Absorption and fluorescence of arylmethylidenoxindoles and isoindigo

Article abstract of DOI:10.1016/j.dyepig.2009.11.001

A series of arylmethylidenoxindoles was synthesized by acid or base catalysed condensation of substituted aldehydes with oxindole. The products were formed as mixtures of Z- and E-isomers that were separated using column chromatography and identified by ¹H and ¹³C NMR. The effect of substituents on absorption maxima was investigated both experimentally and theoretically, based on time dependent density functional theory. Low temperature absorption spectroscopy in a solvent glass enabled identification of the position of the 0-0 vibronic bands. The spectral features of the parent Z-benzylidenoxindole were compared to those of trans-isoindigo and trans-stilbene. Fluorescence of arylmethylidenoxindoles in solution was not observed, as E-Z isomerization represents a dominant deactivation channel after irradiation; when this geometrical isomerization was sterically hindered at low temperature solvent glass or in solid, fluorescence of planar Z-isomers was observed.

Full text of DOI:10.1016/j.dyepig.2009.11.001

Dyes and Pigments 85 (2010) 171e176
Contents lists available at ScienceDirect
Dyes and Pigments
journal homepage: www.elsevier.com/locate/dyepig
Absorption and fluorescence of arylmethylidenoxindoles and isoindigo
a
,
a
b,c
*
ꢀ
ꢀ
Stanislav Lunák Jr. , Petra Horáková , Antonín Lycka
a Department of Organic Technology, Faculty of Chemical Technology, University of Pardubice, Studentská 95, Pardubice CZ-532 10, Czech Republic
b Research Institute for Organic Syntheses, Rybitví 296, Rybitví CZ-533 54, Czech Republic
c University of Hradec Králové, Faculty of Education, Rokitanského 62, CZ 500 03 Hradec Králové 3, Czech Republic
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of arylmethylidenoxindoles was synthesized by acid or base catalysed condensation of substituted
aldehydes with oxindole. The products were formed as mixtures of Z- and E-isomers that were separated
using column chromatography and identified by 1H and 13C NMR. The effect of substituents on absorption
maxima was investigated both experimentally and theoretically, based on time dependent density func-
tional theory. Low temperature absorption spectroscopy in a solvent glass enabled identification of the
position of the 0-0 vibronic bands. The spectral features of the parent Z-benzylidenoxindole were compared
to those of trans-isoindigo and trans-stilbene. Fluorescence of arylmethylidenoxindoles in solution was not
observed, as EeZ isomerization represents a dominant deactivation channel after irradiation; when this
geometrical isomerization was sterically hindered at low temperature solvent glass or in solid, fluorescence
of planar Z-isomers was observed.
Received 20 August 2009
Received in revised form
29 October 2009
Accepted 1 November 2009
Available online 12 November 2009
Keywords:
Absorption
Fluorescence
Density functional theory
Isoindigo
Ó 2009 Elsevier Ltd. All rights reserved.
Arylmethylidenoxindole
1. Introduction
theory (TD DFT) is not supported by our experience with related
molecular structures [8], so we decided to reinvestigate this feature.
Z-Benzylidenoxindole (I) is a
p
-isoelectronic N-analogue of iso-
The aim of this paper is to study in detail the relation between
structure and absorption spectra of arylmethylidenoxindoles and
isoindigo and to look for their so far unknown fluorescence prop-
erties in various environments, i.e. in room temperature dimethyl
sulfoxide (DMSO) and 2-methyltetrahydrofuran (MTHF) solutions,
low temperature MTHF glass and polycrystalline solid state.
We have synthesized a set of model compounds shown in Fig. 1.
The biphenylyl compound (II) was synthesized for the first time
in order to investigate an effect of a conjugated chain elongation
on the spectral properties. The derivative with strong electron-
donating group in para position of pendant phenyl (III) is known in
both its isomeric forms [3]. There are several studies on the anti-
tumour activity of III, that we will not review here. The compound
III also appeared as an example of a dye for mass colouration of
plastics [9], but it is not clear whether the Z- or E-isomer or their
mixture was used.
Structural calculations of arylideneoxindoles were reported [4].
No quantum-chemical study of spectral properties of these chro-
mophores was found. That is why DFT was used to geometry
calculations and TD DFT to calculate the vertical excitation energies,
with the same hybrid B3LYP exchange e correlation functional. The
solvent effect was involved through polarized continuum model
(PCM). No experimental absorption spectra of the compounds under
study have been reported other than in solution and fluorescence
has not been investigated up until now.
aurone, a parent chromophore of naturally occurring yellow pigments
[1]. The compound I was prepared in both Z- and E-isomeric forms [2]
and the X-ray structural data are available for them [3,4] (file LOHZAJ
in Cambridge structural database (CSD) for Z-isomer). The compound
I can be also formally considered as a hemistilbene/hemiisoindigo
hybrid (Fig. 1).
