Solvent polarity scale on the fluorescence spectra of a dansyl monomer copolymerizable in aqueous media

Article abstract of DOI:10.1016/S0009-2614(99)00495-9

A copolymerizable fluorescent monomer N-[2-[[[5-(N,N-dimethylamino)-1-naphthalenyl]sulfonyl]-amino]ethyl]-2- propenamide (DANSAEP) was synthesized, which exhibits dual fluorescence due to the twisted intramolecular charge transfer in the excited state. The emission maximum λ_em shifts from 463.3 nm in n-hexane to 530.0 nm in water, showing solvent polarity dependence. The relations between λ_em and the conventional solvent polarity parameters E_T(30) or Z are linear, dividing solvents into protic and aprotic groups. Kamlet's linear solvation energy relationship gives a good description for λ_em as a solvent polarity scale. The increment of dipole moment Δμ at the excited state was estimated as 5.09 D with the solvatochromic analysis.

Full text of DOI:10.1016/S0009-2614(99)00495-9

2
5 June 1999
Chemical Physics Letters 307 Ž1999. 55–61
Solvent polarity scale on the fluorescence spectra of a dansyl
monomer copolymerizable in aqueous media
)
Biye Ren, Feng Gao, Zhen Tong , Yu Yan
Research Institute of Materials Science, South China UniÕersity of Technology, Guangzhou 510 641, PR China
Received 5 January 1999; in final form 28 April 1999
Abstract
A
copolymerizable fluorescent monomer N-w2-www5-ŽN,N-dimethylamino.-1-naphthalenylxsulfonylx-aminoxethylx-2-
propenamide ŽDANSAEP. was synthesized, which exhibits dual fluorescence due to the twisted intramolecular charge
transfer in the excited state. The emission maximum lem shifts from 463.3 nm in n-hexane to 530.0 nm in water, showing
solvent polarity dependence. The relations between lem and the conventional solvent polarity parameters E Ž30. or Z are
T
linear, dividing solvents into protic and aprotic groups. Kamlet’s linear solvation energy relationship gives a good
description for lem as a solvent polarity scale. The increment of dipole moment Dm at the excited state was estimated as
5
.09 D with the solvatochromic analysis. q 1999 Elsevier Science B.V. All rights reserved.
1
. Introduction
and intermolecular interactions and chain conforma-
tion. Fluorescence of polyelectrolyte is expected to
be able to reveal the changing pattern of the polarity
in the solvation layer.
Fluorescence label techniques have been widely
used in polymer research w7,8x. For our purpose, the
The finding of volume phase transition in gels
raises an important question on the solvation behav-
ior of polyelectrolytes, since all gels showing discon-
tinuous volume phase transition in response to in-
finitesimal stimulation usually contain more or less
ionizable groups on the chains w1x. For highly charged
strong polyelectrolytes in mixed organic solvents, the
composition of solvent at which the volume phase
transition occurs in crosslinked gel is found to be
similar to that inducing the corresponding linear
polyelectrolyte to precipitate from the homogeneous
solution w2–6x. This fact implies that the macro-
label should be easy to copolymerize with ionic
monomers in aqueous media and definitely represent
the polarity change around the chromophore. Pyrene
and dansyl probes are most desirable candidates
w8–10x. The pyrene scale for the solvent polarity is
well established on the relative intensity I rI w9x.
1
3
However, this probe exhibits a strong tendency to
form exciplexes in water at relatively low concentra-
tion w8x, which limits its application in polyelec-
scopic volume phase transition and precipitation ap-
pear to be caused by the change in the solvation of
the macromolecular chains, which alters the intra-
trolyte. The emission maximum of dansyl groups
covalently linked to polymer shifts with solvent po-
larity w7,10,11x. A universal relation between the
)
Corresponding author. E-mail: mcztong@scut.edu.cn
emission maximum of our dansyl monomer and the
0
009-2614r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.
