Solvatochromism in aqueous micellar solutions: Effects of the molecular structures of solvatochromic probes and cationic surfactants

Article abstract of DOI:10.1039/a809244c

Solvatochromic behavior of 2,6-diphenyl-4-(2,4,6-triphenyl-1-pyridinio)- 1-phenolate (RB); 1-methyl-8-oxyquinolinium betaine (QB); sodium 1-methyl-8- oxyquinolinium betaine-5-sulfonate (QBS); and 1-methyl-3-oxypyridinium betaine (PB) was studied spectrophotometrically in micellar solutions of the following cationic surfactants: cetyltrimethylammonium chloride, cetyldimethylbenzylammonium chloride, dodecyltrimethylammonium chloride, and dodecyldimethylbenzylammonium chloride. Microscopic polarity of water at the (average) solubilization site of the solvatochromic probe, E(T) in kcal mol^{- 1}, was calculated from the position of the longest-wavelength absorption band of the probe. The visible spectrum of PB, the most hydrophilic probe, is not affected by surfactants because it is not included in the micellar pseudo phase. For the other three solvatochromic probes, calculated E(T) values depend on the structures of both the probe and the surfactant, namely, its headgroup and long-chain alkyl group. RB, the most hydrophobic probe, samples a much lower microscopic polarity than QB and QBS because it penetrates deeper into the cationic micelle. This conclusion has been confirmed by ¹H NMR. Polarities measured by (zwitterionic) QB and (anionic) QBS differ because the latter probe exchanges with the surfactant counterion. Calculated E(T) values refer to micelle-bound probes and are, therefore, different from those reported in the literature, typically determined at [surfactant] ≤ 0.05 mol L^-1. Effective water concentrations at the solubilization sites of these solvatochromic probes has been calculated by using as references mixtures of water with each the following organic solvents: n-propanol and dioxane (RB); ethanol, n-propanol, acetonitrile and dioxane (QB and QBS).

Full text of DOI:10.1039/a809244c

View Online / Journal Homepage / Table of Contents for this issue
Solvatochromism in aqueous micellar solutions: e†ects of the
molecular structures of solvatochromic probes and cationic
surfactants
Luzia P. Novaki and Omar A. El Seoud*
Instituto de Qu•mica, Universidade de Sao Paulo, C.P. 26077, 05599-970, Sao Paulo, S.P., Brazil.
E-mail: elseoud=iq.usp.br; Fax: ]55 11 818 3874
Received 26th November 1998, Accepted 1st February 1999
Solvatochromic behavior of 2,6-diphenyl-4-(2,4,6-triphenyl-1-pyridinio)-1-phenolate (RB); 1-methyl-8-
oxyquinolinium betaine (QB); sodium 1-methyl-8-oxyquinolinium betaine-5-sulfonate (QBS); and 1-methyl-3-
oxypyridinium betaine (PB) was studied spectrophotometrically in micellar solutions of the following cationic
surfactants: cetyltrimethylammonium chloride, cetyldimethylbenzylammonium chloride,
dodecyltrimethylammonium chloride, and dodecyldimethylbenzylammonium chloride. Microscopic polarity of
water at the (average) solubilization site of the solvatochromic probe, E in kcal mol~1, was calculated from
T
the position of the longest-wavelength absorption band of the probe. The visible spectrum of PB, the most
hydrophilic probe, is not a†ected by surfactants because it is not included in the micellar pseudo phase. For
the other three solvatochromic probes, calculated E values depend on the structures of both the probe and
T
the surfactant, namely, its headgroup and long-chain alkyl group. RB, the most hydrophobic probe, samples a
much lower microscopic polarity than QB and QBS because it penetrates deeper into the cationic micelle. This
conclusion has been conÐrmed by 1H NMR. Polarities measured by (zwitterionic) QB and (anionic) QBS di†er
because the latter probe exchanges with the surfactant counterion. Calculated E values refer to micelle-bound
T
probes and are, therefore, di†erent from those reported in the literature, typically determined at [surfactant]
O0.05 mol L~1. E†ective water concentrations at the solubilization sites of these solvatochromic probes has
been calculated by using as references mixtures of water with each the following organic solvents: n-propanol
and dioxane (RB); ethanol, n-propanol, acetonitrile and dioxane (QB and QBS).
this being the reason for designating their discrete protons,
vide infra.
Introduction
Organized assemblies a†ect rates of chemical and photo-
chemical reactions and the position of chemical equilibria.
Knowledge of the properties of water at interfaces of
organized assemblies, at the microscopic level, is a prerequisite
for understanding the above mentioned e†ects because inter-
facial water is involved in, inter alia, solvation of reactants and
transition states, solvation of surfactant headgroup and
counterion, and in proton transfers.1h3 An important property
of interfacial water is its microscopic polarity, measured by
Our objective is to determine the microscopic polarity of
water at the average solubilization site of each probe as a
function of its structure and charge, as well as the structure of
the surfactant (length of the hydrophobic tail, and structure of
the headgroup). We show that these variables a†ect the solu-
bilization site of the probe in the micellar pseudo phase, and
hence calculated E . Our results are used to mimic micro-
T
scopic polarities at micellar solubilization/reaction sites of
hydrophobic (modeled by RB), polar (modeled by QB) as well
as anionic substrates (modeled by QBS).
