Solute-solvent and solvent-solvent interactions in the preferential solvation of 4-[4-(dimethylamino)styryl]-1-methylpyridinium iodide in 24 binary solvent mixtures

Article abstract of DOI:10.1016/j.saa.2005.12.005

The molar transition energy (E_T) polarity values for the dye 4-[4-(dimethylamino)styryl]-1-methylpyridinium iodide were collected in binary mixtures comprising a hydrogen-bond accepting (HBA) solvent (acetone, acetonitrile, dimethyl sulfoxide (DMSO), and N,N-dimethylformamide (DMF)) and a hydrogen-bond donating (HBD) solvent (water, methanol, ethanol, propan-2-ol, and butan-1-ol). Data referring to mixtures of water with alcohols were also analyzed. These data were used in the study of the preferential solvation of the probe, in terms of both solute-solvent and solvent-solvent interactions. These latter interactions are of importance in explaining the synergistic behavior observed for many mixed solvent systems. All data were successfully fitted to a model based on solvent-exchange equilibria. The E_T values of the dye dissolved in the solvents show that the position of the solvatochromic absorption band of the dye is dependent on the medium polarity. The solvation of the dye in HBA solvents occurs with a very important contribution from ion-dipole interactions. In HBD solvents, the hydrogen bonding between the dimethylamino group in the dye and the OH group in the solvent plays an important role in the solvation of the dye. The interaction of the hydroxylic solvent with the other component in the mixture can lead to the formation of hydrogen-bonded complexes, which solvate the dye using a lower polar moiety, i.e. alkyl groups in the solvents. The dye has a hydrophobic nature and a dimethylamino group with a minor capability for hydrogen bonding with the medium in comparison with the phenolate group present in Reichardt's pyridiniophenolate. Thus, the probe is able to detect solvent-solvent interactions, which are implicit to the observed synergistic behavior.

Full text of DOI:10.1016/j.saa.2005.12.005

Spectrochimica Acta Part A 65 (2006) 535–542
Solute–solvent and solvent–solvent interactions in the preferential
solvation of 4-[4-(dimethylamino)styryl]-1-methylpyridinium
iodide in 24 binary solvent mixtures
Tharly Bevilaquaa, Thaini F. Gonc¸alvesa, Cristina de G. Venturinib, Vanderlei G. Machadob,∗
a Centro Tecnolo´gico, Universidade do Sul de Santa Catarina, CP 370, Tubara˜o, SC 88704-900, Brazil
b Departamento de Qu´ımica, Universidade Regional de Blumenau, FURB, CP 1507, Blumenau, SC 89010-971, Brazil
Received 26 September 2005; received in revised form 2 December 2005; accepted 4 December 2005
Abstract
The molar transition energy (ET) polarity values for the dye 4-[4-(dimethylamino)styryl]-1-methylpyridinium iodide were collected in binary
mixtures comprising a hydrogen-bond accepting (HBA) solvent (acetone, acetonitrile, dimethyl sulfoxide (DMSO), and N,N-dimethylformamide
(DMF)) and a hydrogen-bond donating (HBD) solvent (water, methanol, ethanol, propan-2-ol, and butan-1-ol). Data referring to mixtures of water
with alcohols were also analyzed. These data were used in the study of the preferential solvation of the probe, in terms of both solute–solvent and
solvent–solvent interactions. These latter interactions are of importance in explaining the synergistic behavior observed for many mixed solvent
systems. All data were successfully fitted to a model based on solvent-exchange equilibria. The ET values of the dye dissolved in the solvents show
that the position of the solvatochromic absorption band of the dye is dependent on the medium polarity. The solvation of the dye in HBA solvents
occurs with a very important contribution from ion–dipole interactions. In HBD solvents, the hydrogen bonding between the dimethylamino group
in the dye and the OH group in the solvent plays an important role in the solvation of the dye. The interaction of the hydroxylic solvent with the
other component in the mixture can lead to the formation of hydrogen-bonded complexes, which solvate the dye using a lower polar moiety, i.e.
alkyl groups in the solvents. The dye has a hydrophobic nature and a dimethylamino group with a minor capability for hydrogen bonding with
the medium in comparison with the phenolate group present in Reichardt’s pyridiniophenolate. Thus, the probe is able to detect solvent–solvent
interactions, which are implicit to the observed synergistic behavior.
© 2006 Elsevier B.V. All rights reserved.
Keywords: Preferential solvation; Binary mixtures; Merocyanines; Solvent effects; Solvent–solvent interactions; Solute–solvent interactions
1. Introduction
triphenylpyridinium-1-yl)phenolate (1), which is one of the
most commonly employed solvatochromic dyes [1,2,4].
The rate and course of chemical processes can be influenced
by the medium in which they are carried out [1,2]. These
effects of the medium are commonly interpreted as resulting
from changes in its polarity, a general term that is related to the
overall solvating capability of the solvent [3]. The polarity of
the medium provokes changes in the absorption and emission
spectra of many organic dyes [1–4]. Merocyanines comprise a
class of solvatochromic compounds that are commonly used
in the investigation of solvent properties [3,4]. An important
example is Reichardt’s betaine dye, 2,6-diphenyl-4-(2,4,6-
In studies performed in order to better understand these sol-
vent effects, considerable research has been carried out on the
physical organic chemistry of mixed solvents [4–27], due to their
∗
Corresponding author. Tel.: +55 47 221 6081; fax: +55 47 221 6001.