PII: S0009-2614Ž99.00495-9

5
6
B. Ren et al.rChemical Physics Letters 307 (1999) 55–61
and the aqueous layer was extracted with CH Cl
2
2
Ž4= 50 ml.. The combined organic layer was dried
with MgSO and evaporated to leave pale yellow-
4
green oil. Crystallization from ethanolrwater gave
Scheme 1.
light-green needles Ž0.5653 g, mp 149–1518C, yield
1
6
0.6%.. H-NMR Ž400 MHz, CDCl .: d 8.51 Žd, 1
3
H, Js8.4 Hz, ArH.; 8.29 Žd, 1 H, Js8.4 Hz,
ArH.; 8.22 Žd, 1 H, Js8.4 Hz, ArH.; 7.51 Žm, 2 H,
ArH.; 7.15 Žd, 1 H, Js7.6 Hz, ArH.; 2.90 Žm, 8 H,
NŽCH . and SO NHCH .; and 2.68 Žm, 2 H,
medium polarity is required to interpret the polyelec-
trolyte solvation in mixed solvents.
Several empirical solvent polarity parameters have
been proposed to characterize and rank the polarity
of solvent w12x. Among the more well known are
3
2
2
2
CH NH ..
2
2
Dimroth’s E Ž30. w13,14x and Kosower’s Z scales
2.3. Preparation of N-[2-[[[5-(N,N-dimethylamino)-
1-naphthalenyl]sulfonyl]-amino]ethyl]-2-propenamide
[22] (DANSAEP)
T
w15–17x. Both are based on the transition energy of
specified compounds excited in objective solvent.
Moreover, the Kamlet–Taft linear solvation energy
relationship ŽLSER. w12,18–20x containing multiple
To a solution of DANSA Ž0.3007 g, 1.04 mmol.
in THF Ž30 ml. at room temperature was added
acryloyl chloride Ž0.085 ml, 1.04 mmol. and trieth-
ylamine Ž0.145 ml, 1.04 mmol.. The reaction was
solvatochromic parameters for individual interaction
has been successfully utilized in quantitative struc-
ture-propertyractivity relationships study.
To follow the polarity change in the solvation
layer of polyelectrolytes, we synthesized a dansyl
monomer, N-w2-www5-ŽN,N-dimethylamino.-1-naph-
thalenylxsulfonylx-aminoxethylx-2-propenamide
ŽDANSAEP. ŽScheme 1.. The relationship between
the observed emission maximum and solvent polarity
parameters was established from less-polar hexane to
water.
stirred at room temperature overnight. The salts were
then filtered and washed with THF. The combined
filtrates were evaporated, and the residue was chro-
matographed on silica with etherrhexane Ž4r1. as
eluent to give light-green solid or needle DANSAEP
1
Ž0.2790 g, m.p. 105–1078C, yield 78.2%.. H-NMR
Ž400 MHz, CDCl .: d 8.50 Žd, 1 H, Js8.4 Hz,
3
ArH.; 8.25 Žd 1 H, Js8.8 Hz, ArH.; 8.18 Žd, 1 H,
Js7.2 Hz, ArH.; 7.47 Žm, 2 H, ArH.; 7.13 Žd, 1 H,
Js7.2 Hz, ArH.; 6.60 Žbr, 1 H, CONH.; 6.25 Žbr, 1
2
. Experimental
H, SO NH.; 6.12 Žd, 1 H, Js17.2 Hz, vinyl.; 5.88
2
Ždd, 1 H, Js10 and 17 Hz, vinyl.; 5.48 Žd, 1 H,
2
.1. Materials
Js10 Hz, vinyl.; 3.35 Žm, 2 H, COCNCH .; 3.03
2
Žm, 2 H, SO NHCH .; and 2.84 Žs, 6 H, NŽCH . .
2
2
3 2
1
3
Dansyl chloride, acryloyl chloride ŽAldrich., and
C-NMR Ž400 MHz, CDCl .: d 166.3, 151.8, 134.2,
3
other reagents were all from commercial sources and
were further purified just before use according to
standard procedure w21x. All reactions were per-
130.5, 130.3, 129.7, 129.4, 129.3, 128.4, 126.5,
123.1, 118.6, 115.2, 45.3, 42.8, 39.2.
formed under the nitrogen atmosphere.
2.4. Spectral measurements
2
.2. Preparation of N-(2-aminoethyl)-5-(dimethyl-
amino)-naphthalene sulfonamide [22] (DANSA)
Fluorescence spectra of DANSAEP were recorded
y5
in 12 solvents at concentration of 1=10 molrl at
room temperature with a Perkin-Elmer LS50B Lumi-
nescence spectrometer, using an excitation wave-
length of 330 nm. UV absorption spectra were mea-
To a solution of ethylenediamine Ž2.18 ml, 32.24
mmol. in THF Ž150 ml. at 08C was added dropwise
dansyl chloride Ž0.8875 g, 3.224 mmol. in THF Ž100
ml.. The reaction was stirred at 08C for 4 h, and 1 N
KOH Ž9 ml. was added. The THF was evaporated,
1
sured with a Hitachi UV-240 spectrophotometer. H-
1
3
and C-NMR was recorded on a Bruker DRX-400
spectrometer.

