Charge-transfer interactions in 4-donor 4′-acceptor substituted 1,1-diphenylethenes

Article abstract of DOI:10.1002/ejoc.200400140

Two 1,1-diphenylethenes bearing either a 4-dimethylamino or 4-methoxy group as electron donor and a 4′-cyano group as electron acceptor, as well as compounds containing only a donor or acceptor functionality, were synthesized. The observation of strong fl

Full text of DOI:10.1002/ejoc.200400140

FULL PAPER
Charge-Transfer Interactions in 4-Donor 4Ј-Acceptor Substituted
1,1-Diphenylethenes
Cornelis A. van Walree,*[a] Veronica E. M. Kaats-Richters,[a] Sandra J. Veen,[a]
Birgit Wieczorek,[a] Johanna H. van der Wiel,[a] and Bas C. van der Wiel[a]
Keywords: Charge transfer / Conjugation / Donor-acceptor systems / Solvatochromism
Two 1,1-diphenylethenes bearing either a 4-dimethylamino
or 4-methoxy group as electron donor and a 4Ј-cyano group
as electron acceptor, as well as compounds containing only
a donor or acceptor functionality, were synthesized. The ob-
flects modest ground state donor-acceptor coupling. These
results show that charge transport through a branched con-
jugation path is possible. 1-[4-(Dimethylamino)phenyl]-1-
phenylethene was found to fluoresce from two different
servation of strong fluorescence solvatochromism (originat- states. In nonpolar solvents the source is a local aniline-like
ing from a biradicaloid twisted intramolecular charge-trans-
fer state) of the donor-acceptor compounds reveals that
photoinduced charge separation through the cross-conjug-
ated 1,1-diphenylethene spacer occurs. The presence of
weak charge-transfer absorption bands in the UV spectra re-
state, whereas in polar solvents the fluorescence originates
from a twisted intramolecular charge-transfer state.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2004)
Introduction
charge-transfer absorption band is observed, and the actual
charge-separation process takes place from a locally excited
donor or acceptor chromophore. In this case a full elemen-
tary charge is transferred from the donor to the acceptor
site.
In the field of molecular electronics and optoelectronics,
charge and electron transport play important roles as they
form the foundation of a large number of (opto)electronic
phenomena. Examples are nonlinear optics,[1,2] molecular
rectification,[3,4] transport through molecular wires[5] and
the photovoltaic effect.[6] One of the ways to generate
charge transport is by photoinduced electron or charge
transfer, which takes place in compounds consisting of an
electron donor and an electron acceptor separated by a
bridge. The types of bridges investigated include an abun-
dant number of linear π-conjugated systems, but saturated
σ-systems,[3,7Ϫ10] σ-π systems,[11,12] spiro-conjugated com-
pounds[13] and monoatomic sp3-hybridized spacers[14Ϫ16]
have also been explored. Although linear π-conjugation is
generally most beneficial for donor-acceptor interaction,
photoinduced electron or charge transfer through the other
type of bridges has been found to occur as well. It should be
realised that strong interaction through a π-bridge results
in substantial ground state donor-acceptor mixing and the
appearance of a charge-transfer (CT) band in the electronic
absorption spectrum. However, this also means that charge
separation in the excited state is not complete. When a
ground state donor-acceptor interaction is absent, no
Hitherto, relatively little attention has been paid to com-
pounds in which the donor and acceptor chromophores are
separated by a single sp2-hybridized carbon atom. Work on
cross-conjugated donor-acceptor tetraethynylethenes[17Ϫ19]
revealed that charge-transfer interactions and optical non-
linearities are substantially smaller than in linear π-conju-
gated systems. Eckert et al. reported intramolecular charge-
transfer fluorescence of donor-acceptor 1,1-diarylethene de-
rivatives.[20] Furthermore, a carbonyl spacer has been used
in order to develop optically transparent second-order non-
linear optically active compounds.[21] Although it is clear
that charge-transfer interactions in cross-conjugated sys-
tems are not particularly strong, it is of interest to gain
more insight into their nature, since they reflect the tend-
ency of charge carriers to take a nonlinear conjugated path
during their transport, which may be of great value for the
development of materials containing branched or multiple
conduction channels.[22]
Here we report on the nature of charge-transfer interac-
tions through a spacer consisting of a single sp2-hybridized
carbon atom. To this end, 1,1-diphenylethenes func-
tionalized with an electron donor and/or an electron ac-
ceptor at the 4- and 4Ј-positions are studied. Dimethylam-
ino and methoxy groups are used as electron donors, while
a cyano group is applied as electron acceptor (Scheme 1).
These functionalized 1,1-diphenylethenes can be regarded
[a]
Debye Institute, Department of Physical Organic Chemistry,
Utrecht University,
Padualaan 8, 3584 CH Utrecht, The Netherlands
Fax: (internat.) ϩ 31-302534533
E-mail: c.a.vanwalree@chem.uu.nl
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
3046
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejoc.200400140
Eur. J. Org. Chem. 2004, 3046Ϫ3056

Charge-Transfer Interactions in Substituted 1,1-Diphenylethenes
FULL PAPER
as cross-conjugated systems, in which the donor and ac-
ceptor moieties are both linearly conjugated to the vinylid-
enic double bond, but not to each other. This study gives
an impression as to what extent a charge can be directed
to follow such a cross-conjugated path instead of a linear
conjugation path. In this picture, the interactions of the do-
nor and acceptor chromophores with the vinylidene
spacer and with the second phenyl group are also of inter-
est, and so the properties of the 4-methoxy-, 4-dimethylami-
no- and 4-cyano-functionalized styrenes ST-D1, ST-D2 and
ST-A are taken into account. The degree of conjugation in
a series of oligomers based on the 1,1-diphenylethene unit
has recently been discussed.[23]
Scheme 2
It is noteworthy that 1,1-diphenylethene derivatives are
not planar, as a result of steric interactions between the
hydrogen atoms ortho to the vinylidene bridge. An X-ray
structure of 1,4-bis(1-phenylvinyl)benzene revealed torsion
angles of 39.5° or 30.5° around the vinylidene-phenylene
and vinylidene-phenyl carbon-carbon single bonds, respec-
tively.[23] For all compounds under investigation, AM1 cal-
culations gave torsion angles of approximately 39° around
the vinylidene-phenylene/phenyl single bonds. It should be
realized that the deviation from planarity is an intrinsic
property of 1,1-diphenylethenes.
Scheme 1
Electronic Absorption Spectra
The charge-transfer interactions in the substituted 1,1-
diphenylethenes were examined by UV and fluorescence
spectroscopy, and the experimentally obtained results have
been interpreted by use of AM1 and PPP/SCF semi-empiri-
cal calculations. Investigation of the (steady-state) photo-
physics of the compounds under study is also relevant for
another reason: although the photophysics of 1,1-dipheny-
lethene is well documented, since the compound is fre-
quently employed in photochemical studies,[24,25] knowl-
edge of the photophysics and photochemistry of substituted
1,1-diphenylethenes is less well founded.[26] It is shown here
that 1D2 exhibits intramolecular CT emission, and the nat-
ure of the emitting state is discussed.
Electronic absorption spectra of 1D1, 1D2 and 1A in
cyclohexane, along with the spectra of styrenes ST-D1, ST-
D2, and ST-A, are shown in Figure 1, and maxima and mo-
lar absorption coefficients are compiled in Table 1. PPP/
SCF calculations were performed in order to establish the
nature of the observed transitions. Results are given in
Table 2. Although the PPP/SCF calculated values, which
should formally be regarded as gas-phase data, are not al-
ways accurate in their details, they provide a reliable and
illustrative general picture.[28] The UV spectra of donor
styrenes ST-D1 and ST-D2 exhibit weak bands located at
the red edge of a more intense band. These weak bands are
1
benzene Lb-type transitions, which are perturbed by the
presence of the donor functionality and the double bond.[29]
1
1
The Lb bands are partly obscured by La-type transitions
located at 259.5 (ST-D1) and 293.0 (ST-D2) nm. As can be
inferred from the configuration interaction data in Table 2,
Results and Discussion
Synthesis and Structure
1
the La transitions of both ST-D1 and ST-D2 involve pro-
The donor- and/or acceptor-substituted 1,1-diphenyleth- motion of an electron from the HOMO to the LUMO. Both
enes were synthesized in a few steps (Scheme 2). The appro- the HOMO and the LUMO of the donor-styrenes are delo-
priately substituted phenylmagnesium bromide or phe- calized over the entire compounds (not shown). However,
nyllithium reagent was treated either with acetophenone or the HOMO receives large contributions from the donor,
with 4-bromoacetophenone to give the 1,1-diphenylethan- while the LUMO has higher density at the double bond.
