Conformationally controlled intramolecular charge transfer complexes

Article abstract of DOI:10.1021/jo020164t

Trans-1-acceptor-2-donor-substituted cyclohexanes (1), as well as their 4- (or 5-)methyl-substituted homologues (2), have been prepared and are shown to form intramolecular charge-transfer (donor-acceptor) complexes. These weak complexes are turned on and off by the chair-chair interconversion of the cyclohexane ring. The CT absorptions have been measured and the equilibrium constants for the ring reversal have been determined by UV/vis spectroscopy at 298 K, as well as by NMR spectroscopy at two temperatures: at 183 K, by direct comparison of signals due to the two chair conformations, and at 300 K, by comparison of calculated and measured widths of the α-proton signals. The Gibbs free energies assigned to the donor-acceptor interactions range between 0 and -1 kcal mol^-1. A crystal structure of one of the complexes (1b) confirms the intramolecular donor-acceptor alignment and interaction. The regioisomers of the methyl-substituted complexes were characterized by NOE interaction between the methyl and an α-proton cis to it.

Full text of DOI:10.1021/jo020164t

Con for m a tion a lly Con tr olled In tr a m olecu la r Ch a r ge Tr a n sfer
Com p lexes
Daniel Kost,*,† Na’ama Peor,† Gali Sod-Moriah,† Yifat Sharabi,† David T. Durocher,‡ and
Morton Raban‡
Department of Chemistry, Ben Gurion University of the Negev, Beer-Sheva 84105, Israel, and
Department of Chemistry, Wayne State University, Detroit, Michigan 48202
kostd@bgumail.bgu.ac.il
Received March 8, 2002
Trans-1-acceptor-2-donor-substituted cyclohexanes (1), as well as their 4- (or 5-)methyl-substituted
homologues (2), have been prepared and are shown to form intramolecular charge-transfer (donor-
acceptor) complexes. These weak complexes are turned on and off by the chair-chair interconversion
of the cyclohexane ring. The CT absorptions have been measured and the equilibrium constants
for the ring reversal have been determined by UV/vis spectroscopy at 298 K, as well as by NMR
spectroscopy at two temperatures: at 183 K, by direct comparison of signals due to the two chair
conformations, and at 300 K, by comparison of calculated and measured widths of the R-proton
signals. The Gibbs free energies assigned to the donor-acceptor interactions range between 0 and
-1 kcal mol-1. A crystal structure of one of the complexes (1b) confirms the intramolecular donor-
acceptor alignment and interaction. The regioisomers of the methyl-substituted complexes were
characterized by NOE interaction between the methyl and an R-proton cis to it.
In tr od u ction
In the present paper we describe the preparation and
spectroscopic study of a series of analogous trans-1-
acceptor-2-donor-cyclohexane (1) molecules, and their
4-(or 5-)methyl-substituted analogues (2, 2′). UV/vis and
NMR spectroscopies are utilized to measure the CT
complex strengths and spectroscopic properties.
Intermolecular charge transfer (CT) complexes arising
from the interaction of aromatic π-donor and acceptor
molecules have been studied extensively.1 Far fewer
intramolecular CT analogues have been reported, pri-
marily of the cyclophane type, in which the donor and
acceptor portions are locked together in a rather rigid
arrangement.2 More flexible intramolecular CT com-
plexes have been reported in a recent communication,
whereby a cyclohexane skeleton is substituted at adjacent
trans positions with aromatic donor and acceptor groups
(1).3 The appearance of a weak new absorption band in
The cyclohexane ring system is known to undergo rapid
chair-chair interconversion (eq 1). This dynamic process
causes the donor and acceptor moieties to be either
relatively far from each other in the diaxial geometry
(1a a ) or near each other in the diequatorial conformation
(1ee). In the former, the substantial distance between
the donor and acceptor rings prevents CT interaction and
formation of an intramolecular complex; in the latter,
however, the inter-ring distance is small enough and the
geometry favorable to permit the formation of a CT
complex.4
After the first demonstration of conformationally con-
trolled intramolecular complexes,3 it became of interest
to evaluate the complex strengths, measure the molar
extinction coefficients, and study possible substituent
effects on complex properties: strength, λmax, and ꢀ. The
long-range goal of this project is to develop on-off-
the UV/vis spectra of these materials, relative to the
spectra of the separate donor and acceptor groups,
provided evidence for the formation of an intramolecular
CT complex.