B. Ren et al.rChemical Physics Letters 307 (1999) 55–61
57
dual fluorescence: One at high energy Žb) band. is
3
. Results and discussion
from the non-charge-transfer state and the other at
)
3
.1. Absorption spectra and fluorescence spectra
The absorption and fluorescence spectra of
low energy Ža band. from the twisted intramolec-
ular charge-transfer ŽTICT. state, in which the
dimethylamino group plane is nearly perpendicular
DANSAEP in several select solvents are illustrated
as examples in Fig. 1. There appear three absorption
bands in the UV spectra, being identical for all the
solutions. In most solvents the DANSAEP exhibits
to naphthalene plane w23,24x. The relative intensity
)
)
of a and b emission bands varies with the solvent
polarity due to the different quantum yields. In so
dilute solutions, the intensity of two bands varies
with the chromophore concentration without any
change in the band position.
We summarize the absorption l
cence maximum lem Ža band. of DANSAEP in 12
and fluores-
abs
)
tested solvents in Table 1 with solvent bulk proper-
ties. One can find that the lowest energy absorption
band labs is almost constant in 12 solvents while the
emission maximum lem obviously varies with sol-
vents. According to the TICT hypothesis w23,24x the
)
wavelength of a band is sensitive to the local
polarity of the medium. In water the emission maxi-
mum lem of DANSAEP is 530 nm and in less-polar
n-hexane lem is blue-shifted to ;460 nm. The fact
suggests that lem can be used as a polarity scale for
characterizing solvents. In Section 3.2 we will
demonstrate the establishment of relations between
lem and solvent polarity parameters.
There may be an argument on lem of dansyl and
its derivatives. Shea et al. w26x reported that for
5
-Ždimethylamino.-1-www pŽm.-vinyl benzylxaminox-
sulfonylxnaphthalene l is 453.9 nm in n-hexane
em
and 537.8 nm in water. Strauss and Vernewe w10x
observed that the lem of 580 nm for 1-dimethyl-
amino-5-Žb-aminoethyl.sulfonamidonaphthalene in
water was blue-shifted to 520 nm when the probe
was conjugated with the butyl polyacid. The sub-
stituent to the chromophore alters its solvation, which
in turn changes the energy of excited states. Seliskar
and Brand w27x found that lem varied from 414 to
5
44 nm for a variety of substituted 2-aminonaphtha-
lenes in water. These results indicate that each of
these dansyl molecules should be treated as an inde-
pendent polarity scale.
3
.2. The dansyl scale of solÕent polarity
In order to build the dansyl scale of DANSAEP
for ranking the solvent polarity, we correlate the lem
with some effective solvent polarity parameters. The
Fig. 1. Absorption spectra Ža. and fluorescent spectra Žb. of
DANSAEP at room temperature in solvents indicated by numbers
as the same as those in Table 1.

5
8
B. Ren et al.rChemical Physics Letters 307 (1999) 55–61
Table 1
Refractive index n, dielectric constant ´, Dimroth’s E Ž30., Kosower’s Z, the absorption l
and fluorescence maximum lem of
abs
T
DANSAEP in tested solvents at room temperature
na
´ a
b
Zc
labs
Žnm.
lem
Žnm.
No.
Solvent
E Ž30.
T
ŽJ moly1.
1
2
3
4
5
6
7
8
9
n-hexane
diethyl ether
benzene
1.372
1.35
1.89
4.20
2.27
2.21
7.58
129.3
144.8
144.4
150.6
156.5
172.0
176.6
217.2
183.3
232.3
188.3
264.3
–
–
–
–
333.7
332.5
–
463
478.5
483
488.5
489.5
496
498.5
506.5
505.6
512
1.498
1.420
1.405
1.421
1.356
1.359
1.428
1.327
1.477
1.333
dioxane
336.7
335.7
337.8
336.5
334.9
337.3
336.3
–
tetrahydrofuran
dichloromethane
acetone
ethanol
dimethylformamide
methanol
–
8.93
64.2
65.7
79.6
68.5
83.6
71.1
94.6
20.56
24.55
36.71
32.66
46.45
78.30
1
1
1
0
1
2
dimethyl sulfoxide
water
515
530.0
329.3
a
From Refs. w21,25x.