There were several reasons for our interest on this class of
compounds. We are generally interested in the dyes arising
from a relatively simple condensation of aldehydes with heterocy-
cles containing an activated methylene group [5]. N-acylated I was
reported to show solid state fluorescence [4], which is also of our
interest, because of its potential use in OLED devices [6]. While the
spectral properties and excited state processes occurring in trans-
stilbene were widely studied, there is a limited knowledge of the
effect of its cyclization into one (I) or two (isoindigo) pyrrolinone
rings on the absorption spectra and excited state behaviour. A
recent theoretical study of the indigo isomers has found a signifi-
cant discrepancy between theoretical excitation energy and a posi-
tion of long wavelength absorption maximum of trans-isoindigo [7].
An explanation based on underestimation of excitation energies of
charge-transfer transitions by time dependent density functional
* Corresponding author. Tel.: þ420 602437863; fax: þ420 466635292.
ꢀ
E-mail address: stanislav.lunak@seznam.cz (S. Lunák Jr.).
0143-7208/$ e see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.dyepig.2009.11.001

ꢀ
S. Lunák Jr. et al. / Dyes and Pigments 85 (2010) 171e176
172
N
2.2. DFT ground state geometry and energy
N
O
trans-isoindigo
Restricted B3LYP/6-311þþG(d,p) DFT calculations in vacuum
resulted in a strictly planar geometry for Z-isomers of all three
arylmethylidenoxindoles (except for the phenylephenyl torsion in
Z-II) and trans-stilbene. On the other hand, all three E-isomers were
computed to be non-planar with pendant phenyl ring rotated out of
oxindole plane. The experimental [3] dihedral angles for both Z-I
(8.0ꢀ) and E-I (49.7ꢀ) are rather higher than the theoretical ones (0ꢀ,
resp. 37.0ꢀ), that can be ascribed to packing effects in the crystal, not
being taken into account in these calculations. The corresponding
dihedral angles of E-II and E-III were computed to be 35.2ꢀ and
28.6ꢀ, respectively. The dihedral angle between both phenyls rings
was 38.1ꢀ for Z-II and 38.9ꢀ for E-II.
O
N
H
N
N
O
Z-I (R=H)
Z-II (R = Ph)
Z-III (R=N(CH3)2)
R
Z-isomers are generally only a bit more stable than E-isomers
by calculation. The energy differences are 0.32 kcal molꢁ1 (I),
0.60 kcal molꢁ1 (II) and 1.35 kcal molꢁ1 (III). The thermodynamic
stability of both isomers is thus very similar, that can be one of the
reasons of their simultaneous rise in a reaction mixture.
trans-stilbene
Fig. 1. Chemical structure of synthesized compounds (drawn as of Z-isomers) together
with trans-isoindigo and trans-stilbene formulas.
The central ethylenic bond is significantly elongated when going
from trans-stilbene (theor.1.346 Ǻ, exp. [12] 1.326 Ǻ) to trans-isoindigo
(theor. 1.377 Ǻ, exp. [13] 1.369 Ǻ), which agrees with a significant
decrease of the frequency of the corresponding C]C stretching mode
[10]. Thus Z-I is their true hybrid, as the corresponding bond length
lies between those ones of its two symmetrical parent compounds
both theoretically (1.359 Ǻ) and experimentally [3] (1.350 Ǻ).
There is a known discrepancy between the calculated non-
planarity [7,10] and experimental planarity [13] of trans-isoindigo.
So we have focused on this problem in a more detailed manner.
The dependence of the total energy of trans-isoindigo on the central
double bond torsion, as computed in vacuum and DMSO solution
involved by PCM, is shown in Fig. 2. The energy minimum shifts from
10.0ꢀ in vacuum to 14.6ꢀ in DMSO, as the polar environment more
efficiently stabilizes the non-planar and thus polar structure through
soluteesolvent interaction. The minimum is shallow and the energy
of a transition state (180ꢀ) and the structure with a dihedral angle
150ꢀ is almost the same in DMSO.
2. Results and discussion
2.1. Syntheses
trans-Isoindigo was synthesized by a condensation of oxindole
and isatin [10]. Compounds I and II were synthesized by p-tolue-
nesulfonic acid catalysed condensation of benzaldehyde (4-biphe-
nylcarboxaldehyde) with oxindole. Compound III was synthesized
by base (piperidine) catalysed condensation of p-dimethylamino-
benzaldehyde and oxindole. The reaction mixture always contained
both the E- and Z-isomers. Compounds I and III crystallized as
E-isomers and were converted to Z-isomers by an exposure of
their solutions to daylight. Pure Z-II crystallized from a hot reaction
mixture, while E-II was obtained from cooled filtrates. Final prod-
ucts were obtained by flash column chromatography on silica
(acetone e hexane 2:3). The identity of both isomers of I and III was
checked by a comparison of their melting points with the published
data [3] and the purity was monitored by elemental analysis and
mass spectrometry.