† Ben Gurion University.
‡ Wayne State University.
(1) (a) Foster, A. Organic Charge-Transfer Complexes; Academic
Press: New York, 1969. (b) Mulliken, R. S.; Person W. B. Molecular
Complexes; Wiely, New York, 1969. (c) Abdel-Kader, M. H.; Issa, R.
M.; Ayad, M. M.; Abdel-Mottaleb, M. S. Z. Naturforsch 1984, 39a, 1274.
(d) Bundi, M. L.; J ayadevappa, E. S. Spectrochim. Acta 1988, 44A, 607.
(2) (a) Cram, D. J .; Reeves, R. A. J . Am. Chem. Soc. 1958, 80, 3094.
(b) Cram, D. J .; Bauer, R. H. J . Am. Chem. Soc. 1959, 81, 5971. (c)
Cram, D. J .; Wilkinson, D. L. J . Am. Chem. Soc. 1960, 82, 5721. (d)
Staab, H. A.; Krieger, C.; Wahl, P.; Kay, K.-Y. Chem. Ber. 1987, 120,
551.
(4) This assessment has been tested and verified recently by a
theoretical study: Kost, D.; Frailich, M. Theochem J . Mol. Struct. 1997,
399, 265.
(3) Preliminary results were communicated: Raban, M.; Durocher,
D. T. Tetrahedron Lett. 1990, 31, 5125.
10.1021/jo020164t CCC: $22.00 © 2002 American Chemical Society
Published on Web 09/07/2002
6938
J . Org. Chem. 2002, 67, 6938-6943

Intramolecular Charge Transfer Complexes
CHART 1
SCHEME 1. Syn th etic Rou te for CT Com p lexes
(Sh ow n for Ben zen e-Der ived Don or s)
switchable CT complexes that will function as color
indicators for molecular switches. The requirements from
such CT complexes are that their absorptions fall well
within the visible range and at significant absorptivities.
and hence
∆Gcomp ) ∆Geq + AOPh + AOCOPh - AMe ) ∆Geq
+
0.65 + 0.5 - 1.74 ) ∆Geq - 0.59 (4)
Resu lts a n d Discu ssion
Meth od s
(a ) Syn th esis. The general synthetic route leading to
all of the complexes is shown in Scheme 1. Epoxidation
using three different reagent systems (perbenzoic acid
freshly prepared from benzoyl peroxide,6 hydrogen per-
oxide in the presence of trichloroacetonitrile,7 and N-
bromosuccinimide8) proceeded with essentially the same
results: yields between 60 and 75% of a cis and trans
mixture (for series 2) of 4-methylcyclohexene oxide. The
mixtures were further reacted with substituted pheno-
lates or thiophenolates, producing nearly equal amounts
of the regioisomers (4 and 4′) resulting from axial attack
on the epoxides. The overall 1,2-trans (in 3, 4, and 4′)
and 1,4-cis (in 4, 2,5-cis in 4′) stereochemistry was
obtained exclusively in all cases.
The mixtures of regioisomer pairs 4 and 4′ were
separated to their individual components in three cases
(4a , 4′a ; 4b, 4′b; and 4d , 4′d ) by fractional crystallization
from ethyl acetate/n-hexane mixtures. All other cyclo-
hexanols 4 were further used as mixtures, from which
regioisomeric CT complex mixtures, 2 and 2′, were
obtained and studied as mixtures. Conversion of 3, 4, and
4′ to 1, 2, and 2′ proceeded smoothly by reaction with
3,5-dinitrobenzoyl chloride.
