Electronic and steric substituent influences on the conformational equilibria of cyclohexyl esters: The anomeric effect is not anomalous!

Article abstract of DOI:10.1002/chem.200390155

The cyclohexyl esters of a series of carboxylic acids, RCO₂H, spanning a range of electronegativities and quotients of steric hindrance for the R substituent (R = Me, Et, iPr, tBu, CF₃, CH₂Cl, CHCl₂, CH₂Br, CHBr₂, and CBr₃) were prepared. Their conformational equilibria in CD₂Cl₂ were examined by low-temperature ¹H NMR spectroscopy to study the axial or equatorial orientation of the ester functionality with respect to the adopted chair conformation of the cyclohexane ring. The ab initio and DFT geometry-optimized structures and relative free energies of the axial and equatorial conformers were also calculated at the HF/ 6-311G**, MP2/6-311G, and B3LYP/ 6-31G** levels of theory, both in the gas phase and in solution. In the latter case, a self-consistent isodensity polarized continuum model was employed. Only by including electron correlation in the modeling calculations for the solvated molecules was it possible to obtain a reasonable correlation between ΔG°_calcd and ΔG°_exp. Both the structures and the free energy differences of the axial and equatorial conformers were evaluated with respect to the factors normally influencing conformational preference, namely, 1,3-diaxial steric interactions in the axial conformer and hyperconjugation. It was assessed that hyperconjugative interactions, σ_C-C/σ_C-H and σ_C-O*, together with a steric effect - the destabilization of the equatorial conformer with increasing bulk of the R group - were the determinant factors for the position of the conformational equilibria. Thus, because hyperconjugation is held responsible as the mitigating factor for the anomeric effect in 2-substituted, six-membered saturated heterocyclic rings, and since it is also similarly responsible, at least partly, in these monosubstituted cyclohexanes for a preferential shift towards the axial conformer, the question is therefore raised: can the anomeric effect really be construed as anomalous?