The syntheses of both isomers of a new compound II are
described in detail in Experimental. The isomers were preliminary
assigned according to the shifts of their absorption spectral maxima
after irradiation (Z / E hypsochromic and hypochromic, E / Z
bathochromic and hyperchromic) and checked by recently verified
methodology based on J(13C, 1H) coupling constants [5]. We utilised
a method proposed by Cho et al. [11], based on a stereochemical
dependence of 3J(13C, 1H) coupling constants on a C]C double
bond. Cho et al. found that coupling constant of carbon trans to
the olefinic proton is considerably larger to that of carbon cis to the
same proton utilising single crystal X-ray data in the solid state and
heteronuclear Overhauser effect in solution in suitable model
compounds. Stereochemical dependence of the 3J(13C, 1H) coupling
constants were studied in the fragment HNeC(]O)eC(]CHe)e
having measured proton-coupled 13C NMR spectra of chromato-
graphically separated isomers of compound II at digital resolution
better than 0.5 Hz/point. The NHeC]O resonances gave a dd in
the proton-coupled 13C NMR spectra. The small coupling constant
in the dd in both cases corresponds to the interaction of carbonyl
carbon with NH proton (2J(13C(O)N1H) ¼ 2.5 Hz). The coupling
constant 3J(13C(O)C]C1H) ¼ 12.1 Hz belongs to Z form on double
bond (compound Z-II) while 3J(13C(O)C]C1H) ¼ 6.9 Hz belongs to
E form on double bond (compound E-II).
2.3. Absorption and fluorescence of arylmethylidenoxindoles
All three arylmethylidenoxindoles easily undergo a photo-
chemical ZeE isomerization in both directions in solution. The
absorption spectra of non-planar E-isomers show hypsochromic
and hypochromic shift relative to the Z-isomers (Table 1) caused by
an imperfect conjugation.
No fluorescence of any arylmethylidenoxindole was observed in
DMSO or MTHF solution at room temperature. On the other hand,
all three Z-isomers of I, II and III show fluorescence both in rigid
-875,90
-875,91
-875,92
-875,93
-875,94
-875,95
-875,96
-875,97
105
120
135
150
165
180
(°)
Fig. 2. The dependence of a total energy of trans-isoindigo on a central dihedral angle
in vacuum (C) and DMSO (-).

ꢀ
S. Lunák Jr. et al. / Dyes and Pigments 85 (2010) 171e176
173
Table 1
Theoretical excitation energies of the first two pp* transitions (converted to l00) and experimental absolute absorption maximum lmax in DMSO (room temperature) and
a position of 0e0 vibronic band in MTHF (low temperature).
PCM (DMSO) TD-DFT
Exp. lmax DMSO (l00 MTHF)
S0eS1
S0eS1
S0eS2
S0eS2
lmax l00
l00
fosc
coeff.
l00
fosc
coeff.
lmax
(l00
)
3max
(
)
3max
Z-I
421
0.179
0.63 (1-10)
0.22 (2-10)
0.64 (1-10)
0.18 (2-10)
0.63 (1-10)
0.21 (2-10)
0.64 (1-10)
0.18 (2-10)
0.63 (1-10)
0.12 (2-10)
0.65 (1-10)
352
0.663
0.18 (1-10)
0.64 (2-10)
0.15 (1-10)
0.65 (2-10)
0.18 (1-10)
0.65 (2-10)
0.16 (1-10)
0.66 (2-10)
0.10 (1-10)
0.67 (2-10)
0.10 (1-10)
0.67 (2-10)
w400
w4000
341 (355)
19200
E-I
403
434
415
434
431
0.164
0.441
0.381
0.973
0.803
339
378
366
377
375
0.422
0.869
0.509
0.206
0.044
w388
w2100
321
8600
32500
18150
Z-II
E-II
Z-III
E-III
359 (380)
345
444 (478)
436
37500
15100
lmax [nm] e measured absolute absorption maxima, 3max [l molꢁ1 cmꢁ1] e measured molar absorptivity, l00 [nm] e computed or measured absorption maximum of 0e0
vibronic band, f osc e computed oscillator strength of 0e0 vibronic band, coeff. e expansion coefficient of the monoexcited configurations in TD DFT.
MTHF glass and in the solid state. The fluorescence of Z-III is strong,
observable by naked eye under UV irradiation of the solid sample.
The fluorescence of Z-II is measurable without problems, while for
Z-I it is only detectable with low signal/noise ratio. E-arylmethyli-
denoxindoles do not fluoresce under these conditions, as is usual
for less planar isomers [14].