From Ref. w13x.
From Refs. w15,16x.
b
c
first attempt is based on the E Ž30., which is defined
as the transition energy of the intramolecular
charge-transfer band of pyridinium-N-phenolbetaine
beside the solvent polarity and polarizability, the
proton donating capability in forming hydrogen bond
between the solute and solvent is indispensable for
characterizing solvent polarity.
Another empirical solvent polarity parameter we
adopted is Kosower’s Z obtained from the position
of the absorption maximum of 1-ethyl-4-carbo-
T
w13,14x. Fig. 2 presents the plot of l
T
against
em
E Ž30. listed in the fifth column of Table 1. We can
see that the data are divided into two groups, one for
aprotic solvents with E Ž30. values lower than 200 J
T
y1
mol , the other for protic solvents of ethanol,
methoxypyridinium oxide w15–17x. The plots of l
em
methanol, and water. Within each group, lem in-
against Z ŽFig. 3. also manifest the similar behavior
creases linearly with E Ž30.. These results mean that
of dual linear dependence as found for E Ž30. though
T
T
Fig. 2. Plot of fluorescence emission maximum lem against
Fig. 3. Plot of fluorescence emission maximum lem vs. Kosower’s
Z for DANSAEP probe in tested solvents Žsolid lines: linear
fitting with the equations indicated..
Dimroth’s E Ž30. for DANSAEP probe in tested solvents Žsolid
T
lines: linear fitting with the equations indicated..

B. Ren et al.rChemical Physics Letters 307 (1999) 55–61
59
In view of all contributions to the free energy, the
linear solvation energy relationship ŽLSER. devel-
oped by Kamlet, Taft and co-workers w18–20x is
adopted to correlate the lem for our observed data.
The resulting regression equation is shown below:
)
lem
Ž
nm
.
s465.20y6.99V r100q40.33p
i
q16.59a q12.07b ,
Ž .
1
m
m
2
with a correlation coefficient of r s0.9865 and a
standard deviation of ss2.64. Here, V is the intrin-
i
sic Žvan der Waals. molecular volume and represents
the free energy required to separate the solvent
molecule and provide a suitably configured cavity
)
for the solute molecule. p is an index of solvent
polarityrpolarizability as a measure of the molecular
ability to stabilize a neighboring charge or dipole
through nonspecific dielectric interactions. am and
bm are the indices of hydrogen bond donor strength
Fig. 4. Plot of the observed emission maximum lem against the
cal
em
l
calculated with Eq. Ž1. for DANSAEP probe in tested sol-
vents numbered as in Table 1 Žsolid line: the diagonal..
and acceptor strength, respectively. The observed
cal
only a few points due to the lack of Z values. Dong
and Winnik w28x also reported that the pyrene scale
lem is plotted against the calculated lem in Fig. 4,
where the diagonal provides a perfect fitting to the
data. Thus, we can conclude that the lem of
DANSAEP can be used as a linear scale to rank the
solvent polarity following the LSER procedure.
was divided into the similar two groups according to
the E Ž30. or Z.
T
In fact, the E Ž30. and Z are based on the
T
experimental spectral energy level of specified com-
pounds. Therefore, the above finding addresses that
the solvent effect on the DANSAEP is different from
3.3. SolÕatochromic analysis
that on the compounds for E Ž30. and Z, especially
in the protic solvents due to the hydrogen bonding.
The intramolecular charge-transfer state can be
investigated through the difference in dipole mo-
T
Table 2
Kamlet–Taft’s multiple solvatochromic parameters, D f, and the Stokes shift DÕ of DANSAEP probe in tested solvents at room temperature
) a
am a
bm a
D f b
No.
Solvent
Vi a
Ž=10
p
DÕ
y2
Ž10 cmy1.
y3
.
