The aggregation of the merocyanine dyes, depending of the type of the counterions

Article abstract of DOI:10.1016/j.saa.2007.10.017

Counterions affect on the substructures formation in the case of the merocyanine dye, 1-methyl-4-[2-(4-hydroxyphenyl)ethenyl)]piridinium] hydrogensquarate both in gas and condense phase. Spectroscopically and structural elucidation of these aggregates have been performed, using solid-state conventional and linear-polarized IR-spectroscopy of oriented colloids as a nematic liquid crystal suspension, UV-vis spectroscopy, HPLC tandem ESI mass spectrometry, ¹H and ¹³C NMR, TGV and DSC. Quantum chemical DFT calculations have been carried out as well. Experimental and theoretical data are compared with analogous ones of corresponding iodide salt of dye studied.

Full text of DOI:10.1016/j.saa.2007.10.017

Available online at www.sciencedirect.com
Spectrochimica Acta Part A 70 (2008) 1087–1096
The aggregation of the merocyanine dyes, depending
of the type of the counterions
Tsonko Koleva, Bojidarka B. Kolevab,∗, Stanimir Stoyanovc,
Michael Spitellera, Ivan Petkovc
a Institut fu¨r Umweltforschung, Universita¨t Dortmund, Otto-Hahn-Strasse 6, 44221 Dortmund, Germany
b Lehrstuhl fu¨r Analytische Chemie, Ruhr-Universita¨t Bochum, Universita¨tsstraße 150, 44780 Bochum, Germany
c Sofia University “St. Kl. Ohridsky”, Department of Organic Chemistry, 1, J. Bouchier, Blvd., 1164 Sofia, Bulgaria
Received 12 September 2007; received in revised form 12 October 2007; accepted 15 October 2007
Abstract
Counterions affect on the substructures formation in the case of the merocyanine dye, 1-methyl-4-[2-(4-hydroxyphenyl)ethenyl)]piridinium]
hydrogensquarate both in gas and condense phase. Spectroscopically and structural elucidation of these aggregates have been performed, using
solid-state conventional and linear-polarized IR-spectroscopy of oriented colloids as a nematic liquid crystal suspension, UV–vis spectroscopy,
1
HPLC tandem ESI mass spectrometry, H and 13C NMR, TGV and DSC. Quantum chemical DFT calculations have been carried out as well.
Experimental and theoretical data are compared with analogous ones of corresponding iodide salt of dye studied.
© 2007 Elsevier B.V. All rights reserved.
Keywords: Merocyanine dye; Aggregation; Solid-state IR-LD; UV–vis; ESI-MS; HPLC-MS/MS; TGV; DSC; 1H and 13C NMR; Quantum chemical calculations
1. Introduction
but in most cases neither long range ordering nor advanced
functionality could be achieved [23–28]. In our paper we are
described the role of the counterion in the merocyanine dyes,
1-metyl-4-[2-(4-hydroxyphenyl)ethenyl)]piridinium]·X, where
X = I− and hydrogensquarate anion (Schemes 1–3). Spectro-
scopic and structural elucidation of association processes in
gas and condense phase are studied by means of solution and
solid-state conventional and linear-polarized IR-spectroscopy
(IR-LD) of oriented colloids as a nematic liquid crystal suspen-
Dye assemblies offer important functionality in nature and
technology. Prominent examples from nature are the efficient
light harvesting dye aggregates found in purple [1] and green
bacteria [2–5]. The latter are especially remarkable as they
are composed solely of chlorin dyes. The main component
being bacteriochlorophyll that self-assemble into extended rod
aggregates called chlorosomes. Within these rod aggregates
the dyes are packed in a cylindrical fashion by means of
coordinative, hydrogen bonding and ␲–␲-stacking noncova-
lent interactions [2–5]. If such control of dye–dye aggregate
formation could be realized in artificial systems, very inter-
esting properties might arise which will open new avenues
for electronic and photonic materials. However, rational con-
trol of dye–dye interactions in supramolecular architectures
is still a difficult task [6,7]. Most of the known dye assem-
blies are constructed as simple one-dimensional stacks through
␲-system stacking [8–12]. In other examples, hydrogen bond-
ing or metal–ligand coordination [13–22] has been applied,
1
sion, UV–vis, HPLC tandem ESI mass spectrometry, H and
13C NMR, TGV and DSC methods. Quantum chemical DFT
calculations at B3LYP level of theory and 6-311++G** basis
set are applied for structural optimization and predicting of the
spectroscopic properties.
2. Experimental
2.1. Materials and methods
Conventional and polarized IR-LD spectra were recorded
on a Thermo Nicolet 7000 FTIR-spectrometer equipped
with a Perkin-Elmer wire-grid polarizer (4000 and 400 cm−1
range, 150 scans and 1 cm−1 resolution). A 4-cyano-4ꢀ-
alkylbicyclohexyl mixture (ZLI-1695, Merck) was used as
∗
Corresponding author.
E-mail address: BKoleva@chem.uni-sofia.bg (B.B. Koleva).