While no vibronic progressions can be observed for the shoulder
in the visible region in MTHF glass, the stilbene-like [15] vibronic
structure of the UV band is significantly sharper, showing that
the absolute absorption maximum corresponds to 0e1 vibronic
transition (Fig. 4). It is not surprising [16], that the low temperature
MTHF glass behaves as an even more polar environment than DMSO
at room temperature: the absolute absorption maxima in DMSO
(341 nm for Z-I and 359 nm for Z-II) are slightly (1 nm for Z-I and
3 nm for Z-II) hypsochromically shifted with respect to the 0e1
maximum in MTHF glass (342 nm for Z-I and 362 nm for Z-II). Thus,
it is reasonable to compare the theoretical excitation energies,
computed in DMSO by PCM model, with well resolved 0e0 vibronic
sub-bands in MTHF glass rather than with hardly detectable shoul-
ders in DMSO spectra. The agreement is excellent e the difference is
3 nm for Z-I and 2 nm for Z-II (Table 1). The absorption spectra of E-I
(Fig. 3) and E-II are qualitatively similar, but no significant vibronic
progressions of either the visible shoulder or the UV band can be
observed in MTHF glass, so the exact quantitative comparison with
the theoretical values is impossible.
The position of 0e0 vibronic sub-band of a long wavelength
shoulder of Z-I and Z-II is not clear even from the low temperature
absorption. Rather better insight on a position of 0e0 vibronic sub-
band can be indirectly derived from a shape of low temperature
fluorescence spectra. The fluorescence emission spectrum of Z-II
clearly shows, that its maximum corresponds to 0e2 vibronic sub-
band (Fig. 5). This effect is usually connected with a significant
difference between equilibrium geometries in S0 and S1 states. Thus,
although the fluorescence excitation spectrum of a long wavelength
transition is not resolved into the vibronic sub-bands, we can
reasonably suppose, that the 0e0 vibronic band is hidden at the long
The absorption spectra of Z-I and Z-II in DMSO are qualitatively
similar, only the latter is bathochromically shifted as a consequence
of enlarged conjugated system. They show two main spectral
features: structureless long wavelength band (usually observed as
a shoulder) in a visible region, tailing up to about 500 nm, and more
intense band in UV region (Fig. 3). PCM TD DFT results (Table 1)
qualitatively agree with all these spectral characteristics of E- and
Z-isomers of I and II. The long wavelength shoulder represents
S0 / S1 pp* transitions of 1-10 (HOMO / LUMO) charac0ter, while
the UV band corresponds to S0 / S2 pp* transitions of 2-1 (HOMO-
1 / LUMO) character. Both transitions are generally allowed, but
weak, especially for I. The spectral positions of 0e0 vibronic sub-
bands of both transitions, necessary for a quantitative correlation of
theoretical and experimental values, were not generally available
from room temperature spectra.
25000
20000
15000
10000
5000
0
1
0,9
0,8
0,7
0,6
0,5
0,4
0,3
0,2
0,1
0
360
260
310
410
460
510
560
300
350
400
450
500
550
600
[nm]
λ
λ [nm]
Fig. 3. Absorption spectra of Z-I (black line) and E-I (grey line) in DMSO at room
temperature.
Fig. 4. Room (DMSO e grey line) and low (MTHF e black line) temperature absorption
spectra of Z-II.

ꢀ
174
S. Lunák Jr. et al. / Dyes and Pigments 85 (2010) 171e176
1
0,8
0,6
0,4
0,2
0
1
0,8
0,6
0,4
0,2
0
500
550
600
650
700
750
800
850
[nm]
λ
Fig. 7. Normalized solid state fluorescence emission spectra of Z-II (grey line) and Z-III
(black line) at room temperature.
300
350
400
450
500
550
600
650
[nm]
λ
Fig. 5. Fluorescence excitation and emission spectra of Z-II in MTHF at low
temperature.
manifests itself mainly by hypochromic shift e relative decrease of
3max with respect to Z-III is comparable with that of the parent E- and
Z-I pair (Table 1).
wavelength spectral tail. This is the reason, why the computed
absorption maxima are generally red shifted with respect to the
observed maxima (or inflections) of I and II corresponding to their
S0 / S1 transition.
The solid state fluorescence emission spectra of Z-isomers of II
and III are shown in Fig. 7. The emission of Z-III shows a maximum
in the red region (665 nm). Such a strong solid state luminescence is
remarkable, as it is not a common phenomenon in the area of organic
dyes and pigments [17e19]. The absence of solid state fluorescence
of all three E-isomers is a consequence of their non-ability to fluo-
resce as isolated molecules, that is always the necessary condition of
solid state emission [17,18].