E-mail address: gageiro@furb.br (V.G. Machado).
1386-1425/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.saa.2005.12.005

536
T. Bevilaqua et al. / Spectrochimica Acta Part A 65 (2006) 535–542
important applications in many fields, such as chromatography,
hydrometallurgy, organic synthesis, and the kinetics of chemical
processes [5]. The analysis of the behavior of solutes in mixed
solvent systems is more complex than in pure solvents due to
thepossibilityforpreferentialsolvation(PS), whichoccurswhen
the solute has in its microenvironment more of one solvent than
the other, in comparison to the bulk composition. A compre-
hension of phenomena related to PS may help in the elucida-
tion of kinetic, spectroscopic, and equilibrium data for mixed
solvents.
An important aspect in the studies performed on mixed sol-
vents is the role of solvent–solvent interactions in competition
to the solvation of solutes [5,13,25–27]. In these interactions,
one component of the binary mixture preferentially binds to a
molecule of the same type in its neighborhood, forming clus-
ters. Although the solute cannot be proved to be responsible for
the clustering, it can be solvated in these solvent aggregates.
In these molecular associations, hydrogen bonding can play an
important role and has been widely investigated because it is
present in a large variety of chemical, biochemical, and phar-
macological events.
Althoughtheliteratureshowsmanystudiesonmixedsolvents
using 1 as a dye [4,6–12,17–20], relatively few investigations
of these solvent systems have been carried out using other
classes of solvatochromic probes [13–15,21–24]. It is consid-
ered that structurally different probes should be able to detect
different aspects of the physical chemistry of mixed solvents.
Thus, in this paper we describe an investigation into the behav-
ior of 4-[4-(dimethylamino)styryl]-1-methylpyridinium iodide
(2) in binary mixtures comprising a hydrogen-bond accept-
ing (HBA) solvent [acetone, acetonitrile, dimethyl sulfoxide
(DMSO), and N,N-dimethylformamide (DMF)] and a hydrogen-
bonddonating(HBD)solvent(water, methanol, ethanol, propan-
2-ol, and butan-1-ol) solvent. We also show data relating
to mixed systems comprising water with alcohols (methanol,
ethanol, propan-1-ol, and propan-2-ol). The obtained data are
discussed in relation to solute–solvent and solvent–solvent inter-
actions.
J = 16.1 Hz), 7.60 (2H, d, J = 8.8 Hz), 7.92 (1H, d, J = 16.1 Hz),
8.05 (2H, d, J = 6.8 Hz), 8.69 (2H, d, J = 6.7 Hz).
2.2. UV–vis measurements
The following procedure was typical for all experiments per-
formed. A 1.0 × 10−3 mol L−1 stock solution of 2 was prepared
in dichloromethane. From this stock solution 100 ␮L aliquots
were transferred to 10 mL volumetric flasks. After the evapo-
ration of the dichloromethane, the probe was solubilized in the
pure solvents, resulting in solutions with final concentrations
of 1.0 × 10−5 mol L−1. The UV–vis experiments were recorded
on a Cary 1G spectrophotometer at 25 ◦C. The maxima on the
UV–vis spectra were calculated from the first derivative of the
absorption spectrum. The λmax values thus obtained in nanome-
ters were transformed into ET values (kcal mol−1), according
to the expression ET(dye) = 28,590/λmax. Binary mixtures were
prepared by weighing the solutions prepared as described above
in a quartz cuvette closed with a rubber septum (the cosolvent
solution was added with a 1 mL microsyringe) and the final val-
ues are expressed in terms of the hydroxylic cosolvent mole
fraction (X2) in the mixtures. The λmax values thus obtained were
transformed into ET(dye) values (see above). All experiments
with the mixed solvent systems were carried out in duplicate
and the reproducibility of the curve features was confirmed.
2.3. Calculations
The parameters ET(dye)1, ET(dye)2, ET(dye)12, f2/1, and f12/1
were calculated from non-linear regressions using the ORIGIN
5.0 program.
