Fundamental studies on the structure and spectroscopic behavior of Phenol Blue

Article abstract of DOI:10.1021/jp992610v

The structure and spectroscopic properties of Phenol Blue [N-(4-dimethylaminophenyl)-1,4-benzoquinoneimine (4)], which can exist in principle either as a zwitterion (A) or as a quinoneimine (B), have been assessed both experimentally and theoretically usi

Full text of DOI:10.1021/jp992610v

11442
J. Phys. Chem. A 1999, 103, 11442-11450
Fundamental Studies on the Structure and Spectroscopic Behavior of Phenol Blue
John O. Morley* and Ann L. Fitton
Chemistry Department, UniVersity of Wales Swansea, Singleton Park, Swansea, SA2 8PP, United Kingdom
ReceiVed: July 27, 1999; In Final Form: October 20, 1999
The structure and spectroscopic properties of Phenol Blue [N-(4-dimethylaminophenyl)-1,4-benzoquinoneimine
(4)], which can exist in principle either as a zwitterion (A) or as a quinoneimine (B), have been assessed both
experimentally and theoretically using molecular orbital methods. 13C NMR evidence on the more soluble
diethylamino derivative (6) strongly suggests that the molecule exists purely as the quinoneimine (B) both in
protic solvents and in aprotic solvents of low and high dielectric constant. Theoretically, the solvatochromic
shift of the dye in aprotic solvents, calculated using the PM3/COSMO method, shows a good correlation
with the experimental data and arises because solvents with large dielectric constants exert a much greater
stabilizing influence on the more polar excited state than they do on the ground state. In protic solvents, the
larger bathochromic shifts observed are attributable to both a dielectric effect and a separate hydrogen bonding
contribution from the solvent. In water, Phenol Blue is predicted to form a stable trihydrate which is calculated
at the CNDOVS level of theory to absorb at a significantly longer wavelength than the unsolvated structure
in line with the experimental data.
Introduction
scales of solvent polarity1,17,18 which attribute the solvatochromic
shift to the sum of a number of effects measured by the
parameters such as π*, R, and â, where the π* parameter is a
measure of the exoergic effects of solute/solvent, dipole/dipole,
and dipole/induced dipole interactions, and the R and â
parameters are empirical measures of the ability of the bulk
solvent to act either as a hydrogen bond donor or hydrogen
bond acceptor, respectively. Other more complex scales of
solvent polarity have been described,1 but all of these are
essentially empirical and generally rely on the measured
interactions or properties of the given solvent with some other
substance.
Solvent has a pronounced effect on the visible absorption
spectra of many polar conjugated organic molecules. In some
cases, the absorption moves to shorter wavelength with increas-
ing solvent polarity, while in other cases the reverse is true.1
For example, both Brooker’s merocyanine [1-(4-aza-4-meth-
ylphenyl)-2-trans-(4-oxyphenyl)ethene (1)]2,3 and Reichardt’s
betaine [(2,4,6-triphenyl-1-pyridinium)-N-(2,6-diphenyl-4-phe-
noxide) (2)4-7 exhibit negatiVe solVatochromism with the
absorption bands moving from 620 and 731 nm in chloroform
respectively to 442 and 453 nm in water. In contrast, both Nile
Red [7-diethylamino-3H-1,2-benzophenoxazine-3-one (3)]8-10
and Phenol Blue [N-(4-dimethylaminophenyl)-1,4-benzo-
quinoneimine (4), also called indoaniline],8,11-16 exhibit positiVe
solVatochromism with the absorption moving in the opposite
direction from 484 and 552 nm in hexane to 593 and 684 nm
in water, respectively.
In contrast to this multiparameter physical chemistry ap-
proach, we have theoretically modeled the solvatochromic shift
of Brooker’s merocyanine (1) using a simple molecular orbital
method combined with a continuum solvation model, which
does not invoke a range of empirical solvent parameters.19 In
this approach, we have proposed that the solvatochromic shift
can be represented by two distinct effects. The first effect arises
from interactions between the molecule and the dielectric field
of the solvent, while the second arises from specific hydrogen
bonding interactions between the solvent and the lone pair
electrons of the exocyclic oxygen. Justification for this treatment
is provided experimentally by (a) the smooth change in the
position of the low-energy absorption maximum of (1) with
increasing dielectric constant of aprotic solvents19a and (b) the
presence in the crystal structure of two distinct water molecules
which are hydrogen bonded to the exocyclic oxygen.20 Although
the merocyanine can in principle exist in two canonical forms,
it has been universally proposed that the large negative
solvatochromic shift observed is attributable to the zwitterionic
form (1A) in polar solvents and the quinone form (1B) in
nonpolar solvents. However, our recent experimental and
theoretical studies have demonstrated that the molecule exists
mainly as a zwitterion (1A) in all solvents.19a
The large shifts observed for these molecules have been
rationalized in physical organic chemistry using multiparameter
Similar arguments have been proposed to explain the origin
of the large positive solvatochromic shift of Phenol Blue where
10.1021/jp992610v CCC: $18.00 © 1999 American Chemical Society
Published on Web 12/03/1999

Behavior of Phenol Blue
J. Phys. Chem. A, Vol. 103, No. 51, 1999 11443
it has been suggested that the contribution of the zwitterion (4A)
increases with increasing solvent polarity at the expense of the
neutral quinoneimine (4B). There have been a number of
theoretical studies on the structure and spectra of Phenol Blue
in the gas phase using the HMO,21 CNDO/2, and PPP
methods,22-23 and more recently24 using a combination of the
AM1 method25 for the structure, coupled with the INDO/S
method26 for the calculation of transition energies. In this later
study,24 the relationship of the torsion angle between the aniline
and quinone rings and the position of the absorption maximum
was explored, where it was found that the absorption shifts to
longer wavelength with increasing torsion angle. The same
procedure has been applied to calculate the spectra of 23
derivatives of indoanilines and a good correlation has been
claimed with experimental data.27 A similar coupled approach
has been used to calculate the molecular hyperpolarizability of
Phenol Blue containing either donor or acceptor groups in the
quinone ring.28 Hyperpolarizability calculations have been
reported also on two conformers of 4-aminoindoaniline (5)29
but this time using both the PM3 method30 and density
functional theory.31 The relationship between the π-conjugation
size of indoaniline dyes and the spectra has been calculated using
the INDO/S method where it has been shown that the absorption
moves to shorter wavelength with an increasing number of
aromatic rings at the quinone end of the molecule.32
SCHEME 1: Numbering Convention for the
Indoanilines 4-6
hydrogen bonding interactions between protic solvents and the
lone pair electrons at the heteroatoms of the dye have been
ignored.
The present work has been carried out to experimentally
assess the structure of the dye in solution by synthesizing the
more soluble diethylamino derivative (6) and analyzing the 1H
and 13C NMR spectra in range of protic and aprotic solvents
with varying dielectric constants in an attempt to identify the
canonical form(s) responsible for the absorption. Theoretical
calculations have been performed also on the indoaniline (4) at
several levels of theory to investigate the large effect of solvation
on the structure both in terms of the dielectric effect of the
solvent and its hydrogen bonding ability.