The absorption and fluorescence spectra of electron donor
substituted Z-III show significant differences from parent Z-I. The
intensities of both spectral transitions are inverted by theory (Table 1)
and the stronger long wavelength band shows well resolved vibronic
structure at low temperature (Fig. 6). Both absorption and emission
maxima correspond to 0e1 vibronic transition, so the difference
between equilibrium geometries of S0 and S1 is probably not so
pronounced. The second S0 / S2 absorption band is not observable,
as it is overlapped by stronger S0 / S1 band. Z-III (4.04 D) is more
polar than parent Z-I (2.04 D) by computations, so it is more sensitive
to environment polarity. So the 0e1 vibronic band in MTHF is
significantly (20 nm) bathochromically shifted with respect to room
temperature maximum in DMSO and the wavelength of 0e0 sub-
band in MTHF at 77 K is thus dramatically higher than the computed
one in DMSO at 20 ꢀC. The computed excitation energy is over-
estimated by PCM TD DFT in this case. Hypsochromic shift of E-III
with respect to Z-III is less pronounced both by theory and experi-
ment, as S0 / S1 transition is of charge transfer (CT) character and
E-III is even more polar (6.00 D) than Z-III. The non-planarity of E-III
2.4. Absorption of trans-isoindigo
No spectral changes upon irradiation were observed for trans-
isoindigo, thus there is no evidence for the formation of the cis
isomer. trans-Isoindigo does not fluoresce at all. The room and low
temperature absorption spectra of trans-isoindigo (Fig. 8) show
a remarkable similarity with those of Z-I (and Z-II). Its long wave-
length band (490 nm/3900 l molꢁ1 cmꢁ1 in DMSO, 498 nm in MTHF
glass) is of lower intensity and unresolved even at low temperature,
while the more intense UV band shows resolved vibronic progres-
sions. The absolute absorption maximum in DMSO (367 nm/10
600 l molꢁ1 cmꢁ1) is slightly hypsochromically shifted with respect
to the maximum in MTHF glass (370 nm), so the parallel with Z-I is
evident. Nevertheless, it is not clear, whether this maximum corre-
sponds to 0e1 or 0e2 vibronic transition. It is not obvious, whether
the vibronic sub-band in MTHF glass at 393 nm is of 0e0 character,
1
0,8
0,6
0,4
0,2
0
1
0,8
0,6
0,4
0,2
0
300
350
400
450
500
550
600
650
700
300
400
500
600
700
800
[nm]
[nm]
λ
λ
Fig. 6. Fluorescence excitation and emission spectra of Z-III in MTHF at low temper-
ature (black line), together with absorption in DMSO at room temperature (grey line).
Fig. 8. Room (DMSO e grey line) and low (MTHF e black line) temperature absorption
spectra of trans-isoindigo.

ꢀ
S. Lunák Jr. et al. / Dyes and Pigments 85 (2010) 171e176
175
as in Z-I, or 0e1, as the there is a remarkable shoulder at about
415 nm, which is not observable neither in Z-I, nor in Z-II.
intensity. Low temperature absorption spectroscopy in solvent glass
enabled the identification of the position of the 0e0 vibronic bands
in some cases. These 0e0 vibronic bands of S0 / S2 transitions show
excellent agreement with PCM TD DFT computed excitation energies.
Z-benzylidenoxindole is a hemistilbene/hemiisoindigo hybrid from
a structural point of view and its spectral features lie among those ones
of trans-isoindigo and trans-stilbene. Fluorescence of arylmethylide-
noxindoles in solution was not observed, as EeZ isomerization
represents a dominant deactivation channel after irradiation. If this
isomerization was sterically hindered in a low temperature solvent
glass or in the solid state, fluorescence of the Z-isomer was observed.
PCM TD DFT calculations of trans-isoindigo in DMSO resulted
in three transitions covering the investigated spectral region. Their
positions depend significantly on a central torsion angle being:
S0 / S1 (518 nm, fosc ¼ 0.139), S0 / S2 (493 nm, fosc ¼ 0.000) and
S0 / S3 (395 nm, fosc ¼ 0.565) for planar structure and S0 / S1
(537 nm, fosc ¼ 0.175), S0 / S2 (505 nm, fosc ¼ 0.000) and S0 / S3
(420 nm, fosc ¼ 0.293) for the equilibrium minimum at 165ꢀ. Quali-
tatively the situation is clear: S0 / S1 is responsible for a long
wavelength band and S0 / S3 for a more intense band near 400 nm.
S0 / S2 is symmetry forbidden, thus not observable in absorption. Its
eventual role in a deactivation processes after excitation, leading to the
absence of both fluorescence and ZeE isomerization, is out of the
scope of this paper.