the E scale, which is calculated from the position of the
T
longest-wavelength absorption band of a solvatochromic
probe (hereafter denoted as ““probeÏÏ) by the relationship:4
Experimental
(a) Materials
E (kcal mol~1) \ 28591.5/j
(nm)
(1)
T
max
Recently, we have studied solvatochromism in pure organic
solvents,5 in binary mixtures of water with alcohols,6 and
water with dipolar aprotic solvents.7 The following probes
have been employed: 2,6-diphenyl-4-(2,4,6-triphenyl-1-pyridin-
io)-1-phenolate, (RB), 1-methyl-8-oxyquinolinium betaine
(QB), sodium 1-methyl-8-oxyquinolinium betaine-5-sulfonate
(QBS) and 1-methyl-3-oxypyridinium betaine (PB), see struc-
tures opposite. The corresponding solvatochromic scales are
E (30), E (QB), E (QBS), and E (PB), for RB, QB, QBS and
The reagents were obtained from Aldrich, Fluka and Merck,
and were puriÐed as given elsewhere.8 CMe ACl (aqueous
3
T
T
T
T
PB, respectively.
We have now extended this study to aqueous micellar solu-
tions of cationic surfactants whose structures are shown
overleaf.
Using 1H NMR spectroscopy, we have studied solu-
bilization of the probes in micelles of the last two surfactants,
Phys. Chem. Chem. Phys., 1999, 1, 1957È1964
1957

View Online
meter (operating at 300.13 MHz for proton), at a digital
resolution of 0.1 Hz per data point. Chemical shifts were mea-
sured from internal dioxane (5 ] 10~3 mol L~1) and then
converted into the TMS scale, by using d
\ 3.53.11
dioxane
Results and discussion
(a) Choice of solvatochromic probes and surfactants
We used RB because it has been most extensively studied in
pure solvents, and in binary solvent mixtures.4h7,10,12,13 Sol-
vatochromic data are also available in aqueous micellar solu-
tions.14,15 RB is a very hydrophobic probe, solubility in
water \ 7.2 ] 10~6 mol L~1,4 and is expected, therefore, to
bind strongly to cationic micelles. QB is more hydrophilic
than RB, the negative charge of QBS makes it still more
hydrophilic than QB, without a†ecting its solvatochromic
behavior both in pure solvents, and in binary solvent mix-
tures, i.e., E (QB) B E (QBS).5h7 Anionic QBS is expected to
T
T
exchange with the surfactant counterion, Cl~. This may lead
to di†erent (average) micellar solubilization sites of QB and
QBS. Finally, PB is the most hydrophilic of these probes.
We are interested in investigating e†ects of headgroup
structure and length of the hydrophobic tail of the surfactant
on the properties, e.g., microscopic polarity of interfacial
solution, 25% surfactant) was dried under reduced pressure,
then recrystallized from acetoneÈmethanol. DMe ACl was
3
recrystallized from the same solvent mixture.
CMe BzACl and DMe BzACl were prepared by reacting
2
2
water.16 Although CMe BzACl is extensively used in phar-
N,N-dimethylalkylamine (alkyl group; cetyl or dodecyl) with
benzyl chloride in anhydrous acetone, under reÑux for four
hours. The solvent was evaporated and the remaining solid
was extracted with dry ethyl ether, then recrystallized from
acetoneÈmethanol.
2
maceutical preparations as a topical antiseptic,17 very little
work has been done on its e†ects on rates and equilibria of
chemical reactions.1h3 It was deemed important to include
CMe BzACl and DMe BzACl in the present study in order to
2
2
investigate how the change of the headgroup from tri-
methylammonium to dimethylbenzylammonium may a†ect
the results, especially because 1H NMR spectroscopy has indi-
PuriÐed surfactants were dried under reduced pressure, over
P O to a constant weight. Their (nonhygroscopic) perchlor-
2
5
ates gave satisfactory microanalysis (Perkin Elmer CHN-2000
apparatus, the Microanalysis Laboratory, Instituto de Qui-
mica, Universidade de Sao Paulo). Their critical micelle con-
centrations, c.m.c., were determined at 25 ¡C by surface
tension measurement (Lauda TE 1C digital ring-tensiometer)
and were in agreement with literature values.9 The probes QB,
QBS, and PB were those used in previous studies.5h7
cated that the benzyl moiety of CMe BzACl lies more or less
2
parallel to the micellar interface.18 Finally, substrateÈmicelle
association constants usually increase as a function of increas-
ing the chain length of the surfactant,1 a factor whose impor-
tance can be determined by varying the length of the
surfactant hydrophobic group, e.g., CMe BzACl and
2
DMe BzACl.
All-glass, double-distilled water was used throughout. Den-
sities of aqueous dioxane mixtures were determined at 25 ¡C
by a PAAR DMA 40 digital densimeter.
2
(b) Solvatochromic behavior in micellar solutions
In principle, the transfer of a substrate from bulk water to the
micellar pseudo phase a†ects its physicoÈchemical properties,
(b) Spectrophotometric determination of E values
T
e.g., its pK and absorption spectra. Meaningful analysis of
All solutions were prepared in 10~3 mol L~1 NaOH instead
of water. The probe stock solution was 0.009 mol L~1, and its
Ðnal concentration in the micellar solution was 2.1 ] 10~4
mol L~1. RB was used as a saturated surfactant solution. A
Beckman DU-70 UVÈvis spectrophotometer was used,
equipped with a thermostated cell holder. After thermal equili-
a
experimental data, e.g., the dependence of UVÈvis absorbance
on [surfactant] requires knowledge of concentrations of the
species present. The problem is simpliÐed if the micellar solu-
tion contains only one probe species, the zwitterionic form in
the present case. Consequently, we address RB because it has
the highest pK in water (8.63),14 and is the most hydrophobic
bration, the spectra were recorded at 25 ^ 0.1 ¡C (CMe ACl,
a
3
probe, i.e., it is expected to be most a†ected by the above men-
DMe ACl, and DMe BzACl) or 30 ^ 0.1 ¡C (CMe BzACl).