ols 3, 5, 7, 8 and 9. Acid-catalysed dehydration of these Upon HOMO-LUMO excitation there is some net charge
compounds afforded the donor compounds 1D1 and 1D2, transfer from the donor to the double bond.
as well as the bromides 4, 6 and 10. The bromides were
The UV spectra of 1D1 and 1D2 are in large part similar
converted into the corresponding cyanides 1D1A, 1D2A to those of ST-D1 and ST-D2, respectively, but also exhibit
and 1A through Rosenmund-von Braun reactions with cop- additional features. New transitions are visible at 239.5
per() cyanide in DMF.[27]
(1D1, calculated value 233 nm) and 259.0 nm (1D2, calcu-
Eur. J. Org. Chem. 2004, 3046Ϫ3056
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3047

C. A. van Walree et al.
FULL PAPER
sition is predominantly based on the transition from the
HOMOϪ1 to the LUMO, which bear a strong resemblance
to the HOMO and LUMO of 1,1-diphenylethene.[23] How-
ever, the 259.0 nm transition of 1D2 corresponds to a
HOMO Ǟ LUMOϩ3 transition and involves other types
of orbitals. Furthermore, according to the calculations, for-
bidden 1,1-diphenylethene-type transitions are present at
262Ϫ263 nm.
In comparison to ST-D1 and ST-D2, the HOMO Ǟ
LUMO (1La) transitions of 1D1 and 1D2 are slightly red-
shifted and substantially less intensive (Figure 1, Table 1).
As depicted for 1D2 in Figure 2, the HOMO is localized at
the donor site and the vinylidenic double bond. The LUMO
of 1D1 and 1D2, resembling the LUMO of 1,1-diphenyle-
thene,[23] is spread over the entire compound. Delocaliz-
ation of the LUMO lowers its energy and explains why the
1La transitions are red-shifted in relation to those of the
styrenes. Although delocalized over the entire compound,
the LUMO receives the largest contributions from the se-
cond phenyl group and, in particular, the vinylidenic double
bond. The concentration of the HOMO and the LUMO in
different parts of the molecules accounts for the decrease in
1La transition dipole moment (and consequently oscillator
1
strength). In addition, it shows that the La transitions of
the donor chromophores effectively involve charge transfer
from the donor site to the double bond, and, to a lesser
extent, to the second phenyl group.
1
4-Cyanostyrene ST-A exhibits an intensive La band at
1
263.5 nm and a weak La band at 288Ϫ298 nm (Figure 1,
c). The former is the HOMO Ǟ LUMO transition.[30] On
going from ST-A to 1,1-diphenylethene 1A, a small ba-
thochromic and a hypochromic shift of the HOMO Ǟ
LUMO transition are again observed. As can be seen in
Figure 2, the HOMO is 1,1-diphenylethene-like, while the
LUMO is largely centred on the cyanophenyl moiety. The
HOMO of 1A is more spread out than that of ST-A, affect-
ing the HOMO Ǟ LUMO transition in a way similar to
that outlined above for 1D1 and 1D2. From Figure 2 it can
be seen that this transition involves charge transfer from the
1-phenylvinyl group to the cyano functionality. The
240.5 nm band in the spectrum of 1A is based on a tran-
sition from the HOMOϪ1 to the LUMO.
The UV spectra of the donor-acceptor compounds 1D1A
and 1D2A are characterized by quite intensive bands at 255
and 280 nm, respectively, and absorption tails extending to
well beyond 300 or 350 nm (Figure 3). Second derivative
spectra indicate that actual peak maxima are situated at 258
and 294 nm for 1D1A and at 256, 281 and 339 nm for
1D2A. A difference spectrum (1D1A Ϫ ST-1D Ϫ ST-A)
gives a maximum of 292 nm with ε ϭ 2600 L·molϪ1cmϪ1
for the absorption tail of 1D1A. For the low-energy band
of 1D2A, values of λmax ϭ 341 nm and ε ϭ 2800
L·molϪ1cmϪ1 are obtained from the difference spectrum
(1D2A Ϫ ST-2D Ϫ ST-A). These data are in fine agreement
with the second derivative maxima. No significant solvent
Figure 1. UV spectra of 4-donor- or 4-acceptor-functionalized 1,1-
diphenylethenes and styrenes in cyclohexane; (a) methoxy-substi-
tuted compounds 1D1 and ST-D1, (b) dimethylamino-substituted
compounds 1D2 and ST-D2, (c) cyano-substituted compounds 1A
and ST-A; in each graph the solid line represents the spectrum of
the 1,1-diphenylethene and the dashed one the spectrum of the
styrene
Table 1. Main UV maxima (in nm) of donor/acceptor-substituted
styrenes and 1,1-diphenylethenes in cyclohexane; molar absorption
coefficients, in 103 L·molϪ1cmϪ1, are given between parentheses
1La
1Lb
ST-D1
ST-D2
ST-A
1D1
259.5 (19.3)
293.0 (23.4)
263.5 (24.1)
261.0 (11.6)
298.5 (15.1)
270[a]
293.0, 304.5 (1.4)
325[a]
288.0, 298.0 (1.0)
239.5 (18.6)
259.0 (15.6)
240.5 (20.7)
292[a]
[b]
1D2
1A
1D1A
1D2A
283[a]
294[a]
339[a]
255.0 (24.7)
280.0 (24.6)
257.0 (18.6)
[a]
[b]
Shoulder. Hidden.
lated value 256 nm), which are reminiscent of the main effect on the UV spectra of 1D1A and 1D2A was detected.
transition in the UV spectrum of 1,1-diphenylethene.[23] For
The experimentally recorded spectra of the donor-ac-
1D1 this is confirmed by the fact that the 239.5 nm tran- ceptor systems are satisfactorily reproduced by the PPP/
3048
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3046Ϫ3056

Charge-Transfer Interactions in Substituted 1,1-Diphenylethenes
FULL PAPER
Table 2. PPP/SCF-calculated electronic transitions of donor/acceptor substituted styrenes and 1,1-diphenylethenes and donor-acceptor
stilbene 2D2A[a][b]
[c]
λmax/nm
f
CI[d]
ST-D1
ST-D2
ST-A
1D1
280
257
311
299
284
282
274
262
254
233
307
302
263
256
241
230
279
275
262
242
284
279
274
261
237
342
308
288
280
266
382
310
288
275
264
0.03
0.82
0.08
0.91
0.02
1.07
0.02
0.00
0.38
0.55
0.08
0.48
0.00
0.48
0.01
0.32
0.01
0.79
0.00
0.43
0.45
0.01
0.03
0.76
0.29
0.09
0.07
0.95
0.03
0.58
1.33
0.07
0.00
0.00
0.53
0.81(1-2Ј) Ϫ 0.53(2-1Ј) ϩ 0.20(2Ϫ3Ј)
0.98(1-1Ј)
0.93(1-2Ј) ϩ 0.32(2-1Ј)
0.98(1-1Ј)
Ϫ0.56(1-2Ј) ϩ 0.79(2-1Ј)
0.98(1-1Ј)
0.79(1-3Ј) Ϫ 0.24(2-3Ј) ϩ 0.37(3-1Ј) ϩ 0.40(3-4Ј)
Ϫ0.36(1-2Ј) Ϫ 0.59(2-2Ј) Ϫ 0.37(4-4Ј) ϩ 0.58(4-1Ј)
0.96(1-1Ј)
Ϫ0.59(1-4Ј) ϩ 0.75(2-1Ј)
0.93(1-3Ј) Ϫ 0.27(3-4Ј)
1D2
0.31(1-4Ј) ϩ 0.93(1-1Ј)
0.32(1-2Ј) Ϫ 0.64(2-2Ј) Ϫ 0.30(4-4Ј) ϩ 0.57(4-1Ј) ϩ0.24(5Ϫ2Ј)
0.88(1-4Ј) Ϫ 0.20(1-5Ј) Ϫ 0.29(1-1Ј) ϩ 0.23(2-1Ј)
0.92(1-2Ј) ϩ 0.22(5-2Ј) Ϫ 0.25(4-1Ј)
0.92(2-1Ј) Ϫ 0.19(1-4Ј)
Ϫ0.42(1-3Ј) ϩ 0.37(2-3Ј) ϩ 0.80(4-1Ј)
0.90(1-1Ј) Ϫ 0.38(2-1Ј)
Ϫ0.54(1-4Ј) Ϫ 0.40(2-4Ј) ϩ 0.23(3-1Ј) ϩ 0.61(3-2Ј) Ϫ 0.24(3-5Ј)
0.50(1-2Ј) ϩ 0.33(1-1Ј) ϩ 0.20(2-2Ј) ϩ 0.73(2-1Ј)
0.87(1-1Ј) ϩ 0.43(2-1Ј)
Ϫ0.25(1-3Ј) Ϫ 0.49(2-3Ј) ϩ 0.79(4-1Ј)
0.79(1-4Ј) Ϫ 0.21(2-4Ј) Ϫ 0.26(3-5Ј) Ϫ 0.47(3-2Ј)
Ϫ0.48(1-2Ј) Ϫ 0.33(1-1Ј) ϩ 0.77(2-1Ј)
0.80(1-2Ј) ϩ 0.38(2-1Ј) Ϫ 0.23(1-1Ј) ϩ 0.28(1-5Ј)
0.95(1-1Ј) ϩ 0.23(1-2Ј)
1A
1D1A
1D2A
2D2A
0.93(1-4Ј) Ϫ 0.24(3-2Ј)
0.71(1-2Ј) ϩ 0.31(1-3Ј) ϩ 0.29(1-5Ј) ϩ 0.50(2-1Ј) Ϫ 0.16(1-1Ј)
0.20(1-3Ј) ϩ 0.25(2-2Ј) Ϫ 0.45(2-3Ј) ϩ 0.78(4-1Ј)
Ϫ0.42(1-2Ј) ϩ 0.81(2-1Ј) ϩ 0.21(5-1Ј)
0.90(1-1Ј) Ϫ 0.32(1-2Ј) Ϫ 0.26(2-1Ј)
0.91(1-4Ј) ϩ 0.26(3-2Ј)
Ϫ0.45(1-3Ј) ϩ 0.48(2-3Ј) ϩ 0.71(4-1Ј)
0.13(1-1Ј) ϩ 0.78(1-2Ј) Ϫ 0.52(2-1Ј)
0.36(1-1Ј) ϩ 0.41(1-2Ј) ϩ 0.75(2-1Ј)
[a]
Calculations were run on geometries with torsion angles of 40° between the CϭC and phenylene groups for the diphenylethenes and
[d]
[b]
[c]
of 20° for stilbene 2D2A; styrenes were assumed to be planar.