Methods of Calculation
Molecular orbital calculations were carried out on empirical
structures for Phenol Blue (4) using the AM125 and PM330
methods of the MOPAC 93 Program37 with full optimization
of all bond lengths, angles, and torsion angles except where
stated otherwise. The atom numbering convention used is shown
above (Scheme 1). The effect of solvents of varying dielectric
constant (ꢀ) on the structures and energies of the indoaniline
(4) were assessed at the AM1 and PM3 levels using the COSMO
method38 incorporated in the MOPAC 93 program. In this
solvation model, the solute molecule is embedded in a cavity
constructed from the intersecting van der Waals spheres of the
component atoms surrounded by a dielectric continuum of
permittivity ꢀ. The surface between the continuum and the solute
is then partitioned into a large number of segments and the
interaction between the charge density at each segment polarizes
the surrounding medium and produces a reaction field which
in turn acts on the solute. Typical solvents used in the
calculations included tetrahydrofuran (ꢀ ) 7.52), acetone (ꢀ )
21.0), and dimethyl sulfoxide (ꢀ ) 47.2) (keywords for the
calculation in tetrahydrofuran: prec pm3 ef xyz geo-ok eps )
7.52). Spectroscopic calculations were carried out on the fixed
ground state geometry using the multielectron configuration
interaction (MECI) treatment in MOPAC 93 which considers
100 configurations (50 singlets, 45 triplets, and 5 quintets)
between the three highest occupied and two lowest unoccupied
molecular orbitals (keywords for the ground state calculation
in water: pm3 prec 1scf. xyz eps)80.1 ef geo-ok meci singlet
c.i.)5 root)1 vectors), and the transition energy evaluated from
the difference between the heats of formation of the modified
ground state energy and the first excited singlet state energy
(same keywords except: root)2 and open(2,2) added). The
spectra were calculated also using the CNDOVS method, which
has been specifically developed for dyes and pigments, with
the spectroscopic constant adjusted to 0.45 and the core
coefficient of 0.33 retained for the heavy atoms.39 Molecules
and crystal structures were displayed and analyzed using the
SYBYL molecular modeling package.40
Solvent effects on the spectra of Phenol Blue have been
explored using the QCFF/SOL method, which incorporates both
permanent and induced dipoles for the solvent.33 The results
suggest that the quinoneimine (4B) is favored over the zwitterion
(4A) (which was constructed from optimized fragments rather
than the complete molecule) in both polar and nonpolar solvents,
but the predicted absorptions at 437 and 384 nm, respectively,33
are both too short by comparison with experimental data (see
later). More thorough molecular orbital calculations using the
INDO/S method coupled with a self-consistent reaction field
lead to essentially the same conclusions with the results best
interpreted if the quinoneimine is assumed to be favored over
the zwitterion in both water and in the gas phase with predicted
absorptions at 456 and 383 nm, respectively.34 However, recent
ab initio calculations on the structure of Phenol Blue, coupled
with spectroscopic calculations at the INDO/S level, are
contradictory and suggest that the zwitterion (4a) is more stable
than the quinoneimine (4b) in polar solvents, with the reverse
true in nonpolar solvents.35
A number of these studies can be criticized, however, either
because the optimized gas-phase structure used for the subse-
quent INDO/S spectroscopic calculations does not correlate
particularly well with experimental data available in the
Cambridge Structural Database,36 especially in relation to the
key C1-N7, C4-N8, N8-C9, and C12-O15 bond lengths of
the dye (Scheme 1), which will have a marked effect on the
calculated transition energies (see later) or because specific

11444 J. Phys. Chem. A, Vol. 103, No. 51, 1999
Morley and Fitton
TABLE 1: 13C NMR Data Obtained for Structure 6a
chemical shift (ppm)
solvent
C6D6
CDCl3
C5D5N
(CD3)2CO
CD3OD
ꢀ
C1
C2
C3
C4
C9
C10
C11
C12
C16
C18
2.28
4.81
13.3
21.0
33.0
47.2
147.5
148.3
149.0
149.1
149.0
148.2
149.3
111.7
111.7
112.3
112.6
111.7
111.8
112.8
126.4
127.9
127.6
128.4
127.5
127.9
127.8
131.3
131.1
131.3
131.6
153.5
152.8
153.2
153.4
151.5
151.8
153.1
139.4
138.9
139.8
139.7
138.4
138.4
139.9
142.5
142.4
142.9
143.2
187.0
188.0
188.1
187.8
187.9
186.9
188.1
44.22
44.73
44.82
45.12
44.24
44.15
45.19
12.68
12.68
12.81
12.85
11.51
12.60
12.98
(CD3)2SO
CH3CONDCH3
130.6
130.8
142.3
143.3
179.0
a Carbons at the 5, 6, 13, 14, 17, and 19 positions are essentially equivalent to those at the 3, 2, 11, 10, 16, and 18 positions. ꢀ is the dielectric
constant of the nondeuterated solvent.
TABLE 2: 1H NMR Data Obtained for Structure 6a
chemical shift (ppm)
coupling constant (Hz)
solvent
C6D6
CDCl3
C5D5N
(CD3)2CO
CD3OD
ꢀ
H2
H3
H10
H11
H13
H14
H16
H18
J2,3
J10,11
J13,14
2.28
4.81
13.3
21.0
33.0
47.2
6.37
6.73
6.79
6.70
6.82
6.79
6.74
6.98
7.07
7.28
6.94
7.12
7.06
7.15
7.18
7.33
7.34
7.16
7.35
7.30
7.37
6.50
6.65
6.62
6.43
6.60
6.57
6.61
6.47
6.57
6.56
6.38
6.58
6.53
6.58
7.20
7.38
7.46
7.30
7.46
7.38
7.43
2.86
3.44
3.30
3.37
3.50
3.44
3.48
0.82
1.23
1.08
1.07
1.23
1.16
1.20
9.86
9.10
9.86
9.12
9.10
8.88
9.11
9.83
9.86
9.86
9.87
b
9.84
b
10.05
10.15
10.06
10.19
b
10.15
b
(CD3)2SO
CH3CONDCH3
179.0
a Protons at the 5, 6, 17, and 19 positions are essentially equivalent to those at the 3, 2, 16, and 18 positions. b Peaks at the 10, 11, 13, and 14
positions in deuterated methanol are inadequately resolved to measure the coupling constants. ꢀ is the dielectric constant of the nondeuterated
solvent.
Discussion
(10), the carbon attached to the nitrogen atom resonates at 147.8
ppm in CDCl3 and those in the two ortho positions of the ring
resonate at 111.9 ppm.43a
Most of the experimental studies published on Phenol Blue
(4) have attributed the large solvatochromic shift to a distinct
structural change in moving from nonpolar to polar solvents.
Indeed, recent resonance Raman spectra of the dye in various
solvents ranging from protic solvents such as ethylene glycol
to nonpolar solvents such as ethane near their critical densities41
are highly supportive. For example, both the positions and
intensities of the CdN and CdO stretching bands show large
solvent-induced shifts which appear to correlate well with the
shift in absorption peak reflecting the dependence of the
structure on the solvent used for measurement.41 Further support
is provided also by recent theoretical calculations on Phenol
Blue at the ab initio MP2/STO-3G level,35 which have suggested
that the solvatochromic shift arises because the zwitterion (A)
is more stable in polar solvents than the quinoneimine (B), with
the reverse being true in nonpolar solvents. However, as our
experimental and theoretical studies appear to conflict with these
recent studies and with much of the established literature, we
will discuss each piece of our evidence in turn.
