Two heterocyclic merocyanine classes and their optical properties in relation to the donor-acceptor strength of end substituents

Article abstract of DOI:10.1007/s10593-012-0900-y

The relationship between the first electronic transition energy of heterocyclic merocyanines and donor-acceptor properties of their end substituents has been quailtatively treated and analytically substantiated in terms of the classical valence bond model and the biradicaloid theory. A simple graphical technique has been suggested to classify donor-acceptor-substituted compounds in relation to their electronic structure and to predict the effect of donor-acceptor strength on their optical behavior. In this context, merocyanines can be divided into two classes which differ in charges and electron populations of their donor and acceptor moieties and exhibit mirror-like responses to the variation of donor-acceptor strength. Two families of new heterocyclic merocyanines have been synthesized and their experimentally observed spectroscopic properties have been rationalized using the approach developed.0009-3122/12/4710-12442012 Springer Science+Business Media, Inc.

Full text of DOI:10.1007/s10593-012-0900-y

Chemistry of Heterocyclic Compounds, Vol. 47, No. 10, January, 2012 (Russian Original Vol. 47, No. 10, October, 2011)
TWO HETEROCYCLIC MEROCYANINE CLASSES AND
THEIR OPTICAL PROPERTIES IN RELATION TO THE
DONOR-ACCEPTOR STRENGTH OF END SUBSTITUENTS
W. Rettig1*, M. L. Dekhtyar2*, A. I. Tolmachev2, and V. V. Kurdyukov2
The relationship between the first electronic transition energy of heterocyclic merocyanines and donor-
acceptor properties of their end substituents has been quailtatively treated and analytically
substantiated in terms of the classical valence bond model and the biradicaloid theory. A simple
graphical technique has been suggested to classify donor-acceptor-substituted compounds in relation to
their electronic structure and to predict the effect of donor-acceptor strength on their optical behavior.
In this context, merocyanines can be divided into two classes which differ in charges and electron
populations of their donor and acceptor moieties and exhibit mirror-like responses to the variation of
donor-acceptor strength. Two families of new heterocyclic merocyanines have been synthesized and
their experimentally observed spectroscopic properties have been rationalized using the approach
developed.
Keywords: heterocyclic merocyanines, cyanine limit, donor-acceptor strength, first electronic transition,
graphical hyperbolic model.
Quasi-linear conjugated polyenes (merocyanines) terminated by donor-acceptor (DA), mostly
heterocyclic residues can boast of the nearly one-hundred-year-long research history started by the legendary
works of König [1] in the 1920's and then remarkably contributed to by Brooker [2] and Kiprianov [3] in the
1940–60's and their abundant fruitful successors in the late past century [4, 5]. The perennial and consistent
interest in these compounds rests, first, on their surprisingly simple structure, affording a convenient testing
ground for theoretical conjectures, and, second, on their outstanding significance in optical engineering [6–9].
Notwithstanding the thorough and extensive theoretical studies of the past decades [10–14], the
inexhaustibility of merocyanines leaves enough space for new attempts to rationalize their structure–property
relationships. As an example, here we address the trends in merocyanine S0–S1 transition energies (E01) in
relation to the interplay of DA strength (SDA) of end substituents and the mesomeric coupling between them. As
was previously found by one of the authors [15] using ab initio calculations on various derivatives of the
_______
*Dedicated to the memory of M. O. Lozinskii.
**To whom correspondence should be addressed, e-mail: rettig@chemie.hu-berlin.de.
1Institut für Chemie, Humboldt-Universität zu Berlin, Brook-Taylor Str. 2, Berlin D-12489, Germany.
2Institute of Organic Chemistry, National Academy of Sciences of Ukraine, Murmanskaya St. 5, Kyiv 02094,
Ukraine; е-mail: mdekh@mail.ru.
__________________________________________________________________________________________
Translated from Khimiya Geterotsiklicheskikh Soedinenii, No. 10, pp. 1506-1520, October, 2011.
Original article submitted May 26, 2011.
1244
0009-3122/12/4710-1244©2012 Springer Science+Business Media, Inc.