1386-1425/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.saa.2007.10.017

1088
T. Kolev et al. / Spectrochimica Acta Part A 70 (2008) 1087–1096
gan Surveyor). Using the Excalibur 1.4 software processed
the data obtained. The molecule weight was determined using
FAB mass spectra, measured on a Fusion VG Autospect instru-
ment employing 3-nitrobenzyl alchohol as a matrix. 1H and 13
C
NMR measurements, referenced to sodium 3-(trimethylsylyl)-
tetradeuteriopropionate, were made at 298 K with a Bruker
DRX-400 spectrometer using 5 mm tubes and D2O as solvent.
Quantum chemical calculations were performed using GAUS-
SIAN 98 program package [33]. The visualization is done
by ChemCraft program [34]. Geometries of (1) and (2) were
optimized at two levels of theory: second-order Moller-Pleset
perturbation theory (MP2) and density functional theory (DFT)
using 6-311++G** basis set. In second case B3LYP method,
which combines Becke’s three-parameter non-local exchange
functional with the correlation function of Lee, Yang and Parr
is applied. Molecular geometries of the studied species were
fully optimized by the force gradient method using Bernys’
algorithm. For every structure the stationary points found on
the molecule potential energy hypersurphases were character-
ized using standard analytical harmonic vibrational analysis.
Absence of the imaginary frequencies and negative eigenvalues
of the second-derivative matrix, confirmed that the stationary
points correspond to minima of the potential energy hyper sur-
faces. The calculation of vibrational frequencies and infrared
intensities were checked for which kind of calculations per-
formed agree best with the experimental data. In cases of (1)
and (2) the DFT method provide more accurate vibrational data,
as far as the calculated standard deviations are about 9 cm−1
(B3LYP) and 22 cm−1 (MP2), respectively. B3LYP/6-31G* data
are presented for above discussed modes, where for the better
correspondence between the experimental and theoretical val-
ues, a modification of the results using the empirical scaling
factor 0.9614ismade. In thecaseof MP2calculationsthescaling
factor of 0.8929 is used.
Scheme 1. Reaction scheme and atom numbering of (1).
Scheme 2. Reaction scheme for obtaining of (2).
orientation host, applying the technique, colloid suspension in
nematic liquid crystal [29–32]. The elemental analysis was per-
formed according to the classical methods: C and H as CO2
and H2O, N – through the Duma’s method, chlorine – by titra-
tion with Hg(NO3)2 after a wet digestion of the sample. The
thermogravimetric (TGV) study was performed using Perkin-
Elmer TGS2 apparatus. The EP data are obtained in X-band on
Bruker ER 420 spectrometer. The calorimetric (DSC) ones are
performed on DSC-2C Perkin-Elmer equipment in argon. The
UV-spectra (1 cm quartz cell and concentration range within
1 × 10−5 to 2.5 × 10−6 M) were recorded on Perkin-Elmer
Lambda 15 UV-VIS spectrophotometer. HPLC-ESI MS/MS
analysis. The analyses of the samples were performed with a
Thermo Finnigan surveyor LC-Pump. Compounds were sep-
arated on a Luna C18 column (150 mm × 2 mm, 4 ␮m particle
size) from Phenomenex (Torrance, CA, USA). The mobile phase
consisted of water + 0.1% HCOOH (A) and acetonitrile + 0.1%
HCOOH (B) using a gradient program. The compound was
detected via UV and a TSQ 7000 (Thermo Electron Corpo-
ration, Dreieich, Germany) mass spectrometer. The spectra
were obtained using the TSQ 7000 equipped with an ESI
ion source and operated with the following conditions: capil-
lary temperature 180 ◦C; sheath gas 60 psi and spray voltage
4.5 kV. 1 mg/ml of the sample was dissolved in acetonitrile
and injected into the ion source by an auto sampler (Finni-
2.2. Synthesis
Substances for synthesis of the 1-methyl-4-[2-(4-hydroxy-
phenyl)ethenyl)]pyridinium iodide (1), 1,4-dimethylpyridinium
iodide and 4-hydroxybenzaldehyde were Merck (Germany)
products.
2.2.1. 1-methyl-4-[2-(4-hydroxyphenyl)ethenyl)]pyridinium
iodide (1)
The reaction scheme for obtaining of the (1) is 2.3500 g
(10.0 mmol) 1,4-dimethylpyridinium iodide and 1.2200 g
(10.0 mmol) 4-hydroxybenzaldehyde were mixed in 50.0 ml
toluene (Scheme 1). Then, 5.00 ml acetic acid and 0.77 g
(10.00 mmol) ammonium acetate were added. The obtained sus-
pension is stirring off during 24 h at 298 K. 0.50 ml K HI and
10.00 ml ethanol are added and thus obtained reaction mixture,
is stayed 16 h at room temperature. The received orange precip-
itate was filtered off, washed with C2H5OH and dried on P2O5
at 298 K. Yields 73%. m.p. 274–275 ◦C. TLC factor of Rf = 0.16
(Al2O3 /CH3OH) was obtained. Found: C, 49.6; H, 4.2; N, 4.16;
[C14H14NOI] calcd.: C, 49.0; H, 4.2; N, 4.16%.
Scheme 3. Chemical diagram of (3).