1. NMR Analysis. In these studies, the diethylamino deriva-
In contrast, in model compounds for the right-hand ring of
the zwitterion (A), such as the phenoxide anion, the carbon
attached to oxygen now resonates at 168.1 ppm.42c However,
there are few published spectra of model compounds for the
left-hand ring of the zwitterion with the exception of thionin
perchlorate (11),43c though it is difficult here to absolutely assign
the resonance of the carbon attached to the exocyclic nitrogen.
In aniline, the carbon attached to nitrogen resonates at 146.4
ppm in CDCl3, but on protonation to form the hemisulfate, the
resonance of the same carbon moves upfield to 139.8 ppm in
2H6 dimethyl sulfoxide.43b There are no resonances between 140
and 150 ppm in the spectrum of thionin perchlorate (11) in 2H6
dimethyl sulfoxide,43c but we assign the single distinct resonance
at 139.0 ppm to the carbon attached to the exocyclic protonated
nitrogen of the ring.
1
tive (6) of Phenol Blue was synthesized for both H and 13C
NMR analysis (Tables 1 and 2) as it was found to be much
more soluble in a wide range of solvents than the parent (4). In
the postulated change from the zwitterion (A) to the quinon-
eimine (B), the largest electronic effect would be expected at
those atoms which are directly connected to, or in the vicinity
of, either the positive nitrogen or negative oxygen atoms. 13C
NMR chemical shift data on atoms, C1, C2, and C12, therefore,
should provide a good indication of the structural changes
induced by increasing solvent polarity and/or hydrogen bonding
effects. In model compounds for the right-hand ring of the
quinoneimine (B), such as 1,4-benzoquinone (7) and 4,4-
dimethylcyclohexa-2,5-dien-1-one (8), the carbonyl carbon
resonates at 187.142a and 185.8 ppm,42b respectively, in CDCl3,
while the imino carbon of 1,4-benzoquinone dioxime (9)
resonates at 150.4 ppm in (CD3)2SO (this work). In model
compounds for the left-hand ring, such as N,N-diethylaniline
An analysis of the 13C chemical shifts for the indoaniline (6)
shows the resonances for carbon C1 of the anilino ring vary
from 147.5 to 149.0 ppm and those for carbons C2 and C6 vary

Behavior of Phenol Blue
J. Phys. Chem. A, Vol. 103, No. 51, 1999 11445
Figure 1. 1H NMR spectrum of Diethylindoaniline (6) in CDCl3 at 50 °C.
from 111.7 to 112.6 with no significant solvent-induced change
produced in aprotic solvents either on moving from a low
dielectric constant in deuterated benzene (ꢀ ) 2.28) to deuterated
acetone (ꢀ ) 21.0), dimethyl sulfoxide (ꢀ ) 47.2), or N-
methylacetamide (ꢀ ) 179), or in moving to protic solvents such
as deuterated methanol (ꢀ ) 33.0) where hydrogen bonding
effects might have been expected to contribute to a structural
change (Table 1). These chemical shifts imply that the anilino
ring of the indoaniline (6) has very similar aromatic character
to that of N,N-diethylaniline (10). Furthermore, the correspond-
ing 13C chemical shift for carbon C12 in the other ring of the
indoaniline (6) also appears to be unaffected by large changes
in the dielectric constant of the solvent used for measurement
or the presence of a hydrogen bonding donor and resonates
between 186.9 and 188.0 ppm (Table 1) and appears to be
identical in character to the carbonyl carbon in benzoquinone.
A similar picture emerges for the 13C chemical shift of C9,
which appears to be similar to the imino carbon in 1,4-
benzoquinone dioxime (9) and resonates between 151.5 and
153.5 ppm (Table 1).
6.57 ppm which we assign to protons H11 and H13 on the basis
of the shift found for the four equivalent protons in 1,4-
benzoquinone (7) which resonate at 6.83 ppm.43d There are two
further doublets in the spectrum of the indoaniline (6) centered
at 7.33 and 7.38 ppm (Figure 1) which we assign to protons
H10 and H14 on the basis of the shift found for the protons in
the poorly resolved spectrum of 1,4-benzoquinone dioxime
(11).43e However, a more detailed examination of the spectrum
of (11) in (CD3)2SO (this work) shows that both trans and cis
forms are present in solution, in a ratio of around 2:1; the two
unshielded ring protons of each form resonate at similar field
to the four equivalent protons in 1,4-benzoquinone (7) at 6.70
and 6.73 ppm, respectively, while the two other ring protons
which are shielded by the hydroxyl groups resonate at 7.17 and
7.13 ppm, respectively. An analysis of the coupling constants
of 6 show that the resonance at 7.38 ppm is clearly coupled to
that at 6.57 ppm with J13,14 ) 10.15 Hz, and the resonance at
7.33 ppm is clearly coupled to that at 6.65 ppm with J10,11
)
9.86 Hz.
The clear separation of the resonances of H11 and H13, and
H10 and H14, relative to the resonances of H2/H6 and H3/H5
which are not separated and equivalent, may arise from a number
of factors. Recent calculations using density functional theory
have claimed that there are two stable conformers of 4-ami-
noindoaniline (5),29 where the quinone ring is displaced below
the plane of the anilino ring with a torsion angle C5-C4-N8-
C9 of either 31.5 or 45.2°, and which differ in energy by only
0.5 kcal mol-1. It is possible therefore that the separate
resonances observed for H11 and H13, and those for H10 and
H14 in diethylindoaniline (6), arise because of the presence of
two distinct conformers in solution. However, this possibility
appears to be unlikely for two reasons: first, the visible
absorption spectrum of 6 shows a single band in all solvents,
whereas other conjugated molecules which are known to have
different conformations in solution, such as the merocyanine
The 1H NMR spectrum of the indoaniline (6) shows a number
of surprisingly features (Figure 1). Six distinct resonances are
present in the aromatic region corresponding to protons H2, H3,
H5, H6, H10, H11, H13, and H14 located at the C2, C3, C5,
C6, C10, C11, C13, and C14 ring positions (Scheme 1), all of
which are influenced by changes in the dielectric constant of
the solvent used for measurement (Table 2). In CDCl3, there
are two sharp absorptions centered at 6.73 and 7.07 ppm,
corresponding to four protons of an AA′BB′ system, which are
assigned to the two equivalent protons H2/H6 and the two
equivalent protons H3/H5, respectively, on the basis of the shifts
found for the ortho and meta protons in N,N-diethylaniline (10)
which resonate at 6.65 and 7.20 ppm, respectively, in the same
solvent.43a The protons in the other ring, however, are not