Dimroth-Reichardt dye ET-30 [16], a situation can be reached for a certain DA strength where the transition
energy reduces to zero in a 90°-twisted system. We can call this the "critically heterosymmetric" situation
[17, 18], which is equivalent to the "cyanine limit" (CL) [19] for these systems.
This equivalence becomes clear if coupling between the subunits is introduced (e.g., for planar systems).
For the "critically heterosymmetric" situation, this interaction creates an energy gap between the previously
isoenergetic S0 and S1 states, and the energy gaps for other SDA values are also increased. Nevertheless, this gap
is smallest for the DA strength corresponding just to the critically heterosymmetric point (Fig. 1). The large-
interaction case is experimentally found in symmetric and unsymmetric cyanines, with the symmetric cyanines
corresponding to the CL and having the lowest transition energy and equal contribution of the two main
resonance structures for both S0 and S1 states [2–4]. But a similar situation is also present for planar uncharged
merocyanines [5].
The aim of the present study is not merely to describe the apparent regularity on the qualitative intuitive
basis but also to incorporate it into the classical valence bond theory and the biradicaloid theory [17, 18]. It is
also challenging to use the resulting unified approach for getting an insight into the spectroscopic properties of
newly synthesized heterocyclic merocyanines.
Fig. 1. A schematic representation of the S0–S1 transition energy as a function of DA
strength in the limits of zero interaction between -decoupled molecular subunits
(90°-twisted systems) and strong interaction (planar conjugated systems).
Qualitative treatment of the relation between the first electronic transition energy and the DA
strength. In order to develop the dependence of the transition energy E01 on the DA strength SDA, we define
the latter as the differrence of the highest occupied and lowest unoccupied MO energies in the respective -
decoupled (or fully separated) donor (D) and acceptor (A) moieties of the merocyanine molecule [20, 21].
Taking into account that the first electronic transition implies a DA electron transfer, it is clear that the
interaction between the end substituents dictates the shape and the vertical offset of the curve in Fig. 1. Indeed,
if the D and A moieties are not -conjugated, then the transition energy E01 is substantially determined through
SDA for one branch of the curve and -SDA for the other, i.e., the plot reduces to two symmetric straight lines
inclined at 45o and -45o to the abscissa and crossing it at a certain point (the dashed line in Fig. 1). Any
interaction of the moieties causes their frontier levels to move apart from each other thus raising the electronic
transition energy; correspondingly, the stronger the conjugation between the end substituents, the greater the
increase in the transition energy. At the same time, the bottom of the plot becomes smoothly curved. The
asymptotes to the resulting curve are the lines standing for the non-interaction case, as the transition energy at
large absolute magnitudes of SDA is governed mainly by the energy difference between the transition-relevant
MOs. Thus, it is merely by formal criteria that the dependence of the transition energy on the DA strength is
found to correspond to a hyperbolic function. In what follows, this inference will be verified in terms of the
classical valence bond model.
1245

Treatment in terms of resonance structures and biradicaloid theory. In the framework of the
valence bond approach and restricting ourselves to two main resonance structures, we can divide (het)aryl-
containing stilbenoid merocyanines into two classes, each demonstrating the characteristic resonance
equilibrium of a zwitterionic (Z) and a non-charge-separated form (N). Class I compounds possess an aromatic
N structure, and class II compounds an aromatic Z structure, as exemplified by two well-known representatives
of DA-merocyanines (Scheme 1). The two resonance structures differ by the transfer of one electron between
the D and A ends.
Scheme 1
Class I
Me
Me
Me
N+
N
N
X
N
Me
Z
N
N, ar
Z, ar
Class II
O
O
+
X
hp-dd model
.. . .
hole-pair (hp)
dot-dot (dd)
To generalize this picture, we invoke the biradicaloid theory of Michl and Koutecky [17, 18] describing
such pairs of electron-transfer related structures. The simplest way is to cut off all interactions between the D
and A moieties. This can be done by omitting the -conjugating bridge (in our case the ethylene bridging unit),
or by twisting the subsystems so that -interaction is switched off. Then, we get two limiting resonance
situations: a closed-shell resonance structure with paired electron distribution (called a hole-pair structure or hp)
shown in the left bottom panel of Scheme 1 and a biradical structure (dot-dot or dd) with charged moieties Z for
class I and uncharged moieties N for class II shown in the right bottom panel. The dd-structures are formally
derived from their hp-counterparts by taking an electron from the D moiety and placing it on the A moiety.
The Z state of moieties of class I is very well known as the prototype of twisted internal charge-transfer
(TICT) states [22, 23]. Although stilbenoids like the one shown in Scheme 1 do not form TICT-states because
valence bond resonance stabilization leads to a planar geometry, the closely related biphenyl derivatives do form
TICT-states with very high excited state dipole moments, if sterically hindering substituents favor the
orthogonal conformation [23, 24].
Starting from the hp- and dd-resonance structures, we can then introduce the interaction through a
chemical bridge. Considering only the two resonance structures and neglecting all other interactions (e.g., with
locally excited states) leads to a situation where the ground and excited states S0 and S1 possess a mixed hp-dd
character. Which of the two states is of more dd- or hp-nature depends on the electronegativity difference
between the D and A (or, essentially, on SDA) and the interelectronic repulsion [17, 18, 22].
For uniformity, the donor-acceptor strength SDA is defined for the aromatic form of both classes. As a
consequence, classes I and II exhibit opposite relations between SDA and the dipole moment of the ground and
excited states. If SDA is small, the ground state of class I compounds (aromatic valence bond structure with the
closed shell hp-pattern) is less polar than the excited state of dd-character (on the left side of the left panel in
Fig. 2), but for very strong SDA, the more polar state with the predominant Z contribution can also become S0 (on
the right side of the left panel). At the crossing of the two lines, S0 and S1 have equal contributions of Z and N.
1246