T. Kolev et al. / Spectrochimica Acta Part A 70 (2008) 1087–1096
1089
2.2.2. 1-methyl-4-[2-(4-oxocyclohexadienilydene)
3. Result and discussion
ethylidene]-1,4-dihydropyridine (2)
The obtaining of (2) is shown in Scheme 2. 10.0 mg of (1)
were dissolved in 10.00 ml ethanol and were added 5 ml 1.0 M
KOH. The obtained lilac solution is heated for 2 h at 70 ◦C and
then cooled at 4 ◦C for 16 h. The obtained violet precipitate is
filtered off, dried on P2O3 at 298 K. Yield 98%. m.p. 222–224 ◦C
and Rf = 0.45 (cellulose/CH3OH). Found: C, 75.6; H, 6.2; N, 6.7;
[C14H13NO] calcd.: C, 75.5; H, 6.2; N, 6.7%.
Independently of the intensively study of the merocyanine
dyes, the crystallographic data for (1)–(3) absent in the litera-
ture, giving a motive for their structural elucidation, using the
possibilities of IR-LD spectroscopy of oriented colloids. Last
approach is appeared to be unique for obtaining of the structural
in solid state, allowing investigation of amorphous samples, like
here studied. The adequate assignment of the experimental IR-
band and structural elucidation are supported by the quantum
chemical calculations and vibrational analysis both of cationic
and quinoide form of merocyanine dye (1).
2.2.3. 1-methyl-4-[2-(4-hydroxyphenyl)ethenyl)]pyridinium
hydrogensquarate (3)
Equimolar amounts of (2) and squaric acid are mixed in
20 ml ethanol. Thus obtained mixture was stirred for 10 h at
40 ◦C (Scheme 3). The obtained solution is cooled at 4 ◦C
for 10 h. The obtained orange precipitate was filtered off,
washed with methanol and dried on P2O5 at 298 K. Yield
86%. m.p. 281–283 ◦C and Rf = 0.55 (Al2O3/ethanol). Found:
C, 66.6; H, 4.6; N, 4.2; [C18H15NO5] calcd.: C, 66.5; H, 4.7; N,
4.3%.
3.1. Quantum chemical calculations
Optimized geometry of both (1) and (2) are shown in
Scheme 4. Two type of stable conformers in (1) at both level
of theoretical approximation are obtained. Depending of the
C–O(H) bond rotation, i.e. the dihedral angle C13–C14–O15–H
(Scheme 4A), two conformers with values of 0.1◦ and 177.6◦
are obtained. When the energies are corrected with the zero
point vibrational energy, the first conformer is more stable
TGV and DSC analysis in the range 300–500 K showed that
no solvent was included in all of the obtained compounds.
Scheme 4. Optimized molecular geometry of (1) (A) and (2) (B) at B3LYP/6-311++G** level of theoretical approximation.
Scheme 5. Visualizing of selected transition moment directions of (1) (A) and (2) (B).

1090
T. Kolev et al. / Spectrochimica Acta Part A 70 (2008) 1087–1096
Table 1
Electron charge values of (1) and (2) using atom numbering Scheme 1
Atom
(1)
(2)
N
−0.074
0.006
−0.199
−0.012
−0.211
0.062
−0.246
−0.003
−0.139
−0.056
−0.236
0.281
C2
C3
C4
C7
C8
C1ꢀ
C2ꢀ
C3ꢀ
C4ꢀ
O
−0.175
0.135
−0.255
0.064
−0.164
−0.036
−0.199
0.148
−0.228
−0.317
with Erel of 4 kJ/mol. The optimized geometry parameters
of (1), i.e. bond lengths and angles are typical, comparing
with the data for analogous compounds as trans-4-(4ꢀ-
(N,N-diphenylamino)styryl)-Nꢀ-methylpyridinium iodide, 4-[4-
methoxy-␣-styryl]-N-isopentylpyridinium bromide sesquihy-
drate as est. [35–38]. The obtained bond lengths and angles
differences are less then 0.067 A and 3.8 . The molecule of (1)
is flat with a deviation of total planarity of 0.4◦ (Scheme 4A).
Trans-configuration of the double bond is preferable then cis-
one. Both aromatic fragments (phenyl and pyridinium rings) are
with a local C2␯, symmetry assuming the collinear dispositions
of corresponding A1, B2 and B1 transition moments. Illustra-
tively, the transition moments of 8a, 8b and 11-␥CH are shown
in Scheme 5A.
Fig. 1. Calculated IR-spectra of (1) and (2) at B3LYP/6-311++G** level of
theoretical approximation.
and benzene rings: 1638 cm−1 (8apy), 1599 cm−1 (8a, phenyl).
The symmetric stretching C C mode is at 1552 cm−1. ␥C C
at 989 cm−1; B1 (11-␥CH) at 734 cm−1 (py) and 833 cm−1
(ph). The theoretical IR-spectrum of (2) is characterized with
a disappearance of the typical aromatic bands between 1700
and 1500 cm−1 range (Fig. 1), supposing the stabilization of
quinoide for in (2). The observed bands are with mixed charac-
ter and are assigned as follow: 1657 cm−1 (␯C O), 1618 cm−1
◦
˚
(␯
) and 1580 cm−1 (␯C C), respectively. The o.p. modes
C Cpy
of the double bonds are predicted at 998 cm−1 (␥ CH, cis-
Quantum chemical calculations of (2) shows the stabiliza-
tion of only one conformer with Erel < 5 kJ/mol (0.3 kJ/mol)
(Scheme 4B). The molecule is absolutely flat with a deviation
of the total planarity of 0.3◦. Like in previous case the good cor-
relation between theoretical and experimental crystallographic
data for analogous compounds is obtained [35–38]. The bond
disubstituted), 972 cm−1 (␥ CH, trisubstituted). The bands for
␥
and δs modes are at 774 and 1335 cm−1, respectively.