3. Results and discussion
Table 1 summarizes the polarity and hydrogen-bonding prop-
erties of the pure solvents used in this study, in the form of
the Reichardt normalized ENT values [1] as well as α (solvent
hydrogen-bond donor acidity), β (solvent hydrogen-bond accep-
tor basicity), and π* (dipolarity/polarizability) Kamlet–Taft
parameters [2,30]. On analysis of these parameters, the solvents
can be separated into two classes. The first class is comprised
2. Experimental
2.1. Materials
Table 1
Polarity parameters of pure solvents at 25 ◦C
All solvents were HPLC grade and were purified following
methodology described in the literature [28]. Deionized water
was used in all measurements. This solvent was boiled and bub-
bled with nitrogen and kept under a nitrogen atmosphere to avoid
the presence of carbon dioxide. Dye 2 was synthesized, follow-
ing a reported procedure [29], refluxing 1-methyl-4-picolinium
iodide with 4-(dimethylamino)benzaldehyde and piperidine in
ethanol for 4 h in a water bath (80–90 ◦C). The reaction mix-
ture was left to stand overnight, the red-purple crystals were
filtered off and washed with ethanol. The crystals were washed
with cold diethyl ether and were recrystallized twice from hot
water and dried under vacuum. Yield: 45%; mp 254.5–255.5 ◦C
(literature [29], 251–255 ◦C). 1H NMR (200 MHz, DMSO-d6):
3.02 (6H, s), 4.18 (3H, s), 6.79 (2H, d, J = 8.8 Hz), 7.17 (1H, d,
Solvent
π*a
αa
βa
ENT b
ET(dye)c
Water
Methanol
Ethanol
Propan-1-ol
Butan-1-ol
Propan-2-ol
DMSO
DMF
Acetonitrile
Acetone
1.09
0.60
0.54
0.52
0.47
0.48
1.00
0.88
0.75
0.71
1.17
0.93
0.83
0.78
0.79
0.76
0.00
0.00
0.19
0.08
0.47
0.62
0.77
0.90
0.88
0.95
0.76
0.69
0.31
0.48
1.000
0.762
0.654
0.617
0.586
0.546
0.444
0.386
0.460
0.355
63.8
60.1
59.7
59.0
58.8
59.4
60.7
60.9
60.8
60.7
a
References [2,30].
Reference [4].
This study, in kcal mol−1
b
c
.

T. Bevilaqua et al. / Spectrochimica Acta Part A 65 (2006) 535–542
537
maintained without alteration. From these data the possibility of
this dye being an ion-pair solvatochromic compound, such as the
Kosower’s 1-ethyl-4-(methoxycarbonyl)pyridinium iodide [31],
can be discarded. In this respect, Al-Ansari [29] has demon-
strated that the solvatochromic band of 2 disappears in acidified
methanol, due to the protonation of the nitrogen lone-pair in the
dye. This experiment helped to reveal the charge-transfer nature
of the solvatochromic band of 2. Another aspect to be consid-
ered regarding the solvatochromism of dye 2 is that, in a salt, the
cation and the anion can be preferentially solvated by the com-
ponents of the mixed system. In the solutions of 2, only the PS of
the cation of 2 can be investigated by means of solvatochromic
measurements.
The elements discussed above suggest that different
solute–solvent interactions are responsible for the solvation
of the dye in HBD and HBA solvents. The solvation of 2
in HBA solvents occurs with a very important contribution
from ion–dipole interactions. These interactions are responsi-
ble for the stabilization of the electronic ground state of the
dye and, consequently, for the increase in its charge-transfer
transition energy in solution with an increase in the polarity of
the medium. In HBD media, simultaneously with ion–dipole
interactions, hydrogen bonding between the dimethylamino
hydrogen-bonding receptor group in the dye and the solvent can
occur [29,32]. It has been reported that the pyridiniophenolate
1, the basis for the ENT polarity scale, is able to interact through
hydrogen bonding with proton donors by means of its pheno-
late group [33,34]. The formation of hydrogen bonds leads to
a stabilization of the ground state of the dye, and a consequent
increase in its charge-transfer transition energy. Thus, the greater
the capability of the solvent to form hydrogen bonds with the
dye the greater the increase in the ET(dye) values.
Fig. 1. UV–vis spectra of 2 in water (a), acetonitrile (b), DMSO (c), propan-2-ol
(d), and propan-1-ol (e) at 25 ◦C.
by the HBD solvents, which show strong hydrogen-bond donor
capabilities. The second class comprises the HBA solvents, for
which high β values (except for acetonitrile) and low α values
were obtained, indicating that these solvents are good hydrogen-
bond acceptors and poor hydrogen-bond donors.
The UV–vis absorption spectra of dye 2 were recorded in
several HBD and HBA solvents. Fig. 1 shows the spectra of dye
2 in water, propan-1-ol, propan-2-ol, acetonitrile, and DMSO.
Spectra were collected for the solvents with different dye con-
centrations and no changes in the shape and λmax position were
observed for the solvatochromic band, indicating that there is no
formation of dye aggregates. The maxima in absorbance for all
spectra were obtained with a precision of 0.5 nm. The repro-
ducibility of the λmax values was verified by their determination
from five spectra of the dye in each pure solvent. All data were
transformed to molar transition energies for 2 (ET(dye) values),
given in kcal mol−1 with a precision of 0.1 kcal mol−1 and are
shown in Table 1. The data show that the position of the solva-
tochromic band of the dye is dependent on the medium polarity,
having a variation of ꢀλ = 38 nm for the solvents used.
