Solvent effects on hydrogen bonding and rotation barriers in α-alkyl-substituted 2-alkoxybenzyl alcohols

Article abstract of DOI:10.1039/b102416g

2-Alkoxyphenyl(α,α-dialkyl)methanols exist in two conformations, where the hydroxy hydrogen is either intramolecularly hydrogen-bonded to the alkoxy oxygen (syn rotamer) or is remote from the alkoxy group (anti rotamer) and therefore free . The anti and syn rotamers of 2-anisyldi(tert-butyl)methanol, and of derivatives where one or both tert-butyls are replaced by 1-adamantyl, can be separated by chromatography. The activation energy for anti→syn rotation of 2-anisyldi(tert-butyl)methanol (about 28 kcal mol^-1 at 373 K) varies insignificantly with the solvent, while that for the reverse reaction decreases in the order: chloroform ≈ toluene > pyridine > DMSO. Stabilization of the anti rotamer and the rotation transition state by hydrogen-bonding solvents would appear to be of equal importance, whereas the syn rotamer has no requirement for solvation of the polar OH group. Very similar solvent effects on equilibrium constants and rotation barriers are found for 2-anisyl(isopropyl)(tert-butyl)methanol, the rotamers of which are separable on the NMR time-scale. The free energy differences for the rotamers of this alcohol in a variety of solvents correlate with those for 3,4-(ethylenedioxy)-2-thienyldi(tert-butyl)methanol and with solute hydrogen bond basicity parameters.

Full text of DOI:10.1039/b102416g

View Article Online / Journal Homepage / Table of Contents for this issue
Solvent effects on hydrogen bonding and rotation barriers
in ꢀ-alkyl-substituted 2-alkoxybenzyl alcohols†
John S. Lomas* and Alain Adenier
2
ITODYS, Université de Paris 7, associé au C.N.R.S., 1 rue Guy de la Brosse, 75005 Paris,
France
Received (in Cambridge, UK) 14th March 2001, Accepted 27th April 2001
First published as an Advance Article on the web 12th June 2001
2-Alkoxyphenyl(α,α-dialkyl)methanols exist in two conformations, where the hydroxy hydrogen is either
intramolecularly hydrogen-bonded to the alkoxy oxygen (syn rotamer) or is remote from the alkoxy group
(anti rotamer) and therefore “free”. The anti and syn rotamers of 2-anisyldi(tert-butyl)methanol, and of derivatives
where one or both tert-butyls are replaced by 1-adamantyl, can be separated by chromatography. The activation
energy for anti→syn rotation of 2-anisyldi(tert-butyl)methanol (about 28 kcal molϪ1 at 373 K) varies insignificantly
with the solvent, while that for the reverse reaction decreases in the order: chloroform ≈ toluene > pyridine > DMSO.
Stabilization of the anti rotamer and the rotation transition state by hydrogen-bonding solvents would appear to be
of equal importance, whereas the syn rotamer has no requirement for solvation of the polar OH group. Very similar
solvent effects on equilibrium constants and rotation barriers are found for 2-anisyl(isopropyl)(tert-butyl)methanol,
the rotamers of which are separable on the NMR time-scale. The free energy differences for the rotamers of this
alcohol in a variety of solvents correlate with those for 3,4-(ethylenedioxy)-2-thienyldi(tert-butyl)methanol and
with solute hydrogen bond basicity parameters.
Introduction
Intramolecular hydrogen bonds stabilize the three-dimensional
structures of biological and other molecules. These hydrogen
bonds may be in competition with intermolecular hydrogen
bonds to solvent molecules, of which water is the most
important for biological systems. Consequently, structure
depends on solvent. In extreme cases change of solvent may
result in complete structural change; in less extreme cases the
equilibrium between different conformations may be modified.
In heteroaryl(α,α-dialkyl)methanols the balance between
intermolecular and intramolecular hydrogen bonding is a func-
tion of the heteroaryl group and of the solvent, two conformers
being readily distinguished by IR and NMR spectroscopy.1
Similar behaviour is expected of species where the heteroatom
is in a substituent close to the OH group. The product isolated
from the reaction of 2-anisyllithium with di(1-adamantyl)
ketone has the syn structure, 1S, with the hydroxy hydrogen
intramolecularly hydrogen-bonded to the methoxy oxygen.2
This conformation is clearly favoured in that it also minimizes
steric interactions between the methoxy group and the adam-
antyls, whereas the anti isomer, 1A, would not only lose the
benefit of hydrogen bonding but also suffer greater steric strain.
However, recent work on 3,4-ethylenedioxythiophene (EDOT)
analogues, 6–8,3 has shown that these can exist in both anti and
syn forms, and that the isomer with the “free” OH group can be
stabilized by solvation in a hydrogen-bonding solvent such as
DMSO or pyridine.
It has been reported that for 2-anisyl(α-alkyl)methanols and
2-anisyl(α,α-dialkyl)methanols, with rather smaller alkyl sub-
stituents [up to di(tert-butyl)], both forms occur and that they
are in equilibrium at 293–323 K.4–6 From the temperature
dependence of the IR spectra in carbon tetrachloride Ito and
Hirota5 determined “enthalpies of hydrogen bond formation”
(i.e. ∆HЊ for the equilibrium between the free and hydrogen-
bonded forms). In this and more recent work6 molecular mech-
anics calculations were used to determine the steric energy
profile for rotation about the sp2–sp3 bond in these alcohols,
the main conclusions being that the α,α-dialkyl derivatives
adopt stable conformations in which the C–OH bond is close
to the plane of the benzene ring, while in the rotation transition
state it is approximately orthogonal to this plane. These con-
clusions agree with crystallographic studies2,7,8 and previous
calculations9 on analogues without the 2-methoxy group.
According to a dynamic 1H NMR study of α,α-dialkyl-3,4,5-
trimethoxybenzyl alcohols in DMSO the rotation barrier for
the di(tert-butyl) derivative is substantially higher than that
for the (isopropyl)(tert-butyl) analogue: 21 and 13 kcal molϪ1
,
respectively.‡10 Now, that for 2-anisyl(isopropyl)(tert-butyl)-
methanol, 4, is about 18 kcal molϪ1,6 which suggests that, if the
above difference were reproduced in the 2-anisyl derivatives,
2-anisyldi(tert-butyl)methanol, 3, would be in the range where
the rotamers can be separated by chromatography or crystal-
lization. If this were the case it would be difficult to understand
how they could be in equilibrium at 293–323 K.
† Electronic supplementary information [ESI] available: Tables 4 and 5
with 13C NMR spectral data and details of the hydroxy proton shift in
solvents other than chloroform. See http://www.rsc.org/suppdata/p2/
b1/b102416g/
‡ 1 cal = 4.184 J.