The chair-chair equilibrium reaction of the cyclohex-
ane ring can be utilized to study the complex strength,
by comparison between complexes of type 1 with their
methyl-substituted analogues 2 (Chart 1). In series 1, the
equilibrium is shifted predominantly toward the 1ee
conformation, because both of the factors present, steric
and CT interaction, favor this conformation. In dilute
solution, CT interaction is essentially all intramolecular
(as will be shown below), and hence, the stoichiometric
concentrations represent actual CT complex concentra-
tions. The molar extinction coefficients of the CT com-
plexes can therefore be evaluated directly from the
absorbances and stoichiometric concentrations.
The introduction of a methyl substituent at the 4-posi-
tion (2) strongly affects the equilibrium constant. The
substantial preference of the methyl group to occupy an
equatorial position partly offsets both the steric prefer-
ence of the donor and acceptor substituents to occupy
equatorial positions and the CT interaction. The equi-
librium constant Keq, which can be measured either by
electronic or by NMR spectroscopy, can lead to the
assessment of ∆Gcomp, the free energy of the CT interac-
tion, as follows:
(b) Ch a r a cter iza tion of Regioisom er s. Identifica-
tion of the regioisomers 2 and 2′ was accomplished by
observation of a nuclear Overhauser enhancement (NOE)
between the methyl group and the R-proton in position
3 relative to it. Examination of the conformations in
Scheme 2 reveals that in the 2ee conformation the
protons in the R-position to the oxygens assume axial
positions. In this conformation, the axial methyl group
∆Geq ) -RT ln Keq
(2)
The free energy ∆Geq is the sum of the free energy
components resulting from the steric preferences of the
three substituents and the complex-binding term ∆Gcomp
(eqs 3 and 4). The steric free energies are simply the
steric A values for methyl, phenoxy, and benzoyloxy
groups.5
(5) Eliel, E. L.; Wilen, S. H. Stereochemistry of Organic Compounds;
Wiley: New York 1994; p 695.
(6) Braun, G. Org. Synth. Coll. Vol I, 431.
(7) Arias, L. A.; Adkins, S.; Nagel, C. J .; Bach, R. D. J . Org. Chem.
1983, 48, 888.
(8) Craig, T. W.; Harvey, G. R.; Berchtold, G. A. J . Org. Chem. 1967,
32, 3743.
∆Geq ) ∆Gcomp + ∆GOPh + ∆GOCOPh - ∆GMe
)
∆Gcomp - AOPh - AOCOPh + AMe (3)
J . Org. Chem, Vol. 67, No. 20, 2002 6939

Kost et al.
SCHEME 2. P r oxim ity of r-Hyd r ogen s to th e
if ee were the only conformation present, the R-protons
would be axial and each one would be coupled to two axial
and one equatorial neighbors. The maximum expected
width of this multiplet would be 26.5 Hz, based on the
sum of the three coupling constants, calculated for ideal
180° and 60° dihedral angles and taking into account the
different substitution patterns for each proton [calculated:
Meth yl Gr ou p in 2a a a n d 2eea
3J (HRHR) ) 12.4 Hz, J (HRHâc) ) 2.8 Hz, J (HRHât) )
11.3 Hz; these values are not significantly different for
the sulfur-substituted compound 2f].9 In any other equi-
librium situation the weighted average of ee and a a
conformations would lead to significantly smaller cou-
pling constants. Indeed, in all of the compounds of series
1 the R-proton multiplets are 24.0-25.5 Hz wide, corre-
sponding to nearly exclusive occupation of the ee con-
formation, and thus justifying the assumptions.
3
3
a
NOE expected only in 2ee, between the methyl and the proton
at position 3 to it.
In accord with this analysis, the widths of the R-proton
signals of compounds of series 2 are substantially smaller
(Figure 1), reflecting the much greater proportion of the
a a conformation in these compounds due to the effect of
the methyl group. For the pure a a conformation the three
calculated coupling constants are 2.7-2.8 Hz,9 and hence,
the R-protons should give rise to a quartet of total width
8.2 Hz. The actual spectra for all but one pair of the
compounds (2f and 2′f) feature broad quartets or unre-
solvable multiplets for the R-protons, with a total width
ranging between 13.1 and 14.5 Hz. From these band-
widths (W) the conformational equilibrium can be calcu-
lated using eq 5 (Wee, J ee, Xee correspond to the calculated
width of the quartet, calculated coupling constant, and
mole fraction, respectively, of the ee conformation):
F IGURE 1. The R-proton region of the 200 MHz 1H NMR
spectra of 1b (left) and 2b (right).