3
2
2
tioned transfer. Micelle-induced pK shift is due to ““mediumÏÏ
Stoppered cells with adequate path length (0.2È4 cm) were
a
and electrostatic e†ects. Interfacial water is known to be less
used to obtain convenient absorbances. Accurate j
determined from the Ðrst derivative of the absorption spec-
were
max
polar than bulk water, and is akin to an electrolyte salt solu-
tion, e.g., the ionic strength in the Stern layer of 0.025 mol
trum, then used to calculate the corresponding E (kcal
T
mol~1) from eqn. (1). The uncertainties in E values are 0.1,
T
L~1 CMe ACl is ca. 1.4 mol L~1.19 The combined e†ects of
3
lower polarity of interfacial water (a pK increase), and elec-
0.03, and 0.6 kcal mol~1, for QB (or QBS), PB, and RB,
a
trostatic stabilization of the zwitterionic form of RB by the
respectively.5h7,10
cationic interface (a pK decrease) is a net decrease which
a
ranges from 1.2 (DMe ACl) to 1.7 pK units (CMe ACl).14
(c) 1H NMR study of solubilization of the probes in micellar
solutions
3
a
3
Deprotonation of micellar RB is caused by micelle-bound
hydroxide ions. Interfacial pH, however, is lower than bulk
solution pH because the equilibrium constant for ion
exchange between bulk solution hydroxide ions and the sur-
The surfactant stock solutions were 0.1 mol L~1 in 10~3 mol
L~1 NaOD/D O. Aqueous QB and QBS stock solutions were
2
added to a 1 mL volumetric tube, the solvent evaporated, and
factant counterion is small (K for [OH~ ]/[Cl~
\
(bulk)
(counterion)
the surfactant stock solution added. Accurately weighed RB
was directly dissolved in the surfactant stock solution, 24 h
were required to obtain complete probe dissolution. Proton
NMR spectra were obtained with a Bruker DPX-300 spectro-
0.02).20 Therefore, RB is expected to be present in its zwitter-
ionic form at bulk pH P 10. Indeed, we have found that the
absorbance of RB is independent of the acidity of DMe ACl
solution, at bulk pH P 11. Use of a 0.001 M NaOH solution
3
1958
Phys. Chem. Chem. Phys., 1999, 1, 1957È1964

View Online
Fig. 1 Dependence of the visible spectra of probes on [surfactant] for: (A) RB in DMe BzACl, (B) QB in CMe ACl, and (C) QBS in
2
3
DMe ABzCl. The maximum surfactant concentrations were 0.08, 0.35 and 0.30 mol L~1, respectively.
2
instead of water for preparation of all stock solutions should
not a†ect the results because the accompanied change in c.m.c.
is negligible.21
At the outset, we discuss the results of PB. The UVÈvis
spectrum of this probe is not a†ected by increasing
[surfactant]. This shows that it is not incorporated in the
micelle, in agreement with our results for other small, highly
hydrophilic acidÈbase indicators.22 Therefore, we need not
consider this probe any further.
present study, our discussion, however, will most certainly
apply to the dichloroderivative of RB.24 A corollary to the
above raised question is: at what [surfactant] is the probe
totally incorporated in the micellar pseudo phase? This con-
centration can not be readily estimated for QB because of its
high water solubility. The problem is solved, however, by
determining the fraction of micelle-bound probe, hence the
associated micellar E , as given in the Appendix.
T
In order to show the dependence of observed microscopic
Fig. 1 shows typical spectra for RB, QB, and QBS at di†er-
ent surfactant concentrations. The isosbestic points show the
presence of one probe form at two sites, namely, bulk water,
and micellar pseudo phase. The spectra of RB at low sur-
factant concentrations are probably due to the probe associ-
ated with either monomers, or small surfactant aggregates.
Substrate-induced formation of pre-micellar aggregates has
been shown for several hydrophobic species.1 The visible
spectra of RB ““bundleÏÏ at [surfactant] [ 0.003 mol L~1,
because the strong association between this probe and the
micelle means that the probe is totally micelle-bound.
polarities, Eobs, on [surfactant], we have applied eqn. (1) to
T
our absorbance data, typical graphs are shown in Fig. 2. The
points are experimental, and the curves have been calculated
by regression analysis, by using equations which gave the best
data Ðt, as indicated by sums of the squares of the residues,
;Q, Table 1. The dependence of the microscopic polarity on
structures of the probe and the surfactant is clear, and is dis-
cussed in detail in the next section.