DOI: 10.1039/b102416g
J. Chem. Soc., Perkin Trans. 2, 2001, 1051–1057
This journal is © The Royal Society of Chemistry 2001
1051

View Article Online
Table 1 Rate constants (sϪ1) and rotation barriers (kcal molϪ1) for 2-
anisyldi(tert-butyl)methanol, 3, in various solvents
ition state) the polar OH group in the initial state is intra-
molecularly hydrogen-bonded and has no specific requirement
for solvation.6 The solvent is therefore much more important
in the syn→anti than in the anti→syn reaction. A similar
result has been reported for the rotamerization of EDOTdi(1-
adamantyl)methanol, 6, in toluene and pyridine.3
‡
‡
Solvent
Temp./K
106kS
106kA
∆GS
∆GA
Chloroform
358
372
385
401
3.84
18.6
76.3
63.0
253
900
29.99
30.01
30.08
30.17
27.99
28.08
28.18
28.27
Inspection of the activation enthalpies and entropies, despite
their relative imprecision, reveals some significant features. For
toluene and chloroform the activation entropies for the two
299
3260
reactions are similar and average about Ϫ5 cal molϪ1
K
Ϫ1, the
Toluene
Pyridine
DMSO
348
358
373
383
1.40
4.31
22.8
61.2
17.2
49.6
249
29.80
29.88
29.93
30.00
28.07
28.14
28.15
28.25
greater stability of the syn isomer residing almost totally in the
enthalpy term. Conversely, for pyridine and toluene hydrogen
bonding favours the anti isomer by a little over 1 kcal molϪ1 in
both cases, whereas the activation entropies are now quite
different, about 6 cal molϪ1 KϪ1 more negative for the syn→anti
reaction than for anti→syn. The overall result is that, because
of this entropy term, at room temperature the syn rotamer still
predominates. Analogous findings for the EDOT derivative, 6,
were attributed to solvent structuring in the rotation transition
state.3
604
348
358
373
383
4.45
12.6
61.4
19.5
60.9
294
29.00
29.11
29.19
29.30
27.98
27.99
28.03
28.04
153
795
348
358
373
388
8.29
25.0
109
433
18.4
58.2
279
28.57
28.63
28.76
28.89
28.02
28.03
28.07
28.11
1190
(ii) 2-Anisyl(1-adamantyl)(tert-butyl)methanol, 2
Rotamer equilibria and rotation barriers. The alcohol, as
isolated by column chromatography, is the syn isomer, 2S.
Samples were equilibrated in various solvents at 423 K, the
equilibrium constants being little different from those for the
previous alcohol, 3, the anti isomer, 2A, representing 12, 16 and
23% for toluene, pyridine and DMSO, respectively.
We now report the separation of the syn and anti rotamers
of alcohols 1–3. The kinetics of the interconversion of the
rotamers of 3 have been studied in several solvents. Some less
congested 2-alkoxyphenyl(α,α-dialkyl)methanols, which are
separable on the NMR time-scale, will also be discussed.
Rotation barriers, measured in toluene, are 33.9 and 32.2 kcal
molϪ1 for the syn→anti and anti→syn reaction, respectively,
both values being 4 kcal molϪ1 higher than for the corre-
sponding reactions of the less crowded 2-anisyldi(tert-butyl)-
methanol, 3. This is closely similar to the effect (4.1 kcal molϪ1
at 473 K) of replacing one tert-butyl by 1-adamantyl in the
anti→syn reaction of 2-tolyldi(tert-butyl)methanol.9b
Results
(i) 2-Anisyldi(tert-butyl)methanol, 3
Rotamer equilibria. 2-Anisyldi(tert-butyl)methanol, 3, was
synthesized by reaction of 2-anisyllithium (from anisole,
n-butyllithium and TMEDA in diethyl ether at room temper-
ature) with di(tert-butyl) ketone. Distillation of the product at
reduced pressure gives a mixture of two isomers in a ratio of
about 10 : 1, the major component being the syn isomer, 3S.
They were separated by column chromatography on alumina,
the syn isomer being eluted first in light petroleum–diethyl ether
mixtures. The less stable, anti isomer, 3A, held at the highest
temperature used by Ito and Hirota,5 320 K, for several hours in
chloroform or in hydrogen-bonding solvents, such as DMSO
and pyridine, gives no syn isomer. Samples of this material were
heated for 2 h at 423 K in tubes sealed under vacuum either neat
or with a deuteriated solvent: chloroform, benzene, DMSO or
pyridine. The neat sample and those in chloroform or benzene
show about 9% of the anti isomer, while those in pyridine and
DMSO contain rather more (16 and 26%, respectively).
(iii) 2-Anisyldi(1-adamantyl)methanol, 1
Rotamer equilibria. An attempt to attain rotamer equilibrium
by heating the syn isomer, 1S, in DMSO at 473 K led to no
product which could be identified as the anti rotamer. It is
possible that in this solvent a carbocation is formed and that
this undergoes a 1,5-hydride shift, with the formation of a
carboxonium ion and subsequent reaction with a nucleophile.11
This reaction was not investigated further. On the other hand,
equilibration in toluene and pyridine proceeds normally to give
mixtures in which the anti isomer represents about 13.5 and
14%, respectively. That these figures should be so similar is
explained in part by the fact that the amount of the less stable
isomer increases with temperature in toluene, whereas it
decreases in pyridine, as can be seen from the data on 2-
anisyldi(tert-butyl)methanol, 3. A small amount of the anti
isomer, 1A, was isolated by reaction of the syn in a xylene
mixture in a sealed tube at 473 K, followed by alumina chrom-
atography. No attempt was made to study the kinetics of its
rotation.
Rotation barriers. Rates of equilibration of the alcohol,
starting with the anti isomer, 3A, were determined in several
solvents. In all cases the activation energy for anti→syn rotation
(associated with kA) is high, about 28 kcal molϪ1, as it must be
for separation to be possible at room temperature. A remark-
able feature is that for four very different solvents the anti→syn
barriers are closely similar, averaging 28.1, 28.2, 28.0 and 28.1
kcal molϪ1 for chloroform, toluene, pyridine and DMSO,
respectively (Table 1). The differences in the equilibrium
constants are due almost entirely to differences in the barrier
for the syn→anti reaction (associated with kS), these decreasing
in the order: chloroform (30.1 kcal molϪ1) ≈ toluene (29.9 kcal
molϪ1) > pyridine (29.15 kcal molϪ1) > DMSO (28.7 kcal
molϪ1). It is therefore possible to speak of the syn→anti
reaction as being “solvent-driven”, insofar as it is accelerated
by hydrogen-bonding solvents.
(iv) 2-Alkoxyphenyl(isopropyl)(tert-butyl)methanols, 9–11
Rotamer equilibria. 2-Anisyl(isopropyl)(tert-butyl)methanol,
9, was previously investigated by Suezawa et al.6 In contrast to
its more congested analogues, in this case the rotamers cannot
be separated by chromatography or crystallization, but the
These results are consistent with the idea that in the anti→syn
reaction the initial and the transition states are similarly
solvated, whereas in the reverse reaction (via the same trans-
1052
J. Chem. Soc., Perkin Trans. 2, 2001, 1051–1057

View Article Online
Table 2 Free energy differences and MMFF94-calculated steric energy differences (both in kcal molϪ1) for 2-anisyl(α,α-dialkyl)methanols: solvent
effects
Compound
Solvent
[anti]/[syn]
∆GЊa,e
∆SEb
∆solv.c
∆∆solv.d
2-Anisyl(isopropyl)(tert-butyl)methanol, 9 (298 K)
Chloroform
Benzene
Toluene
Acetonitrile
Methanol
Acetone
THF
0.17
0.23
0.24
0.52
0.67
0.68
1.03
1.35
2.93
1.04
0.88
0.85
0.40
0.24
1.65
1.65
1.65
1.65
1.65
1.65
1.65
1.65
1.65
0.61
0.77
0.80
1.25
1.41
1.43
1.67
1.81
2.29
0
0.16
0.19
0.64
0.80
0.82
1.06
1.20
1.68
0.22
Ϫ0.02
Ϫ0.16
Ϫ0.64
Pyridine
DMSO
2-Anisyldi(tert-butyl)methanol, 3 (348 K)
Chloroform
Toluene
Pyridine
DMSO
0.050f
0.081
0.23
0.45
2.07 (2.17)
1.73 (1.69)
1.02 (0.74)
0.55 (0.28)
2.64
2.64
2.64
2.64
0.57
0.91
1.62
2.09
0 (0)
0.34 (0.48)
1.05 (1.43)
1.52 (1.89)
a Free energy difference (anti Ϫ syn). b Steric energy difference (anti – syn).13 c “Solvation effect”. d Relative solvation effect, referenced to chloro-
form. e Data in parentheses are extrapolated to 298 K from data at higher temperatures. f Extrapolated from data at higher temperatures.