4. Experimental
4.1. Syntheses and analyses
The quantitative comparison of the theoretical excitation energies
and experimental spectral features is complicated, because the
uncertainty lies both on experimental (position of 0e0 vibronic
bands) and theoretical (geometry in DMSO solution) aspects. We tend
to the following explanation: the shoulder at about 415 nm in MTHF
glass really corresponds to 0e0 vibronic sub-band, that quite well
relates with the value 420 nm computed for non-planar geometry
in DMSO. Thus, trans-isoindigo is, because of relatively long central
double bond, unable to retain a planar geometry as in crystal. The
position of a 0e0 vibronic sub-band of a long wavelength band is not
clear from the absorption spectra. It is evident from a broad spectral
shape, that its maximum does not correspond to 0e0, but probably to
0e2 vibronic transition, the same way as for more intense band near
400 nm. In this case the computed value 537 nm is not overestimated
[7], but may be even underestimated, as this band is tailing up to
600 nm. The usual energy difference between vibronic sub-bands,
corresponding to carbonecarbon stretching, is about 1500 cmꢁ1. If
a double of this value is subtracted from the spectral maximum
relating to 0e2 vibronic sub-band (490 nm ¼ 20 408 cmꢁ1), the
position of 0e0 vibronic sub-band reaches 574 nm (17 408 cmꢁ1),
i.e. the value near to the spectral onset of this broad band (Fig. 8). At
least, there is no reason to doubt the ability of the PCM TD DFT method
to describe trans-indigo, that is only slightly polar (1.39 D by compu-
tation) for equilibrium C2 non-planar geometry. The discrepancy
between calculated and experimental values could mainly be ascribed
to the difficulties associated with locating the correct spectral posi-
tions of 0e0 vibronic transitions.
Oxindole, 4-biphenylcarboxaldehyde, dimethyl sulfoxide, 2-
methyltetrahydrofurane, piperidine and trans stilbene were purchased
from SigmaeAldrich. Benzaldehyde, p-toluenesulfonic acid and
p-dimethylaminobenzaldehyde came from Lachema.
4.1.1. (3Z)-3-(1,10-biphenyl-4-ylmethylen)-1,3-dihydro-
2H-indol-2-on (Z-II)
Oxindole (2.7 g, 0.02 mol), 4-biphenylcarboxaldehyde (3.64 g,
0.02 mol), p-toluensulfonic acid (0.5 g) and toluene (50 ml), were
placed in a 100 ml flask and the reaction mixture was stirred and
heated under reflux. After three hours, the hot, orange suspension
of Z-II was filtered off and washed with hot toluene. Yield 3.8 g
(64%), m.p. 272e275 ꢀC.
Calculated: C(84.82),H(5.08),N(4.71), Found: C(84.56),H(4.94),N
(4.59)
GC/MS: M (297), RT (16.53 min), 100%
1H NMR (500 MHz, DMSO-d6)
d 10.72 (1H, bs, NH), ]CHe: 7.90
(1H, s), phenyl group: 8.56 (2H, d, J ¼ 8.1 Hz) 7.56 (2H, t, J ¼ 7.6 Hz), 7.45
(1H, tt, J ¼ 7.6 and 1.3 Hz), 1,4-C6H4-: 7.84 (2H, m), 7.79 (2H, m),
1,2-C6H4-: 7.78 (1H, d, J ¼ 7.5 Hz), 7.27 (1H, t, J ¼ 7.6 Hz), 7.05 (1H, t,
J¼ 7.5 Hz), 6.88 (1H, d, J¼ 7.5 Hz), 13CNMR(125MHz,DMSO-d6)
d167.3
(3J(13C(O)C]C1H) ¼ 12.1 Hz), 141.8, 140.8, 139.3, 136.3, 133.3, 132.8,
129.1, 129.0, 128.1, 126.9, 126.8, 126.4, 125.1, 121.2, 119.9 and 109.5.
4.1.2. (3E)-3-(1,10-biphenyl-4-ylmethylen)-1,3-dihydro-
2H-indol-2-on (E-II)
The position of 0e0 vibronic sub-band of trans-stilbene in MTHF
glass is 327 nm, rather lower with respect to the value computed in
DMSO for equilibrium planar geometry (335 nm). This difference can
be explained on the often discussed phenyl torsions in solution [20].
So the difference between experimental (theoretical) excitation
wavenumbers of a UV transition with pronounced vibronic structure
between Z-I and trans-stilbene is about 2400 cmꢁ1 (1450 cmꢁ1),
while the same differences between Z-I and trans-isoindigo are
considerably higher 4100 cmꢁ1 (4600 cmꢁ1). From this point of
view Z-I can be considered to resemble a perturbed stilbene, rather
than true hemistilbene/hemiisoindigo hybrid. The long wavelength
CT transitions of Z-I and trans-isoindigo in a visible region have no
counterpart in trans-stilbene spectrum and can be considered as
a demonstration of their (coupled) merocyanine character [21].
The hot filtrate obtained from the preparation of Z-II was cooled
to room temperature. Yellow crystals of E-II were filtered and dried.