Table 2 shows polarities of the micellar interfacial regions
as a function of the above mentioned structural variables,
along with Eobs, calculated at 0.05 mol L~1 surfactant, and
T
The polarities of interfacial regions of monolayers, aqueous
micelles, and vesicles have been investigated by several
authors. Examples are the following (probe; single, or
maximum cationic surfactant concentration used): RB, 0.05
mol L~1;14,23 2,6-dichloro-4-(2,4,6-triphenyl-N-pyridinio)-phe-
probeÈmicelle association constants, K
. We use Emic to
assoc
T
denote the microscopic polarity as determined by the micelle-
bound probe, see Appendix. In order to make the results for
RB and QB (or QBS) readily comparable, we have converted
the corresponding Emic to
a dimensionless, normalized
polarity scale, Enom, whose value for any probe is deÐned by:
T
nolate [polarity scale E (33)], 0.01 mol L~1;24 ketocyanine
T
T
dyes, 0.01 mol L~1;25 3H-indole derivatives, 0.01 mol L~1 26
Enorm \ (Emic [ EDMSO)/(Ewater [ EDMSO)
(2)
derivatives of pyrene, 0.01 mol L~1;27 esters of 9-anthracene
carboxylic acids 0.10 mol L~1;28 and methylene blue, 0.02
mol L~1.29 The following question now arises: do the report-
ed microscopic polarities correspond to that of the interfacial
region? We will address this question for probes used in the
T
T
T
T
T
DMSO was used as a reference solvent rather than TMS,4e
because QBS is only soluble in highly polar solvents. E
T
values of the di†erent probes in water and DMSO are given
elsewhere.5
Fig. 2 Dependence of Eobs on [CM ACl] for (A) RB, (B) QB, and (C) QBS.
T
3
Phys. Chem. Chem. Phys., 1999, 1, 1957È1964
1959

View Online
Table 1 Regression parameters for the dependence of Eobs on the concentration of surfactanta
T
Surfactant
Probe
Nb
Equationc
A
B
C
D
E
F
; Qd
CMe ACl
RB
QB
QBS
RB
QB
QBS
RB
QB
QBS
RB
QB
15
14
14
9
15
15
15
22
22
14
18
18
Expon. [2
Polynom.
Expon. [2
Expon. [2
Polynom.
Expon. [2
Expon. [1
Polynom.
Polynom.
Expon. [2
Polynom
52.71
64.51
63.48
48.39
64.5
63.15
53.7
64.6
65.0
53.23
64.6
61.25
10.376
0.00089
1.52
0.0058
0.96
1.046
0.0915
1.72
0.01
3
[3.29
1.006
0.468
1.15
6.98
0.56
0.0817
0.019
0.002
0.041
0.008
0.014
2.33
0.054
0.058
2.59
CMe BzACl
4.23
2
[3.62
0.92
1.13
0.0014
0.017
2.96
113.72
0.011
81.62
0.021
0.031
DMe ACl
10.98
3
[4.06
[19.55
11.21
[316.51
[0.222
[657.8
2.05
411.81
0.033
2394.1
17.98
[197.53
[3128.9
DMe BzACl
2
[7.43
0.007
0.013
QBS
Expon. [2
1.762
a At 25 ¡C for CMe ACl, DMe ACl, DMe BzACl, and 30 ¡C for CMe BzACl. b Number of surfactant concentrations used in the regression
3
3
2
2
analysis. The maximum [surfactant] ranged from 0.3 to 0.5 mol L~1. c Equation that best Ðts the data: Expon. [1 \ exponential decay of the
type: E \ A ] B exp(~*surfactant+@C).; Expon. [2 \ exponential decay of the type: E \ A ] B exp(~*surfactant+@C) ] D exp(~*surfactant+@E);
T
T
Polynom. \ polynomial equation of the type: E \ A ] B [surfactant] ] C [surfactant]2 ] D [surfactant]3 ] E [surfactant]4 ] F
T
[surfactant].5 d ; Q \ sum of the squares of the residues.
E on the chain length of the surfactant, the headgroup being
constant, is similar to that observed for RB, namely it is
higher for DMe ACl and DMe BzACl than for CMe ACl
We now address the results of Table 2. Eobs was calculated
T
T
at [surfactant] \ 0.05 mol L~1, i.e., the maximum concentra-
tion employed elsewhere.14,23 Even for RB, the most hydro-
phobic probe, Eobs [ Emic! That is, at this surfactant
3
2
3
and CMe BzACl.
2
The preceding results can be explained in terms of the
microscopic polarity at the average solubilization site of the
probe in the micellar pseudo phase, which is determined by its
hydrophobic/hydrophilic character, and electrostatic inter-
actions with the cationic interface. 1H NMR spectroscopy is
an appropriate technique to determine the average solu-
bilization site of each probe. The desired information is
obtained by examining the dependence of observed chemicals
T
T
concentration there is contribution from the fraction of probe
which resides in bulk water. It is clear, therefore, that use of
Eobs is unjustiÐed because interfacial microscopic polarity is
T
most probably overestimated due to the fraction of probe
present in bulk water.
Considering RB Ðrst, and Enorm, we note that: (i) as
T
expected, this hydrophobic probe associates with micelles
much stronger than QB or QBS; (ii) for each surfactant, RB
samples a microenvironment which is much less polar than
that sampled by QB or QBS; (iii) the order of polarity
decreases in going from the trimethylammonium to the
shifts, d , of the discrete surfactant protons on [probe]. The
obs
dynamic nature of the micelle, and of probe solubilization has
to be born in mind.30
A note on the discrete groups of DMe BzACl which are
dimethylbenzylammonium
headgroup,
CMe ACl [
2
observable by 1H NMR spectroscopy is in order. The peak of
3
CMe BzACl, and DMe ACl [ DMe BzACl; (iv) the polarity
the main methylene moiety of this surfactant is split into two
2
3
2
order, however, increases in going from a surfactant with a
long chain hydrophobic tail to its short chain counterpart,
CMe ACl \ DMe ACl, and CMe BzACl \ DMe BzACl.
apparent singlets, the Ðrst, D, consists of Ðve CH groups, and
2
the second, D@, consists of four CH groups. The reason for
2
this splitting is diamagnetic shielding of the D protons by the
3
3
2
2
Emic of Table 2 shows that microscopic polarities measured
benzyl moiety which lies more or less parallel to the micelle
interface.18,31 Also, we were unable to follow protons C
because their peak is buried under that of protons D.