exchange rate at room temperature is low on the NMR time-
scale. Equilibrium constants ([anti]/[syn]) at 298 K are about 0.7
in acetone and methanol,6 and 0.2, 0.2, 0.5, 1 and 1.3 in chloro-
form, toluene, acetonitrile, THF and pyridine, respectively. In
DMSO the (Z)–(E) (syn–anti) ratio is not 77 : 23 (OCH3) or
72 : 28 (i-PrCH),6 but the inverse, the anti isomer, 9A, being
preferred in this solvent. A plot of the 8 data-points, expressed
as log K, where K is the equilibrium constant for the syn anti
reaction, against the solute hydrogen bond basicity parameter,
β2H,12 is approximately linear (intercept Ϫ0.84 0.08; gradient
1.58 0.16; correlation coefficient 0.9716). These data are also
well correlated with those for EDOTdi(tert-butyl)methanol, 8,
in the same solvents (intercept –0.64 0.05; gradient 0.85
0.07; correlation coefficient 0.9794).3
For the four solvents for which we have comparable data, the
equilibrium constants are about 5 times lower (i.e. there is a
preference for the syn isomer) for the di(tert-butyl), 3, than for
the (isopropyl)(tert-butyl) derivative, 9 (Table 2). Extrapolating
the data for 9 to 298 K gives ∆GЊ values which can be compared
with those for 3: the average difference is 0.96 0.12 kcal molϪ1
which is very close to the variation (0.99 kcal molϪ1) in the
anti Ϫ syn steric energy difference on going from 3 to 9,
calculated on the basis of the MMFF94 force field in Sybyl.13
Two other alkoxy derivatives were examined: the 2-ethoxy
and 2-isopropoxy analogues, 10 and 11, respectively. Regardless
of the solvent (chloroform, benzene, pyridine or DMSO),
increasing the size of the alkoxy group lowers the anti–syn ratio.
The corresponding free energy differences at 298 K, compared
to the 2-methoxy derivative, 9, are about 0.1 and 0.4 kcal molϪ1
for 10 and 11, respectively. There are two possible explanations
for the increase in the relative stability of the syn rotamer. The
first is that increasing the bulk of the alkoxy group destabilizes
the anti more than the syn isomer. This can be tested by mole-
cular mechanics calculations.13 These indicate that the difference
between the steric energies of the anti and syn rotamers varies
very little as the size of the alkoxy group is increased and that,
contrary to expectation, it is 0.1 kcal molϪ1 less for 10 and 11
than for 9, i.e. by this token the anti–syn ratio should be higher
for 10 and 11. A second possibility is that in the syn isomer
hydrogen bonding is stronger because the oxygen atom bears a
more electron-donating alkyl group. This latter possibility
would appear to be supported by the increasingly downfield
shift of the hydroxy proton in the NMR spectrum, whatever the
solvent, and the concomitant change in the IR frequency of the
OH stretching vibration to lower wavenumbers (3528, 3517 and
3493 cmϪ1 for 9, 10 and 11, respectively). If no other factors
have been overlooked, after correction for the steric energy
increment, these changes correspond to hydrogen bond energy
increments of about 0.2 and 0.5 kcal molϪ1
.
Correlation of the free energy differences between the rotam-
ers in DMSO against temperature gives very similar ∆HЊ and
∆SЊ values for alcohols 9–11. The enthalpy term (1.5–1.7 kcal
molϪ1) is close to that for 3 in the same solvent (1.5 kcal molϪ1),
while the entropy term is somewhat smaller, being 3–5 cal molϪ1
KϪ1 in favour of the syn isomer as compared to 5.7 cal molϪ1
KϪ1 for 3. This difference presumably reflects structural
dependence of the solvation of the various species involved in
rotamerization.
Rotation barriers. Dynamic 1H NMR was used to determine
rotation barriers for 2-anisyl(isopropyl)(tert-butyl)methanol, 9,
the exchange rates being calculated by gNMR simulation14 of
the spectra at different temperatures. Values for the anti→syn
reaction are 17.4, 17.7 and 17.3 kcal molϪ1 for chloroform (308–
343 K), toluene (328–353 K) and pyridine (303–333 K), respect-
ively, while the corresponding values for the syn→anti reaction
are 18.3, 18.5 and 17.3 kcal molϪ1. In DMSO the hydroxy pro-
ton signal of anti-2-anisyl(isopropyl)(tert-butyl)methanol, 9A,
moves into the region of the methoxy signals as the temperature
is increased, making it difficult to determine the rotation
barrier, though Suezawa et al. reported a value of 17.8 kcal
molϪ1 for the anti→syn reaction.6,15 By 13C DNMR we obtain
1
17.6 kcal molϪ1, in good agreement with the H NMR value,
and 17.0 kcal molϪ1 for the syn→anti reaction (298–328 K). The
variation in the anti→syn rotation barriers, though greater than
for 3, is of the same order of magnitude as the experimental
error on their determination, while the syn→anti reaction
barrier is again much reduced in the hydrogen-bonding solv-
ents, such as pyridine and DMSO, as compared to chloroform
and toluene.
Alcohols 10 and 11 also were studied by 13C DNMR in an
attempt to determine the effect, if any, of the size of the alkoxy
group. In fact, increasing the size of the alkyl group merely
raises both barriers slightly, the effect being slightly greater for
the syn→anti (17.0, 17.3 and 17.9 kcal molϪ1 for 9, 10 and 11,
respectively) than for the anti→syn reaction (17.6, 17.7 and 18.1
kcal molϪ1). This result contrasts with the much greater effects
of 3-alkoxy groups on the rotation of anti-3-alkoxy-2-thienyl-
di(1-adamantyl)methanes, a total variation of 2.3 kcal molϪ1
on going from methoxy to isopropoxy.8 This suggests that
in the transition state for rotation of the 2-alkoxyphenyl(iso-
propyl)(tert-butyl)methanols the smaller group, isopropyl, is
brought the closer to the alkoxy group (though not eclipsed
with it; conformations with either alkyl group eclipsed with the
J. Chem. Soc., Perkin Trans. 2, 2001, 1051–1057
1053

View Article Online
Table 3 Free energy differences (kcal molϪ1) for EDOT(α,α-dialkyl)-
methanols: solvent effects3
benzene ring do not constitute maxima on the rotation energy
profile). Since isopropyl has a sterically undemanding orienta-
tion, strain can be minimized by turning the CH bond towards
the alkoxy group, which means that the nature of this group
is of less importance than when both α-alkyl substituents are
sterically demanding in all orientations, as is 1-adamantyl.