In contrast, for class II compounds at small SDA, the ground hp-state possesses charged moieties and is more
polar than the excited dd-state (on the left side of the right panel in Fig. 2). Again, for sufficiently large SDA, the
CL is reached with equal dipole moments for S0 and S1, and if SDA is further increased, the situation is reversed,
with S0 of predominantly dd-character (quinoid valence bond structure for planar systems) and the dipole
moment of S1 being larger than that of S0 (on the right side of the right panel in Fig. 2).
Fig. 2. State classification and energetic ordering derived from the two-state hp–dd
interaction model for two classes of DA compounds in relation to the donor-acceptor
strength of their substituents.
The situation in which the Z form becomes predominant in the ground state is much less probable for
class I than for class II compounds. This trend is attributable to aromatic stabilization which energetically favors
aromatic forms, N for class I and Z for class II. In other words, the horizontal line referring to the aromatic hp
state energy in both panels of Fig. 2 is lowered and the normal region for classes I and II is effectively extended.
It is for this reason that TICT-compounds always belonging to class I have never been encountered to show
zwitterionic structures in the ground state, whereas zwitterionic class II compounds exist in abundance (among
well-known examples there are Brooker’s merocyanine (4-[(1-methyl-4(1H)-pyridinylidene)ethylidene]-2,5-
cyclohexadien-1-one [25]) and Dimroth-Reichardt's betaine ET-30 dye (2,6-diphenyl-4-(2,4,6-triphenyl-
N-pyridino)phenolate [16]).
Analytical approach in terms of the valence bond model. To derive the relationship between the
charge transfer transition energy and SDA of the merocyanines concerned, we follow the analytical approach of
the valence bond model developed by Förster [26] and later Platt [10]. In this notation, the ground state
wavefunction of the DA system DA(S0) is a mixture of the two valence bond wavefunctions, Z and N, and
depends on the square of a mixing coefficient c (Eqs. 1 and 2).
DA(S0) = c Z + (1 – c2)1/2 N
DA(S1) = (1 – c2)1/2 Z – c N
(1)
(2)
The value of the transition energy ΔE01 can be obtained by solving the 22 secular determinant (Eq. 3)
involving the energies bZ and bN of the two resonance structures Z and N on the main diagonal and the energy F
of their interaction as off-diagonal elements.
bZ  E 0.5F
0
(3)
0.5F bN  E
1247

The excitation energy is then given by Eq. 4
E01  F2  b2
(4)
where b = bZ – bN is the energy difference between the two resonance structures, and the squared coefficient c2
of the Z structure in the ground-state wavefunction appears as follows:
F2  b2 b
2 F2  b2
c2 
(5)
We can readily see that Eq. 4 represents a hyperbolic dependence E01 as a function of b as depicted
graphically in Fig. 3. The shape of the plot and the offset of the minimum above the abscissa are specified by the
interaction parameter F. If the interaction between two resonance structures or, in our terms, between D and A
moieties, is absent (F = 0), then E01 = b, i.e., the plot represents two straight lines with the common point at b
= 0 where the transition energy is formally zero. At a nonzero F value, the energy gap widens so that E01 > 0
throughout the range of b variation, but the minimum is found again at b = 0. This mimimum is CL
characterized by the equal energy of N and Z resonance structures and, accordingly, by c2 = 0.5, as follows from
Eq. 5.
Thus, both wavefunctions contribute equally to S0 and S1 at the CL point. Importantly, the transition
energy in CL is merely equal to F, thus affording a direct estimate of the DA interaction. At large absolute
magnitudes of b, the transition energy asymptotically approaches b, since it is then dominated by the energy
difference between two resonance structures and the contribution from their interaction can be neglected (Eq. 4).
Here we call the left branch of the hyperbola (where the hp-state is energetically preferred) the "normal region"
and the right branch (where the dd-state becomes a lower one) the "overcritical region". These definitions are
consistent with the situation arising from the perpendicular twist of the D/A moiety: for the resulting twisted
molecule, the ground hp-state (in the normal region) is much more typical than the biradical dd-state.
Fig. 3. S0–S1 transition energy as a function of the energy difference between
resonance structures (or DA strength) for two classes of DA compounds: going from
the zero-interaction 22 model to stronger interaction cases.
1248

It is evident that b and SDA, in essence, are very close, since either quantity provides the main
contribution to the excitation energy. On the other hand, the two parameters, one belonging to the valence bond
model and the other to the MO approach, can hardly be superposed in the strict sense. Among the distinctions to
be noted, there are the electron repulsion effects: they are implicitly included in the excitation energy as
specified by Eq. 4 but should be explicitly taken into account on going from b to SDA. Notwithstanding
limitations of this kind (to be discussed in more detail elsewhere), b and SDA represent much the same energy
difference, and it is readily apparent that the empirically found hyperbolic-type dependence in Fig. 1 is fully
substantiated in the framework of the valence bond model.
It follows from the Z–N and hp–dd-interrelations for class I and II compounds (Scheme 1) that the two
classes exhibit mirror-like representations on the hyperbolic curve E01 versus SDA discussed above. Thus,
ground-state molecular polarity and, accordingly, the c2 value increase with DA strength for class I compounds
and decrease for class II compounds, as shown in Fig. 3. As a result, class I has the less polar N structure-
dominated ground state (and hence positive solvatochromism) in the normal region, and the more polar Z
structure-dominated ground state (negative solvatochromism) in the overcritical region. The opposite is true for
class II: the ground state is dominated by the less polar N structure in the overcritical (positively
solvatochromic) region, and by the more polar Z structure in the normal (negatively solvatochromic) region. In
both cases, the CL point with c2 = 0.5 implies the minimum energy separation of the ground and excited states
(and corresponds to the formal S0/S1 state crossing in Fig. 2). The CL is therefore characterized by the most red-
shifted absorption/emission and the absence of solvatochromism because the dipole moments of the ground and
excited states are equal.
Optical properties of heterocyclic merocyanines in terms of the hyperbolic approach. We now use
the model developed to locate a number of DA compounds relative to CL starting from their spectroscopic data.
The application of the "hyperbolic approach" is illustrated well by the following two newly synthesized families
of heterocyclic merocyanines, both analogous to the well-known Brooker’s merocyanine [25] and differing only
by the substituents on the heterocyclic ring:
Scheme 2
S
t-Bu
t-Bu
O
O
+
N Me
N Me
+
t-Bu
t-Bu
S
4 (NO-DT)
1 (NO-DP)
t-Bu
S
t-Bu
O
O
+
O
O
+
t-Bu
t-Bu
S
5 (OO-DT)
2 (OO-DP)
1249