CH3
C O
The last value is typical for δsCH in N–CH3 group.
3
3.2. UV-spectra
◦
˚
lengths and angles values are differ less then 0.033 A and 5.9 .
The calculated geometries show a co-linearity of the transition
The possible redistribution of the electronic density, typ-
ical push–pull systems, depends of the solvent polarity. As
far as the electron transition is connected with intramolecular
charge transfer (CT), a significant difference between the dipole
moment in ground and exited state is expected, thus determining
the solvatochromism or NLO properties in solution. Depending
of the solvent polarity the CT band, λmax, in (1) is bathochromic
shifted with 70 nm from H2O to 1,2-dichloromethane solvents.
The obtained negative solvatochromic effect has been explained
both with intra- and intermolecular charge transfer. A classi-
cal example is 1-ethyl-4-methoxycarbonylpyridinium iodide,
which underlined negative solvatochromism, has been explained
with that in the ground state, is observed ion pairs and CT band
corresponds to intermolecular charge transfer from iodide ion to
pyridinium cation [39]. In the ground state the dipole moment
is significant, contrary to excited state, where μe have a lower
value due to the radical formation. Therefore, in (1) could be
proposed the observation of the similar processes.
moments of C O (␯C O), C C (νC C) and δs modes. Same
CH3
can be drown for the bands, corresponding to C O and C C
out-of-plane modes ␥C O, ␥
(tri- and disubstituted double
CH
bonds) (see Scheme 5B)
The theoretical electron charge values (Table 1) shows that
the deprotonation of OH group in (1) and obtaining of (2) leads
to partial electron charge redistribution (Scheme 6).
Theoretical calculated IR-spectra of (1) and (2) are shown in
Fig. 1. The highest frequency peak in the IR-spectrum of (1) at
3490 cm−1, corresponds to stretching ␯OH mode. The low inten-
sive bands within 3055–3000 cm−1 range are assigned to CH
stretching vibrations, while the maxima at 2965 and 2892 cm−1
to CH3 stretching vibrations. Within whole 1650–1500 cm−1
range are observed in-plane A1 and B2 modes of pyridine
Looking for developing of the new materials with poten-
tial NLO application, compounds with quinoide like forms as
(2) are more reasonable to be studied, where the significant
Scheme 6. Electron density redistribution of (2).

T. Kolev et al. / Spectrochimica Acta Part A 70 (2008) 1087–1096
1091
Scheme 7. Resonance structure in (1) and (2) depending of the pH.
Table 2
UV-spectra of (2)
Solvent
H2O
CH3OH
C2H5OH
2-propanole
CH2Cl2
λmax [nm]
log ε
444
3.808
482
4.430
514
4.592
525
4.543
614
3.211
bathochromic effect is observed (Table 2). The negative solva-
tochromism with lowering of the solvent polarity and a shifting
of the λmax with 170 nm is obtained for (2). The resonance
structures depend of pH are illustrated in Scheme 7.
Scheme 8. H-aggregates in (2) in acetonitrile solvent.
characterized with two bands at 514 and 621 nm. The presence of
two maxima in (3) like in (1) in acetonitrile again manifested the
H-type of aggregation similar of other merocyanine dye. How-
ever, the different position of the absorption maxima comparing
with (1) proposed the participation of hydrogensquarate anion in
different manner (Fig. 4) (see Scheme 9A). These assumptions
are supported with the HPLC ESI MS/MS data of both discussed
compounds.
Using the parameter of the solvent polarity Z as a molar
energy of the transition in ET, in kcal/mol for CT band the ET
values of sodium salt of (3) are shown in Table 3.
As could be see of the data in Table 3 and Fig. 5, the
analogy between the Z values of sodium salt of (3) with stan-
dard, used by Kosower [39] is obtained. The Z values are
practically equal in solvents acetone, pyridine and cyclohex-
The UV–vis spectrum of (2) in acetonitrile is depicted in
Fig. 2. In contrast to (1) at same conditions, where only one
band λmax = 392 nm is observed, in (2) two bands at 388 and
566 nm are recorded. The observed characteristics in acetonitrile
and well-defined isosbestic points at 455 nm provide good evi-
dence for equilibrium between monomeric and dimeric species
[15,16,18,19,23–26]. It is noteworthy that in addition to the very
intense hypsochromically shifted absorption band for the H-
dimmer (Scheme 8), a weak band appears at longer wavelength
which can be ascribed to the forbidden transition to the lower
energy excitation state.
The UV–vis spectra of (3) and its sodium salt are given in
Fig. 3. Analogously of (1), the (3) is characterizes with a solva-
tochromic effect of 33 nm depending of the solvent used. The
corresponding salt is characterized with a solvatochromism of
167 nm. Similar to (2) in acetonitrile, the salt of (3) in ethanol is
Fig. 2. UV–vis spectra of (1) (line) and (2) (dot line) in acetonitrile
(2.5 × 10−5 M).