It would have been interesting to compare 2-alkoxyphenyldi-
(tert-butyl)methanols with the same alkoxy groups, but only
the 2-ethoxy derivative, 12, is readily accessible. The yield of
the 2-isopropoxy compound is poor (less than 2%) and the
equilibrium concentration of the anti rotamer, even in DMSO,
will be lower than that of 12 and, therefore, too low for the
synthesis of adequate amounts of this material for kinetic
investigation. The anti–syn ratios for 12 in various solvents at
423 K, as for 9 and 10, are lower than for 3, corresponding to a
relative free energy difference of 0.55 0.07 kcal molϪ1. The
increase in steric energy on going from 12S to 12A is 0.3 kcal
molϪ1 higher than for 3S to 3A, which suggests that a small
part of this difference can be attributed to stronger hydrogen
bonding in 12S than in 3S, again correlated with a downfield
shift in the OH proton NMR signal and a lower IR OH
stretching frequency for 12S as compared to 3S.
Compound
Solvent
[syn]/[anti]a
∆GЊb
∆∆solv.c
EDOTdi(tert-butyl)methanol, 8 (298 K)
Chloroform
Benzene
Acetonitrile
Methanol
Acetone
THF
0.94
0.84
3.3
3.85
3.4
4.1
9.2
Ϫ0.04
Ϫ0.10
0.70
0.80
0.73
0.84
1.31
1.71
0
Ϫ0.06
0.74
0.84
0.77
0.88
1.35
1.75
Pyridine
DMSO
18
EDOT(1-adamantyl)(tert-butyl)methanol, 7 (298 K)
Chloroform
Benzene
Pyridine
DMSO
1.08
1.22
10
21
0.12
0.05
1.36
1.80
0
Ϫ0.07
1.24
1.68
a According to the Cahn–Ingold–Prelog rules,16 the syn and anti
designations are reversed with respect to those for the 2-anisyl deriv-
atives. b Free energy difference (syn Ϫ anti). c Relative solvation effect,
referenced to chloroform.
similar to the values for alcohol 9 at the same temperature
(Table 3).
The equilibrium between two conformers in a given solvent
reflects several factors, including the difference in the steric
energies, the strength of the intramolecular hydrogen bond,
general solvation phenomena and the specific effect of the
solvent on the “free” hydroxy group. The first two factors may
be considered to be independent of solvation, and the major
effect of replacing one solvent by another clearly lies in the last
of these factors, differences in the solvation of the rest of the
molecule being apparently negligible for the closely related
species considered in this work.
Comparison of the activation energies for the rotation of 12
in DMSO with those for 3 at 358–388 K reveals again a small
effect on the anti→syn reaction (0.1 kcal molϪ1) and a rather
greater effect upon the syn→anti reaction (0.6 kcal molϪ1). This
increase is greater than for the corresponding (isopropyl)-
(tert-butyl) derivatives, and is consistent with the fact that
tert-butyl has no sterically undemanding orientation, but is less
demanding than 1-adamantyl.
Rotamer equilibria and IR measurements
The data given above indicate that the rate of equilibration of
the two 2-anisyldi(tert-butyl)methanol rotamers, 3A and 3S, in
chloroform would be very slow at 293–323 K. The half-life for
equilibration at 293 K would be something over a year, and
about a week at 323 K. It seems unlikely that in the IR spectro-
scopic work5 the system was allowed to equilibrate several
weeks or years between spectra. Moreover, our data indicate
that there would be less than 4% of the anti isomer at 323 K
whereas Hirota’s data suggest a higher figure. Insofar as chloro-
form is a solvent very similar to carbon tetrachloride, it is
difficult to believe that the reported changes in the IR spectrum
are due to variations in the equilibrium constant.
The question is whether the intensities of the bands can be
taken as a measure of conformer populations. In our hands,
distillation of 3 affords a mixture containing about 9% of the
anti isomer, according to the NMR measurement. The isomer
ratio based on the IR measurement, with the simple assumption
that the peak height is proportional to concentration and that
the two isomers have the same extinction coefficient, gives
a somewhat higher value, about 15%. For the di(isopropyl)
derivative, 4, the IR measurement in carbon tetrachloride
indicates 15% of the anti isomer at 293 K, whereas NMR in
chloroform gives 17% at 223 K.15 Since our data show that the
anti isomer fraction increases with temperature in chloroform,
the IR measurement this time underestimates the amount of
anti. Finally, for the (isopropyl)(tert-butyl) derivative, 9, there is
good agreement between Hirota’s IR and our NMR data (15%
anti in both cases). On the basis of this evidence the IR inten-
sities do not appear to be a reliable guide to rotamer ratios.
Temperature increase reduces the intensity of the OH
absorption for the hydrogen-bonded, syn isomer, while that of
the free OH changes very little, the sum of the intensities
Discussion
Solvent effects on equilibrium constants
According to molecular mechanics calculations (MM3 or
MMFF94) the anti rotamer of 2-anisyl(α,α-dialkyl)methanols
has a higher steric energy than the syn form, the difference
increasing with the size of the alkyl groups.6 Since these force
fields do not take intramolecular hydrogen bonding into
account explicitly, the preference for the syn form should be
even greater than that indicated by MM. In fact, the data for
alcohols 3 and 9 show that, even in non-hydrogen-bonding
solvents, the syn form is somewhat less favoured than expected.
Given that MM calculates gas-phase energies, it may be argued
that this is a solvation effect, whether the solvent be hydrogen-
bonding or not. However, it is possible that the difference is due
to inadequacies in the force field. The data in Table 2 have been
organized to show the magnitude of any such solvation effect
or error, neglecting, however, the contribution of the intra-
molecular hydrogen bond. The relative solvation effects (which
are free of any error in the force field) follow the same trend for
the two alcohols considered. The values for the second set, 3,
are slightly higher, but since these data are extrapolated from
higher temperatures and the precision on the thermodynamic
parameters, particularly the activation entropy, is low, the
difference is not significant. Data on the equilibrium constants
for the EDOT derivatives, 7 and 8, in chloroform, pyridine and
DMSO at 298 K give relative solvation effects averaging 1.3 and
1.7 kcal molϪ1 for pyridine and DMSO, respectively, closely
1054
J. Chem. Soc., Perkin Trans. 2, 2001, 1051–1057

View Article Online
decreasing as the temperature increases.5 This is wrongly attrib-
uted to a slight increase in the proportion of the anti isomer.
The data presented do not allow us to determine whether or
not the integrated peak area remains the same. An interesting
feature of the IR spectra, not previously reported for 2-
alkoxybenzyl alcohols, is that the free hydroxy group is always
associated with two bands of similar intensity and separated,
for the alcohols examined here, by an average of 34 cmϪ1. This
is a common phenomenon with aryl(dialkyl)methanols and has
been attributed to the existence of two conformations with
respect to rotation about the C-O bond.1,9a,17
to understand why there is not a more regular variation. This
brings us back to the problem of the discrepancy between
the calculated steric energy difference, which neglects intra-
molecular hydrogen bonding, and the experimental free
energy difference for the 2-anisyldi(tert-butyl)methanol and
2-anisyl(isopropyl)(tert-butyl)methanol rotamers. High-level ab
initio calculations with a good solvation model could help to
resolve this problem.