ee
ee
W ) J 1 + J 2 + J 3 ) (J 1 + J 2 + J 3ee)Xee
+
is very close to the axial proton cis to it, in position 3
relative to the methyl. Consequently, selective low-power
irradiation at the methyl resonance frequency should
produce a NOE of the R-proton in the 3-position. This is
borne out by NOE difference experiments (shown for
complex 2′b in the Supporting Information). The differ-
ence spectra show significant (4.4% for 2′b) enhancement
of the low-field (5.4 ppm) R-proton resonance, and proves
that the acceptor group in this compound is in the
3-position relative to the methyl. In the corresponding
regioisomer 2b, in which the acceptor is in the 4-position
relative to methyl, the NOE experiment results in a
similar enhancement of the high-field proton R to the
donor group.
This experiment was also run for the other pairs of
isolated regioisomers 2a , 2′a and 2d , 2′d . It was con-
cluded that in the 2 regioisomers the R-H resonances
were the “inner” signals out of the total four (for both
regioisomers), while the 2′ isomers gave rise to the “outer”
signals (see the Supporting Information for sample NMR
spectra).
aa
aa
(J 1 + J 2 + J 3aa)(1 - Xee)
W - Waa
Wee - Waa
Xee
)
(5)
Substitution of the sums of calculated coupling con-
stants in each conformation for Wee and Waa, respectively,
and the measured spectral widths for W in eq 5 results
in a range of Xee ) 0.26-0.34, corresponding to 0.35 <
Keq < 0.52. It is concluded that, on the basis of coupling
constants to the R-protons and on the modified Karplus
equation, in all of the complexes 2 (except 2f) the
population of the uncomplexed conformation (a a ) slightly
exceeds that of the intramolecular CT complex (ee)
conformation. In all of these complexes the width of the
proton signal R to the donor fell within 1 Hz of the width
of the proton signal R to the acceptor. Also, in both
regioisomers (2 and 2′) these widths were equal, within
experimental error.
For the pair 2f, 2′f the width of the R-multiplet (a
(c) Con for m a tion a l An a lysis. The method is based,
as mentioned earlier, on the assumption that the con-
formational equilibrium in series 1 is shifted predomi-
nantly toward the 1ee conformation. Support for this
3
3
doublet of triplets, J (d) ) 3.7 Hz, J (t) ) 6.8 Hz) was
17.3 Hz. Using eq 5 with this value yields Xee ) 0.48,
i.e., Keq ) 1, corresponding to equal populations of the
two conformations.
1
initial expectation is found in the H NMR signals of the
R-protons in series 1. Figure 1 shows the expanded region
1
of the H NMR spectrum for 1b. Each R-proton has three
(9) The modified Karplus data are based on the following: Altona,
C.; Francke, R.; Dehaan, R.; Ippel, J . H.; Daalmans, G. J .; Hoekzema,
A. J . A. W.; Vanwijk, J . Magn. Reson. Chem. 1994, 32, 670, and
references therein. We used a program for calculating 3J values written
by J . W. Blunt, University of Canterbury.
vicinal neighboring protons that are spin-coupled to it.
The magnitudes of the coupling constants depend on the
molecular conformation in a Karplus-type relationship:
6940 J . Org. Chem., Vol. 67, No. 20, 2002

Intramolecular Charge Transfer Complexes
TABLE 1. UV/Vis Absor p tion Ma xim a a n d Extin ction
Coefficien ts for In tr a m olecu la r CT Com p lexes 1 in
CH2Cl2 Solu tion s
complex
λ λ
max, nm ꢀ, M-1 cm-1 complex max, nm ꢀ, M-1 cm-1
1a
1b
1c
1e
380
388
382
434
60 ( 3
72 ( 4
68 ( 3
227 ( 5
1f
1g
1h
480
366
370
410 ( 5
166 ( 7
244 ( 7
for the corresponding complexes of series 2 to represent
the absorptions of the 2ee conformation.