Only transitions above 230 nm are given.
Oscillator strength.
Configuration interaction data. With the exceptions of the 288 nm transition of 1D2A and the 275 nm transition of 2D2A (see text), only
coefficients larger than 0.20 are listed. For all compounds a full CI treatment was applied.
SCF calculations. The lowest-energy bands are predicted to ization is less pronounced. PPP/SCF orbitals calculated for
lie at 284 (1D1A) and 342 (1D2A) nm. This is in good 1D1A are similar to those of 1D2A.
agreement with the experimentally determined maxima of
According to the configuration interaction data in
294 nm and 339 nm, albeit that the intensity is over- Table 2, the 255 and 280 nm bands (calculated values 261
estimated for 1D1A. The low-energy bands correspond to and 288 nm, respectively) are for the major part composed
almost pure HOMO Ǟ LUMO transitions. It appears from of HOMO Ǟ LUMOϩ1 and HOMOϪ1 Ǟ LUMO tran-
Figure 2 that the HOMO of 1D2A is very similar to the sitions. In both 1D1A and 1D2A the LUMOϩ1 essentially
HOMO of donor compound 1D2, while the LUMO is corresponds to the LUMO of the donor compound, while
highly reminiscent of the LUMO of the acceptor chromo- the HOMOϪ1 strongly resembles the HOMO of 1A. The
1
phore 1A. Evidently, the absorption tail represents a direct 255 and 280 nm bands thus correspond to local La tran-
charge-transfer transition. Here it is interesting that the sitions in the donor and acceptor chromophores. From the
HOMO-LUMO mixing mediating the CT interaction oc- configuration interaction data it becomes clear that the CT
curs exclusively at the double bond. In this connection it is bands steal intensity from these transitions. For 1D1A this
1
relevant to consider the frontier orbitals of donor-acceptor mainly seems to concern the acceptor La transition (which
1
stilbene 2D2A (Scheme 1) and related linear conjugated is calculated at a lower energy than the anisole type La
compounds.[31] PPP/SCF calculations for 2D2A revealed transition),[32] while in 1D2A intensity is taken from the ani-
1
that dimethylamino donor effects extend beyond the double lino La transition. Without configuration interaction the
bond, indicating that in cross-conjugated 1D2A the delocal- calculated CT transitions are λmax ϭ 277 nm, f ϭ 0.17 for
Eur. J. Org. Chem. 2004, 3046Ϫ3056
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3049

C. A. van Walree et al.
FULL PAPER
Although the presence of CT bands in the UV spectra
of 1D1A and 1D2A demonstrates that ground state donor-
acceptor interactions are operative, the interactions are not
very strong. This is illustrated by comparison of the inten-
sities of the CT bands with those of linear conjugated do-
nor-acceptor substituted stilbenes. The CT band in 2D2A
is found at 380 nm in diethyl ether and at 383 nm in aceto-
nitrile, with ε ϭ 28500Ϫ31000 L·molϪ1cmϪ1 [33]
Although
.
this shows directly that the probability of photoinduced
charge transfer is much larger in stilbene 2D2A than in
1D2A, it is formally more appropriate to consider the oscil-
lator strength f of the CT absorptions, which is obtained
from [Equation (1)]
Ϫ9
f ϭ 4.32 ϫ 10
ε ∆ν1/2
max
(1)
in which εmax is the absorption coefficient at the band
maximum and ∆ν1/2 is the band width at half height. Taking
εmax ϭ 2800 L·molϪ1cmϪ1 and ∆ν1/2 ϭ 1960 cmϪ1 from
the difference spectrum, f ϭ 0.024 is obtained for 1D2A.
Here it should be born in mind that the band width is rather
small for a CT band, which suggests that the difference
spectrum does not uncover the complete CT band. From
Figure 2. PPP/SCF-calculated HOMO and LUMO plots of com-
pounds 1D2, 1A and 1D2A; eigenvectors are scaled as the radii of
the circles
the spectrum of 2D2A, values of εmax
ϭ 28500
L·molϪ1cmϪ1 and ∆ν1/2 ϭ 4320 cmϪ1 were extracted,[33] so
f ϭ 0.53. The difference in oscillator strength reflects the
difference in electronic coupling between the ground and
the CT states in the two compounds. Unfortunately, a more
thorough treatment is complicated by the facts that quanti-
tative expressions are not available for systems with such
different CT band intensities and that the CT transitions
borrow intensity from the local transitions. While it was
seen above that for 1D2A 60% of the CT band intensity is
stolen from the local transitions, for 2D2A this amounts
to 40%.
It is worthwhile to relate the intensity of the CT band of
1D2A with those of compounds containing an dimethylam-
ino-cyanophenyl donor-acceptor combination and
a
monoatomic sp3-hybridized carbon or silicon spacer.[15] The
intensities were 2100 (Si) and 2400 (C) L·molϪ1cmϪ1, not
very much smaller than that of 1D2A. Apparently the hy-
bridization of a monoatomic spacer does not affect the
magnitude of the CT interaction very strongly.
