055375.
(12) Kappaun, S.; Scheiber, H.; Trattnig, R.; Zojer, E.; List, E. J.
W.; Slugovc, C. Chem. Commun. (Camb.) 2008, 5170.
doi:10.1039/b808407f. PMID:18956058.
Published by NRC Research Press

180
(13) Becker, K.; Lupton, J. M.; Feldmann, J.; Nehls, B. S.; Gal-
brecht, F.; Gao, D. Q.; Scherf, U. Adv. Funct. Mater. 2006,
16 (3), 364. doi:10.1002/adfm.200500550.
(14) Suzuki, K.; Matsu-Ura, N.; Horii, H.; Sugita, Y.; Sanda, F.;
Endo, T. J. Appl. Polym. Sci. 2002, 83 (8), 1744. doi:10.
1002/app.10101.
(15) Meng, Q.; Thibblin, A. J. Am. Chem. Soc. 1997, 119 (6),
1224. doi:10.1021/ja9624681.
(16) Dai, W.; Srinivasan, R.; Katzenellenbogen, J. A. J. Org.
Chem. 1989, 54 (9), 2204. doi:10.1021/jo00270a034.
(17) O’Donnell, M. J.; Polt, R. L. J. Org. Chem. 1982, 47 (13),
2663. doi:10.1021/jo00134a030.
Can. J. Chem. Vol. 89, 2011
Org. Chem. 2007, 72 (7), 2631. doi:10.1021/jo070100o.
PMID:17343421.
(28) Bourgeaux, M.; Skene, W. G. J. Org. Chem. 2007, 72 (23),
8882. doi:10.1021/jo701515j. PMID:17956121.
(29) Saroja, G.; Pingzhu, Z.; Ernsting, N. P.; Liebscher, J. J. Org.
Chem. 2004, 69 (3), 987. doi:10.1021/jo035204n. PMID:
14750836.
(30) Meng, Q.; Thibblin, A. J. Am. Chem. Soc. 1997, 119 (6),
1224. doi:10.1021/ja9624681.
(31) Liao, W.-Y.; Wang, S.-S.; Yao, H.-H.; Chen, C.-W.; Tasi,
H.-P.; Chu, W.-P.; Chen, J.-H.; Lee, S.-B. IEEE Trans.
Magn. 2005, 41, 1019. doi:10.1109/TMAG.2004.842039.
(32) He, F.; Xia, H.; Tang, S.; Duan, Y.; Zeng, M.; Liu, L.; Li,
M.; Zhang, H.; Yang, B.; Ma, Y.; Liu, S.; Shen, J. J. Mater.
Chem. 2004, 14 (18), 2735. doi:10.1039/b407262f.
(33) Liu, L.; Qiao, L.-X.; Liu, S.-Z.; Cui, D.-M.; Zhang, C.-M.;
Zhou, Z.-J.; Du, Z.-L.; Wong, W.-Y. J. Polym. Chem. A
2008, 46 (10), 3193. doi:10.1002/pola.22654.
(18) Pickard, P. L.; Tolbert, T. L. Org. Synth. 1964, 5, 51.
´
`
(19) Dufresne, S.; Perez Guarın, S. A.; Bolduc, A.; Bourque, A.
N.; Skene, W. G. Photochem. Photobiol. Sci. 2009, 8 (6),
796. doi:10.1039/b819735k. PMID:19492107.
(20) Dufresne, S.; Callaghan, L.; Skene, W. G. J. Phys. Chem. B
2009, 113 (47), 15541. doi:10.1021/jp907391y.
(21) Bourque, A. N.; Dufresne, S.; Skene, W. G. J. Phys. Chem.
C 2009, 113 (45), 19677. doi:10.1021/jp907263p.
(22) Tsang, D.; Bourgeaux, M.; Skene, W. G. J. Photochem.
(34) Carmichael, I.; Helman, W. P.; Hug, G. L. J. Phys. Chem.
Ref. Data 1987, 16 (2), 239. doi:10.1063/1.555782.
(35) Carmichael, I.; Hug, G. L. J. Phys. Chem. Ref. Data 1986,
15 (1), 1. doi:10.1063/1.555770.
Photobiol.
A
2007, 192 (2–3), 122. doi:10.1016/j.
jphotochem.2007.05.013.
(36) Scaiano, J. C. CRC Handbook of Organic Photochemistry;
CRC Press: Boca Raton, FL, 1989.
´
`
(23) Perez Guarın, S. A.; Skene, W. G. Mater. Lett. 2007, 61
(29), 5102. doi:10.1016/j.matlet.2007.04.015.
(37) Signals of identically matched intensity are not expected ow-
ing to the different triplet–triplet molar absorptivity coeffi-
cients of 4 and xanthone.
(38) Turro, N. J.; Ramamurthy, V.; Scaiano, J. C. Principles of
Molecular Photochemistry: An Introduction; University
Science Books: Sausalito, CA, 2009.
`
(24) Guarın, S. A. P.; Dufresne, S.; Tsang, D.; Sylla, A.; Skene,
W. G. J. Mater. Chem. 2007, 17, (27), 2801. doi:10.1039/
b618098a.
(25) Bolduc, A.; Dufresne, S.; Hanan, G. S.; Skene, W. G. Can.
J. Chem. 2010, 88 (3), 236. doi:10.1139/V09-166.
(26) Dufresne, S.; Skene, W. G. J. Org. Chem. 2008, 73 (10),
3859. doi:10.1021/jo8002503. PMID:18410143.
(39) Gilbert, A.; Baggott, J. Essentials of Molecular Photochem-
istry; CRC Press: Boca Raton, FL, 1991.
`
(27) Guarın, S. A.; Bourgeaux, M.; Dufresne, S.; Skene, W. G. J.
Published by NRC Research Press