T
by QB and QBS are clearly dependent on the structure of the
probe, and surfactant headgroup. Relative to QB, QBS
samples
a
more polar environment in alkyltrimethyl-
Fig. 3 shows the dependence of d
obs
DMe ACl and DMe BzACl. For clarity, we have plotted our
on [probe] for
ammonium micelles, and a medium of similar polarity in
alkyldimethylbenzylammonium micelles. The dependence of
3
2
data as chemical shift di†erences, *d (Hz) \ d
(surfactant)
Table 2 Dependence on surfactant structure of the probe-surfactant association constant, K
, observed E at 0.05 mol L~1 surfactant, Eobs,
assoc
T
T
interfacial water polarities measured by micelle-bound probes, Emic, and their normalized counterparts, Enorm a,b
T
T
Surfactant
Probe
K
/M~1
assoc
Eobs/kcal mol~1
Emic/kcal mol~1
Enorm c
T
T
T
CMe ACl
RB
QB
QBS
540
3.1
150
53.5
64.3
63.5
52.9
62.6
63.5
0.433
0.806
0.857
3
CMe BzACl
RB
QB
QBS
d
52.4
64.3
63.3
52.1
63.1
63.2
0.389
0.857
0.827
2
8.4
379
DMe ACl
RB
QB
QBS
7500
2.5
18
54.2
64.4
64.2
53.8
62.8
63.7
0.483
0.827
0.878
3
DMe BzACl
RB
QB
QBS
22200
5.7
60
53.6
64.4
63.5
53.4
63.2
63.3
0.461
0.867
0.837
2
a At 25 ¡C for CMe ACl, DMe ACl, DMe BzACl, and 30 ¡C for CMe BzACl. b See Appendix for details. The uncertainty in K
is typically
3
3
2
2
assoc
^5%. c Calculated from Enorm \ (Emic [ EDMSO)/(Ewater [ EDMSO). Emic was calculated from eqn. (7), EDMSO and Ewater are from ref. 5. d Insolubility
T
T
T
T
T
T
T
T
of RB in dilute slutions of CMe BzACl precluded determination of K
.
2
assoc
1960
Phys. Chem. Chem. Phys., 1999, 1, 1957È1964

View Online
hydroxyquinoline-derived probes, Table 3. The order of the
slopes for DMe ACl protons is: A, B, and C [ N`(Me) , D
3
3
and E. Thus, the probe penetrates the micelle, and is probably
anchored to the cationic interface by its phenoxide oxygen,14
i.e., it does not extend into the aqueous pseudo phase, as sug-
gested elsewhere.32 The slopes for the D and D@ groups of
DMe BzACl are not far from those of the corresponding B
protons. This shows that either the penetration of the probe is
2
deeper into DMe BzACl than into DMe ACl, or that di†u-
2
3
sion of the probe within the former micelle is slower due to
the presence of the bulky benzyl group in the interfacial
region. Either, or both explanations agree with the obser-
vation that RB samples
a less polar environment in
DMe BzACl than in DMe ACl, Table 2.
2
3
The most noticeable result in the case of the 8-
hydroxyquinoline-based indicators is that the order of o *d o is
QBS [ QB, for each surfactant segment followed! Because the
molecular volumes of both probes are not very di†erent (206
and 259 Ó3 for QB and QBS, respectively,5,33 the di†erence in
slopes does not necessarily mean a deeper penetration of the
micelle by QBS, it may well reÑect a longer probe residence
time in the micelle. Solubilization of QBS in micelles of
DMe ACl or DMe BzACl a†ects *d of the protons at the
3
2
interface, as well as protons A, B and C (D for the latter
surfactant). That is, the solubilization site of this probe, hence
Emic, shows little dependence on headgroup structure, prob-
T
ably because it exchanges with the surfactant counterion. QB,
on the other hand, shows a di†erent behavior, *d is largest for
protons C of DMe ACl, whereas it is largest for the protons
3
at the interface of micelles of DMe BzACl. These results show
2
that QB migrates toward the interface on going from
DMe ACl to DMe BzACl, most probably due to interaction
of its quaternary ammonium nitrogen with the interfacial
benzyl group. This explanation agrees with the observed
Fig. 3 Plots of *d \ d
[ d
on [probe] for dis-
3
2
(surfactant)
(surfactant`probe)
crete surfactant protons, for (A) RB and (B) QBS in DMe ACl. For
3
clarity, we show selected surfactant protons.
increase in K
, and Emic on going from DMe ACl to
assoc
T
3
DMe BzACl, Table 2.
[ d
. All plots are linear, the corresponding
2
(surfactant`probe)
For all probes, the polarity increases on going from sur-
factants carrying the cetyl tail to corresponding ones with a
dodecyl hydrophobic group. This may reÑect the micellar
changes that accompany a decrease in the length of the sur-
factant hydrophobic tail, namely, an increase of c.m.c. and the
degree of micelle dissociation, a, a decrease of the micellar
aggregation number and the thickness of its Stern layer.1,2,34
As a result, the dodecyl chain micelles are smaller, less
compact, and their Stern layers are more ““ionicÏÏ (due to
higher a) than their cetyl chain counterpart. Our results show
slopes are given in Table 3. Designation of the discrete sur-
factant protons is given in the Introduction. Most *d are posi-
tive, showing upÐeld shifts (toward TMS) by these
heterocyclic probes. The magnitude of shielding/deshielding
depends on the average orientation of the probe relative to a
discrete surfactant group, the extension of the p electron cloud
of the probe, and its residence time in the micelle.30 In the
following discussion we are concerned with o *d o.