Fig. 3. UV–vis spectra of the sodium salt of (3) in water, methanol and ethanol
(line) and of neutral form in water (dots line).

1092
T. Kolev et al. / Spectrochimica Acta Part A 70 (2008) 1087–1096
Scheme 9. Aggregation in (3) (A) and its sodium salt (B).
Fig. 4. UV-spectra of compounds (1) and sodium salt of (3) in water.
Fig. 5. Dependence of Z vs. solvent polarity in (2).
3.3. 1H and 13C NMR data
ane, which means that in a difference of the values of ε, DN
and AN of 35.3, 33.1 and 18.9 [39], the energy of the tran-
sition is keep equal. In the case of sodium salt of (3) only
solvents with significant level of AN leads to a stabilization of
quinoide like structure and determines the solvatochromism of
167 nm. Schematically the phenomenon could be explained with
dipole–dipole solute–solvent interactions with participation of
hydrogensquarate ion. As far as Z depends of the temperature we
are obtained a dependence of λmax vs. temperature for sodium
salt of (3), shown in Fig. 6. With the decreasing of the tempera-
ture, a hypsochromic shifting of CT band is obtained (Table 4).
The decreasing of the Z with an increasing of the temperature
shown that the solute–solvent interactions weaker less at higher
temperature.
The stabilization of quinoide structure in (2) after the depro-
tonation of OH-group in (1) or (3) is elucidated by means of the
1H and 13C NMR. Using the atom numbering Scheme 1, the
1H-chemical signals in (1) are assigned: chemical shifts within
8.50–9.21 ppm (C2, C6), 8.00, 7.50 and 7.70 ppm—C3, C5, C7
and C8, respectively. The aromatic protons are characterized
with a signals at 7.00 (H-2ꢀ), 6.80 ppm (H-3ꢀ), 7.24 ppm (H-
5ꢀ) and 7.50 ppm (H-6ꢀ). The 13C NMR signals of (1) are at
132.6 ppm (C2, C6), 122.45 ppm (C3, C5), 152.55 ppm (C4),
118.12 ppm (C7) and 131.85 ppm (C8), respectively. In (2)
the stabilization of the quinoide structure, leads to a shifting
of the proton chemical signals of C2, C6, C3, C5, C7 and
C8 with ꢀδ of 1.24, 0.77, 0.70 and 1.00 ppm. The aromatic
proton chemical signals are shifted as well and are observed
1
Table 3
within 0.55–1.03 ppm. The data are HꢀNMR, δ = 3.93 (s, 3H,
ET values kcal/mol of sodium salt of (3) at standard ET(30) [39]
ꢀ
ꢀ
ꢀ
N–CH3), AA‘BB’ (δAA = 6.54, 2H, H-2 , H-6 , δBB = 7.18, 2H,
H-3ꢀ, H-5ꢀ), AB-signals (δA = 6.53, 1H, H-8, dB-7.39, 1H, H-
Solvent
ET(30) [39] [kcal/mol]
Sodium salt of (3)
H2O
63.1
55.4
51.9
48.4
39.1
42.2
40.5
30.9
64.4
59.3
55.6
54.5
46.6
48.6
48.0
48.3
Table 4
CH3OH
C2H5OH
2-Propanol
CH2Cl2
AIIeTOH
C5H5N
Difference of ꢀλmax vs. ꢀT in the case of sodium salt of (3)
Solvent
ꢀλmax [nm] ꢀT [◦C] Solvent
ꢀλmax [nm] ꢀT [◦C]
H2O
C2H5OH
(CH3)2C
9
36
40
75
95
65
CH3OH
C3H7OH 40
CH2Cl2 37
19
65
105
65
Cyclo-C6H12
O

T. Kolev et al. / Spectrochimica Acta Part A 70 (2008) 1087–1096
1093
Fig. 6. λmax depending of temperature and type of solvents in the case of sodium salt of (3).
7, 3JAB = 16.1 Hz), AA‘BB’ signals (δAA = 7.46, 2H, H-2, H-6,
The difference IR-LD spectra of (1) and (2), oriented as a
colloid suspension in nematic liquid crystals (Figs. 7.2 and 7.4)
show significant degree of orientation of suspended particles
allow the adequate IR-LD data interpretation. The elimination
of the bands at 1641 cm−1 in the IR-LD spectra of (1) leads to
a disappearance of the bands at 1599 and 1560 cm−1 (Fig. 8A).
ThisresultconfirmsthepresentedinScheme4Ageometryof(1),
due to the presented elimination is possible when the discussed
transition moments are mutual collinear disposed. An addi-
tional confirmation of this statement follows from the observed
elimination of the bands at 833, 736 and 989 cm−1 at same
dichroic ratio (Fig. 8B). The presented in scheme geometry of
(2) is experimentally supported by the application of reducing
ꢀ
δBB = 8.30, 2H, H-3, H-5). The 13C NMR spectrum is also sig-
ꢀ
nificant differ and the data are δ = 46.58 (q, N-CH3), 115.11 (d,
C7), 120.78 (d, C-3ꢀ, C-5ꢀ), 121.55 (s, C-1ꢀ), 122.39 (d, C-3, C-
5), 132.23 (d, C-2ꢀ, C-6ꢀ), 143.18 (d, C-2, C-6), 144.56 (d, C-8),
155.65 (s, C-4) и 174.76 ppm (s, C O). Comparing the magnetic
resonance spectra of (1) and (3) it is observed a significant dif-
ference only of the aromatic fragments proton signals, which are
downfield shifted in (3) with ꢀδ of 0.4–0.6 ppm. In correspond-
ing 13C NMR spectrum are observed four new signals within
190–170 ppm, corresponding to carbons in hydrogensquarate
anion.