Fluorescence
The fluorescence maximum of compound 1A in cyclo-
hexane is situated at 329 nm, while in acetonitrile it is at
360 nm. This small but distinct solvatochromism indicates
that the lowest excited state is slightly dipolar in nature and
that a limited extent of charge separation occurs. This was
already deduced from the UV spectrum of 1A. With quan-
tum yields of 9 ϫ 10Ϫ4 in cyclohexane and 2.0 ϫ 10Ϫ3 in
acetonitrile, the fluorescence of 1A is weak. The fluor-
escence maximum of 1D1 varies from 332 nm in cyclohex-
Figure 3. (a) UV spectrum of 1D1A (Ϫ) and its reference chrom-
ophores ST-D1 (Ϫ Ϫ Ϫ) and ST-A (···); (b) UV spectrum of 1D2A
(Ϫ) and its reference chromophores ST-D2 (Ϫ Ϫ Ϫ) and ST-A (···)
1D1A and λmax ϭ 315 nm, f ϭ 0.04 for 1D2A, suggesting ane to 358 nm in THF (with a vibrational spacing of ca.
that in both cases about 60% of the CT band intensity is 1300 cmϪ1) to 360 nm in acetonitrile. Here as well, a small
taken from the local transitions.
degree of charge separation occurs. The fluorescence quan-
3050
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org Eur. J. Org. Chem. 2004, 3046Ϫ3056

Charge-Transfer Interactions in Substituted 1,1-Diphenylethenes
FULL PAPER
tum yield is an order of magnitude higher than for 1A. In
cyclohexane it is 9.4 ϫ 10Ϫ3, while in acetonitrile it is 8.6
ϫ 10Ϫ3. The fluorescence quantum yields of 1D1 and 1A
are much lower than those of cyanobenzene (0.23 in cyclo-
hexane[15]) and anisole (0.45 in cyclohexane[15]), respec-
tively. This can be explained by the presence of excited
states localized at the double bond which are close in energy
1
to the emitting Lb state. Rapid radiationless deactivation,
as in 1,1-diphenylethene,[23,34] then occurs through twisting
of the double bond. For 4-cyanostilbene similar reasoning
has been used.[35]
The fluorescence behaviour of 1D2 is quite complex. As
can be seen from the data in Table 3 and the Supplementary
Information, the fluorescence is fairly solvatochromic; upon
going from cyclohexane to acetonitrile a bathochromic shift
of 6090 cmϪ1 is observed. Figure 4 displays a plot in which
Figure 4. LippertϪMataga plot of the fluorescence solvatochro-
mism of 1D2; squares, nonpolar solvents; circles, polar solvents;
the lines are results of linear fits to Equation (2)
the wavenumber at maximum fluorescence νfl is plotted vents, the best-fit results were: slope Ϫ26.4 Ϯ 5.0 ϫ 103
against the solvent polarity parameter ∆f. A linear relation- cmϪ1, intercept 32.4 Ϯ 1.7 ϫ 103 cmϪ1 and correlation co-
ship is expected according to the LippertϪMataga relation efficient Ϫ0.967. Fluorescence band widths ∆ν1/2 (Table 3)
[Equation (2) and Equation (3)][36]
also suggest differences between nonpolar and polar sol-
vents. In the former they range from 3580 to 4130 cmϪ1
,
whereas in the latter the bands abruptly become much
broader, with widths between 4810 and 5310 cmϪ1. The
small increase in the band width on going from dibutyl
ether to diethyl ether suggests that a second band gradually
emerges. This is corroborated by close inspection of the
spectral shape, which indicates a broadening at the red side
(see Supporting Information).
(2)
(3)
with
From the fluorescence data in polar solvents it can be
concluded that 1D2 possesses a first excited state with a
large dipole moment. The observation of different solva-
2
tochromic sensitivities 2µe /hcρ3 and different band widths
Here νfl(0) represents the (hypothetical) gas-phase fluor-
escence wavenumber, µe the excited state dipole moment, h
the Planck constant, c the velocity of light and ρ the solute
cavity radius. The solvent polarity parameter ∆f is a func-
tion of the dielectric constant ε and the refractive index
n.[37] In the case of 1D2 two separate linear relationships
are found. One holds for the nonpolar solvents cyclohexane
to diethyl ether, the other is valid for the polar solvents
ethyl acetate, THF, butyronitrile and acetonitrile. For the
nonpolar solvents a linear regression analysis to Equa-
for nonpolar and polar solvents moreover strongly suggests
that different emitting species are present as a function of
solvent polarity. In most solvents only a single species is
present, but in diisopropyl ether and diethyl ether the two
species coexist. This kind of behaviour is not unusual for
a 4-substituted dimethylaniline; dual fluorescence has been
observed for a large number of dialkylaniline deriva-
tives.[35,38] The short-wavelength component emanates from
a slightly polar locally excited (LE) aniline-1Lb-like state,
while the long-wavelength band originates from a highly
polar intramolecular charge-transfer (ICT) state. In nonpo-
lar solvents the LE state is lower in energy than the ICT
state, whereas in polar solvents the ICT state is more stable.
The exact nature of the ICT state such as observed for 1D1
has given rise to controversy. Neglecting some other opi-
nions, one view is that the anilino group (or another moi-
ety) is rotated 90° out of the conjugation plane by twisting
along a ground state formal single bond. This is referred
to as a twisted ICT (TICT) state.[35,38] Such a TICT state
possesses biradicaloid character. According to the other
point of view, a planar quinoid-like structure is adopted (a
planar intramolecular CT or PICT state).[39,40] Convincing
evidence for the former hypothesis has recently been put
forward.[41]
2
tions (2) and (3) resulted in a slope 2µe /ρ3 of Ϫ5.8 Ϯ 1.0
ϫ 103 cmϪ1, an intercept νfl(0) of 28.4 Ϯ 0.2 ϫ 103 cmϪ1
and a correlation coefficient of Ϫ0.974. For the polar sol-
Table 3. Fluorescence properties of donor compound 1D2
[a]
Solvent
∆f
νfl /103 cmϪ1 Φfl
∆ν1/2 /103 cmϪ1
Cyclohexane
Di-n-butyl ether
Diisopropyl ether 0.234 27.03
Diethyl ether
Ethyl acetate
THF
Butyronitrile
Acetonitrile
0.100 27.74
0.193 27.36
0.29 3.58
[b]
[b]
3.58
3.96
0.251 26.81
0.292 24.45
0.308 24.39
0.376 22.83
0.392 21.65
0.21 4.13
0.19 5.08
0.21 4.81
[b]
5.14
0.23 5.31
Irrespective of the exact nature of the CT state, the fact
that 1D2 populates this state clearly shows that photoind-
[a]
[b]
Fluorescence quantum yield. Not determined.
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3046Ϫ3056
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3051

C. A. van Walree et al.
FULL PAPER
˚
uced electron transport occurs. If a cavity radius ρ of 4.2 A
Strong fluorescence solvatochromism is also observed for
is used,[42] for the polar excited state a dipole moment of donor-acceptor compounds 1D1A and 1D2A. This is illus-
13.9 D is obtained upon application of Equations (2) and trated for 1D2A in Figure 5, while fluorescence data are
(3). The dimethylamino group must function as an electron given in Table 4. Unlike that of 1D2, the fluorescence solva-
donor, while the 1-phenylvinyl group acts as an electron tochromism for both 1D1A and 1D2A obeys a single linear
acceptor. Interestingly, 1,1-diphenylethene was previously relationship; correlation coefficients are excellent (Table 4).
found to behave as an electron donor in intermolecular For a given compound, in all solvents an identical emitting
photoinduced electron transfer reactions.[26,43] The redox species of highly dipolar nature must be present. With use
[42]
˚
activity of the 1,1-diphenylethene-based unit thus comprises of cavity radii of 4.4 and 4.5 A, excited state dipole mo-
both electron-donating and electron-accepting func- ments of 13.0 and 17.9 D are obtained for 1D1A and 1D2A,
tions.
respectively. This shows that a larger degree of charge sep-
The question arises as to what extent the cross-conju- aration occurs in 1D2A than in 1D1A. In fact, the µe value
gated phenyl group plays a role in the electron-accepting of 1D1A is even smaller than µe of the polar CT state of
behaviour of the 1,1-diphenylethene spacer. To find an 1D2. Apparently, the presence of a strong donor is of more
answer, a comparison with the fluorescent properties of 4- importance for the generation of a large excited state dipole
(dimethylamino)styrene ST-D2 and 4-(dimethylamino)stil- moment than the presence of both a donor and a acceptor.
bene 2D2 is illustrative. The fluorescence maximum of ST- In spite of the reduced donor-acceptor distance (AM1 opti-
˚
D2 shows only a very limited shift with the solvent polarity. mized geometries reveal that the NϪN distance is 10.9 A
˚
It is situated at 358 nm in cyclohexane, at 365 nm in both in 1D2A vs. 13.4 A in stilbene 2D2A), the solvatochromic
2
THF and ethyl acetate, and at 371 nm in acetonitrile. sensitivity 2µe /ρ3 and the excited state dipole moment of
Hence, the presence of the second phenyl group in 1D2 is 1D2A are larger than those of 2D2A.[33] This is in accord-
essential for photoinduced charge transfer to occur. Stil- ance with the less pronounced ground state donor-acceptor
bene 2D2 exhibits fluorescence solvatochromism, but it is interaction in 1D2A; the weaker the coupling in the ground
less pronounced than that of 1D2. By fitting literature state, the larger the degree of charge separation in the ex-
data[44] to Equations (2) and (3),[45] a single correlation with cited state.