Yield 0.64 g (10.7%), m.p. 223e225 ꢀC.
Calculated: C(84.82), H(5.08), N(4.71) Found: C(84.77), H(5.25),
N(4.96)
1H NMR (500 MHz, DMSO-d6)
d
10.69 (1H, bs, NH), ¼ CH-: 7.72
(1H, s), phenyl group: 7.82 (2H, d, J ¼ 7.8 Hz), 7.56 (2H, t, J ¼ 7.6 Hz),
7.46 (1H, t, J ¼ 7.6 Hz),1,4-C6H4-: 7.90 (2H, m), 7.86 (2H, m),1,2-C6H4-
: 7.70 (1H, d, J ¼ 7.5 Hz), 7.29 (1H, t, J ¼ 7.6 Hz), 6.95 (1H, d, J ¼ 7.7 Hz),
6.92 (1H, t, J ¼ 7.6 Hz), 13C NMR (125 MHz, DMSO-d6)
d C
168.8 (3J(13
(O)C]C1H) ¼ 6.9 Hz), 143.1, 141.3, 139.3, 135.5, 133.6, 130.3, 130.2,
129.2, 128.3, 127.6, 127.0, 126.8, 122.6, 121.3, 121.0, and 110.3.
4.2. Instrumental
3. Conclusions
4.2.1. Absorption and fluorescence spectra
Perkin Elmer Lambda 9 was used for the measurements of
absorption spectra. Home made cryostat was used for low tempera-
ture measurements (103 K). Perkin Elmer LS 35 furnished with
a commercial low temperature accessory was used for measurements
of fluorescence emission and excitation spectra at low temperature
(77 K). The room temperature solid state fluorescence emission
spectra were measured with Perkin Elmer LS 35, too.
All three Z-arylmethylidenoxindoles are batho- and hyperchromi-
cally shifted with respect to their E-isomers. The enlarged conjugated
system of an aryl moiety causes a bathochromic shift without signifi-
cant change of the mutual intensity of both main spectral features
corresponding to S0 / S1 and S0 / S2 transitions, while a strong
electron-donating substituent on the aryl ring inverts their mutual

ꢀ
176
S. Lunák Jr. et al. / Dyes and Pigments 85 (2010) 171e176
4.2.2. Mass spectrometry
Acknowledgements
The ion trap mass spectrometer MSD TRAP XCT Plus system
(Agilent Technologies, USA) equipped with APCI probe was used.
Positive-ion and negative-ion APCI mass spectra were measured in
the mass range of 50e1000 Da in all the experiments. The ion trap
analyzer was tuned to obtain the optimum response in the range of
the expected m/z values (the target mass was set from m/z 289 to m/z
454). The other APCI ion source parameters: drying gas flow rate
7 L minꢁ1, nebulizer gas pressure 60 psi, drying gas temperature
350 ꢀC, nebulizer gas temperature 350 ꢀC. The samples were dissolved
in a mixture of DMSO/acetonitrile and methanol in various ratios. All
the samples were analyzed by means of direct infusion into LC/MS.
Authors acknowledge the support of the grant project No.
2A-1TP1/041 sponsored by the Ministry of Trade and Industry of
the Czech Republic.
References
[1] Venkateswarlu S, Panchagnula GK, Guraiah MB, Subbaraju GV. Isoaurones:
synthesis and stereochemical assignments of geometrical isomers. Tetrahe-
dron 2006;62:9855e60.
[2] Ruveda EA, Gonzalez HA. Geometric isomerism in benzylidenoxindoles.
Spectrochim Acta A 1970;26:1275e7.
[3] Coda AC, Invernizzi AG, Righetti PP, Tacconi G, Gatti G. (Z)- and (E)-Arylidene-
1,3-dihydroindol-2-ones: configuration, conformation, and infrared carbonyl
stretching frequencies. J Chem Soc Perkin Trans 1984:615e9.
[4] Milanesio M, Viterbo D, Albini A, Fasani E, Bianchi R, Barzaghi M. Structural
study of the solid-state photoaddition reaction of arylideneoxindoles. J Org
Chem 2000;65:3416e25.
4.2.3. Elemental analysis
Perkin Elmer PE 2400 SERIES II CHNS/O and EA 1108 FISONS
instruments were used for elemental analyses.
ꢀ
ꢀ
ꢀ
[5] Vynuchal J, Lunák Jr S, Hatlapatková A, Hrdina R, Lycka A, Havel L, et al.
The synthesis, absorption, fluorescence and photoisomerisation of 2-aryl-4-
arylmethylidene-pyrroline-5-ones. Dyes Pigm 2008;77(2):266e76.