3.4. Conventional and linear-polarized IR-spectroscopic
data
Conventional IR-spectra of (1) and (2) are shown in
Figs. 7.1 and 7.3, respectively. In (1) the observed maxima at
3150 and 3021 cm−1 can be assign to stretching vibrations of
OH and CH groups, in 4-position and trans-disubstituted dou-
ble bond. In 1700–1400 cm−1 spectroscopic range are observed
significant differences comparing with the IR-spectrum of (2)
(see Figs. 7.1 and 7.3) assuming the stabilization of quinoide
structure in (2) after deprotonation of OH-group. The IR-
spectroscopic bands are assigned: (1), 1641 cm−1 (8a, py),
1599 cm−1 (8a, ph), 1560 cm−1 (␯C C); (2) 1648 cm−1 (␯
,
C C
py), 1616 cm−1 (␯C C, ph) and 1560 cm−1 (␯C C), respectively.
The theoretical and experimental data are differing with 9 cm−1
,
illustrating a good correlation. The δsCH banding mode is at
3
1343 cm−1 in (1) and at 1334 cm−1 in (2). As far as the positive
charge in N-atom in (1) is redistributed in (2), the observed low
frequency shifting of the discussed maximum is expectable and
is typical for other merocyanine dyes. For the adequately struc-
tural elucidations of the compounds studied the assignment of
the out-of-plane (o.p.) modes is also important to be done. (1)
is characterized with bands at 989 cm−1 (␥ CH), 736 cm−1 (B1,
py) and 833 cm−1 (11-␥CH, ph) and at 993, 972 and 774 cm−1
in corresponding IR-spectrum of (2).
Fig. 7. Non-polarized IR-(1) and difference IR-LD (2) spectra of compounds
(1)–(3).

1094
T. Kolev et al. / Spectrochimica Acta Part A 70 (2008) 1087–1096
Fig. 8. Non-polarized IR-(1) and reduced IR-LD spectra of (1) after elimination of the band at 1641 cm−1 (A) and 833 cm−1 (B).
difference procedure to polarized IR-spectra of (2). The simul-
taneously elimination of the bands at 1648, 1618, 1560 and
1334 cm−1 in equal dichroic ration could be realized namely
in the frame of the molecular geometry shown in Scheme 4.
The other elimination of the bands at 993, 972 and 774 cm−1
in same dichroic ratio is an additional proves for the presented
geometry.
The IR-spectroscopic pattern of (3) in solid state are sig-
nificant complicated (Fig. 7.5) comparing with the data of (1)
and (2) because of the presence of hydrogensquarate anions is
characterized with a series of absorption peaks within whole
1800–400 cm−1 region. Surprisingly, thedifferenceIR-LDspec-
tra of (3) is characterized with a significant degree of macro
orientation of the suspended particles, which is typical for lay-
Fig. 9. Non-polarized IR-(1) and reduced IR-LD spectra of (3) after elimination of the band at 1644 cm−1 (A) and 846 cm−1(B).

T. Kolev et al. / Spectrochimica Acta Part A 70 (2008) 1087–1096
1095
C O, observed in (2) at 1648 cm−1 is low frequency shifted with
ered structures [36–38] (Fig. 7.6). It is well known that is typical
␯
10 cm−1. This means that in last suspend could be proposed the
formation of dimmers with participation of merocyanine dyes
and hydrogensquarate anion (Scheme 9B).
for merocyanine dyes [36–38], but is it unusual for hydrogen-
squarates. Absence of the ␯OH bands in 3500–3200 cm−1 region
and the observation of broad band within 3000–2600 cm−1 of
␯OH of hydrogensquarate ions intermolecular interacted assume
the participation of ␯OH of cation in strong intermolecular inter-
actions with anions. In 1850–1500 cm−1 are observed three new
bands comparing with the IR-spectra of (1): at 1805, 1662 and
3.5. ESI-MS data
The most intensive signal in the mass spectrum of (3) is at m/z
326.88 corresponding to singly charged [C18H16NO5]+ cation
with molecular weight of 326.32 or the data indicate the cation-
anion associate in gas phase. The next peak at 213.34 is assigned
to [C14H15ON]+ with molecule weight of 213.22 of merocya-
ninedye. Thesedatesupportedthestatementabouttheformation
of stable associates between cation and anions included in (3)
which forms the H-types aggregates in solution. Moreover, the
ESI-MS data of (1) show the signal at m/z of 213.88, corre-
sponding only to [C14H15ON]+ cation. In the case of sodium
salt of (3) the observed results also proves the statement about
the presence of stable associates in gas phase between merocya-
nine dyes and hydrogensquarate species. The signal of negative
ESI-MS spectrum at m/z 325.66 corresponds to [C18H15NO5]−
anion.