equivalent, with two doublet of doublets centered at 6.65 and

11446 J. Phys. Chem. A, Vol. 103, No. 51, 1999
Morley and Fitton
TABLE 3: Calculated Geometries of Phenol Blue (4) versus Crystallographic Dataa
parameter
AM1
PM3
AM1/COSMO
PM3/COSMO
AM1/COSMO#
PM3/COSMO#
TEYGOT (12)
C1-N7
1.404
1.418
1.385
1.414
1.411
1.386
1.417
1.406
1.300
1.481
1.340
1.473
1.475
1.340
1.476
1.238
1.440
117.3
117.6
114.9
115.4
124.0
117.6
-18.5
-165.9
-139.5
4.3
1.444
1.404
1.386
1.400
1.400
1.386
1.404
1.426
1.300
1.472
1.388
1.481
1.484
1.338
1.469
1.219
1.481
118.8
119.8
117.4
116.4
125.8
116.5
-25.0
-160.1
-126.4
1.8
1.410
1.418
1.385
1.415
1.411
1.387
1.415
1.404
1.300
1.484
1.342
1.469
1.470
1.342
1.476
1.248
1.450
117.5
117.8
115.0
116.2
124.8
115.5
-21.8
-161.7
-145.6
5.3
1.448
1.404
1.387
1.400
1.400
1.387
1.403
1.429
1.303
1.473
1.340
1.476
1.478
1.339
1.469
1.231
1.485
119.0
120.1
117.7
117.4
125.6
116.0
-26.1
-159.5
-126.1
1.4
1.380
1.424
1.381
1.418
1.413
1.383
1.422
1.399
1.300
1.484
1.342
1.469
1.470
1.343
1.475
1.248
1.437
117.2
117.6
114.8
115.9
125.9
120.2
0.0
1.400
1.414
1.381
1.409
1.403
1.382
1.413
1.412
1.305
1.473
1.342
1.472
1.474
1.341
1.464
1.234
1.474
118.3
118.9
117.1
117.2
131.4
121.5
0.0
1.350
1.417
1.366
1.437
1.432
1.368
1.416
1.366
1.334
1.433
1.344
1.406
1.443
1.374
1.426
1.302
1.470
117.6
116.7
117.5
120.4
125.8
121.6
-3.3
C1-C2
C2-C3
C3-C4
C4-C5
C5-C6
C1-C6
C4-N8
N8-C9
C9-C10
C10-C11
C11-C12
C12-C13
C13-C14
C9-C14
C12-O15
N7-C16
C2-C1-C6
C3-C4-C5
C10-C9-C14
C11-C12-C13
C4-N8-C9
C1-N7-C16
C2-C1-N7-C16
C2-C1-N7-C17
C3-C4-N8-C9
C4-N8-C9-C14
180.0
-152.1
6.3
180.0
-165.3
2.2
-176.2
160.6
-12.4
a Bond lengths in angstroms, angles in degrees. All COSMO results were obtained using a dielectric constant of 80.1 in the calculation. The #
sign indicates that torsion angles C2-C1-N7-C16 and C2-C1-N7-C17 were constrained during the optimization (see text). Data for structure
12 is taken from the Cambridge Structural Database (Refcode TEYGOT, ref 45).
(1), generally show a number of separate absorption bands,
observed for the ring protons of many other polar organic
molecules, which have only one formal structure42-44 and arise
because the protons are in direct contact with the solvent. The
coupling constants between protons at the 2- and 3-positions
of the molecule, J2,3, have similar values to those found in the
same positions of N,N-diethylaniline (10), while those for 10-
and 11-positions, J10,11, are comparable to those found in 4,4-
dimethylcyclohexa-2,5-dien-1-one (8). Solvent induces very little
change to the values and there appears to be no correlation with
varying dielectric constant (Table 2).
2. Structural Considerations. The Cambridge Structural
Database34 reveals a number of structures related to Phenol Blue,
but only two of these contain the 1,4-benzoquinoneimine moiety
coupled with a donor substituent in the left-hand phenyl ring.
The first of these is a calixarene derivative, 5-(4′-diethylamino-
2′-methylphenylimino)-25,26,27-trihydroxycalix(4)aren-28-
one (12)45 (R-factor ) 8.5%), which contains a ring of
2-hydroxybenzyl groups attached at both the C11 and C13
positions of the indoaniline structure as shown. This crystal
structure shows C1-N7, C4-N8, N8-C9, and C12-O15 bond
lengths of 1.350, 1.366, 1.334, and 1.302 Å, respectively (Table
3), with an estimated standard deviation in bond length between
the heavy atoms of 0.01-0.03 Å.34 The relatively long C12-
O15 bond length appears to lie between the CdO double bond
length of around 1.22 Å found in benzoquinone46 and the C-O
single bond length of around 1.33 Å found in hydrated sodium
phenoxide.47 Furthermore, the C4-N8 bond length appears to
lie between the single C-N bond length of 1.42 Å and NdC
double bond length of 1.26 Å found in the central positions of
4-nitrobenzylidene-4′-N,N-dimethylaminoaniline (13).48
particularly in solvents of low dielectric constant;19b second,
1
an analysis of the H NMR spectra of 6 at both +50 °C and
-65 °C shows no change to the relative intensities of the pair
of the absorptions either at 6.5-6.7 ppm or those at 7.3-7.4
ppm relative to the intensities found at room temperature, though
there is a small downfield shift of around 0.04-0.07 ppm with
decreasing temperature. There is no change observed either for
the resonances of the methylene or methyl protons which
resonate at 3.45 and 1.25 ppm, respectively, with decreasing
temperature. Overall, these results effectively rule out the
possibility that the aromatic resonances are due to more than
one stable conformer in solution as the relative populations
would be expected to show a marked change in moving from
-65 °C to +50 °C.
The most likely explanation for the separate resonances of
the protons in the quinoneimine ring lie with the nonplanar
character of the molecule. Because the quinoneimine ring is
twisted around the C4-N7 bond (see later), H14 points
downward and lies below the plane of the anilino ring and H10
points upward and lies above the plane so that both will
experience a deshielding effect from the π-electrons of the
anilino ring resulting in a downfield shift, with the former more
effected than the latter because it lies closer. However, H13
and H11 are too far away from the anilino ring to experience a
deshielding effect by the π-electrons and these resonances fall
in the same region as those for benzoquinone. Overall, the
variable temperature results strongly suggest that the quinone
ring is rigidly locked into one conformation only as there is no
apparent broadening of the resonances of either H10, H11, H13,
or H14 on moving from -65 to +50 °C.
The anilino ring is approximately planar with a trigonal sp2
hybridized nitrogen at N7 with the other nitrogen N8 lying in
the same ring plane. The quinoneimine ring is twisted, however,
with torsion angles C3-C4-N8-C9 and C4-N8-C9-C14 of
160.6° and -12.4° respectively (Table 3). The carbonyl oxygen,
As expected, solvent has a pronounced effect on the chemical
shifts of all the ring protons (Table 2), and although the doublet
for H2/H6 varies from 6.37 ppm in deuterated benzene to 6.82
ppm in deuterated methanol, these shifts are similar to those

Behavior of Phenol Blue
J. Phys. Chem. A, Vol. 103, No. 51, 1999 11447
O15, is strongly hydrogen bonded to the two hydroxy groups
of the 2-hydroxybenzyl substituents which form the calixarene
ring at distances of 1.661 and 1.714 Å, respectively.