S
t-Bu
t-Bu
O
O
S
S
+
+
t-Bu
t-Bu
S
3 (SO-DP)
6 (SO-DT)
They were obtained by the general scheme (Scheme 3) [4]:
Scheme 3
O
For 1, 4:
1) Py, piperidine;
Me
R
Bu-t
2) NH4OH
X
+
O
X+
R
For 2, 3, 5, 6:
1) fusion;
R
Bu-t
t-Bu
R
–
ClO4
Bu-t
2) Et3N
1–6
1, 4 X = NMe; 2, 5 X = O; 3, 6 X = S; 1–3 R = Ph; 4–6 R = 2-thienyl
Compounds 1–6 all have the trans conformation of the ethylene spacer, as evidenced by the spin-spin
coupling constants of about 14 Hz for vicinal protons (see EXPERIMENTAL). For convenience of
consideration, all the compounds are presented in the hp-Z aromatic form and seen to belong to class II (as they
are seemingly betainic structures in this form). The negative end of all compounds is the same strong electron
donor, and the positive end is an acceptor of varied strength; hence SDA variations are connected with the
acceptor. The change to the non-charge-separated form is formally effected by transferring an electron from the
D to the A moiety.
For the relative ordering of the compounds in Scheme 2 on the SDA axis, we invoke the previously
reported polarographic [27] and spectroscopic data [28–30]. They show that the phenyl-substituted
merocyanines should be ranked by the acceptor strength (and hence by the total SDA value) as follows: NO-DP <
OO-DP < SO-DP, so that the NO-DP with the weakest acceptor is at the leftmost position. As a consequence, the
ground-state contribution of the Z form, c2, is the largest for NO-DP and the smallest for SO-DP.
If protonated at the negatively charged oxygen atom, all these dyes should strongly shift to the left
towards the normal region along the SDA axis, as their common very strong oxyphenyl donor is converted to a
significantly weaker hydroxyphenyl donor, and all SDA values accordingly decrease, accompanied by an increase
in c2 values. The relative order of the compounds remains, however, unchanged on protonation. The same
reasoning is true for thienyl-substituted merocyanines, with the note that, all factors being equal, they are
assumed to lie somewhat to the left of their phenyl counterparts (since electron-donor thienyl groups placed on
the A moieties should weaken the acceptor properties and hence decrease the SDA values).
To correctly place the two heterocyclic merocyanine families relative to CL, it is necessary to formulate
the experimentally observable criteria of this special point on the "spectroscopic hyperbola":
(i) The transition energy should be at a minimum due to the smallest energy distance between the
transition-relevant MOs (or, in other words, as a result of the vanishing energy difference between the hp- and
dd-forms).
(ii) Since E(dd) = E(hp) and, accordingly, hp- and dd-wavefunctions contribute equally to the S0 and S1
wavefunctions of the system (c2 = 0.5), the dipole moment does not change on excitation, which can be judged
from direct dipole moment measurements.
1250

(iii) As a result of the dipole moment invariance on excitation, the solvatochromism should be
practically absent, i.e., the absorption/emission maxima should exhibit only negligible polarity dependence with a
readily changing sign.
(iv) As another consequence of equal hp- and dd-contributions to the S0 and S1 wavefunctions, the bond
lengths do not change on excitation, and S0 and S1 hypersurfaces should have a very small displacement,
resulting in structured absorption/emission spectra with the 0→0 band being most intensive and carrying most of
the oscillator strength.
In what follows, the above criteria will be juxtaposed to the spectroscopically observed optical behavior of
the merocyanines under study. Figures 4–6 as well as Table 1 summarize the spectral characteristics of the
merocyanines in acetonitrile both in basic and acidic media. In the latter case, the dyes (accordingly designated
as NOH-DP, etc.) are considered to be protonated at the exocyclic oxygen atom. We mainly focus on absorption
spectra as they are most informative. At the same time, the fluorescence spectra are also included here although
they are extremely weak in the case of the basic form of the dyes. In this case, the concentration had to be raised
considerably in order to reduce the solvent background sufficiently. We therefore expect a nonnegligible
contribution of secondary inner filter effect in these spectra. Nevertheless, they are instructive in showing an
approximate mirror-image relationship between absorption and emission.
Fig. 4. Normalized absorption and fluorescence spectra (solid and dashed lines,
respectively) of the diphenyl merocyanines in the basic and acidic (protonated) forms
in acetonitrile at room temperature. Respective dye concentrations in measuring
absorption and fluorescence are 2×10–5 and 2 × 10–6 mol/l.
As seen from Figs. 4 and 5, both phenyl- and thienyl-substituted NO-compounds exhibit the most
pronounced spectral structuring with the 0→0 peak vibronic band as the most prominent one. The lowest energy
absorption bands for OO- and SO-merocyanines are much broader, the former possessing some residual
vibronic structure with the maximum intensity falling at a higher vibronic peak and the latter being completely
unstructured. The spectra of all protonated compounds are practically free of vibronic structure.
1251