Compound 2 is a salt, which can form, in principle, ion pairs
in solution, especially in solvents with a low dielectric constant,
such as acetone. The formation of ion pairs can influence the
absorption spectra of 2 in these solvents. In order to verify this,
UV–vis absorption spectra of the dye in acetone were collected
in the presence of increasing concentrations of tetrabutylammo-
nium iodide. It was verified that the addition of iodide up to
0.01 mol dm−3 did not change the position of the maximum on
the UV–vis spectra of 2. These data suggest that the dispersion of
the positive charge on the nitrogen atom of the dye may prevent
ion pairing. In other experiment, silver tetrafluoroborate was
added to the solution of 2 in acetone, in order to give the precip-
itation of iodide as AgI. It was found that the UV–vis spectrum
of 2 is exactly the same after the addition of AgBF4. Similar
experiments were performed in acetonitrile, DMSO, propan-
1-ol, and water. The results showed that the solvatochromism
observed is not influenced by the presence of iodide as a coun-
teranion. In all cases, the solvatochromic absorption band was
3.1. Preferential solvation in mixed systems
Solutions of dye 2 in mixed solvents comprising aqueous
alcohols and a HBA and a HBD component were studied in
order to investigate the PS of the dye. All results were treated by
considering the two-step solvent-exchange model shown below,
taken from a study by Skwierczynski and Connors [17]:
dye(S1)2 + 2S2 ꢀ dye(S2)2 + 2S1
dye(S1)2 + S2 ꢀ dye(S12)2 + S1
This model has been successfully employed to explain the
solvation of the pyridiniophenolate 1 in many binary mixtures
[18–20]. It has also been used to describe the solvation of pyrene
[22] and Brooker’s merocyanine [23,24] in mixtures compris-
ing HBA and HBD solvents. In this model, S1 and S2 represent
the two solvents in the mixture, with the subscript 2 relating
to the most polar component. The interaction of these two sol-
vents through hydrogen bonding yields a common structure S12.
The dye solvated by S1, S2, and S12 is represented by dye(S1)2,
dye(S2)2, and dye(S12)2, respectively. The ET(dye) value for a
given mixture is considered equal to the average of the ET(dye)
values of solvents S1, S2, and S12 in the solvation shell of 2. The
mathematicaldevelopmentofthismodelresultsinEq. (1), which

538
T. Bevilaqua et al. / Spectrochimica Acta Part A 65 (2006) 535–542
Fig. 3. Influence of the hydroxylic cosolvent on the ET values for 2 in
DMSO–ROH mixtures, the ROH component being water (a), methanol (b),
ethanol (c), butan-1-ol (d), or propan-2-ol (e). (· · ·) Theoretical linear depen-
dence for no PS; (—) curve fitted with Eq. (1); (ꢁ) experimental data.
Fig. 2. Influence of water on the ET values for 2 in aqueous mixtures of alcohols,
the alcohol being methanol (a), ethanol (b), propan-1-ol (c), or propan-2-ol (d).
(· · ·) Theoretical linear dependence for no PS; (—) curve fitted with Eq. (1); (ꢁ)
experimental data.
relates the ET(dye) value in a binary mixture to the ET(dye) val-
ues of the two pure solvents [18–20,22–24].
ET(dye)1(1 − X2)2 + ET(dye)2f2/1X22
+ ET(dye)12f12/1(1 − X2)X2
ET(dye) =
(1)
(1 − X2)2 + f2/1X22 + f12/1(1 − X2)X2
The PS parameters f2/1 and f12/1 measure the tendency of 2 to be
solvated by the solvents S2 and S12 with respect to solvent S1.
The ET(dye) values as a function of X2 (Figs. 2–6) were plot-
ted and the experimental data for each binary solvent mixture
were fitted through a non-linear regression to Eq. (1). The results
are displayed in Table 2 and show very good fits for all binary
solvent mixtures studied (S.D. < 5 × 10−2). Table 2 shows also
the PS parameter f12/2, which is a measure of the tendency of 2 to
be preferentially solvated by the solvent S12 in comparison with
S2. It was calculated for all mixtures studied using the expres-
sion f12/2 = f12/1/f2/1. In addition, the calculated ET(dye) values
in Table 2 for the pure solvents agree well with the experimental
values given in Table 1.
3.2. Alcohol–water mixtures
Fig. 2showsasetofplotsofET valuesfor2asafunctionofthe
water mole fraction (X2) for alcohol–water mixtures. The exper-
imental data show the occurrence of strong PS by the alcoholic
component in these mixtures, similarly to the results obtained
by Ortega et al. [19] using dye 1 as an indicator. The plots in
Fig. 2 also show that, with the lengthening of the alkyl chain of
Fig. 4. Influence of the hydroxylic cosolvent on the ET values for 2 in
acetone–ROH mixtures, the ROH component being water (a), methanol (b),
ethanol (c), butan-1-ol (d), or propan-2-ol (e). (· · ·) Theoretical linear depen-
dence for no PS; (—) curve fitted with Eq. (1); (ꢁ) experimental data.

T. Bevilaqua et al. / Spectrochimica Acta Part A 65 (2006) 535–542
539
Table 2
Calculated parameters for dye 2 in binary solvent mixtures at 25 ◦C
a
Binary mixture
ET(dye)1
ET(dye)2
ET(dye)12
f2/1
f12/1
0.16
f12/2
S.D.