s
as
1544 cm−1 corresponding to ␯C=O, ␯
and ␯
stretches
C C
C=O
of hydrogensquarate. The observed significant low frequency
as
shift of ␯
with 38 cm−1 from the typical for hydro-
C=O(Sq)
gensquarated values, also support the assumption for strong
intermolecular interaction of these species. The values of the
aromatic fragments bands are practically uninfluenced compar-
ing with the data of (1) and are observed at 1644 cm−1 (8a(py)),
1598 cm−1 (8a(ph)). In 850–750 cm−1 range a high frequency
shifting of the B1(py) and 11-␥CH out-of plane modes with 72 and
13 cm−1 are observed, comparing with the data of (1). This result
indicatesaclosedisposedandsignificant␲-interactedmolecules
in the layers, probably formed by dimmers shown in Scheme 8.
The elimination of the 8a(py) band at 1644 cm−1 leads to dis-
appearance (i.e. equal dichroic ratio) of the band at 1598 cm−1
s
(8a(ph)) and ␯C=O(Sq) at 1805 cm−1, unambiguously defined the
co-linearity of corresponding transition moments (Scheme 5) in
the frame of the dimmers (Fig. 9A). This geometry is predicted
theoretically as well to be the most stable for the compound
studied at B3LYP/6-311++G** level of theory (Scheme 4).
Theplanargeometryofthecationsisconfirmsbytheobtained
reduction of the 808 and 846 cm−1 bands at relatively equal
dichroic ratio (Fig. 9B). In parallel the 721 cm−1 band is also
disappeared. Due to it belongs to out-of-plane band of hydro-
gensquarate, this result additionally supported the assumption
for the interactions of type in Scheme 10 in solid (3).
4. Conclusion
On the basis of our spectroscopic and structural elucidation of
dyes 1-metyl-4-[2-(4-hydroxyphenyl)ethenyl)]piridinium]·X,
where X is I− or hydrogensquarate anion, by means of
solution and solid-state conventional and linear-polarized
IR-spectroscopy of oriented colloids in nematic host, UV–vis,
HPLC tandem ESI mass spectrometry, H, 13C NMR, TGV
1
and DSC methods, the role of counterions on the aggregation
of merocyanine dyes is studied. The following essential
conclusions can be drawn: (i) in the solid state, the merocyanine
dyes forms supramolecular structures with significant ␲–␲
interactions. In the case of hydrogensquarate derivative the
anion take part in the ␲-stacking as well; (ii) in solution the
systems studied form H-types of aggregates. Hydrogensquarate
ion takes part in the interactions between two dye–anion
associates; (iii) in gas phase only dye-anion associates are
obtained.
It is interesting to observed the typical multiple character
of out-of-plane modes of aromatic and hydrogensquarare frag-
ments in solid-state IR-spectra of (1) or (3), which is observed in
the cases of significant ␲–␲ interacted systems [15,16,23–26]
in the IR-spectra of acetonitrile solutions. Moreover the dis-
as
cussed significant low frequency shifting of the of ␯C=O(Sq) with
30 cm−1 and an absence of ␯
is also observed in (3). This
OH
means that the conclusion for the dimmer aggregates of type in
Scheme 8 is obtained by IR-spectroscopy as well. It is interest-
ing to note the their-spectrum of sodium salt of (3) obtained,
keep the discussed tendencies. An observation of the band of
Acknowledgements
B.K. whish to thank the Alexander von Humboldt Founda-
tion for the Fellowship. T.K. whish to thank the DAAD for a
grant within the priority program “Stability Pact South-Eastern
Europe” and the Alexander von Humboldt Foundation
References
[1] T. Pullerits, V. Sundstrom, Acc. Chem. Res. 29 (1996) 381.
[2] A.R. Holzwarth, K. Schaffner, Photosynth. Res. 41 (1994) 225.
[3] H. Tamiaki, Coord. Chem. Rev. 148 (1996) 183.
[4] J.M. Olson, Photochem. Photobiol. 67 (1998) 61.
[5] T. Nozawa, K. Ohtomo, M. Suzuki, Y. Morishita, M.T. Madigan, Bull.
Chem. Soc. Jpn. 66 (1993) 231.
Scheme 10. Optimized molecular geometry of interacted species in (3).
[6] H. Engelkamp, S. Middelbeck, R.J.M. Nolte, Science 284 (1999) 785.

1096
T. Kolev et al. / Spectrochimica Acta Part A 70 (2008) 1087–1096
[7] P.A.J. De Witte, M. Castriciano, J.J.M. Cornelissen, L. Monsu-Scolaro,
R.J.M. Nolte, Chem. A Eur. J. 9 (2003) 1775.
[28] G. Ashwell, K. Skjonnemand, G. Paxton, D. Allen, J. Mifflin, X. Li, J.
Mater. Chem. 11 (2001) 1351.
[8] C. van Nostrum, Adv. Mater. 8 (1996) 1027.
[9] J. Vacus, J. Simon, Adv. Mater. 7 (1995) 797.
[29] B.B. Ivanova, M.G. Arnaudov, P.R. Bontchev, Spectrochim. Acta 60A
(2004) 855.
[10] A. Sautter, C. Thalacker, F. Wu¨erthner, Angew. Chem. Int. Ed. 40 (2001)
40.