slope Ϫ13.3 Ϯ 1.4 ϫ 103 cmϪ1, intercept 27.8 Ϯ 0.4 ϫ 103
cmϪ1 and correlation coefficient Ϫ0.990 was obtained. This
indicates that for the stilbene only a single excited species
with a dipole moment substantially smaller than that of
1D2 is present. The presence of different emitting species
for 1D2 and 2D2 is corroborated by the fluorescence band
widths. For the polar excited state of 1D2 these are much
larger than for the stilbene, which range from 3300 to 3690
cmϪ1 [44]
Hence, both the solvatochromic shift and larger
.
band widths indicate that the degree of charge separation
is larger in 1D2. This is in line with the expectation that the
degree of photoinduced charge separation is larger in the
less conjugated system, something which can nicely be seen
from comparison of the HOMOs and LUMOs of 1D2 and
2D2.[31] In 1D2 the HOMO is entirely localized on the anil-
ino site and the vinylidenic double bond, whereas in 2D2 it
is also spread over the second phenyl group. The LUMO of
1D2 is slightly more centred at the unsubstituted phenyl
group than the LUMO of 2D2.
Figure 5. Fluorescence solvatochromism of 1D2A: (a) cyclohexane,
(b) di-n-butyl ether, (c) diethyl ether, (d) ethyl acetate, (e) THF,
(f) acetonitrile
Table 4. Fluorescence properties of donor-acceptor compounds
1D1A and 1D2A
Fluorescence quantum yields of 1D2 are of the same or-
der of magnitude as those of dimethylaniline (0.19 in cyclo-
hexane[15]) and are much larger than the quantum yields
of 1D1 and 1A. In both nonpolar and polar solvents the
1D1A
1D2A
[a]
[a]
[b]
[a]
[b]
[a]
Solvent
∆f
νfl
Φfl ∆ν1/2 νfl
Φfl
∆ν1/2
fluorescence quantum yields of 1D2 are also an order of Cyclohexane
magnitude larger than those of 2D2, which are about
0.100 29.45 0.016 3.79 25.71 0.087 3.66
[c]
Di-n-butyl ether 0.193 27.97
4.28 22.86 0.20 3.93
0.251 27.03 0.031 4.74 21.48 0.19 4.02
0.292 25.91 0.053 4.97 19.03 0.082 4.38
0.308 25.71 0.079 4.86 19.03 0.13 4.23
0.392 23.58 0.12 5.09 16.53 0.0018 4.67
Diethyl ether
0.03.[44] This shows that the deactivation process governing
Ethyl acetate
the decay of 2D2, involving a twist of the double bond,[44]
THF
does not play a significant role in 1D2. Apparently, the ani-
line-centred lowest excited state of 1D2 is so low in energy
that radiationless deactivation through 1,1-diphenylethene-
like states is prevented. It is furthermore remarkable that
the fluorescence quantum yields of 1D2 are hardly affected
by the nature of the solvent.
Acetonitrile
2
Ϫ2µe /hcp3 [a]
20.0 Ϯ 1.2
31.7 Ϯ 0.3
Ϫ0.993
32.2 Ϯ 1.8
29.0 Ϯ 0.5
Ϫ0.994
νfl(0)[a]
r
[a]
[b]
[c]
Units 103 cmϪ1
.
Fluorescence quantum yield.
Not deter-
mined.
3052
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3046Ϫ3056

Charge-Transfer Interactions in Substituted 1,1-Diphenylethenes
FULL PAPER
Fluorescence quantum yields are larger for 1D1A than conjugation, but is still sizeable. As a consequence of the
for 1D1 and 1A, which is in line with expectation when the weaker ground state interaction, the degree of charge separ-
energy of the emitting state becomes more distant from the ation upon excitation is larger than in the stilbene. Because
1,1-diphenylethene-like states. Nonetheless, in common of the cross-conjugated nature of the spacer, compounds
with the trend for donor chromophores 1D1 and 1D2, the 1D1A and 1D2A must adopt biradicaloid TICT states. The
quantum yields of 1D1A are much lower than those of occurrence of photoinduced charge separation is not, how-
1D2A, except in acetonitrile. This means that in 1D1A a ever, restricted to the donor-acceptor substituted 1,1-di-
prominent deactivation pathway is still present. Since fluor- phenylethenes; it is observed for 1D2 as well. This shows
escence quantum yields of 2-(4-cyanophenyl)-2-(4Ј-meth- that the 1,1-diphenylethene moiety does not behave as a
oxyphenyl)propane range from 0.25 to 0.58 in the same set simple spacer. In addition to its already known electron-
of solvents,[15] the double bond must still play a role in the donating behaviour[26,43] it also exhibits electron-accepting
deactivation mechanism. Only in acetonitrile is the ICT behaviour.
state so stabilized that no double bond twisted state is
within reach. For 1D2A this is the case in all solvents
(cyclohexane possibly being an exception), meaning that
fluorescence quantum yields reach their maximum in me-
Experimental Section
dium polarity solvents. This trend has been observed before
General: Reactions involving organolithium or Grignard reagents
were conducted under a nitrogen atmosphere with use of standard
Schlenk techniques. Starting materials and reagents were obtained
from commercial sources and were used as received unless stated
otherwise. Solvents were generally distilled before use; dry diethyl
ether and toluene were obtained by distillation from sodium-benzo-
phenone. Column chromatography was performed with Acros silica
(0.035Ϫ0.070 mm, pore diameter ca. 6 nm). NMR spectra were ob-
tained on Bruker AC 300 or Varian Unity Inova 300 spectrometers,
operating at 300 MHz for 1H and 75 MHz for 13C NMR spec-
troscopy. All NMR spectra were recorded in CDCl3 and are refer-
enced to TMS. Infrared spectra were recorded with a Mattson Gal-
when intensity stealing is involved.[15,46]
It is of interest to speculate on the structure of the fluor-
escent CT state of 1D2A and 1D1A. Although in principle
the negative charge could find itself at the double bond,
there is in practice little doubt that the negative charge is
located on the strong cyanophenyl acceptor. This is indi-
cated both by the larger dipole moment and the much lower
energy of the CT state of 1D2A vs. that of 1D2 and by the
fact that strong fluorescence solvatochromism is observed
for 1D1A but not for 1D1. A bona fide quinoid resonance
structure with a positive charge at the amino group and a axy Series FTIR 5000 instrument in diffuse reflectance mode on
dispersions in KBr. Elemental analyses were carried out at the
Kolbe Microanalytisches Laboratorium, Mülheim an der Ruhr,
Germany. Since some of the diphenylethenes were found to be sus-
ceptible to oxidation,[47] all compounds were stored at Ϫ20 °C un-
der nitrogen atmosphere.
negative charge at the cyano group cannot be drawn, be-
cause of the cross-conjugated nature of the spacer. Conse-
quently, the PICT model does not hold for donor-acceptor
1,1-diphenylethenes. The ICT state must have biradicaloid
character (Figure 6), and 1D2A and 1D1A adopt a TICT
structure. It is not certain which bond is twisted, since twist-
ing around bonds other than the NϪaryl bond can also be
involved in the formation of TICT states.[35,38] By the same
reasoning it is highly probable that the fluorescence of 1D2
in polar solvents also originates from a TICT state.
UV spectra were collected on Cary 1 or Cary 5 spectrophotometers
in spectrophotometric grade solvents. Fluorescence spectra were
obtained on a Spex Fluorolog instrument, equipped with a Spex
1680 double excitation monochromator, a Spex 1681 emission
monochromator and a Spex 1911F detector. Fluorescence emission
spectra were corrected for the detector spectral response with the
aid of a correction file provided by the manufacturer. Fluorescence
quantum yields[48] were determined relative to anthracene (Φfl
ϭ
0.27, excitation wavelength 310 nm) or naphthalene (Φfl ϭ 0.23,
excitation wavelength 270 or 280 nm). An excitation wavelength of
310 nm was used for anilines, while 270 or 280 nm excitation was
employed for other compounds. Solutions were degassed by purg-
ing with argon for 15 minutes. Cyclohexane, ethyl acetate, aceto-
nitrile, diethyl ether and THF used for fluorescence measurements
were of spectrophotometric grade (Acros) and were dried over mo-
lecular sieves prior to use. Butyronitrile (Fluka, 99 ϩ%) was dis-
tilled from calcium hydride. Diisopropyl ether (Acros, 99 ϩ%) and
dibutyl ether (Fluka, p.a.) were purified as described elsewhere.[12]
PPP/SCF/CI calculations were performed with a home-adapted
version of Griffiths’ program.[49] Torsion angles of 40° around the
CϪC formal single bonds were introduced by correction of the
bond resonance energy with cos40°. In all cases a full configuration
interactions was performed. AM1 calculations were run with the
Quantum CAChe package.[50] Geometries were optimized with the
eigenvector following routine and the precise option.