Water–methanol
Water–ethanol
60.1
59.7
59.4
58.9
60.7
60.7
60.7
60.8
60.7
60.5
60.7
60.7
60.7
60.7
60.8
60.8
60.8
60.8
60.8
61.0
61.0
60.9
60.9
60.9
63.8
63.9
63.9
63.9
63.6
60.1
59.7
59.4
58.9
63.8
60.1
59.7
59.5
58.8
63.9
60.1
59.7
59.0
58.8
63.8
60.1
59.7
59.5
58.9
62.2
60.0
60.8
60.6
60.1
60.9
61.0
61.1
61.1
61.3
60.8
59.8
60.1
60.0
61.1
60.4
60.8
60.0
59.4
60.5
61.0
61.0
60.7
60.9
0.027
0.63
0.19
0.089
0.025
0.81
1.7
0.87
0.54
0.12
0.040
0.0015
0.35
1.4
0.57
0.0097
1.5
0.46
0.098
0.014
0.077
4.9
5.9
2.0 × 10−3
1.6 × 10−2
4.6 × 10−2
3.5 × 10−2
1.3 × 10−3
1.5 × 10−4
6.5 × 10−4
7.2 × 10−4
1.2 × 10−3
1.3 × 10−3
3.5 × 10−4
2.4 × 10−4
3.2 × 10−4
1.5 × 10−3
6.8 × 10−3
5.0 × 10−5
9.0 × 10−5
3.0 × 10−4
6.8 × 10−4
1.9 × 10−3
1.9 × 10−4
7.8 × 10−4
3.1 × 10−4
1.6 × 10−3
8.9
4.9
3.7
0.44
5.3
6.9
2.5
1.4
4.1
0.24
0.27
3.2
5.1
13
0.26
4.0
3.0
1.6
0.67
0.37
23
0.75
2.1
14
Water–propan-2-ol
Water–propan-1-ol
DMSO–water
DMSO–methanol
DMSO–ethanol
DMSO–propan-2-ol
DMSO–butan-1-ol
Acetone–water
Acetone–methanol
Acetone–ethanol
Acetone–propan-2-ol
Acetone–butan-1-ol
Acetonitrile–water
Acetonitrile–methanol
Acetonitrile–ethanol
Acetonitrile–propan-2-ol
Acetonitrile–butan-1-ol
DMF–water
26
42
18
6.5
4.1
2.9
2.6
34
6.0
180
9.1
3.6
23
27
2.7
6.5
16
48
4.8
4.7
6.8
2.7
DMF–methanol
DMF–ethanol
DMF–propan-2-ol
DMF–butan-1-ol
0.11
0.79
a
f12/2 = f12/1/f2/1
.
the alcohol, the extent of the PS of the dye cation increases in
the water-rich region and, simultaneously, PS by the alcoholic
component in the alcohol-rich region changes towards an ideal
behavior. These results show that the dye cation is preferen-
tially solvated by the alcoholic component in all mole fractions
in aqueous mixtures with methanol and ethanol. It has been
previously discussed that water makes strong hydrogen-bonded
nets in the water-rich region, which are not easily disrupted
Fig. 5. Influence of the hydroxylic cosolvent on the ET values for 2 in
acetonitrile–ROH mixtures, the ROH component being water (a), methanol (b),
ethanol (c), butan-1-ol (d), or propan-2-ol (e). (· · ·) Theoretical linear depen-
dence for no PS; (—) curve fitted with Eq. (1); (ꢁ) experimental data.
Fig. 6. Influence of the hydroxylic cosolvent on the ET values for 2 in
DMF–ROHmixtures, theROHcomponentbeingwater(a), methanol(b), ethanol
(c), butan-1-ol (d), or propan-2-ol (e). (· · ·) Theoretical linear dependence for
no PS; (—) curve fitted with Eq. (1); (ꢁ) experimental data.

540
T. Bevilaqua et al. / Spectrochimica Acta Part A 65 (2006) 535–542
The term ‘synergism’ was first proposed by Reichardt and
coworkers in a study involving solutions of the pyridiniopheno-
late 1 in binary mixtures of dialkyl ketones, trialkyl phosphates,
or DMSO with chloroform [7]. Synergistic effects are due to the
formation of 1:1 complexes through hydrogen bonding between
the HBD and the non-HBD components of the mixture, yielding
theS12 species[7,18–20,22–24]. Thesespeciescanbedifferenti-
ated by structurally different probes: compound 1 and Brooker’s
merocyanine, for instance, recognize S12 species as more polar
becausetheirmorepolarmoietysolvatesthedyes[18–20,23,24].
On theotherhand, pyrenerecognizestheS12 speciesasless polar
because this non-polar solute is solvated by the less polar groups
in these species [22]. It was noted by Marcus that highly polar,
mutually miscible solvents, such as DMF or DMSO, interact
so strongly with water that the mole fraction of the hydrogen-
bonded DMSO–water associate is greater than the free water
mole fraction and the self-associated water mole fraction [35].