data on related compounds, and furthermore, the nitrogen atom,
N7, is sp3 hybridized, with the methyl groups inclined at average
torsion angles C2-C1-N7-C16/C6-C1-N7-C17 of 22.5°
and 16.3° below the anilino ring plane for the PM3 and AM1
methods, respectively. An even larger overestimation of the C1-
N7 bond length of Phenol Blue has been reported recently at
the MP2/STO-3G level with a calculated value of 1.48 Å,
coupled with similar nitrogen hybridization problems, where
the alkyl carbons, C16 and C17, are calculated to lie 10.4° and
61.1° above and below the ring plane, respectively, at the HF/
STO-3G level.35 At the MP2/6-31G* level, there is no significant
improvement to the nitrogen hybridization, as the alkyl carbons,
C16 and C17, now lie 3.4° and 49.1° above and below the ring
plane, respectively.35 This erroneous sp3 hybridization at N7
will result in reduced conjugation between the dimethylamino
group and the left-hand ring and is probably responsible for
the relatively small calculated dipole moment of 2.7 D at the
MP2/STO-3G level,35 and our values of 4.39 and 5.32 D at the
PM3 and AM1 methods, respectively, versus the reported
experimental value of 5.80D.52
In our previous studies19a on the merocyanine (1), we found
significant improvements in the bond lengths when a dielectric
field was imposed during the structure optimization using the
COSMO routine present in MOPAC 93. Accordingly, we have
repeated the structure optimization of the Phenol Blue using
the dielectric constant of water (Table 3). At the PM3 level,
the calculated bond lengths surprisingly show few differences
from the gas-phase results with values for the C1-N7, C4-
N8, N8-C9, and C12-O15 bond lengths of 1.448, 1.429, 1.303,
and 1.231 Å. The nitrogen atom, N7, remains sp3 hybridized,
with the methyl groups now inclined at an average torsion angle
of 23.3° to the anilino ring plane. The AM1 method also
produces similar small changes over the gas-phase structure
(Table 3). The presence of solvent, therefore, has little effect
on the hybridization of the nitrogen atom, N7, and the C1-N7
bond length is still too long resulting in a substantial underes-
timation of the π-electron-donating ability of the dimethylamino
group, which would be expected to have a marked effect on
the excited state properties of the dye. We found a similar
overestimation of the length of the C-NMe2 bond length in
donor-acceptor azobenzenes which is attributable to the
parametrization adopted for the nitrogen atom in both methods.39
The second relevant crystal structure, 4-(2-hydroxy-4-meth-
ylphenylimino)-2,6-di-tert-butyl-2,5-cyclohexadien-1-one (14),49
(R-factor ) 5.2%) contains three molecules in the asymmetric
unit, with average C4-N8, N8-C9, and C12-O15 bond lengths
of 1.409, 1.302, and 1.222 Å respectively, with an angle C4-
N8-C9 of 124.0°, which are fully consistent with the quino-
neimine form (B) of the molecule. Two further related structures,
4-(2-acetoxy-4-nitrophenylimino)-2,6-di-tert-butyl-2,5-cyclo-
hexadien-1-one (15)50 and 4-(4-bromophenylimino)-2,6-di-tert-
butyl-2,5-cyclohexadien-1-one (16),51 also appear in the quino-
neimine form (B) with C4-N8, N8-C9, and C12-O15 bond
lengths of 1.402, 1.305, and 1.234 Å for the former and 1.417,
1.294, and 1.244 Å for the latter, respectively.
To overcome the hybridization problems associated with both
methods, further optimizations were carried out on Phenol Blue
in water with the structure constrained so that the dimethylamino
group was forced to lie in the plane of the anilino ring by setting
the C2-C1-N7-C16 and C6-C1-N7-C17 torsion angles to
0°. This procedure resulted in a significant improvement in the
C1-N7 bond length to give values of 1.380 and 1.400 Å for
the AM1 and PM3 methods, respectively, which are now much
closer to the expected value of 1.36-1.39 Å for the C-N bond
length of donor-acceptor arenes containing a conjugated
dialkylamino group36 such as the C-NMe2 bond in 4-nitroben-
zylidene-4′-N,N-dimethylaminoaniline (13),48 which has a value
of 1.387 Å. The other calculated distances for the C4-N8, N8-
C9, and C12-O15 bonds of 1.399, 1.300, and 1.248 Å,
respectively, at the AM1 level, and 1.412, 1.305, and 1.234 Å,
respectively, at the PM3 level, now show a very good correlation
with the average experimental data of 1.409, 1.300, and 1.233
Å, respectively, reported for the same bonds in 13, 14, and 15.
The constrained structures are both higher in energy than those
produced by complete optimization of all variables, with the
former typically 1.28 kcal mol-1 above the latter at the AM1
level.
Previous calculations on 4-aminoindoaniline (5)29 in the gas
phase using the PM3 method gave a structure with C1-N7,
C4-N8, N8-C9, and C12-O15 bond lengths of 1.425, 1.421,
1.302, and 1.219 Å, respectively. We have obtained similar
results on Phenol Blue (4) at the PM3 level where the same
bonds have values of 1.444, 1.426, 1.300, and 1.219 Å,
respectively, and also using the AM1 method, where these bonds
now have values of 1.404, 1.406, 1.300, and 1.238 Å,
respectively (Table 3). However, the C1-N7 bond length
appears to be far too long by comparison with crystallographic

11448 J. Phys. Chem. A, Vol. 103, No. 51, 1999
Morley and Fitton
TABLE 4: Effect of the Dielectric Constant of Aliphatic
Aprotic Solvents on the Experimental Absorption Maximum
of Diethylindoaniline (6) Compared with the Experimental
and Calculated Values for Phenol Blue (4)a
TABLE 5: Estimated Effect of the Dielectric Constant and
Hydrogen Bonding Ability of Acidic Solvents on the
Experimental Absorption Maximum of Diethylindoaniline
(6)a
solvent
ꢀ
λ (6)expt λ (4)expt λ (4)calc
solvent
ꢀ
λ
∆λ
∆λꢀ
∆λH
cyclohexane
diethyl ether
ethyl acetate
tetrahydrofuran
cyclohexanone
acetone
2.02
4.27
6.08
7.52
16.1
21.0
36.6
38.3
47.2
66.1
89.8
563
571
583
588
552
554
572
587
587
582
584
595
605
598
461.7
496.7
540.5
551.6
576.1
580.4
585.6
585.8
586.9
588.1
590.2
590.1
589.7
cyclohexane
chloroform
tert-butyl alcohol
propan-2-ol
cyclohexanol
ethanol
ethanediol
water
formamide
N-methylformamide
N-methylacetamide
2.02
4.81
12.5
19.9
15.0
25.3
41.4
80.1
111.0
189.0
179.0
563
609
607
615
623
635
646
684
643
612
609
46
44
52
60
72
83
121
80
49
46
13
24
29
26
32
38
46
46
46
46
33
20
23
34
40
45
75
34
3
593
597
605
614
acetonitrile
N,N-dimethylformamide
dimethyl sulfoxide
propylene carbonateb
ethylene carbonatec
N-ethylacetamide
N-methylacetamide
609
609
609
0
135.0
179.0
a ꢀ is the dielectric constant of the solvent (ref 54); λ is the long-
wavelength absorption maximum; ∆λ is the shift in absorption relative
to that found in cyclohexane; ∆λꢀ is the estimated dielectric component
of the shift; ∆λH is the hydrogen bonding component of the shift (all
in nm).
a ꢀ is the dielectric constant of the solvent (ref 54); λ is the long-
wavelength absorption maximum (in nm). Experimental data for Phenol
Blue (4) is derived from literature sources (refs 8, 11-16). b 4-Methyl-
1,3-dioxolan-2-one. c 1,3-Dioxolan-2-one.
from the dielectric effect of the solvent by interpolation of the
data given in Table 4; the shift due to hydrogen bonding,
therefore, is much larger at 76 nm. This method can be extended
to other solvents such as formamide where the dielectric
component remains at 46 nm but the hydrogen bonding
component is now smaller at 34 nm. The estimated hydrogen
bonding shifts (in nm) for protic solvents follow the expected
order with water (76) > ethanediol (45) > ethanol (40) >
cyclohexanol ≈ formamide (34) > 2-propanol (23) > tert-butyl
alcohol (20) > N-methylformamide (3) (Table 5).