Conclusion
The syn and anti rotamers, with intramolecularly hydrogen-
bonded and “free” hydroxy groups, respectively, of 2-anisyl-
di(tert-butyl)methanol, 3, are readily separated at room
temperature. The anti→syn rotation barrier is of the order of 28
kcal molϪ1, independent of the nature of the solvent, whether it
be hydrogen bonding or not. In hydrogen-bonding solvents the
syn→anti conversion can be spoken of as “solvent-driven”.
Replacing one or both of the tert-butyl groups by 1-adamantyl
makes the rotation barrier even higher. Other less encumbered
alcohols with lower rotation barriers and, therefore, not physi-
cally separable, behave in the same way. The general interpret-
ation is that the anti isomer and the transition state are similarly
solvated whereas the syn isomer has quite different solvation
characteristics, insofar as the hydrogen of the polar OH group
is intramolecularly hydrogen-bonded to the 2-methoxy oxygen.
The fact that the rotamerization of 3 is slow casts doubt upon
the interpretation of IR studies where the intensities of the
absorption bands of the syn and anti isomers of this and other
alcohols are found to be temperature-dependent.5 Though
we have not re-examined the series fully, the fact that all the
alcohols show the same temperature dependence of the IR
spectra as 3 suggests that this phenomenon is in no case attrib-
utable to variations in the equilibrium constant, despite the fact
that rotation is much faster for the less congested alcohols.
Solvent effects on rotation barriers
Our values for the rotation barriers in alcohols 3 and 9 clearly
show that solvent hydrogen bonding has little or no effect on
the anti→syn reaction but marked effects on the reverse reac-
tion, where the initial and transition states have very different
solvation characteristics. In previous work rotational barriers
and free energy differences for 2-anisyl(methyl)(tert-butyl)-
methanol, 5, in acetone–chloroform mixtures were studied.6
The listed activation energy relates to the syn→anti reaction,15
and that for the anti→syn reaction is obtained by subtracting
the free energy difference. There is a very small decrease (0.14
kcal molϪ1) in the anti→syn barrier on going from acetone to
90% chloroform, while that for syn→anti increases by about 1.5
kcal molϪ1, at temperatures which decrease from about 265 to
255 K.15 While these results are consistent with the general
pattern indicated above, the magnitude of the effect upon the
syn→anti reaction for the change from a moderately hydrogen-
bonding solvent, acetone, to chloroform is surprisingly large.
For 9 at 298 K the corresponding change in ∆GЊ is only 0.8 kcal
molϪ1. This may again reflect a structural dependence of
solvation. On the other hand, it should be noted that in the
cited work:6 (i) the temperatures are 30–40 K lower, which will
increase the importance of hydrogen bonding (the anti–syn
ratio in hydrogen-bonding solvents increases as the temperature
falls); (ii) the equilibrium constants were measured at an even
lower temperature, 223 K,15 which apart from the effect on
the ∆GЊ value, must have complicated the evaluation of the
exchange rate.
Experimental
General methods have been described elsewhere (see sup-
plementary information and ref. 8).
Alcohol synthesis
All syn alcohols were synthesized by the reaction of the
appropriate 2-alkoxyphenyllithium (prepared by the reaction of
2-alkoxybenzene with TMEDA and n-butyllithium in diethyl
ether or TFA at room temperature under argon) with the
appropriate ketone. After aqueous work-up and extraction into
light petroleum (boiling range 35–60 ЊC), drying and evap-
oration of the solvents, the product was purified either by
distillation or by column chromatography on alumina in light
petroleum–diethyl ether mixtures containing up to 20% of
diethyl ether. The anti rotamers are either in equilibrium with
the syn rotamer at room temperature (alcohols 9–11) or are
obtained by partial rotation of the syn isomer in a suitable
solvent at higher temperature, followed by chromatographic
separation on alumina (alcohols 1–3, 12).
Rotation barriers in ortho-substituted aryldi(tert-butyl)-
methanols
Only three data are available, the ortho substituent being
hydrogen, methoxy or methyl. For the first, values for several
solvents were determined by Sternhell et al.,18 but these values
vary as much with the method of determination as with the
solvent. Baas et al. give for the 3,4,5-trimethoxy derivative in
DMSO a value of 21.4 kcal molϪ1 at a coalescence temperature
of 421 K.10 For the o-tolyl derivative, only the anti→syn barrier
has been determined,9a the equilibrium constant lying so far in
the direction of the syn rotamer that the reverse reaction cannot
be detected. Molecular mechanics calculations indicate a steric
energy difference of about 7 kcal molϪ1 9a,b,d but this has never
been verified by measuring the heats of formation. However,
acid-catalysed solvolysis rates for the two rotamers in acetic
acid at 298 K differ by a factor of about 104,19 which corre-
sponds to 5.5 kcal molϪ1, in fair agreement with theory. The
anti→syn rotation barrier in dodecane at 373 K is 29 kcal
2-Anisyldi(tert-butyl)methanol, 3. Di(tert-butyl) ketone was
added to 2-anisyllithium in diethyl ether. The as-synthesized
alcohol is the syn isomer, 3S. Distillation (165 ЊC/20 mmHg;
lit.4 155 ЊC/5 mmHg) gave the alcohol as a 10 : 1 mixture of the
syn and anti isomers, which were separated by column chrom-
atography, the syn isomer being eluted first: νOH/cmϪ1 (CCl4)
3510 (lit.5 3508); δH (chloroform) 1.12 (t-Bu), 3.87 (CH3), 6.60
(OH), 6.92 (H5, J 1.4, 7.2 and 8.1), 6.94 (H3, J 0.3, 1.4 and 8.3),
7.19 (H4, J 1.7, 7.2 and 8.3) and 7.56 (H6, J 0.3, 1.7 and 8.1).
molϪ1 9a,b
which makes the syn→anti barrier about 35 kcal
,
molϪ1. To complete the set, we have the results of the present
work in toluene or chloroform: anti→syn, 28 kcal molϪ1; syn→
anti, 30 kcal molϪ1
.
These results show a fairly smooth progression in the value of
the syn→anti barrier as hydrogen (21.4 kcal molϪ1) is replaced
by methoxy (30 kcal molϪ1) and then by methyl (35 kcal molϪ1).
However, the reverse reaction has very similar values for methyl
(29 kcal molϪ1) and methoxy (28 kcal molϪ1), and a much
smaller one for hydrogen (21.4 kcal molϪ1). Intuitively it is hard
anti-2-Anisyldi(tert-butyl)methanol, 3A. Distilled 2-anisyldi-
(tert-butyl)methanol (2 g) was dissolved in dry DMSO (20 cm3)
and held at 150 ЊC for 30 min. After cooling, the alcohol was
extracted into pentane and the organic phase thoroughly
J. Chem. Soc., Perkin Trans. 2, 2001, 1051–1057
1055

View Article Online
washed with water, then dried and the solvent evaporated.