[11] F. Wu¨rthner, C. Thalacker, S. Diele, C. Tschierske, Chem. A Eur. J. 7 (2001)
2245.
[12] F.S. Schoonbeek, J.H. van Esch, B. Wegewijs, D. Rep, M. de Haas, T.
Klapwijk, R. Kellogg, B. Feringa, Angew. Chem. Int. Ed. 38 (1999) 1393.
[13] A. Ajayaghosh, S. George, J. Am. Chem. Soc. 123 (2001) 5148.
[14] A. Ajayagosh, S. George, V. Praveen, Angew. Chem. 115 (2003) 346.
[15] F. Wu¨rthner, C. Thalacker, A. Sautter, Adv. Mater. 11 (1999) 754.
[16] F. Wu¨rthner, C. Thalacker, A. Sautter, W. Schartl, W. Ibach, O. Hollricher,
Chem. A Eur. J. 6 (2000) 3871.
[17] A.P.J. Schenning, P. Jonkheijm, E. Peeters, E. Meijer, J. Am. Chem. Soc.
123 (2001) 409.
[18] F. Wu¨rthner, S. Yao, B. Heise, C. Tschierske, Chem. Commun. (2001)
2260.
[19] S. Yagai, T. Karatsu, A. Kitamura, Chem. Commun. (2003) 1844.
[20] T.N. Milic, N. Chi, D. Yablon, G. Flynn, J. Batteas, C. Drain, Angew. Chem.
Int. Ed. 41 (2002) 2117.
[30] B.B. Ivanova, D.L. Tsalev, M.G. Arnaudov, Talanta 69 (2006) 822.
[31] B.B. Ivanova, V.D. Simeonov, M.G. Arnaudov, D.L. Tsalev, Spectrochim.
Acta 67A (2007) 66.
[32] B.B. Koleva, T. Kolev, V. Simeonov, T. Spassov, M. Spiteller, J. Colloid.
Int. Sci., submitted for publication.
[33] M. Frisch, G. Trucks, H. Schlegel, G. Scuseria, M. Robb, J. Cheeseman,
V. Zakrzewski, J. Montgomery Jr., R. Stratmann, J. Burant, S. Dapprich, J.
Millam, A. Daniels, K. Kudin, M. Strain, O. Farkas, J. Tomasi, V. Barone,
M. Cossi, R. Cammi, R. Mennucci, C. Pomelli, C. Adamo, S. Clifford,
J. Ochterski, G. Petersson, P. Ayala, Q. Cui, K. Morokuma, P. Salvador,
J. Dannenberg, D. Malick, A. Rabuck, K. Raghavachari, J. Foresman, J.
Cioslowski, J. Ortiz, A. Baboul, B.B. Stefanov, G. Liu, A. Liashenko, P.
Piskorz, I. Koma´romi, R. Gomperts, R. Martin, D. Fox, T. Keith, M. Al-
Laham, C. Peng, A. Nanayakkara, M. Challacombe, P. Gill, B. Johnson, W.
Chen, M. Wong, L. Andres, C. Gonzalez, H. Head-Gordon, E. Replogle,
J. Pople, Gaussian 98, Gaussian, Inc., Pittsburgh, PA, 1998.
[34] G.A. Zhurko, D.A. Zhurko, ChemCraft: tool for treatment of chemical data,
Lite version build 08 (freeware), 2005.
[21] M. Arabe, K. Yamashita, K. Sada, S. Shinkai, A. Ikeda, S. Sakamoto, K.
Yamaguchi, J. Org. Chem. 68 (2003) 1059.
[22] F. Wu¨rthner, C. You, C. Saha-Moller, Chem. Soc. Rev. 33 (2004) 133.
[23] S. Yao, U. Beginn, T. Gress, M. Lysetska, F. Wu¨rthner, J. Am. Chem. Soc.
126 (2004) 8336.
[24] S. Rusch, S. Yao, R. Wortmann, F. Wu¨rthner, Angew. Chem. 118 (2006)
7184.
[25] F. Wu¨rthner, S. Yao, B. Beginn, Angew. Chem. Int. Ed. 42 (2003) 3247.
[26] F. Wu¨rthner, S. Yao, Angew. Chem. Int. Ed. 39 (2000) 1978.
[27] G. Ashwell, G. Jefferies, E. Dawnay, A. Kuczynski, D. Lynch, G. Yu, D.
Bucknall, J. Mater. Chem. 5 (1995) 975.
[35] M. Gruselle, B. Malezieux, S. Benard, C. Train, C. Guyard-Duhayon, G.
Gredin, K. Tonsuaadu, R. Clement, Tetrahedron Asy. 15 (2004) 3103.
[36] E. Cariati, R. Ugo, F. Cariati, D. Roberto, N. Masciocchi, S. Galli, A. Sironi,
Adv. Mater. 13 (2001) 1665.
[37] X. Tan, S. Sun, W. Yu, D. Xing, Y. Wang, C. Qi, Acta Crystallogr. 60E
(2004) o1054.
[38] P. Ren, J. Qin, C. Chen, D. Zhang, H. Hu, J. Chem. Crystallogr. 34 (2004)
291.
[39] C. Reihardt, Solvents and Solvent Effects in Organic Chemistry, VCH, NY,
1990.