Figure 6. Biradicaloid structure of the CT state of 1D2A; note that
the radical sites can also be situated at other atoms in the rings or
at the vinylidenic double bond
Conclusion
From the occurrence of strong fluorescence solvatochro-
mism and the presence of an intramolecular charge-transfer
absorption it is evident that photoinduced charge separ-
ation occurs in donor-acceptor substituted 1,1-diphenyleth-
enes. In line with a few previous examples,[17Ϫ21] charge
transport over a spacer consisting of a single sp2-hybridized
carbon atom, and hence in a branched conjugation path,
proves to be possible. The electronic interaction is not as
4-Methoxystyrene (ST-D1) and 4-cyanostyrene (ST-A) were pur-
chased from Acros and were used as received. 4-(Dimethylamino)-
efficient as in an identically substituted stilbene with linear styrene (ST-D2) was prepared as described by Hollywood et al.[51]
Eur. J. Org. Chem. 2004, 3046Ϫ3056
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3053

C. A. van Walree et al.
FULL PAPER
1-(4-Bromophenyl)-1-(4-methoxyphenyl)ethene (4): A solution of 4-
(methoxyphenyl)magnesium bromide was prepared by allowing 4-
bromoanisole (14.14 g, 75.6 mmol) and magnesium (2.19 g, 2.98 [s, 6 H, N(CH3)2], 5.23 and 5.38 (AB, JH,H ϭ 1.2 Hz, 2 ϫ 1
90.1 mmol) to react in diethyl ether (85 mL). A solution of 4-
bromoacetophenone (14.97 g, 75.2 mmol) in diethyl ether (35 mL)
was added dropwise to the Grignard reagent, after which the mix-
was recrystallized from methanol/chloroform, 10:1 v/v. Yield 4.50 g
1
(14.9 mmol, 73% with respect to 5). M.p. 127 °C. H NMR: δ ϭ
2
H, ϭCH2), 6.7 and 7.2 (AAЈBBЈ, 3JH,H ϭ 8.8 Hz, 2 ϫ 2 H, ArϪH),
3
7.24 and 7.45 (AAЈBBЈ, JH,H ϭ 8.5 Hz, 2 ϫ 2 H, ArϪH) ppm.
13C NMR: δ ϭ 40.5, 111.8, 112.0, 121.4, 128.8, 128.9, 130.1, 131.2,
ture was heated at reflux for 1.5 h. Subsequently, sulfuric acid 141.3, 148.8, 150.3 ppm. IR: ν˜max ϭ 2802, 1605, 1522, 1354, 897,
(30%, 20 mL) was added, and the mixture was heated under reflux
for another hour.[52] Water (30 mL) and diethyl ether (120 mL) were
added, the layers were separated, and the aqueous layer was ex-
tracted with diethyl ether (2 ϫ 50 mL). The combined organic ex-
tracts were dried with magnesium sulfate, filtered, and concentrated
to dryness.
824 cmϪ1
.
1-(4-Cyanophenyl)-1-[4-(dimethylamino)phenyl]ethene (1D2A):
mixture of (4.00 g, 13.3 mmol), copper() cyanide (1.37 g,
A
6
15.3 mmol) and DMF (10 mL) was heated at reflux under a nitro-
gen atmosphere for six hours. After workup as described for 1D1A,
the crude product was purified by column chromatography (silica,
eluent chloroform), giving a light yellow solid (1.33 g, 5.36 mmol,
40%). M.p. 161 °C. 1H NMR: δ ϭ 2.98 [s, 6 H, N(CH3)2], 5.31 and
A
1H NMR spectrum of the resulting product showed that dehy-
dration of alcohol 3 to 1,1-diphenylethene derivative 4 was not com-
plete. The product was therefore dissolved in toluene in the presence
of a trace of p-toluenesulfonic acid and boiled for 3 hours in a flask
fitted with a DeanϪStark separator. After removal of the toluene in
a rotary evaporator, recrystallization of part (11.80 g) of the resulting
solid from ethanol gave 4 (7.70 g, 26.6 mmol) as an off-white solid.
M.p. 89 °C. 1H NMR: δ ϭ 3.82 (s, 3 H, OCH3), 5.33 and 5.40 (AB,
2
5.48 (AB, JH,H ϭ 1.1 Hz, 2 ϫ 1 H, ϭCH2), 6.68 and 7.16
3
(AAЈBBЈ, JH,H ϭ 8.9 Hz, 2 ϫ 2 H, ArϪH), 7.45 and 7.62
(AAЈBBЈ, JH,H ϭ 8.5 Hz, 2 ϫ 2 H, ArϪH) ppm. 13C NMR: δ ϭ
3
40.2, 110.8, 111.8, 113.3, 118.8, 127.7, 128.7, 128.9, 131.7, 146.9,
148.3, 150.2 ppm. IR: ν˜max ϭ 2810, 2226, 1607, 1523, 1358, 897,
856, 820 cmϪ1. C17H16N2 (248.3): calcd. C 82.23, H 6.49, N 11.28;
found C 81.88, H 6.55, N 11.16.
2JH,H ϭ 1.1 Hz, 2 ϫ 1 H, ϭCH2), 6.85 and 7.22 (AAЈBBЈ, 3JH,H
ϭ
3
8.8 Hz, 2 ϫ 2 H, ArϪH), 7.20 and 7.46 (AAЈBBЈ, JH,H ϭ 8.6 Hz,
2 ϫ 2 H, ArϪH) ppm. 13C NMR: δ ϭ 55.4, 113.4, 113.7, 121.7,
129.4, 130.0, 131.3, 133.5, 140.8, 148.6, 159.6 ppm. IR: ν˜max ϭ 2835,
1-(4-Methoxyphenyl)-1-phenylethene (1D1): A solution of 4-(meth-
oxyphenyl)magnesium bromide was prepared by allowing 4-bromo-
anisole (9.35 g, 50.0 mmol) and magnesium (1.50 g, 61.7 mmol) to
react in diethyl ether (50 mL). A solution of acetophenone (6.00 g,
50.0 mmol) in diethyl ether (15 mL) was added dropwise, and the
mixture was heated at reflux for two hours. After the mixture had
cooled to room temperature, water (100 mL) and hydrochloric acid
(1 , 50 mL) were added cautiously. After separation of the layers
the organic one was washed with water (3 ϫ 50 mL), dried over
magnesium sulfate and filtered, and the solvents were evaporated
to dryness. Yield 10.64 g (46.6 mmol, 93%) of 7 in the form of a
yellow oil, which was used without further purification. Compound
1D1 was prepared from 7 by Dean-Stark dehydration as described
for 6. The crude product, which was obtained in quantitative yield,
was recrystallized from methanol to afford white crystals in 53%
1604, 1508, 1246, 1027, 901, 829 cmϪ1
.
1-(4-Cyanophenyl)-1-(4-methoxyphenyl)ethene (1D1A): A mixture
of 4 (7.60 g, 26.3 mmol), copper() cyanide (2.78 g, 31.0 mmol) and
DMF (19 mL) was heated at reflux under nitrogen for 20 hours.
Subsequent to cooling to room temperature the mixture was
poured into ammonia (25%, 300 mL), after which air was passed
through for 3 hours. The aqueous system was extracted three times
with chloroform/hexane (90 mL, 1:1 v/v). The combined extracts
were washed with water (200 mL) and dried on magnesium sulfate.