Chromatography yielded 3A (oil, 0.42 g, 21%): νOH/cmϪ1 (CCl4)
3613, 3648 (lit.5 3616); δH (chloroform) 1.16 (t-Bu), 1.90 (OH),
3.80 (CH3), 6.88 (H3, J 0.5, 1.3 and 8.1), 6.98 (H5, J 1.3, 7.3 and
8.1), 7.26 (H4, J 1.8, 7.3 and 8.1) and 7.90 (H6, J 0.5, 1.8 and
8.1) (Found: C, 76.9; H, 10.6. C16H26O2 requires C, 76.75; H,
10.47%).
lithium with (isopropyl) (tert-butyl) ketone in diethyl ether: oil;
11S: νOH/cmϪ1 (CCl4) 3493; δH (chloroform) 0.79 (CH3, J 6.7),
0.98 (t-Bu), 1.15 (CH3, J 6.7), 1.35 (CH3, J 6.1), 1.43 (CH3,
J 6.1), 2.51 (CH, J 1.3 and 6.7), 4.69 (CH, J 6.1), 6.41 (OH,
J 1.3), 6.90 (H5, J 1.3, 7.2 and 8.0), 6.91 (H3, J 0.4, 1.3 and 8.3),
7.14 (H4, J 1.7, 7.2 and 8.3) and 7.22 (H6, J 0.4, 1.7 and 8.0);
11A: νOH/cmϪ1 (CCl4) 3607, 3640; δH (chloroform) 0.61 (CH3,
J 6.8), 0.98 (t-Bu), 1.13 (CH3, J 6.8), 1.35 (CH3, J 6.1), 1.43
(CH3, J 6.1), 1.61 (OH), 3.44 (CH, J 6.8), 4.69 (CH, J 6.1), 6.76
(H3, J 8.2), 6.90 (H5), 7.14 (H4) and 7.70 (H6, J 1.9 and 7.8)
[Found (anti ϩ syn): C, 77.4; H, 10.8. C17H28O2 requires C,
77.22; H, 10.67%].
syn-2-Anisyl(1-adamantyl)(tert-butyl)methanol, 2S. (1-Adam-
antyl) (tert-butyl) ketone was added to 2-anisyllithium in
diethyl ether. The crude alcohol, identified as the syn isomer,
was purified by column chromatography (yield 71%): mp 84 ЊC;
νOH/cmϪ1 (CCl4) 3503; δH (chloroform) 1.15 (t-Bu), 1.59 and
1.8–2.1 (Ad), 3.87 (CH3), 6.56 (OH), 6.93 (H5, J 1.4, 7.2 and
8.2), 6.94 (H3, J 0.2, 1.4 and 8.3), 7.20 (H4, J 1.7, 7.2 and 8.3)
and 7.53 (H6, J 0.2, 1.7 and 8.2) (Found: C, 80.4; H, 9.8.
C22H32O2 requires C, 80.44; H, 9.82%).
syn-2-Ethoxyphenyldi(tert-butyl)methanol, 12S. Di(tert-
butyl) ketone was added to 2-ethoxyphenyllithium in THF. The
crude alcohol, identified as the syn isomer, was purified by
column chromatography (yield 38%): mp 50 ЊC; νOH/cmϪ1
(CCl4) 3493; δH (chloroform) 1.13 (t-Bu), 1.46 (CH3, J 7.0),
4.12 (CH2, J 7.0), 6.78 (OH), 6.90 (H5, J 1.4, 7.2 and 8.2), 6.92
(H3, J 0.3, 1.4 and 8.3), 7.16 (H4, J 1.7, 7.2 and 8.3) and 7.55
(H6, J 0.3, 1.7 and 8.2) (Found: C, 76.7; H, 10.7. C17H28O2
requires C, 77.22; H, 10.67%).
anti-2-Anisyl(1-adamantyl)(tert-butyl)methanol,
2A.
By
treating 2S (1.47 g) in DMSO (100 cm3) at 150 ЊC for 4 h,
followed by extraction into pentane, thorough washing with
water, evaporation of the solvent and alumina chromatography
2A was obtained (yield 0.282 g, 19%): mp 96 ЊC; νOH/cmϪ1
(CCl4) 3611, 3643; δH (chloroform) 1.14 (t-Bu), 1.58 and 1.7–2.1
(Ad), 1.88 (OH), 3.78 (CH3), 6.85 (H3, J 0.2, 1.3 and 8.1), 6.94
(H5, J 1.3, 7.2 and 8.1), 7.22 (H4, J 1.8, 7.2 and 8.1) and 7.81
(H6, J 0.2, 1.8 and 8.1) (Found: C, 80.5; H, 9.8. C22H32O2
requires C, 80.44; H, 9.82%).
anti-2-Ethoxyphenyldi(tert-butyl)methanol, 12A. Compound
12A was obtained by partial rotation of the syn isomer (0.74 g,
2.8 mmol) in dry DMSO (20 cm3) at 150 ЊC for 5 h. Chrom-
atography yielded 12A: (96 mg, 13%): mp 44 ЊC; νOH/cmϪ1
(CCl4) 3613, 3649; δH (chloroform) 1.14 (t-Bu), 1.48 (CH3,
J 7.0), 1.86 (OH), 4.08 (CH2, J 7.0), 6.85 (H3, J 0.1, 1.3 and
8.1), 6.93 (H5, J 1.3, 7.1 and 8.1), 7.20 (H4, J 1.8, 7.1 and 8.1)
and 7.89 (H6, J 0.1, 1.8 and 8.1) (Found: C, 77.2; H, 10.8.
C17H28O2 requires C, 77.22; H, 10.67%).
anti-2-Anisyldi(1-adamantyl)methanol, 1A. Di(1-adamantyl)
ketone was added to 2-anisyllithium, as previously described,2
to give the syn isomer, 1S. Treatment of this material (0.39 g) in
a xylene mixture (5 cm3) in a sealed tube at 200 ЊC for 5 h,
followed by evaporation of the solvent and alumina chrom-
atography gave 1A (yield 48 mg, 12%): mp 190 ЊC; νOH/cmϪ1
(CCl4) 3606, 3640; δH (chloroform) 1.62 and 1.8–2.1 (Ad), 1.94
(OH), 3.81 (CH3), 6.88 (H3, J 0.4, 1.2 and 8.1), 6.95 (H5, J 1.2,
7.2 and 8.0), 7.24 (H4, J 1.7, 7.2 and 8.1) and 7.81 (H6, J 0.4,
1.7 and 8.0) (Found: C, 82.6; H, 9.5. C28H38O2 requires C, 82.71;
H, 9.42%).
Equilibration experiments
Samples of 1S, 2S, 3S and 12S (10–15 mg) in a deuteriated
solvent (0.5 cm3) were sealed under vacuum in small tubes. The
tubes were held at 150 or 200 ЊC for 2 or 5 h, then opened and
1
the contents analysed by H NMR spectroscopy. Results are
given as percent anti in equilibrium with syn. 3S (2 h at 150 ЊC):
9.5, 9, 16 and 26% for chloroform, benzene, pyridine and
DMSO, respectively. No other product was detected. 2S (5 h at
150 ЊC): 12, 16 and 23% for toluene, pyridine and DMSO,
respectively. 1S (5 h at 200 ЊC) 13.5 and 14% for toluene and
pyridine, respectively. No anti isomer could be detected in the
products of the reaction of this alcohol in DMSO. 12 (5 h at
150 ЊC): 5, 5, 9 and 15% for chloroform, benzene, pyridine and
DMSO, respectively.
The other alcohols are equilibrium mixtures at room tem-
perature. 10: 11, 15, 51 and 70% anti and 11: 7, 9, 39 and 61%
anti for chloroform, benzene, pyridine and DMSO, respectively.