In agreement with the order of K
the case of RB are much larger than those of the 8-
, Table 2, the slopes in
assoc
that these di†erences contribute to higher values of Enorm in
micelles of the former surfactant.
T
Table 3 Dependence of the slopes of plots of *d \ d
(surfactant)
Interfacial water has been modeled by binary aqueous sol-
vents, e.g., aqueous dioxane and aqueous alcohols. This raises
the following question: which binary solvent mixtures can be
used as references for comparison with interfacial water? Our
previous results showed that solvatochromism of RB, QB, and
QBS in binary solvent mixtures is not ideal. That is, the
[ d
on [probe] for the discrete surfactant protonsa
(surfactant`probe)
Slopeb/Hz mol~1 L probe
Surfactant/
proton
RB
QB
QBS
DMe ACl
3
dependence of the appropriate E on the mole fraction of
wN`Me
3526
4677
4727
4434
3843
2924
24
36
49.6
82.9
43.1
427.8
T
3
water is not linear, as required by the Onsager model of
A
429.7
406.5
293.5
59.8
B
C
D
E
dielectrics, due to preferential solvation of the probe by one
component of the binary solvent mixture.6,7 Consequently, we
report the molarity of water in binary mixtures with organic
solvents most commonly used to mimic interfacial water,
namely, aqueous methanol, ethanol, n-propanol, dioxane, and
acetonitrile, Table 4.
[41.4
[43.5
DMe BzACl
2
wN`Me
2347
2590
2534
2559
2364
2203
1119
53.6
160.6
43.6
609.5
1154.6
177
2
C H wCH
At the outset, we show the importance of using Emic instead
6
5
2
A
B
T
of Eobs. Compare, e.g., the following interfacial water concen-
T
37.2
252.0
trations for CMe ACl and CMe BzACl micelles, as deter-
D
D@
E
[53.9
[179.5
[63.4
[182.2
3
2
mined from solvatochromic data of RB in two reference
solvents: ethanolÈwater; 6.5 and 1.6 mol L~1 (Table 4); 10.4
and 5.4 mol L~1;14,15,23 dioxaneÈwater; 25.2 and 21.6 mol
L~1 (Table 4); 28.5 and 24.3 mol L~1.14,15,23 That is, use of
9.7
[73.8
a Experiments were carried out at room temperature at 300.13 MHz.
b The uncertainty in these slopes is typically ^4%.
E (30) at [surfactant] O0.05 mol L~1 leads to [water] in the
T
Phys. Chem. Chem. Phys., 1999, 1, 1957È1964
1961

View Online
Table 4 Molar concentrations of water at the average micellar solubilization sites of probes, calculated from solvatochromic data in bulk binary
aqueous solventsa
Water concentration in bulk reference aqueous solvents mol L~1
Surfactant
Probe
Methanol
Ethanol
n-Propanol
Dioxane
Acetonitrile
CMe ACl
RB
QB
QBS
b
6.5
40.4
40.4
25.2
40.8
37.9
25.2
42.7
42.2
7.0
38.9
39.4
3
34.7
36.3
CMe BzACl
RB
QB
QBS
b
1.6
42.4
37.0
12.0
42.8
33.8
21.6
44.2
39.3
4.9
40.9
35.2
2
37.3
32.0
DMe ACl
RB
QB
QBS
b
10.9
39.4
42.6
30.4
39.8
40.6
32.6
41.8
44.2
10.3
37.9
42.0
3
33.4
39.0
DMe BzACl
RB
QB
QBS
b
9.0
43.3
38.2
28.4
43.7
35.2
31.2
45.0
40.2
8.7
42.0
36.6
2
38.6
33.5
a The bulk aqueous solvents are used as reference for water at the solubilization site of the probe, see Results and discussion section for details.
b Calculated microscopic polarity is less than that of anhydrous methanol.
Stern layer which are di†erent from those reported in Table 4,
for the same binary solvent mixture.
Acknowledgements
We thank FAPESP and FINEP for Ðnancial support and the
CNPq for research productivity fellowship to O. A. El Seoud.
L. P. Novaki thanks the FAPESP for a postdoctoral fellow-
ship.
Examination of Table 4 shows that estimation of [water] at
the solubilization site of the probe depends on its structure,
and on the reference aqueous mixture. This dependence is
especially clear for RB, which reports polarities lower than
that of pure methanol! This reÑects the fact that the relation-
ship between the appropriate E scale and the mole fraction of
water is not ideal, the deviation from ideality is binary
Appendix: Calculations
T
solvent-dependent, and is largest for RB for all binary mix-
(a) Solvatochromic probe-micelle association constant, K
assoc
Eqn. (3) describes the observed absorption, A , in terms of
tures investigated.6,7 We recall that the micellar medium e†ect
on pK of several hydrophobic acidÈbase indicators, including
obs
a
absorption of the species present, K
tion of micellized surfactant:35
, and the concentra-
RB, has been conveniently reproduced by 50% aqueous
assoc
dioxane.2,14 This [water] is not far from that reported by RB
in aqueous n-propanol and aqueous dioxane, i.e., it seems
appropriate to use these binary solvent mixtures to mimic
polarity at the solubilization site of this probe.