Evaporation under reduced pressure gave a brown oil (5.75 g,
24.4 mmol, 93%). Purification of 3.0 g of the crude product by col-
umn chromatography on silica with chloroform as eluent afforded
a yellow oil (2.18 g), which very slowly (weeks) solidified on stand-
ing. M.p. 48.5 °C. 1H NMR: δ ϭ 3.82 (s, 3 H, OCH3), 5.42 and
1
yield. M.p. 73.6 °C. H NMR: δ ϭ 3.83 (s, 3 H, OCH3), 5.36 and
5.40 (AB, JH,H ϭ 1.4 Hz, 2 ϫ 1 H, ϭCH2), 6.86 and 7.26
2
2
(AAЈBBЈ, 3JH,H ϭ 8.8 Hz, 2 ϫ 2 H, ArϪH), 7.32 (m, 5 H, ArϪH)
ppm, in agreement with literature data.[53] 13C NMR: δ ϭ 55.3,
112.9, 113.5, 127.6, 128.1, 128.3, 129.4, 134.0, 141.8, 149.4,
159.3 ppm. IR: ν˜max ϭ 2836, 1607, 1510, 1252, 1028, 901, 841, 785,
5.51 (AB, JH,H ϭ 0.8 Hz, 2 ϫ 1 H, ϭCH2), 6.88 and 7.21
3
(AAЈBBЈ, JH,H ϭ 8.9 Hz, 2 ϫ 2 H, ArϪH), 7.43 and 7.61
3
(AAЈBBЈ, JH,H ϭ 8.6 Hz, 2 ϫ 2 H, ArϪH) ppm. 13C NMR: δ ϭ
55.3, 111.3, 113.8, 115.3, 118.8, 128.9, 129.3, 132.0, 132.6, 146.4,
148.2, 159.7 ppm. IR: ν˜max ϭ 2837, 1256, 2230, 1603, 1510, 1030,
901, 839, 766 cmϪ1. C16H13NO (235.3): calcd. C 81.67, H 5.57, N
5.95, O 6.80; found C 81.50, H 5.47, N 5.84, O 6.95.
708 cmϪ1
.
1-[4-(Dimethylamino)phenyl]-1-phenylethene (1D2): Firstly, 1-[4-(di-
methylamino)phenyl]-1-phenylethanol (8) was synthesized from 4-
(dimethylamino)phenyllithium (50.0 mmol) and acetophenone
(6.00 g, 49.9 mmol) as described for 5. Yield 10.38 g (43.0 mmol,
86%) of a yellowish oil. Subsequently 1D2 was obtained from 8 by
the procedure given for 6. Purification was accomplished by kugel-
rohr distillation. The fraction boiling at 110Ϫ150 °C at 0.25 Torr
1-(4-Bromophenyl)-1-[4-(dimethylamino)phenyl]ethene (6): A solu-
tion of 4-bromo-dimethylaniline (5.05 g, 25.2 mmol) in diethyl
ether (50 mL) was added dropwise to finely cut lithium pieces
(0.60 g, 86.4 mmol) in diethyl ether (20 mL). After the mixture had
been stirred overnight at room temperature, a solution of 4-bromo-
acetophenone (5.04 g, 25.3 mmol) in diethyl ether (50 mL) was ad-
ded dropwise. The mixture was next heated at reflux for two hours
and then allowed to cool down to room temperature. Ethanol
(25 mL) and water (120 mL) were added successively and the layers
were separated, upon which the ethereal layer was washed with
water (3 ϫ 25 mL). Drying over magnesium sulfate and concen-
tration under reduced pressure afforded 5 (6.49 g, 20.3 mmol) as a
yellowish oil. Subsequently, a solution of 5 (6.49 g, 20.3 mmol) in
toluene (100 mL) was heated to reflux in a Dean-Stark apparatus
for three hours in the presence of a catalytic amount of p-tolu-
was collected. Yield 78% of a light yellow solid. M.p. 54 °C. 1H
NMR: δ ϭ 2.98 [s, 6 H, N(CH3)2], 5.26 and 5.38 (AB, JH,H
2
ϭ
3
1.2 Hz, 2 ϫ 1 H, ϭCH2), 6.69 and 7.16 (AAЈBBЈ, JH,H ϭ 8.8 Hz,
2 ϫ 2 H, ArϪH), 7.35 (m, 5 H, ArϪH) ppm, in agreement with
literature data.[53] 13C NMR: δ ϭ 40.3, 111.2, 111.8, 127.2, 127.8,
128.3, 128.8, 129.3, 142.1, 149.7, 150.0 ppm. IR: ν˜max ϭ 2805, 1611,
1522, 1356, 889, 820, 777 cmϪ1
.
1-(4-Bromophenyl)-1-phenylethene (10): A solution of phenylmag-
nesium bromide was prepared by treatment of bromobenzene
enesulfonic acid. After removal of the toluene, the remaining solid (4.35 g, 27.7 mmol) with magnesium (0.80 g, 32.9 mmol) in diethyl
3054  2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.eurjoc.org
Eur. J. Org. Chem. 2004, 3046Ϫ3056

Charge-Transfer Interactions in Substituted 1,1-Diphenylethenes
FULL PAPER
[18]
[19]
[20]
[21]
C. Bosshard, R. Spreiter, P. Günter, R. R. Tykwinski, M.
Schreiber, F. Diederich, Adv. Mater. 1996, 8, 231Ϫ234.
R. R. Tykwinski, M. Schreiber, V. Gramlich, P. Seiler, F. Died-
erich, Adv. Mater. 1996, 8, 226Ϫ231.
ether. A solution of 4-bromoacetophenone (5.01 g, 25.2 mmol) in
diethyl ether (20 mL) was added dropwise. Subsequently, the mix-
ture was heated at reflux temperature for three hours. After the
mixture had cooled, water (100 mL) and hydrochloric acid (1 ,
25 mL) were added cautiously, after which the layers were sepa-
rated. The organic one was washed with water (3 ϫ 50 mL), dried
over magnesium sulfate and filtered, and the solvents were evapo-
rated to dryness to give 9 (6.93 g, 5.0 mmol, 99%). Ethene 10 was
then prepared from 9 by the method described for 6. Yield 82% of
´
C. Eckert, F. Heisel, J. A. Miehe, R. Lapouyade, L. Ducasse,
Chem. Phys. Lett. 1988, 153, 357Ϫ364.
´ ˆ
C. Maertens, C. Detrembleur, P. Dubois, R. Jerome, C. Bout-
´
ton, A. Persoons, T. Kogej, J. L. Bredas, Chem. Eur. J. 1999,
5, 369Ϫ380.
[22]
[23]
[24]
[25]
[26]
M. H. van der Veen, M. T. Rispens, H. T. Jonkman, J. C. Hum-
melen, Adv. Funct. Mater. 2004, 14, 215Ϫ223.
M. Klokkenburg, M. Lutz, A. L. Spek, J. H. van der Maas, C.
A. van Walree, Chem. Eur. J. 2003, 9, 3544Ϫ3554.
N. J. Turro, Modern Molecular Photochemistry, Benjamin/
Cummings, Menlo Park, 1978.
A. Gilbert, J. Baggot, Essentials of Molecular Photochemistry,
Blackwell, Oxford, 1991.
D. R. Arnold, X. Du, J. Chen, Can. J. Chem. 1995, 73,
307Ϫ318.
2
a yellow liquid. 1H NMR: δ ϭ 5.45 and 5.47 (AB, JH,H ϭ 1.2 Hz,
3
2 ϫ 1 H, ϭCH2), 7.21 and 7.46 (AAЈBBЈ, JH,H ϭ 8.5 Hz, 2 ϫ 2
H, ArϪH), 7.35 (m, 5 H, ArϪH) ppm. 13C NMR: δ ϭ 114.7, 121.8,
127.9, 128.2, 128.3, 129.9, 131.3, 140.4, 140.9, 149.0 ppm. IR:
ν˜max ϭ 1661, 1588, 1485, 903, 1072, 1011, 831, 777 cmϪ1
.
1-(4-Cyanophenyl)-1-phenylethene (1A): This compound was pre-
pared from 10 by the method described for 1D1A. The crude prod-
uct was purified by crystallization from methanol at Ϫ20 °C, col-
umn chromatography with chloroform as eluent and another crys-
tallization from methanol at Ϫ20 °C. Yield 64% of white crystals.
[27]
[28]
L. Friedman, H. Shechter, J. Org. Chem. 1961, 26, 2522Ϫ2524.
ZINDO calculations were also performed, but the PPP/SCF
results were found to be much more consistent with the exper-
imentally measured data. In particular, ZINDO results for di-
methylaniline derivatives were poor. This has been reported
previously: J.-S. Yang, S.-Y. Chiou, K.-L. Liau, J. Am. Chem.
Soc. 2002, 124, 2518Ϫ2527.