Fig. 5. Normalized absorption and fluorescence spectra (solid and dashed lines,
respectively) of the dithienyl merocyanines in the basic and acidic (protonated)
forms in acetonitrile at room temperature. Dye concentrations are the same as in
Fig. 4.
Fig. 6. Solvatochromic behavior of absorption and fluorescence (solid and dashed
lines, respectively) of NO-DP (redshifted spectra) and NOH-DP (blueshifted spectra)
in solvents of different polarity at room temperature. The spectra are normalized; dye
concentrations are the same as in Fig. 4.
Considering that there are several cases of structureless spectra and given that we must compare the 0→0-
transitions of the dyes and not the maxima if these occur for different vibronic bands, we analyzed the spectra of
Figs. 4 and 5 in the following way. From the structured spectra we can directly see that NO-DP and OO-DP
possess 0→0 bands at approximately 15.8 and 15.6×103 cm–1 (column 0–0 of Table 1). Often, 0→0 transition
energies are assumed to be represented by the crossing of the normalized absorption and fluorescence
1252

TABLE 1. Spectral Characteristics of the Dyes Investigated (Acetonitrile,
298 K) in the Basic (NO-DP, etc.) and Acidic Form (NOH-DP, etc.)
max_ ,
abs
max_
,
abs
_ ,
abs–1/2
,
abs–1/4
,
0–0,
fluo
fluo
Compound
N_max
103 cm–1 103 cm–1 103 cm–1 103 cm–1 103 cm–1 103 cm–1
1 (NO-DP)
2 (OO-DP)
3 (SO-DP)
NOH-DP
OOH-DP
SOH-DP
15.8
18.8
18.5
24.6
20.0
18.9
14.2
13.3
13.7
18.2
16.2
15.0
15.0
15.4
15.6
20.9
17.8
16.6
15.1
16.5
16.2
22.8
18.6
17.5
14.8
15.6
15.4
22.1
18.1
17.0
15.8
15.6
15.4
22.1
18.1
17.0
1
2-3
2-3
2
1-2
1-2
4 (NO-DT)
5 (OO-DT)
6 (SO-DT)
NOH-DT
OOH-DT
SOH-DT
15.4
18.5
18.3
24.2
19.7
18.4
13.2
12.9
12.9
17.4
15.8
14.4
14.5
15.1
15.0
20.5
17.4
16.0
14.6
16.3
16.0
22.2
18.3
16.9
14.3
15.3
15.1
21.6
17.8
16.4
15.4
15.3
15.1
21.6
17.8
16.4
1
2–3
2–3
2
1–2
1–2
_______
max
_
and max
_
– absorption and fluorescence maxima, respectively;
fluo
abs
abs_fluo – crossing of the normalized absorption and fluorescence spectra;
abs–1/2 and abs–1/4 – spectral position of the low-energy absorption tail at
half height and quarter height, respectively; 0–0 – approximate position of
the 0→0 absorption peak; N_max – approximate position of the absorption
maximum with respect to the vibronic progression.
spectra (column abs_fluo). This is an acceptable approximation for OO-DP but bad in the case of NO-DP. Here,
we prefer to use the low-energy wings of the absorption spectra at half- and quarter-height (columns abs–1/2 and


abs–1/4) for the broad spectra of the OO- and SO-compounds. The quarter-height energy of OO-DP (column
abs–1/4), 15.6 × 103 cm-1, is the same as the corresponding 0→0 transition energy extracted from the spectrum
(column 0–0), and we therefore take the quarter-height energy as the best approximation of the 0→0 peak for all
the unstructured spectra. If we take a vibronic progression with a spacing of 1200 cm–1, we can then determine
the approximate vibronic location of the absorption maximum from the 0→0 band (column N_max).
From the values of the 0→0 energies in Table 1, we can draw the following conclusions. The dyes in their
basic (unprotonated) form, with the constant D and varied A, show small to negligible differences in transition
energy, which is generally much lower than for the acidic forms. It should be especially noted that there is nearly
no difference between the NO- and OO-compounds, although their relative electronegativities and polarographic
characteristics [27] indicate that the pyrylium moiety has stronger acceptor properties than the pyridinium one.
Hence, all these dyes must be positioned close to the bottom of the hyperbolic curve, i.e., in the vicinity of CL (Fig.
7). It is also evident from the very weak negative solvatochromism of NO-DP shown in Fig. 6 that this compound
is placed slightly to the left of the CL point, i.e., in the normal region.
As already discussed, the acidic (protonated) forms of the merocyanines concerned have a much weaker
D and the same A moiety; hence, they should be shifted to the left as a whole, in the same order as that
established for basic forms. Placing them on the steep slope of the left hyperbola branch is also consistent with
their rather large and markedly differing transition energies. Solvatochromism should practically vanish for
protonated dyes as is the case for NOH-DP (Fig. 6), because charged molecules exhibit much stronger
monopole than dipole interaction [31]. Apparently, there is no crossing CL in going from NO-DP to NOH-DP,
since even the unprotonated form of the dye is already found in the normal (left) region of the hyperbola. As
regards OO-DP and SO-DP, it remains uncertain (until the corresponding solvatochromic data are available)
whether they lie on the same side of CL as their protonated derivatives.
1253