3. Analysis of Solvatochromic Shifts. The spectrum of
Phenol Blue has (4) been measured in a number of aprotic
solvents, but most of these are clustered in the dielectric constant
range of around 2 (cyclohexane) to 47.2 (dimethyl sul-
foxide).8,11-16 However, the greater solubility of the diethylin-
doaniline (6) allows a more thorough analysis of the effect of
solvent on the spectroscopic shift and we have now extended
the range of experimental measurements to include less common
solvents such as ethylene carbonate, N-ethylacetamide, and
N-methylacetamide which have dielectric constants of 89.8, 135,
and 179, respectively (Table 4). Although each of the N-
alkylacetamides possesses an acidic hydrogen atom and could
be considered as protic solvents, crystallographic data on
N-methylacetamide strongly suggests that it is unlikely to form
hydrogen bonds with the oxygen of the indoaniline (6), because
in the favored trans conformation (either alone53 or in the
presence of metal salts36), the hydrogen is partly shielded by
the adjacent methyl hydrogens on the carbonyl carbon, and the
N-alkylacetamides therefore act essentially as aprotic solvents
as we have previously pointed out.19a
Overall, the effect of the dielectric constants of aprotic
solvents are considerably less marked for this molecule (6) than
the results we obtained for the merocyanine (1), with much
smaller shifts observed, this time in the opposite direction, in
moving from solvents with low dielectric constants to those with
much larger values. Thus, the absorption moves from 563 nm
in cyclohexane (ꢀ ) 2.02) to 593 nm in acetone (ꢀ ) 21.0)
through to a value of 605 nm in N,N-dimethylformamide and
then to 609 nm in ethylene carbonate (ꢀ ) 89.8) (Table 4).
However, the dielectric effect saturates fairly quickly as there
is no further shift observed in moving to N-ethylacetamide (ꢀ
) 135) or to N-methylacetamide (ꢀ ) 179). The result for
dimethyl sulfoxide (Table 4) appears to be anomalous and
possibly arises because of the presence of small amounts of
water which are very difficult to remove completely. Thus, the
maximum shift due to the dielectric effect of the solvent is
around 46 nm (Table 4).
4. Spectroscopic Calculations. A series of spectroscopic
calculations were carried out on the structure of Phenol Blue
(4) where the dimethylamino group was constrained to lie in
the anilino ring plane (as discussed above) at a range of dielectric
values using the PM3/COSMO method, and the results were
then compared with the experimental data. The transition energy
between the ground state and first excited singlet calculated
using the multielectron configuration interaction treatment
(MECI) is found to be highly dependent on the dielectric
constant of the solvent adopted for the two independent
calculations. This arises because of the large change in the
electronic properties of the molecule on excitation as indicated
by the change in the dipole moment in cyclohexane from 8.13
D in the ground state to 19.5 D in the first excited singlet state.
The larger the dielectric constant of the solvent, the greater the
stabilization of the excited state over the less polar ground state
leading to smaller energy gap between the two states. The
absorption bands in aprotic solvents of high dielectric constant
are predicted, therefore, to occur at longer wavelength (i.e.,
lower energy) than those obtained in aprotic solvents of low
dielectric constant, fully in line with the experimental results.
Thus, the calculated absorption at the PM3 level in solvents
with large dielectric constants such as propylene carbonate (ꢀ
) 66.1) at 588 nm, compares favorably with the experimental
value of 598 nm (Table 4). In solvents with very low dielectric
constants such as cyclohexane, the transition energy is over-
estimated, but nonetheless lies in the correct direction, with a
calculated value of 462 nm versus the experimental value of
552, respectively (Table 4). This overestimation almost certainly
arises in solvents of low dielectric constant because COSMO
is primarily a theory for high dielectrics designed for conductors
(ꢀ ) ∞). However, the correlation between theory and experi-
ment improves dramatically with increasing dielectric constant,
and in acetone (ꢀ ) 21.0), the values almost coincide (Table
4). The initial effective dielectric constant in the COSMO routine
In contrast, the shifts observed in protic solvent are much
larger, and arise from a combination of a distinct hydrogen
bonding component, and a dielectric component, which can be
derived in principle from the shifts observed in aprotic solvents
(Table 4). Thus, the observed absorption of the indoaniline (6)
at 684 nm in water (Table 5) represents a total shift of 121 nm
relative to cyclohexane of which 45 nm is estimated to arise

Behavior of Phenol Blue
J. Phys. Chem. A, Vol. 103, No. 51, 1999 11449
which produces a reasonable spectroscopic shift occurs at a value
of around 6.08 (Table 4).
reflects the experimental shift of 76 nm for the diethylindoaniline
(6) (Table 5). At the valence electron level, the indoaniline has
43 occupied orbitals in the ground state, and on excitation, the
first singlet excited state after configuration interaction, S1, is
found to be composed mainly of the expected π-π* HOMO-
LUMO transition but also includes a significant additional
transition from a lower occupied molecular orbital, Ψ38, to the
LUMO. If the contributions of other transitions with coefficients
smaller than 0.1 to the excited singlet state are ignored, then
The origin of the large bathochromic shifts observed with
increasing values of the dielectric constant lies with differential
stabilization of the ground and first excited state by different
solvents. In moving from cyclohexane (ꢀ ) 2.02) to ethylene
carbonate (ꢀ ) 89.6), the heat of formation of the indoaniline
in the ground state after MECI, calculated using the PM3/
COSMO method, falls from 36.97 to 27.99 kcal mol-1, to give
stabilization energy of -8.98 kcal mol-1 for the more polar
solvent. However, the corresponding heat of formation in the
same solvents for the first singlet excited state after MECI falls
from 98.90 to 76.54 kcal mol-1, respectively, giving a net
stabilization energy for the excited state of 22.36 kcal mol-1
for the more polar solvent. The change induced by the increase
in solvent polarity is reflected in the dipole moments which
change from 8.13 in the ground state to 19.5 D in the excited
state in cyclohexane and from 11.1 to 24.9 in ethylene carbonate.
Thus, solvents with large dielectric constants have a much
greater stabilizing effect on the more polar excited state of
Phenol Blue than they do on the ground state.