F IGURE 2. UV/vis absorption spectra of (a) 3b (donor) and
(b) ethyl 3,5-dinitrobenzoate (acceptor), 0.02 M of each in
dichloromethane solution, and their mixture (c) resulting in
CT complex 1b; (d) the corresponding difference spectrum:
donor and acceptor spectra subtracted from CT absorption (d
) c - a - b).
The UV/vis spectra of series 2 served to evaluate the
a a a ee equilibrium constants, Keq. The absorbances of
complexes 2 were substantially smaller than those of the
corresponding 1 at equal concentrations, due to the effect
of the methyl group in 2 to decrease the population of
the ee conformation. Equilibrium constants, free energies
of the conformational equilibrium, and the resulting
binding free energies of the complexes (∆Gcomp) were
determined as detailed in the Methods section and are
listed in Table 2. The data in Table 2 show that in all of
the complexes Keq is smaller than 1, so that the CT
interaction is not sufficiently strong to override the
methyl steric preference. The only exception to this is 2g,
for which Keq is significantly greater, although the donor
group is not stronger than for other members of the
series. This is probably the result of difficulties in
obtaining accurate difference spectra for 2g, since the
overlap between absorption bands was more severe. In
fact, for 2g the λmax value shifted with concentration, as
a result of poor subtraction.
The UV/vis data in Tables 1 and 2 lead to the following
conclusions: (a) in all the compounds 1 and 2 an
intramolecular CT interaction is evident from the char-
acteristic absorption and the associated color of the
substance. (b) The complex binding energies are small
but significant, generally between -0.2 and -0.5 kcal
mol-1. (c) Due to low resolution and difficulties in
subtraction of the spectra, the error associated with this
method may be substantial. (d) There is very good
agreement between equilibrium constants measured by
the UV/vis method and those estimated from the widths
of the NMR signals of the R-protons (Keq ranging between
0.35 and 0.52, except for 2f, for which Keq was 1.0). (e)
The CT complex strengths do not depend critically on the
donor strengths: all of the complexes fall within the same
strength range.
F IGURE 3. The absorbance of 1b at λmax in dichloromethane
plotted against molar concentration.
(d ) UV/Vis Sp ectr oscop y. The CT absorption bands
of the complexes of series 1 and 2 in the UV/vis spectra
were measured as difference spectra, since there was
substantial overlap between the relatively weak CT
absorption and the strong end-absorptions of the donor
and particularly the acceptor molecules (Figure 2). The
difference was taken by subtracting the absorbances of
ethyl 3,5-dinitrobenzoate and the appropriate donor
molecule 3 from the absorption of the complex solution
in the same solvent and concentration.
Determination of the extinction coefficients (ꢀ) of the
CT absorptions is based on the presumption that in
compounds 1 the molecules are predominantly in the 1ee
conformation, and hence, CT concentrations can be
expressed in terms of stoichiometric concentrations.
Evidence to this effect was provided by the analysis of
the NMR signals of the R-protons of compounds 1, and
additional evidence is presented in Figure 3: a linear
correlation is obtained between the absorbance of each
complex 1 and the concentration; i.e., within the given
concentration range the CT absorptions obey Beer’s law.
It is therefore evident that no intermolecular complex-
ation is present (which would be expected to depend on
the concentration squared), and hence, all the absorbance
is due to intramolecular complexation.
The NMR results suggested that 2f was a stronger CT
complex than other compounds 2. This is not borne out
by the UV/vis results, possibly because of the inaccuracy
of the latter, and remains an open question.
(e) Low -Tem p er a tu r e NMR Sp ectr a . The two in-
terconverting conformers ee and a a for each compound
2 and 2′ can be observed directly by NMR spectroscopy
at low temperature, at which the exchange process is
sufficiently slow with respect to the NMR time scale. An
example of such a spectrum at 183 K, at which the
exchange process has “frozen”, is shown in Figure 4.