There still also remain some open questions concerning the vibronic structure for the compounds
studied. Contrary to NO-DP and NOH-DP demonstrating the typical vibronic behavior (Fig. 4 and Table 1),
i.e., a narrow strongly structured first-transition band of the basic form (in the vicinity of CL) and a much
broader structureless band of the acidic form (far from CL), the bands of OO-DP and SO-DP are anomalously
broad for dyes close to CL (even broader than the corresponding bands of their acidic forms). In the context of
the Franck–Condon potential energy curves for the S0 and S1 states, the spectral broadening is explained by their
mutual horizontal displacement, which is the larger, the stronger the difference between the S0 and S1 geometric
structure. Given practically unshifted, in terms of bond lengths, S0 and S1 curves at the CL point, we have to
involve an additional factor which could account for band broadening in OO-DP and SO-DP. This may well be
an additional relaxation coordinate and, accordingly, another couple of the Franck–Condon curves on an
orthogonal structural dimension. They can be mutually shifted due to, e.g., a different degree of planarity of the
pyrylium and thiopyrylium moieties in the ground and excited states.
Fig. 7. Placement of the merocyanine families on the hyperbolic curve (empty squares
designate phenyl-substituted dyes and filled circles their thienyl analogs).
Thienyl-substituted merocyanines reveal essentially the same regularities in spectroscopic data as their
phenyl counterparts (cf. Fig. 4 with Fig. 5 and the corresponding entries in Table 1) and are positioned very
close to them on the hyperbolic curve as shown in Fig. 7.
To conclude, an empirically obtained and numerically simulated relationship between the first electronic
transition energy of merocyanines and the DA properties of their end substituents has been analytically
reproduced and rationalized in terms of the biradicaloid theory and the classical valence bond model. The
graphical implementation of the resulting combined approach, a "spectroscopic hyperbola", represents a
convenient tool to classify substituted DA compounds in relation to their electronic structure and to treat, on this
basis, the effect of DA strength on their optical behavior. As shown, there are two classes of merocyanines
which differ in charges and electron populations of their D and A moieties and exhibit mirror-like responses to
the variation of DA strength. The "hyperbolic" model applied to two families of newly synthesized heterocyclic
merocyanines has enabled elucidation of their spectroscopic properties in various media in terms of the "cyanine
limit" concept. The physical transparency and simplicity of the approach suggested make it relevant to tailor-
making of materials with controllable spectral characteristics such as near-infrared absorbing compounds,
efficient NLO dyes, electrooptical chromophors, polarity probes, etc.
1254