Data for 9 are given in Table 2. Anti–syn ratios for alcohols 9–11
in DMSO were determined over 30–40 K and the correspond-
ing ∆GЊ values plotted against temperature to estimate the ∆HЊ
and ∆SЊ values. Results are as follows (alcohol, ∆HЊ/kcal
molϪ1, ∆SЊ/cal molϪ1 KϪ1): 9, 1.6 0.2, 3.3 0.5; 10, 1.5 0.3,
3.3 0.8; 11, 1.7 0.2, 4.9 0.4.
2-Anisyl(isopropyl)(tert-butyl)methanol, 9. By the reaction of
2-anisyllithium with (isopropyl) (tert-butyl) ketone in diethyl
ether: bp 147 ЊC/20 mm (lit.4 133 ЊC/5 mm); 9S: νOH/cmϪ1 (CCl4)
3528 (lit.5 3532); δH (chloroform) 0.79 (CH3, J 6.6), 0.97 (t-Bu),
1.15 (CH3, J 6.6), 2.52 (CH, J 1.6 and 6.6), 3.87 (CH3), 5.97
(OH, J 1.6), 6.92 (H3, J 0.4, 1.3 and 8.1), 6.94 (H5, J 1.3, 7.3
and 8.0), 7.19 (H4, J 1.7, 7.3 and 8.1) and 7.21 (H6, J 0.4, 1.7
and 8.0); 9A: νOH/cmϪ1 (CCl4) 3606, 3639 (lit.5 3603.5);
δH (chloroform) 0.61 (CH3, J 6.8), 0.96 (t-Bu), 1.14 (CH3, J 6.8),
1.65 (OH), 3.31 (CH, J 6.8), 3.77 (CH3), 6.83 (H3, J 8.0), 6.94
(H5), 7.19 (H4) and 7.69 (H6, J 1.8 and 7.8).
2-Ethoxyphenyl(isopropyl)(tert-butyl)methanol, 10. Com-
pound 10 was obtained by the reaction of 2-ethoxyphenyl-
lithium with (isopropyl) (tert-butyl) ketone in diethyl ether: mp
49 ЊC; 10S: νOH/cmϪ1 (CCl4) 3517; δH (chloroform) 0.80 (CH3,
J 6.7), 0.98 (t-Bu), 1.16 (CH3, J 6.7), 1.47 (CH3, J 7.0), 2.52
(CH, J 1.5 and 6.7), 4.06 (CH, J 7.0 and 9.2), 4.14 (CH, J 7.0
and 9.2), 6.21 (OH, J 1.5), 6.90 (H3, J 0.2, 1.3 and 8.1), 6.92
(H5, J 1.3, 7.2 and 8.1), 7.14 (H4, J 1.7, 7.2 and 8.1) and 7.21
(H6, J 0.2, 1.7 and 8.1); 10A: νOH/cmϪ1 (CCl4) 3606, 3642;
δH (chloroform) 0.62 (CH3, J 6.8), 0.98 (t-Bu), 1.14 (CH3, J 6.8),
1.43 (CH3, J 7.0), 1.64 (OH), 3.42 (CH, J 6.8), 3.97 (CH, J 6.7
and 8.8), 4.04 (CH, J 6.7 and 8.8), 6.79 (H3, J 8.2), 6.92 (H5),
7.14 (H4) and 7.69 (H6, J 1.7 and 7.9) [Found (anti ϩ syn):
C, 76.6; H, 10.6. C16H26O2 requires C, 76.75; H, 10.47%].
Rotation kinetics
Slow rotation. (i) From a solution of 3A (ca. 30 mg) in deuterio-
chloroform (2 cm3) 10 × 0.2 cm3 aliquots were transferred
to small tubes which were sealed under vacuum, the sample
being frozen in liquid nitrogen. Batches of tubes were held
in a thermostat, 8 samples being withdrawn at convenient
intervals, the remaining two being used as “infinities” (ca. 10
half-lives). Each sample was made up in the same solvent
to ca. 0.5 cm3 for 1H NMR spectroscopic analysis. The
same procedure was employed to study the rotation of 2A in
deuteriated toluene.
2-Isopropoxyphenyl(isopropyl)(tert-butyl)methanol, 11. Com-
pound 11 was obtained by the reaction of 2-isopropoxyphenyl-
1056
J. Chem. Soc., Perkin Trans. 2, 2001, 1051–1057

View Article Online
(ii) A sample of 3A (ca. 15 mg) in deuteriated toluene, pyr-
idine or DMSO (0.5 cm3) was placed in an NMR tube which
was then introduced into the apparatus at an appropriate
associated with the isopropyl group being simulated by the
gNMR program. Relative concentrations were determined at
each temperature by integration of the tert-butyl group peaks.
Shifts were optimized at each temperature at the same time as
the exchange rate. Results are as follows [alcohol (temperature
1
temperature. H NMR spectra were then recorded at conven-
ient time intervals over 2–3 half-lives and after approximately
10 half-lives. The same procedure was employed to study the
rotation of 12A in DMSO.
In both cases suitable peaks of the anti and syn isomers were
integrated to determine the relative composition, and the
overall rate constant (kA ϩ kS) was calculated by plotting log
[%anti(t) Ϫ %anti(∞)] vs. time.
‡
range/K, ∆GA‡/kcal molϪ1, ∆GS /kcal molϪ1)]: 9 (298–328,
17.6 0.1, 17.0 0.1); 10 (298–333, 17.7 0.1; 17.3 0.1); 11
(298–338, 18.1 0.2, 17.9 0.1).
Molecular mechanics
Molecular mechanics calculations were performed using
the MMFF94 force field in the Sybyl 6.7 package.13 Values
for 3 and 9 are considerably higher than those reported
previously.16 Steric energies (kcal molϪ1) of the most stable
conformations are: 3A, 90.60; 3S, 87.96; 9A, 70.01; 9S, 68.36;
10A, 67.32; 10S, 65.82; 11A, 71.82; 11S, 70.34; 12A, 88.25;
12S, 85.32. There is no obvious reason why steric energies are
lower for the 2-ethoxy than for the corresponding 2-methoxy
derivatives.
Rate constants for 3A are as follows [solvent (T/K, (kA ϩ
kS)/sϪ1, syn–anti ratio at equilibrium)], the error limits being
the standard deviations on single runs: chloroform (358.3,
6.68 0.03 × 10Ϫ5, 16.4; 372.2, 2.72 0.01 × 10Ϫ4, 13.6; 385.1,
9.76 0.03 × 10Ϫ4, 11.8; 400.9, 3.56 0.02 × 10Ϫ3, 10.9); tolu-
ene (348, 1.86 0.01 × 10Ϫ5, 12.3; 358, 5.39 0.02 × 10Ϫ5, 11.5;
373, 2.72 0.02 × 10Ϫ4, 10.9; 383, 6.65 0.03 × 10Ϫ4, 9.9);
pyridine (348, 2.39 0.01 × 10Ϫ5, 4.4; 358, 7.35 0.05 × 10Ϫ5
4.8; 373, 3.55 0.02 × 10Ϫ4, 4.8; 383, 9.48 0.06 × 10Ϫ4, 5.2);
,
DMSO (348, 2.67 0.03 × 10Ϫ5, 2.2; 358, 8.31 0.04 × 10Ϫ5
,
2.3; 373, 3.88 0.06 × 10Ϫ4, 2.6; 388, 1.62 0.02 × 10Ϫ3, 2.7).
Thermodynamic parameters [solvent (reaction, ∆H‡/kcal
Acknowledgement
We are indebted to Dr H. Suezawa (Yokohama) for helpful
discussions and access to unpublished material, and to Drs K.