S1 ) 0.95ψ43-44 - 0.26ψ38-44
A similar picture emerges for the trihydrate which has 55
occupied orbitals in the ground state, to give an excited state
here of the following composition:
S1 ) 0.95ψ55-56 + 0.27ψ50-56
An analysis of the composition of both of the lower occupied
molecular orbitals of higher energy, Ψ38 and Ψ50, show that
while they are constructed from a number of π atomic orbitals,
they significantly contain large contributions from σ orbitals
on N8. For example, molecular orbital Ψ
following contributions from N7 and N8:
includes the
Similar bathochromic effects on the transition energies would
be expected from increasing dielectric constants of protic
solvents, but here the position is complicated by the formation
of hydrogen bonds with the solvent. In our previous studies on
the merocyanine (1), we found that the PM3 method gave a
much better account of hydrogen bonding of water to the lone
pair electrons at exocyclic oxygen atom than the AM1 method
with calculated O‚‚‚H2O distances of 1.77 Å for the former
versus 2.10 Å for the latter.19a In principle, while there are three
possible sites for hydrogen bonding in Phenol Blue, at N7, N8,
and O15, the sp2 lone pair electrons at the former site are
conjugated with the ring system and are less likely to form
strong hydrogen bonds with the solvent. However, the lone pair
electrons at N8 and O15 (two in this case) readily form
intermolecular hydrogen bonds with water (or any other protic
solvents), though the 13C NMR data in methanol (Table 1)
suggests that the bonding has little effect on the electronic
structure of the molecule. Optimization of a trihydrate structure
with one water molecule attached to N8 and two to O15 using
the PM3/COSMO method in water gives a resulting structure
with two O15‚‚‚H2O hydrogen bond lengths of 1.795 Å and
one N8‚‚‚H2O hydrogen bond length of 1.843 Å with O15-
H-O and N8-H-O angles of 171° and 160°, respectively. The
presence of the three intermolecular hydrogen bonded water
molecules result in very small changes to the overall geometry
of the indoaniline in line with the NMR data, with the length
of the terminal C1-N7 and C12-O15 bonds changing to 1.399
and 1.236 Å, respectively, from their values of 1.400 and 1.234
Å in the absence of the attached water molecules.
38
0.25φpz (N7) + 0.27φs (N8) + 0.72φpy (N8)
It follows that the two transitions, ψ38-44 and ψ50-56, are
essentially n-π* in character, and hydrogen bonding to the σ
orbitals at N8 will have a marked effect on the transition energy.
Overall, therefore, the large positive experimental dielectric and
hydrogen bonding shifts of around 45 and 76 nm, respectively,
found in moving from cyclohexane to water (Table 5), are well
reproduced by the calculated dielectric shift of around 50 nm
modeled by COSMO (relative to the limiting dielectric constant
of 6.08, Table 4) and the hydrogen bonding shift of 56 nm
modeled by the CNDOVS method. The failure of the MECI
routine in our version of MOPAC93 to reproduce the same
trends is due to the restricted configuration space adopted which
only considers 100 configurations (50 singlets, 45 triplets, and
5 quintets) between the three highest occupied and two lowest
unoccupied molecular orbitals.
Conclusions
NMR evidence on the soluble diethylindoaniline (6) in both
protic solvents and aprotic solvents suggests that Phenol Blue
(4) exists solely as the quinoneimine (B). These results contrast
markedly to our previous work on Brooker’s merocyanine (1)
which showed that the structure is weighted toward the
zwitterion (A). In aprotic solvents, the solvatochromic shift of
Phenol Blue, calculated using the PM3/COSMO method, shows
a good correlation with the experimental data, and arises because
solvents with large dielectric constants are able to differentially
stabilize the polar excited state. In protic solvents, such as water,
the PM3/COSMO method using limited MECI calculations
satisfactorily models the dielectric effect of the solvent, but a
larger configuration interaction treatment may be necessary to
reproduce the further bathochromic shift associated with the
formation of a stable hydrogen bonded trihydrate.
MECI calculations on the PM3 structure of the nonhydrated
indoaniline itself in water gives a predicted transition energy
of 589 nm, which is close to the experimental value of 598 nm
expected for an aprotic solvent with the same dielectric constant
as water (Table 4). Surprisingly, however, the presence of three
bonded water molecules has little effect on the predicted
transition energy of the indoaniline trihydrate which only moves
to 596 nm.
However, spectroscopic calculations at the CNDOVS level
on the PM3 structures, using a larger configuration space, but
this time with 50 singly excited configurations, give predicted
transition energies of 596 nm for the indoaniline alone, and 652
nm for the trihydrate. This shift of 56 nm, which is caused solely
by the presence of three hydrogen bonded water molecules,
Experimental
Visible spectra were measured using a Perkin-Elmer Lambda
9 UV/VIS/NIR spectrometer and a matched pair of quartz cells
of 10 mm path length. Solutions of diethylindoaniline (6) were
prepared at a concentration of 20 ppm, and recorded at room

11450 J. Phys. Chem. A, Vol. 103, No. 51, 1999
Morley and Fitton
(23) Yue, C.; Zhen-hua, Z.; Zu-guang, Y.; Su-ying, W. Ganguang Kexue
Yu Kuang Huaxue 1986, 7.
(24) Adachi, M.; Murata, Y.; Nakamura, S. J. Am. Chem. Soc. 1993,
temperature in solvents of the highest purity (spectroscopic grade
and dried by appropriate methods prior to use) where possible.
IR spectra were measured using a Perkin-Elmer FT-IR 1725X
spectrometer, using a KBr disk. 1H and 13C NMR spectra were
measured in deuterated solvents (Sigma-Aldrich) using a Bruker
AC400 spectrometer. N-methylacetamide-d was prepared as we
have described previously.19a
115, 4331.
(25) Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J. J. P. J.
Am. Chem. Soc. 1985, 107, 3902.
(26) Ridley, J. E.; Zerner, M. C. Theor. Chim. Acta 1973, 32, 111. Bacon,
A. D.; Zerner, M. C. Theor. Chim. Acta 1979, 53, 21.
(27) Adachi, M.; Murata, Y.; Nakamura, S. J. Org. Chem. 1993, 58,
5238.
N-(4-Diethyaminophenyl)-1,4-benzoquinoneimine (6). The
preparation was carried out by oxidative condensation of N,N-
diethyl-1,4-phenylenediamine and phenol using silver chloride
as the oxidizing agent according to the literature procedure.55
However, it was necessary to purify the product by chroma-
tography on a silica column using a mixture of petroleum ether
and acetone as eluent, to give green crystals with a metallic
luster after drying over P2O5 in vacuo: mp 107-108 °C; IR
(KBr) 2926, 1627, 1609, 1585, 1510, 1466, 1406, 1373, 1355,
(28) Feng, J. K.; Yu, H. S.; Li, J.; Li, Z. R. Acta Chim. Sin. 1993, 51,
639.
(29) Matsuzawa, N.; Dixon, D. A. J. Phys. Chem. 1994, 98, 11677.
(30) Stewart, J. J. P. J. Comput.Chem. 1989, 10, 209.
(31) Delley, B. J. Chem. Phys. 1990, 92, 508; 1991, 94, 7245.
(32) Adachi, M.; Murata, Y. J. Phys. Chem. A 1998, 102, 841.
(33) Luzhkov, V.; Warshel, A. J. Am. Chem. Soc. 1991, 113, 4491.
(34) Karelson, M. M.; Zerner, M. C. J. Phys. Chem. 1992, 96, 6949.
(35) Serrano, A.; Canuto, S. Int. J. Quantum Chem. 1998, 70, 745.
(36) Cambridge Structural Database; Cambridge Crystallographic Data
Centre: University Chemical Laboratory, Lensfield Road, Cambridge, CB2
2EW, UK.
1274, 1176, 1152, 1074, 872, 823 cm-1
.
(37) MOPAC 93 (J. J. P. Stewart and Fujitsu Limited, Tokyo, Japan,
copyright Fujitsu Limited, 1993) obtained from QCPE, Department of
Chemistry, Indiania University, Bloomington, IN 47405.