EXPERIMENTAL
1H NMR spectra were recorded in CDCl3 solutions on a Varian VXR-300 (300 MHz) spectrometer using
TMS as internal standard. Electronic absorption spectra were recorded on a Shimadzu UV-3100 spectrophotometer
in dichloromethane, acetonitrile, THF, and diethyl ether. TLC was carried out on Silufol UV-254 plates using a
benzene–chloroform, 2:1, mixture as eluent. Column chromatography was performed using aluminum oxide 90
standardized for column chromatographic adsorption analysis (purchased from Merck).
Literature procedures were used to prepare pyridinium [32], pyrylium [33], and thiopyrylium salts [34];
3,5-di(tert-butyl)-4-hydroxybenzaldehyde was purchased from Sigma-Aldrich; piperidine and pyridine were
purchased from Merck.
Synthesis of dyes 1 (NO-DP) and 4 (NO-DP) (General Method). A mixture of 1,4-dimethyl-
2,6-diphenyl- or 1,4-dimethyl-2,6-di(2-thienyl)pyridinium perchlorate (1.4 mmol), 3,5-di(tert-butyl)-4-hydroxy-
benzaldehyde (1.5 mmol), and piperidine (1.4 mmol) in pyridine (10 ml) was boiled for 2 h and evaporated in
vacuo. The residue was dissolved in dichloromethane and shaken in a separating funnel with aqueous ammonia.
The organic layer was separated, washed with water, and dried with anhydrous potassium carbonate. The dye
was chromatographed on aluminum oxide using dichloromethane as eluent and recrystallized from acetonitrile.
Synthesis of dyes 2 (OO-DP), 3 (SO-DP), 5 (OO-DT), and 6 (SO-DT) (General Method). A mixture
of 4-methyl-2,6-diphenyl- or 4-methyl-2,6-di(2-thienyl)pyrylium or -thiopyrylium perchlorate (1 mmol) and
о
3,5-di(tert-butyl)-4-hydroxybenzaldehyde (1.1 mmol) was fused at 185–190 С for 20 min. The cooled fusion
cake was dissolved in a benzene–chloroform, 1:1, mixture, followed by addition of triethylamine (1 ml) and
chromatography of the reaction mixture on aluminum oxide using a benzene–petroleum ether, 3:1, mixture as
eluent. Solvents were removed in vacuo and the residue was recrystallized from acetonitrile. To obtain
protonated dye forms, all merocyanines were treated with 20% HClO4 in acetonitrile at 20 oC.
(E)-2,6-Di-tert-butyl-4-[2-(1-methyl-2,6-diphenylpyridin-4(1H)-ylidene)ethylidene]cyclohexa-2,5-dien-
1-one (1, NO-DP). Yield 48%; mp 242–244°C. 1H NMR spectrum, δ, ppm (J, Hz): 1.36 (18H, s, 2C(CH3)3); 3.10
(3H, s, NCH3); 6.13 (1H, d, J = 14.1, =CH); 6.49 (1H, br. s, H Ar); 6.79 (1H, br. s, H Ar); 6.97 (1H, br. s, H Ar);
7.25 (1H, d, J = 14.1, =CH); 7.44–7.58 (11H, m, H Ar). Found, %: C 85.56; H 7.79; N 2.83. C34H37NO.
Calculated, %: C 85.85; H 7.84; N 2.94.
(E)-2,6-Di-tert-butyl-4-[2-(2,6-diphenyl-4H-pyran-4-ylidene)ethylidene]cyclohexa-2,5-dien-1-one (2,
1
OO-DP). Yield 45%; mp 238–240°C. H NMR spectrum, δ, ppm (J, Hz): 1.36 (9H, s, C(CH3)3); 1.39 (9H, s,
C(CH3)3); 6.38 (1H, d, J = 13.2, =CH); 6.72 (1H, d, J = 1.8, H Ar); 7.02 (1H, d, J = 1.8, H Ar); 7.06 (1H, d,
J = 1.8, H Ar); 7.22 (1H, d, J = 13.2, =CH); 7.47–7.53 (7H, m, H Ar); 7.80–7.85 (2H, m, H Ar), 7.87–7.91 (2H,
m, H Ar). Found, %: C 85.45; H 7.40. C33H34O2. Calculated, %: C 85.68; H 7.41.
(E)-2,6-Di-tert-butyl-4-[2-(2,6-diphenyl-4H-thiopyran-4-ylidene)ethylidene]cyclohexa-2,5-dien-1-one
(3, SO-DP). Yield 42%; mp 212–214°C. 1H NMR spectrum, δ, ppm (J, Hz): 1.34 (9H, s, C(CH3)3); 1.37 (9H, s,
C(CH3)3); 6.56 (1H, d, J = 13.5, =CH); 6.98 (1H, d, J = 1.8, H Ar); 7.01 (1H, s, H Ar); 7.23 (1H, d, J = 13.5,
=CH); 7.39 (1H, s, H Ar); 7.44–7.51 (7H, m, H Ar); 7.61–7.67 (4H, m, H Ar). Found, %: C 82.48; H 7.09; S 6.63.
C33H34OS. Calculated, %: C 82.80; H 7.16; S 6.70.
(E)-2,6-Di-tert-butyl-4-[2-(1-methyl-2,6-dithien-2-ylpyridin-4(1H)-ylidene)ethylidene]cyclohexa-
1
2,5-dien-1-one (4, NO-DT). Yield 31%; mp 230–231°C. H NMR spectrum, δ, ppm (J, Hz): 1.34 (9H, s,
C(CH3)3); 1.37 (9H, s, C(CH3)3); 3.30 (3H, s, NCH3); 6.10 (1H, d, J = 13.5, =CH); 6.52 (1H, s, H Ar); 6.88 (1H,
s, H Ar); 6.95 (1H, s, H Ar); 7.12–7.27 (5H, m, H Ar, =CH); 7.48–7.55 (3H, m, H Ar). Found, %: C 73.91;
H 6.71, N 2.75, S 12.98. C30H33NOS2. Calculated, %: C 73.88; H 6.82, N 2.87; S 13.15.
(E)-2,6-Di-tert-butyl-4-[2-(2,6-dithien-2-yl-4H-pyran-4-ylidene)ethylidene]cyclohexa-2,5-dien-1-one
(5, OO-DT). Yield 55%; mp 244–245°C. 1H NMR spectrum, δ, ppm (J, Hz): 1.35 (9H, s, C(CH3)3); 1.38 (9H, s,
C(CH3)3); 6.33 (1H, d, J = 13.5, =CH); 6.50 (1H, d, J = 1.5, H Ar); 6.86 (1H, d, J = 1.5, H Ar); 7.01 (1H, d,
J = 2.1, H Ar); 7.12–7.17 (3H, m, H Ar, =CH); 7.43–7.52 (4H, m, H Ar); 7.58–7.60 (1H, m, H Ar). Found, %:
C 73.10; H 6.29; S 3.23. C29H30O2S2. Calculated, %: C 73.38; H 6.37; S 13.51.
1255

