tions that may lead to a change in the polarisation of the
molecules. The anomalous behaviour observed in Fig. 3 can be
satisfactorily explained with the formation of intermolecular
H-bonds between dyes and solvents. The hypsochromic shift of
l
A in the row of increasing e(s) of solvents: ethanol, DMF, and
acetonitrile, is surely connected with the different ability of
these solvents to H-bond with the dyes. The proton-donating
ethanol forms H-bonds with the carbonyl groups of the dyes
thus enhancing the electron donor±acceptor interaction in the
chromophoric system. In the case of DMF and acetonitrile, H-
bonds can be formed between the H-atom of the allylamino
group with the nonshared electron pairs of the hetero atoms of
the solvents. The H-bonds decrease the donating ability of the
amino group, and consequently, decrease the electron donor±
acceptor interaction which leads to the hypsochromic shift of
lA. This also explains the close positions of lA in acetone and
chloroform despite the largely different e values of those
solvents.
The Lippert's function, Df, describes the dielectric behaviour
of solvents both by its relative permittivity e(s) and by its
refractive index.22 It was however again not appropriate for the
description of the solvatochromic behaviour of the 1,8-
naphthalimides under study. The plot of the absorption
maxima, in wavenumbers, vs. Df shows only a poor linear
correlation (correlation coef®cient 20.79), decreasing with
increasing polarity Df. The failure has been explained in terms
of the possible H-bonding discussed above, since the under-
lying Lippert's theory22 considers dipole±dipole, dipole±
polarisation, polarisation±inductive, and dispersion forces
but not charge transfer, H-bonding, and related interactions.
In Fig. 4 is shown the dependence of lA of 3a±e on ET(30),
the empirical solvent polarity parameter.23 ET(30) values are
based on the negatively solvatochromic N-phenolate betaine
dye as probe. While e(s) can be related to the solvatochromic
properties of the dyes (Fig. 3) only when no interactions occur,
the parameter ET(30) evidently accounts also for the solvating
effects discussed. The linear course of the lA vs. ET(30)
dependence (Fig. 4) implies positive solvatochromism of the
dyes under study. The analytical form of the lA dependence
upon solvent polarity for that class of substituted dyes is given
by eqn. (2).
Fig. 4 Dependence of the visible absorption maxima of dyes 3a±e,
measured in organic solvents, upon the empirical solvent polarity
parameter ET(30): D~1,4-dioxane, T~toluene, C~chloroform,
A~acetone, E~ethanol, DMF~dimethylformamide and AN~ace-
tonitrile. The symbols for the dyes are: 3a w, 3b &, 3c ©, 3d à and 3e
z.
2. The f values of dyes 3a±e vary from 0.210 to 0.260 in ethanol,
being in accord with the change in e. The slightly larger f values
for dyes 3d and 3e as compared to those for 3a±3c are related to
some increase in their polarisation due to the electron accepting
substituent R1 (dye 3d) and to H-bonding (dye 3e). All the data
found for dyes 3 show that the substituents at the phenyl
moiety have a small effect upon the photophysical character-
istics lA, lF, the Stokes shift, ES1, and f, which is indicative of
their similar polarisation ability.
The substituents R1 and R2 have but a distinct effect upon
WF of the dyes, probably related to the possibility for deviation
from the plane of the naphthalimide structure. All dyes 3
having a secondary amino group exhibit, in accord with our
previous studies,20,21 a very good quantum yield of ¯uorescence
WF (in ethanol), with the exception of dye 3e. The lower value
of WF for dye 3e can alternatively be related to the speci®c
interaction of the R2 (OH group) with ethanol which may ®x
any deviation of the attached phenyl ring out of plane of the
naphthalimide structure.
The large effect that different media have upon the UV±vis
absorption of dyes 3 is illustrated in Figs. 3 and 4, which
display the absorption maxima of the dyes in media of
increasing polarity. The polarity was characterised both by the
relative permittivity, e(s), Lippert's solvent polarity function,
Df,22 and the empirical solvent polarity parameter ET(30).23
On the histogram Fig. 3 are shown the absorption maxima
lA~371:2z1:32ET(30)
(2)
It is derived by the least squares method, with a correlation
coef®cient of R~0.91 and
SD~3.56 nm, N~35.
a
standard deviation of
In LC, all dyes 3 retain their yellow-green colour, with
absorption maxima at 430±432 nm and more intense ¯uores-
cence (Table 2). The absorption maxima lA of the dyes are
hypsochromically shifted with regard to lA in the polar
solvents ethanol and DMF and bathochromically shifted with
regard to the less polar solvents toluene and chloroform. By
means of eqn. (2), one can ®nd an ET(30) value of 45 kcal
mol21 for LC ZLI 1840. The ¯uorescence spectra of the dyes
recorded in surface-stabilised LC cells show that the ¯uores-
cence maxima lF are hypsochromically shifted with regard to
lF in ethanol and the energy of the ®rst excited state ES1 is
accordingly higher. The lower Stokes shift in LC with regard to
ethanol can be connected with the induced disposition of
substituents in the plane of the aromatic structure (by the
surface stabilised LC texture). It should be noted that
intermolecular H-bonding is also possible in LC ZLI 1840
because of the presence of CMN and CLO groups.
lA of dye 3d in solvents of increasing relative permittivity as an
example. There is no unique dependence of lA upon e(s). A
trend to bathochromic shift of lA with increasing e(s) is
observed in the series of the less polar solvents: toluene,
chloroform and acetone, with the exception of 1,4-dioxane.
The absorption maxima in 1,4-dioxane are bathochromically
shifted with regard to toluene, a solvent of almost the same e(s),
being close to those of the more polar chloroform and acetone.
The anomaly is revealed still more clearly in the series of the
more polar solvents: ethanol, DMF, and acetonitrile, where the
trend is for decreasing lA with increasing e(s).
The effect of the medium is due not only to the polarity of
solvents but also to the possibility of intermolecular interac-
Table 3 Molecular aspect ratio l/d, dichroic ratio N(l) (at lmax in LC),
and the orientational parameter SA of 4-allylamine substituted 1,8-
naphthalimide dyes 3a±e
Dyes
3a
3b
3c
3d
3e
Ordering in dye±liquid crystal systems
l/d
N(l)
SA
3.30
6.00
0.63
3.62
6.69
0.65
3.88
7.20
0.67
3.62
4.64
0.55
3.30
3.44
0.45
It has been found that substances added to a nematic LC affect
its orientation order. Particularly convenient for measuring
orientation parameters in LC systems involving dichroic dyes is
1294
J. Mater. Chem., 2000, 10, 1291±1296
Grabchev, Ivo