Comparison of the various signal-pair intensities (by
electronic integration as well as by cutting out and
weighing the signals) provides the population ratio of the
two conformers and hence the equilibrium constant. The
Keq values obtained in this manner for compounds 2 and
2′ are listed in Table 2. The best resolution to major and
As a consequence, the ꢀ values were evaluated for
series 1 in a straightforward manner from the absor-
bances and concentrations and are listed in Table 1. It
is assumed that in the ee conformation of both com-
pounds 1 and 2 the CT interactions are very similar, and
hence, the ꢀ values measured for series 1 have been used
J . Org. Chem, Vol. 67, No. 20, 2002 6941

Kost et al.
TABLE 2. 2a a a 2ee Equ ilibr iu m Con sta n ts, F r ee En er gies, a n d CT-Bin d in g F r ee En er gies for Ser ies 2 fr om UV/Vis
sp ectr a a t 298 K (CH2Cl2) a n d fr om 1H NMR Sp ectr a a t 183 K (CD2Cl2)
UV/vis data at 298 K
NMR data at 183 K
enthalpy and entropyb
∆G°,
∆Gcomp
,
∆G°,
∆H°,
∆S°,
complex
Keq ) ee/aa
kcal mol-1
kcal mol-1
Keq ) ee/aa
kcal mol-1
kcal mol-1
cal mol-1 deg-1
2a
2′a
2b
2′b
2c + 2′c
2d
2′d
2e + 2′e
2f + 2′fa
2g + 2′g
0.68
0.58
0.65
0.66
0.62
0.23
0.32
0.25
0.25
0.28
-0.36
-0.27
-0.34
-0.34
-0.31
0.09 ( 0.02
0.10 ( 0.01
0.13 ( 0.01
0.13 ( 0.01
0.88
0.84
0.74
0.74
1.9
1.7
1.5
1.5
5.6
4.5
4.2
4.3
0.10 ( 0.02
0.10 ( 0.02
0.15 ( 0.02
0.84
0.84
0.69
0.33
0.50
1.88
0.66
0.41
-0.37
0.07
-0.18
-0.96
0.7
0.3
2g: 0.46 ( 0.03
2′g: 0.31 ( 0.05
2h : 0.36 ( 0.02
2′h : 0.25 ( 0.03
0.28
0.43
0.37
0.50
1.3
1.7
0.9
1.2
5.7
6.9
2.7
3.9
2h + 2′h
0.91
0.06
-0.53
a
b
The resolution for 2f + 2′f at low temperature was insufficient to permit measurement. Calculated from the two-temperature data.
F IGURE 4. 1H NMR spectra of 2d in CD2Cl2 at 300 and 183
K, showing the “freezing out” of the conformational exchange.
minor isomer signals can be found in the aromatic
F IGURE 5. X-ray crystal structure of 1b (orange polymorph)
protons of the p-tolyl ring (Figure 4). For other groups
and other compounds, resolution was not always suf-
ficient to permit determination of equilibrium constants.
The best resolved signal pairs were chosen for the data
in Table 2.
Top: molecular geometry. Bottom: stereoview of crystal
packing.
The mixture of the two forms of 1b was subjected to a
differential scanning calorimetry (DSC) measurement
(Figure 6), which confirmed the expectation that these
are two forms of the same material. Interestingly, the
low-melting material slowly transformed upon standing
at room temperature to the high melting (and darker)
polymorph until, after a few weeks, it had completely
disappeared.10 Later recrystallizations no longer suc-
ceeded in producing this polymorph. From the less
intense color it may be speculated that in the less stable
polymorph the molecular geometry did not permit effec-
tive CT interaction. The geometry of this polymorph could
possibly be the a a conformation, in which intramolecular
(f) Cr ysta l Str u ctu r e a n d P olym or p h ism . Upon
recrystallization of 1b from ethyl acetate/hexane, two
different crystal forms (polymorphs) separated, one dark-
orange with needlelike crystals, mp 102.4 °C, and the
other one yellow and amorphous with cottonlike shape,
mp 94.1 °C. The high melting polymorph was subjected
to an X-ray crystallographic analysis, and its molecular
geometry is depicted in Figure 5. The crystal structure
clearly demonstrates that even in the solid-state in-
tramolecular CT interaction is preferred over intermo-
lecular interaction: the molecules are in the ee geometry,
and the aromatic rings of one molecule are mutually
perpendicular to those of the neighboring molecules, so
that intermolecular interactions are practically excluded.