(E)-2,6-Di-tert-butyl-4-[2-(2,6-dithien-2-yl-4H-thiopyran-4-ylidene)ethylidene]cyclohexa-2,5-dien-
1-one (6, SO-DT). Yield 39%; mp 241–242°C. 1H NMR spectrum, δ, ppm (J, Hz): 1.35 (9H, s, C(CH3)3); 1.38
(9H, s, C(CH3)3); 6.54 (1H, d, J = 13.5, =CH); 6.95 (1H, s, H Ar); 6.99 (1H, d, J = 2.0, H Ar); 7.07–7.13 (2H, m,
H Ar); (1H, d, J = 13.5, =CH); 7.33–7.45 (5H, m, H Ar); 7.49–7.51 (1H, m, H Ar). Found, %: C 70.66; H 6.10;
S 19.33. C29H30OS3. Calculated, %: C 70.98; H 6.16; S 19.60.
REFERENCES
1.
2.
W. König, J. Prakt. Chem., 112, 1 (1926).
L. G. S. Brooker, G. H. Keyes, R. H. Sprague, R. H. Van Dyke, E. Van Lare, G. Van Zandt, F. L. White,
H. W. J. Cressman, and S. G. Dent, Jr., J. Am. Chem. Soc., 73, 5332 (1951).
A. I. Kiprianov, Usp. Khim., 29, 1336 (1960) [Russ. Chem. Rev., 29, 618 (1960)].
A. Mishra, R. K. Behera, P. K. Behera, B. K. Mishra, and G. B. Behera, Chem. Rev., 100, 1973 (2000).
A. V. Kulinich and A. A. Ishchenko, Usp. Khim., 78, 151 (2009) [Russ. Chem. Rev. 78, 141 (2009)].
U. Lawrentz, W. Grahn, K. Lukaszuk, C. Klein, R. Wortmann, A. Feldner, and D. Scherer, Chem. Eur.
J., 8, 1573 (2002).
3.
4.
5.
6.
7.
8.
F. Würthner, R. Wortmann, and K. Meerholz, ChemPhysChem, 3, 17 (2002).
A. Abbotto, L. Beverina, S. Bradamante, A. Facchetti, C. Klein, G. A. Pagani, M. Redi-Abshiro, and
R. Wortmann, Chem. Eur. J., 9, 1991 (2003).
9.
10.
11.
12.
13.
F. Würthner, G. Archetti, R. Schmidt, and H.-G. Kuball, Angew. Chem., 120, 4605 (2008).
J. R. Platt, J. Chem. Phys., 25, 80 (1956).
S. Dähne and D. Leupold, Angew. Chem., 78, 1029 (1966). [Angew. Chem. Int. Ed., 5, 984 (1966)].
S. Dähne and F. Moldenhauer, Prog. Phys. Org. Chem., 15, 1 (1985).
G. G. Dyadyusha and A. D. Kachkovskii, Ukr. Khim. Zh., 41, 1176 (1975). [Sov. Prog. Chem., 41, 52
(1975)].
14.
G. G. Dyadyusha and A. D. Kachkovskii, Ukr. Khim. Zh., 44, 948 (1978). [Sov. Prog. Chem., 44, 50
(1978)].
15.
16.
17.
W. Rettig and V. Kharlanov, J. Phys. Chem. A, 113, 10693 (2009).
C. Reichardt, Chem. Rev., 94, 2319 (1994).
J. Michl and V. Bonacic-Koutecký, Electronic Aspects of Organic Photochemistry, Wiley, New York,
1990.
18.
19.
V. Bonacic-Koutecký, J. Koutecký, and J. Michl, Angew. Chem., Int. Ed., 26, 170 (1987).
G. Bourhill, J.-L. Brédas, L.-T. Cheng, S. R. Marder, F. Meyers, and J. W. Perry, B. G. Tiemann, J. Am.
Chem. Soc., 116, 2619 (1994).
20.
21.
22.
23.
24.
25.
M. Dekhtyar and W. Rettig, Phys. Chem. Chem. Phys., 3, 1602 (2001).
W. Rettig and M. Dekhtyar, Chem. Phys., 293, 75 (2003).
W. Rettig, Angew. Chem. Int. Ed., 25, 971 (1986).
Z. R. Grabowski, K. Rotkiewicz, and W. Rettig, Chem. Rev., 103, 3899 (2003).
M. Maus, W. Rettig, D. Bonafoux, and R. Lapouyade, J. Phys. Chem., A 103, 3388 (1999).
L. G. S. Brooker, G. H. Keyes, R. H. Sprague, R. H. Van Dyke, E. Van Lare, G. Van Zandt, and
F. L. White, J. Am. Chem. Soc., 74, 5350 (1951).
26.
27.
Th. Förster, Z. Elektrochem. Angew. Phys. Chem., 45, 548 (1939).
A. D. Kachkovski, M. A. Kudinova, B. I. Shapiro, N. A. Derevjanko, L. G. Kurkina, and A. I. Tolmachev,
Dyes Pigm., 5, 295 (1984).
28.
A. I. Tolmachev and M. A. Kudinova, Khim. Geterotsikl. Soed., 49 (1974). [Chem. Heterocycl. Comp.,
10, 41 (1974)].
1256

29.
30.
A. I. Tolmachev, M. A. Kudinova, and N. A. Derevyanko, Khim. Geterotsikl. Soed., 53 (1974). [Chem.
Heterocycl. Comp., 10, 45 (1974)].
A. I. Tolmachev, N. A. Derevyanko, E. F. Karaban, and M. A. Kudinova, Khim. Geterotsikl. Soed., 612
(1975). [Chem. Heterocycl. Comp., 11, 534 (1975)].
31.
32.
P. Fromherz, J. Phys. Chem., 99, 7188 (1995).
Advances in Heterocyclic Chemistry, in: A. R. Katritzky, and A. J. Boulton (editors), Acad. Press, New
York, 1982, Suppl. 2, vol. XII, p. 432.
33.
34.
G. N. Dorofeenko, and S. M. Lukyanov, Zh. Org. Khim., 7, 419 (1971).
V. G. Kharchenko, V. I. Kleimenova, and A. R.Yakoreva, Khim. Geterotsikl. Soed., 900 (1970). [Chem.
Heterocycl. Comp., 6, 834 (1970)].
1257