1
2
3
4
5
6
7
8
9
n-hexane
diethyl ether
benzene
0.648
0.505
0.491
0.508
0.455
0.278
0.380
0.305
0.381
0.205
0.451
0.102
0
0
0
0
0
0
0.001
0.163
–
8.190
9.177
–
0.27
0.59
0.55
0.58
0.70
0.71
0.54
0.88
0.60
1.00
1.09
0.47
0.14
0.37
0.55
0.09
0.49
0.77
0.69
0.62
0.76
0.18
dioxane
0.021
0.210
0.218
0.285
0.290
0.282
0.309
–
9.229
9.359
9.442
9.658
10.116
9.869
10.204
–
tetrahydrofuran
dichloromethane
acetone
ethanol
dimethylformamide
methanol
0
0.10
0.09
0.83
0
0.93
0
1
1
1
0
1
2
dimethyl sulfoxide
water
1.17
0.320
11.499
a
From Refs. w12,29x or calculated according to Ref. w29x.
Obtained according to Eq. Ž3..
b

6
0
B. Ren et al.rChemical Physics Letters 307 (1999) 55–61
ments between the first excited singlet state mes and
the ground state m , i.e., Dmsm ym . The Stokes
g
es
g
shift ŽDÕ. of the excited state can be determined by
the solvatochromic analysis proposed by Lippert w29x
and Mataga w30x as follows w31x:
DÕsÕabs yÕem
1
2
2
s 4
< m ym < D fqconst. ,
Ž
2
.
3
es
g
p´ hca
0
with
2
´
y1
n y1
D fs 2
y
,
Ž
3
.
2
´q1
2n q1
Fig. 5. Plot of the Stokes shift DÕ vs. the solvent polarity
Ž .
solvents numbered as in Table 1 Žsolid line: linear fitting..
where Õabs and Õem is the wavenumber of the lowest
energy absorption maximum and the fluorescence
maximum, respectively. ´0 and ´ are the dielectric
constant of vacuum and solvent, respectively. n is
the refractive index of the solvent, a the radius of the
Onsager cavity for the solute molecule. h and c are
Planck’s constant and the light velocity, respectively.
Using the values in Tables 1 and 2, we plot DÕ
against D f in Fig. 5. A reasonable linearity is ob-
served from our data except those of water ŽNo. 12.
and dioxane ŽNo. 4.. For water the deviation may be
attributed to the strong hydrogen bonding interaction
which interrupts the reorientation of water molecules
around the probe. While for dioxane, a similar phe-
nomenon was also observed by Horng w32x, who
regarded this as the ‘local’ solvation effect.
The Dm can be estimated from the slop of the
straight line Ž6.1743. in Fig. 5 when the Onsager
parameters D f according to Eq.
2 for DANSAEP probe in
varies with solvents as discussed in Section 3.2.
Seliskar and Brand considered that such a large
change in Dm of 2,6-DNS was due to the fact that
the solvent relaxation could not occur during the
lifetime of the first excited singlet state. They ex-
pected the increase in the probability of intersystem
crossing and appearance of the triplet state induced
by the solvent relaxation. Li et al. w35x argued with
experimental results that intersystem crossing was
relatively unimportant and the 1-Ždimethylamino.-5-
naphthalenesulfonate Ž1,5-DNS, the same as our
starting compound. spectrum was much more sensi-
tive to solvent effects than that of 2,6-DNS.
cavity radius a is determined. The simplest way is to
3
assume a s3Vr4p, where V is the van der Waals
volume of the probe molecule w32x and determined
by the simple additivity of the atomic volume w25x.
Acknowledgements
The value of as0.353 nm so obtained yields a
dipole moment change of Dms5.19 D. The other
approach is to take a as 40% of the long axis of the
chromophore part of the DANSAEP molecule w33x,
The financial support to this work by the National
Natural Science Foundation of China ŽNo. 29725411,
9804003. is gratefully acknowledged.
2
giving as0.366 nm and Dms5.48 D.
These data are close to Dms5.78 D for dansyl-
References
monoaza-18-crown-6 w34x but quite different from
Dms19 D for N,N-dimethyl-2-aminonaphthalene-
w1x M. Shibayama, T. Takana, Adv. Polym. Sci. 109 Ž1993. 1,
6
-sulfonate Ž2,6-DNS, as0.4 nm. w27x, the latter
and references therein.
was determined from ethanolrwater mixtures. In the
w2x Z. Tong, X. Liu, Macromolecules 27 Ž1994. 844.