(10) The phenomenon of “disappearing polymorphs” is well-known
and has recently been addressed in a review: Bernstein, J .; Dunitz,
J . D. Acc. Chem. Res. 1995, 28, 193.
6942 J . Org. Chem., Vol. 67, No. 20, 2002

Intramolecular Charge Transfer Complexes
The calculated ∆H° and ∆S° values are listed in Table
2. While obviously these are crude results, based on only
two temperatures each, they nevertheless provide some
insight to the complex formation: all of the complexes
have small positive reaction enthalpies, which means
that the complex strength is not sufficient to overcome
the conformational preference of the methyl group. On
the other hand, all of the complexes have small positive
reaction entropies. The remarkable significance of the
positive entropy is that it indicates that the CT-bound
(ee) conformation has a higher entropy, and hence is less
ordered, than the corresponding a a conformation. Indeed,
at low temperature the unbound conformation is favored,
since at lower temperature the enthalpy term becomes
dominant. This may be rationalized in terms of the weak
CT interactions that apparently operate between the
molecules 2 and the solvent. Such solvent-solute inter-
actions are more likely for the a a conformation, in which
the donor and acceptor groups cannot interact with each
other, and may be geometrically more exposed to contact
with solvent molecules. The weak solvent-solute interac-
tions involve substantial negative entropy. This is not
as effective at low temperature as it is at high temper-
ature, and as a result at low temperature this negative-
entropy conformation becomes more populated.
F IGURE 6. Differential scanning calorimetry (DSC) spectrum
of a mixture of polymorphs of 1b: heating rate, 5.00 °C min-1
peak temperatures, 94.13 and 102.43 °C; extrapolated peaks,
94.57 and 102.67 °C.
;
interaction is impossible and in which intermolecular
interactions may not be very intense, and hence the
lighter color. However, it could equally well be a different
crystal packing of the same (ee) molecular geometry.
Crystallographic analysis of the low-melting polymorph
could not be carried out to answer these questions,
because no definite crystals had separated.
Con clu sion
(g) Com p a r ison of Resu lts. The conformational
equilibria at room temperature for compounds 2 are all
close to 1, as determined both by the width of the 1H
NMR R-proton signals and by UV/vis spectroscopy. The
resulting complex-binding energies, ∆Gcomp, are small
(around -0.5 kcal mol-1) but negative; i.e., there is
attractive interaction between the donor and acceptor
groups. The equilibrium constants measured by NMR
spectroscopy at 183 K are substantially smaller; i.e., at
this temperature the population of the CT-bound 2ee
conformation is significantly lower. This obviously results
from the large temperature difference. The two Keq
measurements made for each compound over a wide
range of temperatures permit an estimate of the enthalpy
and entropy for the conformational exchange reaction
(eqs 6 and 7, derived from -RT ln K ) ∆H - T∆S):
Intramolecular, on-off-switchable CT complexes have
been prepared and their complex strengths and spectral
properties measured by UV/vis and NMR methods. The
complexes had CT-binding energies in the range from -1
to 0 kcal mol-1. In terms of effectiveness of the color as
an indicator, 1f and 2f seem to be most suitable: their
visible color is more intense because their CT absorption
band is shifted to longer wavelength, and as a result, it
is isolated from the acceptor and donor absorptions and
can be measured more readily (without subtraction).
Ack n ow led gm en t. Financial support by the US-
Israel Binational Science Foundation (BSF) is gratefully
acknowledged.
Su p p or tin g In for m a tion Ava ila ble: The Experimental
Section, including synthetic procedures and characterization
data for all of the compounds, is given in the Supporting
Information, as well as details and data tables for the crystal
R(ln K1 - ln K2)
∆H° )
(6)
(7)
1/T2 - 1/T1
1
structure determination of 1b, and a H NMR difference NOE
spectrum for 2′b. This material is available free of charge via
the Internet at http://pubs.acs.org.
R(T1 ln K1 - T2 lnK2)
T1 - T2
∆S° )
J O020164T
J . Org. Chem, Vol. 67, No. 20, 2002 6943