Since DMSO shows a high β value, it can form strongly associ-
ated complexes through hydrogen bonding with the hydroxylic
solvents and these S12 species may be responsible for the sol-
vation of the dye. Thus, the microenvironment of 2 in these
mixtures is crowded by these DMSO–ROH partners, which ori-
ent their hydrophobic methyl groups towards the probe. This can
be the origin of the hydrophobic microenvironment reported for
2. The behavior of this dye in these mixtures is similar to that
described for pyrene, a hydrophobic probe, which is also pref-
erentially solvated by S12 species in a similar manner [22]. The
values of f12/2 in Table 2 show that 2 is preferentially solvated
by S12 in the following order for the HBD solvents: butan-1-
ol < propan-2-ol < ethanol < methanol < water. This is the same
increasing order of solvent hydrogen bonding donation (see α
values in Table 1), which shows that the change of butan-1-ol
to water strengthens the interaction with DMSO by hydrogen
bonding and these S12 complexes formed are mainly responsi-
ble for the solvation of 2.
Scheme 1.
by the cosolvent [23,24]. This can explain the strong PS by
the alcoholic component in this region since water preferen-
tially interacts with itself rather than with the dye. Thus, it is
expected that the interaction of the alcoholic component with the
dye through both hydrogen bonding and a hydrophobic effect
increases with the lengthening of the alkyl chain of the alco-
hol. The ability of the alcoholic cosolvent to solvate the dye
cation in the water-rich region increases in the following order:
methanol < ethanol < propan-2-ol ≈ propan-1-ol. This is also the
increasing order of the β values, suggesting that the increase in
the basicity of the alcohol makes it able to compete efficiently
with the dye for the water, hindering the interaction between the
water and the dye.
In the alcohol-rich region, the alcohol molecules are freer to
interact with the water and with the dye, since their nets formed
by hydrogen bonds are weaker than in water. In this situation, the
alcohol molecules can, to a greater or lesser extent, interact with
water through hydrogen bonding. The plots in Fig. 2 show that
the extent of PS by the alcoholic component in the alcohol-rich
region decreases in the order methanol > ethanol > propan-1-
ol > propan-2-ol. Thisisalsothedecreasingorderofthesolvent’s
hydrogen-bond donation capability (see α values in Table 1),
suggesting that the increase in the facility of the alcohol for
hydrogen-bond donation makes its interaction with the dimethy-
laminogroupinthedyecationstrongerand, consequently, thePS
by the alcohol in the alcohol-rich region increases. This order is
also the increasing order of the solvent’s hydrogen-bond accep-
tance (see β values in Table 1), indicating that the change in the
alcoholic component in the mixture from methanol to propan-2-
ol increases the capacity of the alcohol to form hydrogen-bonded
complexes with water (Scheme 1), and these complexes can be
responsible for the solvation of the dye cation in the alcohol-
rich region [23,24], reflected in the ideal behavior observed,
e.g., in propan-2-ol mixtures. In addition, it can be seen that the
parameter f12/2 in Table 2 points to the fact that 2 is preferen-
tially solvated by S12 in the following order for the alcohols:
methanol < ethanol < propan-2-ol < propan-1-ol.
3.4. Acetone–ROH mixtures
Fig. 4 depicts the variation in ET values for the longest-
wavelength band of 2 as a function of the HBD component
mole fraction in acetone–ROH mixtures. For all mixtures, ace-
tone, or the less polar moiety of the S12 species (with the
methyl groups of acetone oriented towards the dye), is always
responsible for the PS of 2. However, as the hydroxylic com-
ponent becomes less polar, the intensity of the PS diminishes
and approaches a nearly ideal behavior in butan-1-ol mixtures,
with the solvent composition in the microenvironment of 2
and that in the bulk mixture being almost equal in all mix-
tures. These results can be explained by the fact that 2 is a
hydrophobic probe. Thus, the exchange of the more polar HBD
cosolvent for one which is more hydrophobic may lead to a com-
paratively stronger interaction of the probe with the cosolvent
through hydrophobic interactions. The synergism observed for
the acetone–methanol curve (Fig. 4(b)) suggests that S12 species
play an important role in the PS of the dye. In this situation, the
methyl groups of acetone are responsible for the solvation of the
dye.
3.3. DMSO–ROH mixtures
Fig. 3 shows the ET(dye) values plotted as a function of the
hydroxylic component in DMSO–ROH mixtures. All mixtures
display a strong synergistic behavior, their ET values being lower
than those for the two pure solvents in the aqueous mixtures and
greater than those for the two pure solvents in the alcoholic
mixtures. In addition, in all mixtures dye 2 is preferentially sol-
vated by the less polar component of the binary mixture. It is
interesting to observe that dye 1 is preferentially solvated by the
more polar components in all DMSO–alcohol mixtures, while
PS by DMSO or S12 species is verified in its aqueous mixtures
[19].

T. Bevilaqua et al. / Spectrochimica Acta Part A 65 (2006) 535–542
541
A comparison between our data and those obtained by Ortega
et al. [19] shows that dyes 1 and 2 have a completely differ-
ent behavior: the acetone–water plot for 1 exhibits a sigmoidal
shape while, in acetone–alcohol mixtures, 1 is preferentially sol-
vated by the more polar component of the mixtures. The analysis
of the preferential solvation parameters in Table 2 seems more
complex because these did not indicate a relationship with the
Kamlet–Taft parameters in Table 1. Data in Table 2 show that
for methanol and ethanol mixtures the ET(dye)12 values are very
close to that for acetone in its methanolic mixtures and to that
for ethanol in acetone–ethanol mixtures. This point is important
because it introduces, according to a study by Ortega et al. [19],
an expressive error in the calculation of f12/1 and f12/2 values.