A
\ A
] M(A [ A
)/
obs
water
mic
water
] (1/(K
([surfactant] [ c.m.c.)n ] 1)N (3)
assoc
Compared to RB, QB and QBS show much less dependence
on the nature of the binary solvent mixture. Although the
results are consistently lower when aqueous methanol is used
as a reference solvent, water concentrations estimated from
other aqueous solvents fall within narrow ranges which are
dependent, as discussed above, on the structure of the sur-
factant headgroup. In this regard, in view of the micro-
heterogeneous nature of many aqueous binary solvents, and of
preferential solvation of these probes, it is satisfying that con-
sistent [water] have been obtained by using aqueous alcohols,
or aqueous aprotic binary mixtures as reference solvents.
where A
and A refer to absorbance in the absence of the
water
mic
surfactant, and that due to micelle-bound solvatochromic
indicator, respectively, n is the number of surfactant mono-
mers associated with the probe, and the remaining symbols
are those previously deÐned. The experimentally accessible
terms of eqn. (3) are A , A
Regression analyses were applied to plots of
, [surfactant], and c.m.c.9
obs
water
A
vs.
obs
[surfactant], by using n \ 1 for QB and QBS, and n \ 2 for
RB. To obtain a better Ðt, it was necessary in some cases to
use c.m.c. as an adjustable parameter. The c.m.c. obtained by
iteration, however, were in agreement with those reported else-
where.9 A typical Ðt of eqn. (3) to experimental absorbance
data is shown in Fig. 4, and values of K
2.
are given in Table
Conclusions
assoc
Although microscopic polarity of interfacial water can be con-
veniently measured by solvatochromic probes, only Emic is
T
meaningful. For similarly charged micelles, Emic depends on
T
(b) Calculation of polarities of interfacial water as determined
by micelle-bound probes
the structure and charge of the probe employed, the structure
of the surfactant headgroup, and the length of its hydrophobic
tail. The results are explained in terms of penetration of the
probe into the micellar pseudo phase, as deduced from its
e†ect on 1H NMR chemical shifts of the surfactant discrete
protons. E†ective water concentrations at the solubilization
sites of these probes can be conveniently determined by using
aqueous n-propanol or aqueous dioxane for RB, and binary
mixtures of water with ethanol, n-propanol, acetonitrile, or
dioxane for QB and QBS. Care should be exercised when
using a solvatochromic probe in order to estimate the micro-
scopic polarity at the solubilization/reaction site of a micelle-
bound substrate. The structures of probe and substrate in
question should match as closely as possible.
The following equations apply to the dependence of probe
absorbance on [surfactant] at two di†erent wavelengths:
A
A
\ A
\ A
f
] A
] A
f
(4)
(5)
obs1
obs2
water1 water
mic1 mic
f
f
water2 water
mic2 mic
where the subscripts 1 and 2 refer to two chosen wavelengths,
f
and f
refer to fractions of free, and micelle-bound
water
mic
probe, respectively. Rearrangement of eqns. (4) and (5) yields
eqn. (6):
f
\ [A
obs1
[ (A
A
/A
)]/
mic
water1 obs2 water2
] [A
[ (A
A
/A
)] (6)
mic1 mic2 water1 water2
1962
Phys. Chem. Chem. Phys., 1999, 1, 1957È1964

View Online
Fig. 4 Representative plots for calculation of K
from eqn. (3) for (A) RB, (B) QB in DMe BzACl, and (C) QBS in CMe ACl.
assoc
2
3
Fig. 5 Representative plots of eqn. (7), showing the dependence of Eobs on f , for (A) RB in DMe BzACl, (B) QB in CMe BzACl, and (C) QBS
T
mic
2
2
in CMe BzACl, see Appendix.
2
7
8
9
L. P. Novaki and O. A. El Seoud, Ber. Bunsen-Ges. Phys. Chem.,
1997, 101, 902.
D. D. Perrin and W. L. F. Armarego, in PuriÐcation of L abor-
atory Chemicals, Pergamon Press, New York, 3rd edn., 1988.
K. Mukerjee and K.J. Mysels, in Critical Micelle Concentrations
of Aqueous Surfactant Systems, NSRDS-NBS 36, US Government
Printing Office, Washinton DC, 1971.
f
is then used to calculate Emic from eqn. 7:
mic
T
Eobs \ Ewater f
] Emic f
(7)
T
T
water
T
mic
A plot of Eobs vs. f should be a straight line whose slope is
Emic, as shown in Fig. 5.
T
mic
T
10 J. G. Dawber, J. Ward and R. A. Williams, J. Chem. Soc.,
(c) Calculation of water concentration at the solubilization site
of the probe by using aqueous binary mixtures as reference
solvents
Faraday T rans. 1, 1988, 84, 713.
11 A. Derome, in Modern NMR T echniques for Chemistry Research,
Pergamon Press, Oxford, 1987.
12 O. Pytela, Collect. Czech. Chem. Commun., 1988, 53, 1333.
13 P. Suppan and N. Ghoneim, in Solvatochromism, The Royal
Society of Chemistry, Cambridge, 1997.