X. Liu, Z. Tong, O. Hu, Macromolecules 28 1995 3813.
w4x X. Lui, Z. Tong, X. Cao, O. Hu, Polymer 37 Ž1996. 5947.
methodology behind Eq. Ž2., Dm is assumed to be
w x
3
Ž
.
independent of the solvent polarity, but in fact lem

B. Ren et al.rChemical Physics Letters 307 (1999) 55–61
61
w5x X. Lui, O. Hu, Z. Tong, J. Polym. Sci. B: Polym. Phys. 35
Ž1997. 1433.
w21x N. Chen, S. Hu, Solvent Handbook, Chemical Engineering
Press, Beijing, 1987 Žin Chinese..
w6x X. Lui, Z. Tong, F. Gao, Polym. Int. 47 Ž1998. 215.
w22x K.J. Shea, G.J. Stoddard, D.M. Shavelle, F. Wakui, R.M.
w7x T. Torii, T. Yamashita, K. Horie, Biopolymers 34 Ž1994.
Choate, Macromolecules 23 Ž1990. 449.
1
01.
w23x K. Rotkiewicz, K. Grellmann, Z.R. Grabowski, Chem. Phys.
Lett. 19 Ž1973. 315.
w24x Z.R. Grabowski, K. Rotkiewicz, A. Seimiarczuk, D.J. Cow-
ley, W. Baumann, Nouv. J. Chim. 3 Ž1979. 443.
w25x J.P. Hickey, D.R. Passino-Reader, Environ. Sci. Technol. 25
Ž1991. 1753.
w8x F.M. Winnik, Chem. Rev. 93 Ž1993. 587.
w9x K. Kalyanasundaram, J.K. Thomas, J. Am. Chem. Soc. 99
Ž1977. 2039.
w10x U.P. Strauss, G. Vesnaver, J. Phys. Chem. 79 Ž1975. 2426.
w11x U.P. Strauss, M.S. Schlesinger, J. Phys. Chem. 82 Ž1978.
1
627.
w26x K.J. Shea, D.Y. Sasaki, G.J. Stoddard, Macromolecules 22
w12x M. Kamlet, Prog. Phys. Org. Chem. 13 Ž1982. 485, and
Ž1989. 1722.
references therein.
w27x C.J. Seliskar, L. Brand, J. Am. Chem. Soc. 93 Ž1971. 5414.
w28x D.C. Dong, M.A. Winnik, Can. J. Chem. 62 Ž1984. 2560.
w29x E. Lippert, Z. Naturforsch. A 10 Ž1955. 541.
w30x N. Mataga, Bull. Chem. Soc. Jpn. 36 Ž1963. 654.
w31x Y.V. Il’Ichev, W. Kuhnle, K.A. Zachariasse, Chem. Phys.
211 Ž1996. 441.
w13x K. Dimroth, C. Reichardt, T. Siepmann, F. Bohlmann, Liebigs
Ann. Chem. 661 Ž1963. 1.
w14x K. Dimroth, C. Reichardt, Liebigs Ann. Chem. 727 Ž1969.
9
3.
w15x E.M. Kosower, J. Am. Chem. Soc. 30 Ž1958. 3253.
w16x E.M. Kosower, J. Am. Chem. Soc. 30 Ž1958. 3261.
w17x E.M. Kosower, J. Am. Chem. Soc. 30 Ž1958. 3267.
w18x M.J. Kamlet, J.L. Abboud, R.W. Taft, J. Am. Chem. Soc. 99
Ž1977. 6027.
w32x M.L. Horng, J.A. Gardecki, A. Papazyan, M. Maroncelli, J.
Phys. Chem. 99 Ž1995. 17311.
w33x J.F. Letard, R. Lapouyade, W. Rettig, Pure Appl. Chem. 65
Ž1993. 1705.
w19x M.J. Kamlet, J.L. Abboud, R.W. Taft, J. Am. Chem. Soc. 99
w34x Y. Wu, L. Li, A. Tong, Chin. Sci. Tech. Lett. 4 Ž1998. 196,
Ž1977. 8325.
in Chinese.
w20x M.J. Kamlet, T.N. Hall, J. Boykin, R.W. Taft, J. Chem. Soc.
Perkin Trans. 2 Ž1979. 2559.
w35x Y.-H. Li, L.-M. Chan, L. Tyer, R.T. Moody, C.M. Himel,
D.M. Hercules, J. Am. Chem. Soc. 97 Ž1975. 3118.