If these uncertainties are considered, the same general trends
observed for DMSO mixtures can be visualized for the acetone
mixtures.
(DMF or DMF-like S12 species) of the mixture. In addition,
a strong synergistic behavior for DMF–water mixtures can be
observed (Fig. 6(a)), which diminishes if the HBD solvent is
changed from water to butan-1-ol. A comparison with the data
presented for DMSO–ROH mixtures (Fig. 3) shows that 2 dis-
plays similar trends with both HBA solvents. This results from
the fact that DMSO and DMF present similar Kamlet–Taft
parameters. A view of f12/2 values in Table 2 shows that 2
is preferentially solvated by S12 in an order similar to that
observed for the DMSO mixtures. Again, the discontinuities
in the calculated f12/2 values can be explained as shown above
for the acetone mixtures. Thus, the explanation for the data for
DMF mixtures is the same as that offered for DMSO mixtures
(see above), and the differences in synergistic behavior can be
explained by the fact that it increases when the ET values for
the pure solvents participating in the mixtures are more simi-
lar [18–20,22–24]. Herodes et al. [20] studied the behavior of
1 in binary mixtures of DMF with hydroxylic solvents and also
observed a similarity with the DMSO mixtures [19]: in its aque-
ous mixtures it is preferentially solvated by DMF or S12 species
while in the DMF–alcohol mixtures the PS of 1 occurs via the
alcoholic component or by the most polar moiety of the S12
species.
3.5. Acetonitrile–ROH mixtures
ThevariationinET(dye)valuesasafunctionoftheHBDcom-
ponent mole fraction in acetonitrile–ROH mixtures is shown in
Fig. 5. A simple comparison of these plots with those in Fig. 4
shows that their behavior is very similar: for acetonitrile–ROH
mixtures acetonitrile (or the less polar moiety of the S12
species) preferentially solvates 2. Again, as the hydroxylic
component becomes less polar, PS diminishes in magnitude
and comes close to an ideal behavior in butan-1-ol mixtures.
The only difference observed in the comparison of acetone
and acetonitrile mixtures is that while acetone–methanol mix-
tures are synergistic (Fig. 4(b)) the same was not observed
for acetonitrile–methanol mixtures (Fig. 5(b)). The behavior
of pyridiniophenolate 1 in acetonitrile with HBD solvents is
also similar to the mixtures of acetone with the same cosolvents
[19].
The explanation for the behavior of 2 in these mixtures is
the same as that given for acetone mixtures and the similar
effects observed for acetone and acetonitrile mixtures can be
explained through a comparison of the polarity parameters of
these two solvents (see Table 1). The π* value for the two sol-
vents is very similar and the α values are very low. The difference
observed in the β values can explain the differences observed in
the methanolic mixtures, since acetone is more able than ace-
tonitrile to interact through hydrogen bonding with the alcohol
due to the higher basicity of acetone. In addition, the same expla-
nation for the apparent discontinuities in the calculated f12/2
values for acetone mixtures can explain the data with acetonitrile
mixtures.
4. Conclusion
The spectroscopic studies involving dye 2 dissolved in binary
solvent mixtures show that the behavior of the dye is dictated by
solute–solvent and solvent–solvent interactions. The two-step
solvent-exchange model of Skwierczynski and Connors [17]
was successfully fitted to all mixed systems studied here. The
combination in the mixtures of a solvent with a high β value
and hydroxylic solvents, which display large α values, allows
the formation of complexes formed through hydrogen bond-
ing. The cation of compound 2 is solvated by the less polar
moiety of these complexes, i.e. by the alkyl groups of the sol-
vents, since the dye cation is of hydrophobic nature and has a
dimethylamino group with a minor capability to form hydro-
gen bonds with the medium in comparison with the phenolate
group present in other dyes, such as dye 1 [18–20] and Brooker’s
merocyanine [23,24]. In this aspect, in mixtures comprised of a
HBA solvent and a HBD solvent, the cation of dye 2 behaves in
a similar way to pyrene [22], a highly hydrophobic probe that
interacts with the medium exclusively through dipole–dipole
and dispersion interactions. Thus, these probes are able to detect
solvent–solvent interactions, which are implicit to the observed
synergistic behavior. This is very important in the comprehen-
sion of phenomena related to the solvation of chemical species
and to the importance of hydrophobicity in biological events
[36].
3.6. DMF–ROH mixtures
Fig. 6 depicts the variation in ET(dye) values as a function of
the hydroxylic component mole fraction for DMF–ROH mix-
tures. DMF is an amide that is unable to form hydrogen bonds
with 2 but is capable of interacting with the HBD cosolvent
through hydrogen bonding due to its high β value (β = 0.69),
yielding the S12 species. The plots show that in all cases the dye
cation is preferentially solvated by the less polar component
Acknowledgements
The financial support of Brazilian Conselho Nacional de
´
´
Pesquisa Cientıfica e Tecnologica (CNPq), FURB, and UNISUL
is gratefully acknowledged.