Gao and B. T. Fan (ITODYS) for the MMFF94 calculations.
molϪ1, ∆S‡/cal molϪ1
K
Ϫ1, mean ∆G‡/kcal molϪ1)]: chloroform
(anti→syn, 25.7 0.1, Ϫ6.3 0.2, 28.13; syn→anti, 28.4 0.3,
Ϫ4.4 0.7, 30.06); toluene (anti→syn, 26.5 0.2, Ϫ4.6 1.2,
28.15; syn→anti, 28.0 0.2, Ϫ5.1 0.6, 29.90); pyridine
(anti→syn, 27.3 0.1, Ϫ1.9 0.2, 28.01; syn→anti, 26.3 0.3,
Ϫ8.0 0.8, 29.15); DMSO (anti→syn, 27.2 0.1, Ϫ2.4 0.3,
28.06; syn→anti, 25.7 0.1, Ϫ8.1 0.4, 28.71).
Rate constants for 2A in toluene are as follows [T/K,
(kA ϩ kS)/sϪ1, syn–anti ratio at equilibrium]: 409.1, 6.26 0.04 ×
10Ϫ5, 8.6; 423.1, 2.35 0.02 × 10Ϫ4, 7.8; 438.0, 8.62 0.10 ×
References
1 J. S. Lomas, A. Adenier, C. Cordier and J. C. Lacroix, J. Chem. Soc.,
Perkin Trans. 2, 1998, 2647.
2 J. S. Lomas and J. Vaissermann, J. Chem. Soc., Perkin Trans. 2, 1999,
1639.
3 J. S. Lomas, J. Chem. Soc., Perkin Trans. 2, 2001, 754.
4 A. Yamashita, K. Hara, S. Aizawa and M. Hirota, Bull. Chem. Soc.
Jpn., 1974, 47, 2508.
5 M. Ito and M. Hirota, Bull. Chem. Soc. Jpn., 1981, 54, 2093.
6 H. Suezawa, H. Wada, H. Watanabe, T. Yuzuri, K. Sakakibara and
M. Hirota, J. Phys. Org. Chem., 1997, 10, 925.
7 H. van Koningsveld, Cryst. Struct. Commun., 1973, 3, 407;
E. Hough and J. S. Lomas, Acta Crystallogr., Sect. B, 1984, 40, 1938;
J. S. Lomas and J. Vaissermann, Bull. Soc. Chim. Fr., 1996, 133, 25;
J. Vaissermann and J. S. Lomas, Acta Crystallogr., Sect. C, 1997, 53,
1341; J. S. Lomas and J. Vaissermann, J. Chem. Soc., Perkin Trans. 2,
1998, 1777.
10Ϫ4
, , 6.5. Thermodynamic
6.8; 451.6, 2.33 0.02 × 10Ϫ3
parameters (reaction, ∆H‡/kcal molϪ1, ∆S‡/cal molϪ1 KϪ1
,
mean ∆G‡/kcal molϪ1): (anti→syn, 30.2 0.6, Ϫ4.8 1.4,
32.24; syn→anti, 32.5 0.8, Ϫ3.3 1.9, 33.94). Rate constants
for 12A in DMSO are as follows [T/K, (kA ϩ kS)/sϪ1, syn–anti
ratio at equilibrium]: 358, 6.08 0.06 × 10Ϫ5
,
4.6; 373,
2.86 0.03 × 10Ϫ4
, , 5.4; 398,
5.0; 388, 1.18 0.03 × 10Ϫ3
2.69 0.09 × 10Ϫ3, 5.7. Thermodynamic parameters (reaction,
∆H‡/kcal molϪ1, ∆S‡/cal molϪ1
K
Ϫ1, mean ∆G‡/kcal molϪ1):
anti→syn, 26.4 0.3, Ϫ4.8 0.8, 28.22; syn→anti, 24.9 0.3,
Ϫ12.1 0.9, 29.46.
8 J. S. Lomas, E. Vauthier and J. Vaissermann, J. Chem. Soc., Perkin
Trans. 2, 2000, 1399.
9 (a) J. S. Lomas, P. K. Pham and J. E. Dubois, J. Org. Chem., 1977,
42, 3394; (b) J. S. Lomas and J. E. Dubois, Tetrahedron, 1981, 37,
2273; (c) B. Tiffon and J. S. Lomas, Org. Magn. Reson., 1984, 22, 29;
(d) J. S. Lomas and V. Bru-Capdeville, J. Chem. Soc., Perkin Trans. 2,
1994, 459; (e) J. E. Anderson, V. Bru-Capdeville, P. A. Kirsch and
J. S. Lomas, J. Chem. Soc., Chem. Commun., 1994, 1077; ( f ) J. S.
Lomas and J. E. Anderson, J. Org. Chem., 1995, 60, 3246.
10 J. M. A. Baas, J. M. van der Toorn and B. M. Wepster, Recl. Trav.
Chim. Pays-Bas, 1974, 93, 133.
11 J. S. Lomas and J. Vaissermann, J. Chem. Soc., Perkin Trans. 2, 1996,
1831.
12 M. H. Abraham, Chem. Soc. Rev., 1993, 73.
13 Sybyl 6.7, Tripos Inc., 1699 South Hanley Road, St. Louis, MO,
63144–2917.
14 gNMR from Cherwell Scientific Publishing Limited, The Magdalen
Centre, Oxford Science Park, Oxford OX4 4GA. Now available from
Adept Scientific plc, Amor Way, Letchworth SG6 1ZA.
15 H. Suezawa, personal communication.
Fast rotation. (i) Rotation barriers for 9 in chloroform,
toluene and pyridine were determined by dynamic H NMR,
1
the part of the spectrum associated with the methoxy group
being simulated by the gNMR program.14 A low-temperature
spectrum (no exchange) was first simulated to determine shifts
and line-widths. Variations in the latter parameter at higher
temperatures were neglected, but shifts and relative concen-
trations are temperature-dependent and were, therefore,
optimized at each temperature at the same time as the exchange
rate, which was divided by the species concentrations to obtain
rate constants for the anti→syn and syn→anti reactions, kA and
kS, respectively. The activation energies are not sufficiently
regular for activation enthalpies and entropies to be deter-
mined; the values listed are mean activation energies based on
4–7 measurements at 5 or 10 ЊC intervals in the ranges
indicated. Results are as follows (solvent, temperature range/K,
16 R. S. Cahn, C. K. Ingold and V. Prelog, Angew. Chem., 1966, 78,
‡
∆GA‡/kcal molϪ1, ∆GS /kcal molϪ1): (chloroform, 308–343,
413.
17 F. H. Hon, H. Matsumura, H. Tanida and T. T. Tidwell, J. Org.
Chem., 1972, 37, 1778.
18 R. E. Gall, D. Landman, G. P. Newsoroff and S. Sternhell, Aust. J.
Chem., 1972, 25, 109.
19 J. S. Lomas and J. E. Dubois, Tetrahedron, 1978, 34, 1597.
17.4 0.1, 18.3 0.1); (toluene, 328–353, 17.7 0.1, 18.5
0.1); (pyridine, 303–333, 17.3 0.1, 17.3 0.1).
(ii) Rotation barriers for alcohols 9–11 in DMSO were
determined by dynamic 13C NMR, the part of the spectrum
J. Chem. Soc., Perkin Trans. 2, 2001, 1051–1057
1057