4.2.4. Nuclear magnetic resonance
The 1H (500.13 MHz) and 13C (125.76 MHz) NMR spectra for
compounds Z-II and E-II were recorded on a Bruker Avance 500
spectrometer. The samples were dissolved in hexadeuteriodimethyl
sulfoxide-d6. The 1H and 13C chemical shifts were referenced to the
ꢀ
ꢀ
[6] Vala M, Weiter M, Vynuchal J, Toman P, Lunák Jr S. Comparative studies of
Diphenyl-Diketo-Pyrrolopyrroles for electroluminescent applications. J Fluo-
rescence 2008;18:1181e6.
[7] Perpète EA, Preat J, André JM, Jaquemin D. An ab initio study of the absorption
spectra of indirubin, isoindigo, and related derivatives.
J Phys Chem A
central signal of the solvent (
d
¼ 2.55 (1H) and 39.6 (13C)).
2006;110:5629e35.
ꢀ
ꢀ
[8] Lunák Jr S, Vynuchal J, Vala M, Havel L, Hrdina R. The syntheses, absorption
and fluorescence of polar diketo-pyrrolo-pyrroles. Dyes Pigm 2009;82:102e8.
[9] Roschger P. Massenfarben von Kunststoffen, EP 632102 (Bayer AG, 1994).
4.3. Computational procedures
ꢀ
[10] Lunák Jr S, Frumarová B, Horáková P. Geometry and Raman spectrum of
isoindigo. Chem Phys Lett 2009;477:116e21.
The geometry of trans-stilbene, trans-isoindigo and all three
arylmethylidenoxindoles in both E- and Z-isomeric forms was
optimised using quantum-chemical calculations based on DFT.
The hybrid three-parameter B3LYP functional in combination with
6-311þþG(d,p) basis was used. No constraints were employed. No
imaginary frequencies were found after diagonalization of Hessian
matrix, confirming that the computed geometries were the real
minima on the ground state hypersurfaces.
[11] Cho H, Iwashita T, Hamaguchi M, Oyama Y. Stereochemistry of Knoevenagel
condensation products from cyanoacetates and aromatic aldehydes. Chem
Pharm Bull 1991;39:3341e2.
[12] Bouwstra JA, Schouten A, Kroon J. Structural studies of the system trans-
azobenzene/trans-stilbene. II. A reinvestigation of the disorder in the crystal
structure of trans-stilbene, C14 H12. Acta Crystallogr C 1984;40:428e31.
[13] Yuan MS, Fang Q, Ji L, Yu WT. 1,10-Dibutyl-3,30-biindolinylidene-2,20-dione.
Acta Crystallogr E 2007;63:342.
[14] Krasovitskii BM, Bolotin BM. Organic luminophors. Leningrad: Khimia; 1984.
[15] Itoh T. Franck-Condon analysis of the absorption and fluorescence spectra of
The geometry of trans-isoindigo was also optimised on this level
with an involvement of solvent effect (DMSO) by non-equilibrium
PCM. The dependence of the total energy on a central bond torsion
was computed on fully optimised geometry except the frozen
central dihedral angle in a range 105e180ꢀ with 15ꢀ step. C2
symmetry constraint was used in these calculations.
TD DFT method was used for the computation of vertical exci-
tation energies on the above-mentioned geometries in vacuum or
DMSO for trans-isoindigo. The B3LYP/6-311þþG(d,p) combination
was employed, too. The solvent effect (DMSO) was also involved by
means of the non-equilibrium PCM.
all trans
a,u-diphenylpolyenes with one to seven polyene double bonds.
J Chem Phys 2005;123. 064302 (1e6).
[16] Zoon PD, Brouwer AM. A pushepull aromatic chromophore with a touch of
merocyanine. Photochem Photobiol Sci 2009;8:345e53.
[17] Plötner J, Dreuw A. Solid state fluorescence of Pigment Yellow 101 and
derivatives: a conserved property of the individual molecules. Phys Chem
Chem Phys 2006;8:1197e204.
[18] Langhals H, Potrawa T, Nöth H, Linti G. The influence of packing effect on
the solid-state fluorescence of diketopyrrolopyrroles. Angew Chem Int Ed
1989;28:478e80.
[19] Imoda T, Mizuguchi J. Strikingly different luminescent properties arising
from single crystals grown from solution or from the vapor phase in a diketo-
pyrrolo-pyrrole analog. J Appl Phys 2007:102. 073529 (1e7).
[20] Catalán J. On the nonplanarity of trans-stilbene. Chem Phys Lett 2006;421:134e7.
[21] Dähne S, Leupold D. Coupling principles in organic dyes. Angew Chem Int Ed
1966;5(12):984e93.
All the methods came from the Gaussian03W program suite [22]
and the default values of computational parameters were used. The
results were analyzed using GaussViewW from Gaussian Inc., too.
[22] Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, et al.,
Gaussian, Inc., Wallingford, CT, 2004.