542
T. Bevilaqua et al. / Spectrochimica Acta Part A 65 (2006) 535–542
References
[19] J. Ortega, C. Ra`fols, E. Bosch, M. Rose´s, J. Chem. Soc., Perkin Trans. 2
(1996) 1497–1503.
[20] K. Herodes, I. Leito, I. Koppel, M. Rose´s, J. Phys. Org. Chem. 12 (1999)
109–115.
[21] V.G. Machado, C. Machado, M.G. Nascimento, M.C. Rezende, J. Phys.
Org. Chem. 10 (1997) 731–736.
[22] M.A.R. Silva, D.C. da Silva, V.G. Machado, E. Longhinotti, V.L.A. Fres-
cura, J. Phys. Chem. A 106 (2002) 8820–8826.
[23] D.C. da Silva, I. Ricken, M.A.R. Silva, V.G. Machado, J. Phys. Org. Chem.
15 (2002) 420–427.
[24] T. Bevilaqua, D.C. da Silva, V.G. Machado, Spectrochim. Acta, Part A 60
(2004) 951–958.
[25] Y. Marcus, Chem. Soc. Rev. 22 (1993) 409–416.
[26] Y. Marcus, J. Chem. Soc., Faraday Trans. 91 (1995) 427–430.
[27] F. Rastrelli, G. Saielli, A. Bagno, A. Wakisaka, J. Phys. Chem. B 108 (2004)
3479–3487.
[28] B.S. Furniss, A.J. Hannaford, P.W.G. Smith, A.R. Tatchell, Vogel’s
Textbook of Practical Organic Chemistry, fifth ed., Longman, London,
1989.
[29] I.A.Z. Al-Ansari, Bull. Soc. Chim. Fr. 134 (1997) 593–599.
[30] M.J. Kamlet, J.-L.M. Abboud, M.H. Abraham, R.W. Taft, J. Org. Chem.
48 (1983) 2877–2887.
[1] C. Reichardt, Solvents and Solvent Effects in Organic Chemistry, second
ed., VCH, Weinheim, 1988 (Chapters 6 and 7).
[2] P. Suppan, N. Ghoneim, Solvatochromism, first ed., Royal Society of
Chemistry, Cambridge, UK, 1997 (Chapter 3).
[3] A. Mishra, R.K. Behera, P.K. Behera, B.K. Mishra, G.B. Behera, Chem.
Rev. 100 (2000) 1973–2011.
[4] C. Reichardt, Chem. Rev. 94 (1994) 2319–2358.
[5] Y. Marcus, J. Chem. Soc., Perkin Trans. 2 (1994) 1015–1021.
[6] K. Dimroth, C. Reichardt, Z. Anal. Chem. 215 (1966) 344–350.
[7] Z.B. Maksimovic´, C. Reichardt, A. Spiric´, Z. Anal. Chem. 270 (1974)
100–104.
[8] H. Langhals, Angew Chem. 94 (1982) 739–749;
H. Langhals, Angew. Chem., Int. Ed. Engl. 21 (1982) 724–733.
[9] J.G. Dawber, S. Etemad, M.A. Beckett, J. Chem. Soc., Faraday Trans. 86
(1990) 3725–3729.
[10] N. Ghoneim, Spectrochim. Acta, Part A 57 (2001) 1877–1884.
[11] R.E. Leita˜o, F. Martins, M.C. Ventura, N. Nunes, J. Phys. Org. Chem. 15
(2002) 623–630.
[12] K. Herodes, J. Koppel, C. Reichardt, I.A. Koppel, J. Phys. Org. Chem. 16
(2003) 626–632.
[31] E.M. Kosower, J. Am. Chem. Soc. 80 (1958) 3253–3260.
[32] A.S. El-Dali, M.H. Abdel-Kader, R.M. Issa, E.A. El-Sherbini, Spec-
trochim. Acta, Part A 59 (2003) 405–411.
[33] S. Spange, M. Lauterbach, A.K. Gyra, C. Reichardt, Liebigs Ann. Chem.
(1991) 323–329.
[34] C.A. Coleman, C.J. Murray, J. Org. Chem. 57 (1992) 3578–3582.
[35] Y. Marcus, J. Chem. Soc., Perkin Trans. 2 (1994) 1751–1758.
[36] D. Wade, Chem. Biol. Interact. 117 (1999) 191–217.
[13] P. Chatterjee, A.K. Laha, S. Bagchi, J. Chem. Soc., Faraday Trans. 88
(1992) 1675–1678.
[14] A. Henseler, M. von Raumer, P. Suppan, J. Chem. Soc., Faraday Trans. 92
(1996) 391–393.
[15] S.Y. Bae, B.R. Arnold, J. Phys. Org. Chem. 17 (2004) 187–193.
[16] A. Bagno, J. Phys. Org. Chem. 15 (2002) 790–795.
[17] R.D. Skwierczynski, K.A. Connors, J. Chem. Soc., Perkin Trans. 2 (1994)
467–472.
[18] M. Rose´s, C. Ra`fols, J. Ortega, E. Bosch, J. Chem. Soc., Perkin Trans. 2
(1995) 1607–1615.