The dependence of E on the mole fraction of water is known
T
for the present probes in several binary mixtures.6,7 Emic was
T
converted into mole fraction of water, then into molarity of
14 C. J. Drummond, F. Grieser and T. W. Healy, Faraday Discuss.
Chem. Soc., 1986, 81, 95.
water from known densities.36 For aqueous dioxane, the
relationship between water mole fraction and molarity was
experimentally determined from weights of the two com-
ponent solvents, and density of the binary mixture.
15 F. Grieser and C. J. Drummond, J. Phys. Chem., 1988, 92, 5580,
and references cited therein.
16 O. A. El Seoud, J. Mol. L iq., 1997, 72, 85, and references cited
therein.
17 W. Gump, in Kirk-Othmer Encyclopedia of Chemical T echnology,
Wiley Interscience, New York, 3rd edn., 1979, vol. 7, p. 815.
18 L. T. Okano, O. A. El Seoud and T. K. Halstead, Colloid Polym.
Sci., 1997, 275, 138.
References
1
(a) C. A. Bunton and G. Savelli, Adv. Phys. Org. Chem., 1986, 22,
213; (b) C. A. Bunton, F. Nome, F. H. Quina and L. S. Romsted,
Acc. Chem. Res., 1991, 24, 357; (c) C. A. Bunton, J. Mol. L iq.,
1997, 72, 231.
O. A. El Seoud, Adv. Colloid Interface Sci., 1989, 30, 1, and refer-
ences cited therein.
19 A. Chaudhuri, J. A. Loughlin, L. S. Romsted and J. Yao, J. Am.
Chem. Soc., 1993, 115, 8351.
20 P. S. Araujo, R. M. V. Aleixo, H. Chaimovich, N. Bianchi, L.
Miola and F. H. Quina, J. Colloid Interface Sci., 1983, 96, 293.
21 S. G. Cutler and P. Meares, J. Chem. Soc., Faraday T rans. 1,
1978, 74, 1758.
22 A. M. Chinelatto, M. T. Fonseca, N. Z. Kiyan and O. A El
Seoud, Ber. Bunsen-Ges. Phys. Chem., 1990, 94, 882.
23 (a) P. Plieninger and H. Baumgartel, Ber. Bunsen-Ges. Phys.
Chem., 1982, 86, 161; (b) K. A. Zachariasse, N. V. Phuc and B.
Konzanklewicz, J. Phys. Chem., 1981, 85, 2676; (c) K. A.
Zachariasse, N. V. Phuc, B. Konzanklewicz and W. Kuhnle, in
Surfactants in Solution, ed. K. L. Mittal and B. Lindman, Plenum
Press, New York, 1984, vol. 1, p. 565.
2
3
4
S. Tascioglu, T etrahedron, 1996, 34, 11113.
(a) C. Reichardt, in Solvents and Solvent E†ects in Organic Chem-
istry, VCH, New York, 1988, and references cited therein; (b) C.
Reichardt, Chimia, 1991, 45, 322; (c) C. Reichardt, Chem. Soc.
Rev., 1992, 147; (d) C. Reichardt, S. AsharinÈFard, A. Blum, M.
Eschner, G. Schafer and M. Wilk, Pure Appl. Chem., 1993, 65,
2593; (e) C. Reichardt, Chem. Rev., 1994, 94, 2319.
L. P. Novaki and O. A. El Seoud, Ber. Bunsen-Ges. Phys. Chem.,
1996, 100, 648.
5
6
L. P. Novaki and O. A. El Seoud, Ber. Bunsen-Ges. Phys. Chem.,
1997, 101, 105.
24 M. A. Kessler and O. S. Wolfbeis, Chem. Phys. L ipids, 1989, 50,
51.
Phys. Chem. Chem. Phys., 1999, 1, 1957È1964
1963

View Online
25 D. Banerjee, P. K. Das, S. Mondal, S. Ghosh and S. Bagchi, J.
Photochem. Photobiol. A., 1996, 98, 183.
26 R. S. Sarpal, M. Belletete and G. Durocher, J. Photochem. Photo-
biol. A., 1995, 88, 153.
33 M. H. Abraham and J. C. McGowan, Chromatographia, 1987, 23,
243.
34 S. Barr, R. R. M. Jones and J. S. Johnson, Jr., J. Phys. Chem.,
1992, 96, 5611.
27 G. P. LÏHeureux and M. Fragata, Biophys. Chem., 1988, 30, 293.
28 E. C. C. Melo, S. M. B. Costa, A. L. Macanita and H. Santos, J.
Colloid Interface Sci., 1991, 141, 439.
29 T. Handa, M. Nakagaki and K. Miyajima, Ibid, 1990, 137, 253.
30 C. Chachaty, Prog. Nucl. Magn. Reson. Spectros., 1987, 19, 183,
and references cited therein.
35 J. R. Perussi, V. E. Yushmanov, S. C. Monte, H. Imasato and M.
Tabak, Physiol. Chem. Phys. Med. NMR, 1995, 27, 1.
36 (a) N. van Meurs and G. Somsen, J. Solution Chem., 1993, 22,
427; (b) L. V. Lanshina, Russ. J. Phys. Chem., 1994, 68, 565; (c)
H. W. Bearce, G. Mulligan and M. P. Maslin, in International
Critical T ables, McGraw-Hill, New York, 1930, vol. 3, p. 115.
31 S. Possidonio, F. Siviero and O. A. El Seoud, J. Phys. Org. Chem.,
in the press.
32 P. Plieninger and H. Baumgartel, L iebigs Ann. Chem., 1983, 860.
Paper 8/09244C
1964
Phys. Chem. Chem. Phys., 1999, 1, 1957È1964