(38) Klamt, A.; Schuurmann, G. J. Chem. Soc., Perkin Trans. 2 1993,
799.
(39) Charlton, M. H.; Docherty, R.; McGeein, D. J.; Morley, J. O. J.
Chem. Soc., Faraday Trans. 1993, 89, 1671.
(40) Sybyl, version 6.3; Tripos, Inc.: 1699 S.Hanley Road, St. Louis,
MS 63144-2913.
(41) Yamaguchi, T.; Kimura, Y.; Hirota, N. J. Phys. Chem. A 1997,
References and Notes
(1) (a) Reichardt, C. SolVents and SolVent Effects in Organic Chemistry;
Verlag Chemie: New York, 1988 and references therein. (b) Reichardt, C.
Chem. ReV. 1994, 94, 2319.
(2) Brooker, L. G. S.; Keys, C. H.; Sprague, R. H.; Van Dyke, R. H.;
Van Zandt, E.; White, F. L.; Cressman, H. W. J.; Dent, S. G. J. Am. Chem.
Soc. 1951, 73, 5332.
(3) Brooker, L. G. S.; Keys, C. H.; Heseltine, D. W. J. Am. Chem.
Soc. 1951, 73, 5350.
(4) Dimroth, K.; Reichardt, C.; Siepmann, T.; Bohlmann, F. Liebigs
Ann. Chem. 1963, 661, 1.
101, 9050.
(42) Kalinowski, H.-O.; Berger, S.; Braun, S. Carbon-13 NMR Spec-
troscopy; John Wiley & Sons: New York, 1988; (a) p 310, (b) p 270, (c)
p 315.
(5) Dimroth, K.; Reichardt, C. Liebigs Ann. Chem. 1969, 727, 93.
(6) Reichardt, C. Liebigs Ann. Chem. 1971, 752, 64.
(7) Reichardt, C.; Harbusch-Gornert, E. Liebigs Ann. Chem. 1983, 721.
(8) Kolling, O. W. Anal. Chem. 1978, 50, 212.
1
(43) Pouchert, C. J.; Behnke, J. The Aldrich Library of 13C and H FT
NMR Spectra, 1st ed.; The Aldrich Chemical Co. Inc.: Milwaukee, WI,
1993; (a) Vol. 2, p 455; (b) Vol. 2, p 451; (c) Vol. 3, p 557; (d) Vol. 1, p
713. (e) Pouchert, C. J. The Aldrich Library of NMR Spectra, 2nd ed.; Vol.
2, p 970.
(44) Bothner-By, A. A In AdVances in Magnetic Resonance; Waugh, J.
S., Ed.; Academic Press: New York, 1965; Vol. 1.
(45) Kubo, Y.; Tokita, S.; Kojima, Y.; Osano, Y. T.; Matsuzaki, T. J.
Org. Chem. 1996, 61, 3758.
(46) van Bolhuis, F.; Kiers, C. T. Acta Crystallogr., Sect. B 1978, 34,
1015.
(47) Sielar, J.; Pink, M.; Zahn, G Z. Anorg, Allg. Chem. 1994, 620,
743.
(48) Nakia, H.; Shiro, M.; Ezumi, K.; Sakata, S.; Kubota, T. Acta
(9) Davies, M. M.; Hetzer, H. B. Anal. Chem. 1966, 38, 451.
(10) Deye, J. F.; Berger, T. A.; Anderson, A. G. Anal. Chem. 1990, 62,
615.
(11) Brooker, L. G. S.; Sprague, R. H. J. Am. Chem. Soc. 1941, 63,
3214.
(12) LeRosen, A. L.; Reid, C. E. J. Chem. Phys. 1952, 20, 233.
(13) Figueras, J. J. J. Am. Chem. Soc. 1971, 93, 3255. Figueras, J. J.;
Scullard, P. W.; Mack, A. R. J. Org. Chem. 1971, 36, 3497.
(14) Varma, C. A. G. O.; Groenen, E. J. G. Recl. TraV. Chim. Pays-Bas
1972, 91, 296.
(15) Kolling, O. W.; Goodnight, J. L. Anal. Chem. 1973, 45, 160.
Kolling, O. W.; Anal. Chem. 1981, 53, 54.
Crystallogr., Sect. B 1976, 32, 1827.
(16) Kim, S.; Johnston, K. P. ACS Symp. Ser. 1987, 329, 42. Kim, S.;
Johnston, K. P. Ind. Eng. Chem. Res. 1987, 26, 1206. Johnston, K. P.; Kim,
S. ACS Symp. Ser. 1989, 406, 52.
(49) Rekhlova, O. Yu.; Fumanova, N. G.; Olekhnovich, L. P.; Lyub-
chenko, S. N.; Simakov, V. I.; Andrianov, V. I. Kristallografiya 1991, 36,
1142.
(50) Kompan, O. E.; Simakov, V. I.; Ivakheneko, E. P.; Rekhlova, O.
Yu.; Ryskina, T. A.; Furmanove, N. G.; Struchkov, Yu. T.; Olekhnovich,
L. P. Zh. Obshch. Kim. 1992, 62, 443.
(51) Fraterman, H. A.; Romers, C. Recl. TraV. Chim. Pays-Bas 1971,
90, 364.
(52) Smyth, C. P. Dielectric BehaVior and Structure; McGraw-Hill Book
Co., Inc.: New York, NY, 1955; p 348.
(53) Hamzaoui, F.; Baert, F Acta Crystallogr., Sect. C (Cr. Str. Comm.)
(17) Krygowski, T. M.; Fawcett, W. R. J. Am. Chem. Soc. 1975, 97,
2143. Krygowski, T. M.; Fawcett, W. R. Aust. J. Chem. 1975, 28, 2115.
Krygowski, T. M.; Fawcett, W. R. Can. J. Chem. 1976, 54, 3283. Wrona,
P. K.; Krygowski, T. M.; Galus, Z. J. Phys. Org. Chem. 1991, 4, 439.
(18) Kamlet, M. J.; Taft, R. W. J. Am. Chem. Soc. 1976, 98, 377, 2886.
(19) (a) Morley, J. O.; Morley, R. M.; Docherty, R.; Charlton,M. H. J.
Am. Chem. Soc. 1997, 119, 10192. (b) Morley, J. O.; Morley, R. M.; Fitton,
A. L. J. Am. Chem. Soc. 1998, 120, 11479.
(20) .De Ridder, D. J. A.; Heijdenrijk, D.; Schenk, H.; Dommisse, R.
A.; Lemiere, G. L.; Lepoivre, J. A.; Alderweireldt, F. A. Acta Crystallogr.,
C (Cr. Str. Comm.) 1990, 46, 2197.
(21) Smith, W. F. Tetrahedron 1964, 20, 671.
(22) Hofmann, H.-J.; Hoppner, F.-D.; Weiss. C. J. Signalaufzeichnungs-
mater. 1974, 2, 97.
1994, 50, 757.
(54) See the following: Wohlfarth, C. In CRC Handbook of Chemistry
and Physics, 78th ed.; Lide, D. R., Frederiske, H. P. R., Eds.; CRC Press:
Boca Raton, FL, 1998; 6-139.
(55) Vittum, P. N.; Brown, G. H. J. Am. Chem. Soc. 1946, 68, 2235.