1
2
M.p. 42 °C. H NMR: δ ϭ 5.54 and 5.58 (AB, JH,H ϭ 0.8 Hz, 2
ϫ 1 H, ϭCH2), 7.26 (m, 2 H, ArϪH), 7.35 (m, 3 H, ArϪH), 7.42
and 7.62 (AAЈBBЈ, 3JH,H ϭ 8.5 Hz, 2 ϫ 2 H, ArϪH) ppm, in agree-
ment with literature data.[26] 13C NMR: δ ϭ 111.3, 116.6, 118.8,
128.1, 128.2, 128.4, 128.8, 132.0, 140.2, 146.0, 148.7 ppm, in agree-
ment with literature data.[26] IR: ν˜max ϭ 2226, 1605, 1502, 1491,
[29]
[30]
´
H. H. Jaffe, M. Orchin, Theory and Applications of Ultraviolet
Spectroscopy, Wiley, New York, 1962.
The PPP calculated data of ST-A and 1A do not reflect the
experimental trends very accurately. This may be related to
planarity effects.
912, 853, 783, 706 cmϪ1
.
Supporting Information: Fluorescence spectra of 1D2 in different
solvents (see also the footnote on the first page of this article).
[31]
[32]
C. A. van Walree, O. Franssen, A. W. Marsman, M. C. Flipse,
L. W. Jenneskens, J. Chem. Soc., Perkin Trans. 2 1997,
799Ϫ807.
It would be expected that when the acceptor 1La state is higher
in energy than calculated for 1A, relatively more intensity
[1]
P. N. Prasad, D. J. Williams, Introduction to Nonlinear Optical
1
would be taken from the donor La transition. In addition, the
Effects in Molecules and Polymers, Wiley, New York, 1991.
small energy difference between the 1La state of 1A and the CT
state of 1D1A may account for the large calculated intensity of
the CT transition. The efficiency of intensity stealing depends
on this difference.
[2]
Chem. Rev. 1994, 94, 3Ϫ278.
[3]
R. M. Metzger, J. Mater. Chem. 1999, 9, 2027Ϫ2036.
[4]
R. M. Metzger, Chem. Rev. 2003, 103, 3803Ϫ3834.
[5]
A. Nitzan, M. A. Ratner, Science 2003, 300, 1384Ϫ1389.
[33]
[34]
R. Lapouyade, K. Czeschka, W. Majenz, W. Rettig, E. Gilab-
[6]
H. Imahori, S. Fukuzumi, Adv. Mater. 2001, 13, 1197Ϫ1199.
`
ert, C. Rulliere, J. Phys. Chem. 1992, 96, 9643Ϫ9650.
[7]
P. Pasman, F. Rob, J. W. Verhoeven, J. Am. Chem. Soc. 1982,
M. Brink, H. Möllerstedt, C.-H. Ottosson, J. Phys. Chem. A
2001, 105, 4071Ϫ4083.
W. Rettig, Top. Curr. Chem. 1994, 169, 253Ϫ299.
H. Beens, H. Knibbe, A. Weller, J. Chem. Phys. 1967, 47,
1183Ϫ1184.
Dielectric constants and refractive indices of solvents were
taken from: C. Reichardt, Solvents and Solvent Effects in Or-
ganic Chemistry, 2nd edition, VCH, Weinheim, 1988.
W. Rettig, Angew. Chem. 1986, 98, 969Ϫ986; Angew. Chem. Int.
Ed. Engl. 1986, 25, 971.
104, 5127Ϫ5133.
[8]
M. N. Paddon-Row, Acc. Chem. Res. 1994, 27, 18Ϫ25.
W. D. Oosterbaan, P. C. M. van Gerven, C. A. van Walree, M.
[35]
[36]
[9]
Koeberg, J. J. Piet, R. W. A. Havenith, J. W. Zwikker, L. W.
Jenneskens, R. Gleiter, Eur. J. Org. Chem. 2003, 3117Ϫ3130.
G. Mignani, A. Krämer, G. Pucetti, I. Ledoux, J. Zyss, G.
[37]
[10]
Soula, Organometallics 1991, 10, 3656Ϫ3659.
F. J. Hoogesteger, C. A. van Walree, L. W. Jenneskens, M. R.
[11]
[38]
[39]
Roest, J. W. Verhoeven, W. Schuddeboom, J. J. Piet, J. M. War-
man, Chem. Eur. J. 2000, 6, 2948Ϫ2959.
W. D. Oosterbaan, C. Koper, T. W. Braam, F. J. Hoogesteger,
[12]
K. A. Zachariasse, M. Grobys, T. von der Haar, A. Hebecker,
Y. V. I l Јichev, Y.-B. Jiang, O. Morawski, W. Kühnle, J. Photo-
chem. Photobiol. A: Chem. 1996, 102, 59Ϫ70.
K. A. Zachariasse, Chem. Phys. Lett. 2000, 320, 8Ϫ13.
J. J. Piet, B. A. J. Jansen, C. A. van Walree, H. J. van Rames-
donk, M. Goes, J. W. Verhoeven, W. Schuddeboom, J. M. War-
[40]
[41]
man, L. W. Jenneskens, J. Phys. Chem.
3612Ϫ3624.
P. Maslak, A. Chopra, C. R. Moylan, R. Wortmann, S. Lebus,
A. L. Rheingold, G. P. A. Yap, J. Am. Chem. Soc. 1996, 118,
1471Ϫ1481.
C. A. van Walree, H. Kooijman, A. L. Spek, J. W. Zwikker, L.
W. Jenneskens, J. Chem. Soc., Chem. Commun. 1995, 35Ϫ36.
C. A. van Walree, M. R. Roest, W. Schuddeboom, L. W. Jenne-
skens, J. W. Verhoeven, J. M. Warman, H. Kooijman, A. L.
Spek, J. Am. Chem. Soc. 1996, 118, 8395Ϫ8407.
A 2003, 107,
´
J. Dobkowski, J. Wojcik, W. Kozminski, R. Kolos, J. Waluk, J.
[13]
Michl, J. Am. Chem. Soc. 2002, 2406Ϫ2407.
[42]
Cavity radii were estimated by comparison with the previously
used radius of 2-[4-(dimethylamino)phenyl]-2-(4-cyanophenyl)-
propane (4.5 A, ref.[15]).
˚
[14]
[15]
[43]
[44]
[45]
R. A. Neunteufel, D. R. Arnold, J. Am. Chem. Soc. 1973, 95,
4080Ϫ4081.
´
J. F. Letard, R. Lapouyade, W. Rettig, J. Am. Chem. Soc. 1993,
115, 2441Ϫ2447.
[16]
[17]
´
´
´
The fluorescence maximum in ethanol was omitted from the
fit, since hydrogen bonding interactions cause the fluorescence
to deviate from the behaviour predicted by Equation (2).
J. W. Verhoeven, T. Scherer, B. Wegewijs, R. M. Hermant, J.
A. Garcıa Martınez, J. Osıo Barcina, A. de Fresno Cerezo, G.
´
´
Rojo, F. Agullo-Lopez, J. Phys. Chem. B 2000, 104, 43Ϫ47.
R. R. Tykwinski, M. Schreiber, R. Perez Carlon, F. Diederich,
´
´
[46]
V. Gramlich, Helv. Chim. Acta 1996, 79, 2249Ϫ2281.
Eur. J. Org. Chem. 2004, 3046Ϫ3056
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3055

C. A. van Walree et al.
FULL PAPER
[51]
[52]
[53]
Jortner, M. Bixon, S. Depaemelaere, F. C. De Schryver, Recl.
Trav. Chim. Pays-Bas 1995, 114, 443Ϫ448.
F. Hollywood, H. Suschitzky, R. Hull, Synthesis 1982,
662Ϫ665.
A. McKillop, J. D. Hunt, F. Kienzle, E. Bigham, E. C. Taylor,
J. Am. Chem. Soc. 1973, 95, 3635Ϫ3640.
F. Berthiol, H. Doucet, M. Santelli, Eur. J. Org. Chem. 2003,
1091Ϫ1096.
[47]
Y. Hayashi, M. Takeda, Y. Miyamoto, M. Shoji, Chem. Lett.
2002, 414Ϫ415.
[48]
D. F. Eaton, Pure Appl. Chem. 1988, 60, 1107Ϫ1114.
[49]
J. Griffiths, J. G. Lasch, QCPE Program QCMP054.
[50]
Quantum CAChe, Version 5.04, Fujitsu, 2002.
Received March 2, 2004
3056
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3046Ϫ3056