Full text of DOI:10.1002/chem.200390155

FULL PAPER
Electronic and Steric Substituent Influences on the Conformational Equilibria
of Cyclohexyl Esters: The Anomeric Effect Is Not Anomalous!
Erich Kleinpeter,*[a] Ferdinando Taddei,*[b] and Philipp Wacker[a]
Abstract: The cyclohexyl esters of a
series of carboxylic acids, RCO2H, span-
ning a range of electronegativities and
quotients of steric hindrance for the R
substituent (R ˆ Me, Et, iPr, tBu, CF3,
CH2Cl, CHCl2, CCl3, CH2Br, CHBr2,
and CBr3) were prepared. Their confor-
mational equilibria in CD2Cl2 were ex-
amined by low-temperature 1H NMR
spectroscopy to study the axial or equa-
torial orientation of the ester function-
ality with respect to the adopted chair
conformation of the cyclohexane ring.
The ab initio and DFT geometry-opti-
mized structures and relative free ener-
gies of the axial and equatorial con-
formers were also calculated at the HF/
6-311G**, MP2/6-311G**, and B3LYP/
6-31G** levels of theory, both in the gas
phase and in solution. In the latter case,
a self-consistent isodensity polarized
continuum model was employed. Only
by including electron correlation in the
modeling calculations for the solvated
molecules was it possible to obtain a
reasonable correlation between DG8calcd
and DG8exp. Both the structures and the
free energy differences of the axial and
equatorial conformers were evaluated
with respect to the factors normally
influencing conformational preference,
namely, 1,3-diaxial steric interactions in
the axial conformer and hyperconjuga-
tion. It was assessed that hypercon-
jugative interactions, sCÀC/sCÀ and
H
s* O, together with a steric effect–the
À
C
destabilization of the equatorial con-
former with increasing bulk of the R
group–were the determinant factors for
the position of the conformational equi-
libria. Thus, because hyperconjugation is
held responsible as the mitigating factor
for the anomeric effect in 2-substituted,
six-membered saturated heterocyclic
rings, and since it is also similarly
responsible, at least partly, in these
monosubstituted cyclohexanes for
a
Keywords: ab initio calculations ¥
preferential shift towards the axial con-
former, the question is therefore raised:
can the anomeric effect really be con-
strued as anomalous?
conformation analysis
anes hyperconjugation
spectroscopy
¥
cyclohex-
NMR
¥
¥
quently a reduction in the nonbonded repulsions with the syn-
axial hydrogen atoms in the axial conformation.[2] The
presence of this repulsion, reflecting steric strain within the
cyclohexane ring, is supported by the results of a survey of
appropriate X-ray structures in the Cambridge Structural
Data Base based on inclination angles (q C3, C5, C1-OR)
greater than 908.[3] Seemingly enigmatically, the tendency of
Introduction
It is generally accepted that the chair chair conformational
equilibrium of monosubstituted cyclohexanes with respect to
the axial or equatorial disposition of the substituent (the A
value) is primarily determined by the steric interactions
between the axial substituent and the axial protons in the 3-
and 5-positions.[1] In the case of oxygen as the first atom of the
À
polar substituents, such as OR, at the C2 position of
À
axial substituent, that is OR, this amounts to steric
tetrahydropyrans to occupy the axial position has been
termed the anomeric effect.[4] The hyperconjugative origin
of the anomeric effect has been recently stressed[5] and
quantitatively modeled on the basis of ab initio calculations
and the natural bond orbital (NBO) method.[6] The origin of
this phenomenon is considered to be unfavorable dipole
dipole (electrostatic) interactions[7] that destabilize the equa-
torial conformer and the favorable hyperconjugation between
the ring-oxygen lone-pair electrons and the antibonding
interactions with the oxygen lone pairs.[2] An R group with
electron-withdrawing properties leads to an increase in the
population of the axial conformer, which is consistent with a
decreased electron density on the oxygen atom and conse-
[a] Prof. Dr. E. Kleinpeter, P. Wacker
Institut f¸r Organische Chemie und Strukturanalytik
Universit‰t Potsdam
C-OR orbital (n ! s* H) stabilizing the axial conformer.[4]
Am Neuen Palais 10, 14415 Potsdam (Germany)
Fax : (‡49)331-977-5064
À
C
O
However, although the steric 1,3-diaxial interactions that
destabilize the axial conformer are still present, it is not a
simple matter to separate them from the electronic substitu-
ent effect. The original concept of the anomeric effect has also
E-mail: kp@chem.uni-potsdam.de
[b] Prof. F. Taddei
Dipartimento di Chimica
Universita degli Studi di Modena e Reggio Emilia
1360
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0906-1360 $ 20.00+.50/0
Chem. Eur. J. 2003, 9, No. 6

1360 1368
been extended[8] to various oth-
er substituents and six-mem-
bered saturated heterocycles.
To take into account the differ-
ent steric requirements of vari-
ous substituents in different
heterocycles in comparison to
cyclohexane (and thereby char-
acterize the anomeric effect as
the difference of the A value of
a certain substituent in mono-
substituted cyclohexane and in
the 2-position of a heterocycle),
empirical correlation factors a
for each particular heterocycle
were introduced.[9, 10]
Scheme 1. Structure of the compounds 1 11 and the conformational equilibrium examined.
of the axial and equatorial conformers were calculated by
both ab initio HF (at the HF/6-311G** and MP2/6-311G**
levels of theory) and DFT calculations (at the B3LYP/6-
31G** level of theory), both in the gas phase and in solution,
and determined their free energy differences, DG8. In
addition, NBO analysis at this level of theory was also used
to understand the interaction mechanism. Initial results,[22]
obtained for the cyclohexanol esters with R ˆ CH3, CF3, and
CCl3 were encouraging as calculations correctly provided the
sequence order, but not the amount, of the corresponding A
values. This prompted the current expanded study.
Previously,[11 we studied the influence of increasing the
13]
polarity of substituents on the conformational equilibria of
cyclohexyl esters of acetic acid analogues with respect to the
chair conformation of the cyclohexane ring (i.e., the ring
À
substituents are O2CR) and assessed the observed increasing
preference of the axial conformer with an increase in the
À
polarity of the O2CR substituent within the context of
Bushweller×s steric interaction model.[2] Other reports have
corroborated[14, 15] our finding[11
that increasing electron
13]
withdrawal by the substituent in monosubstituted cyclohex-
anes leads to an increase in the amount of the axial con-
former; however, Kirby and Williams ascertained[15] from
simultaneous bond length variations that sCÀ
hyperconjugation is the factor responsible for this observa-
tion.
In fact, workers have long been aware of the presence of
hyperconjugation in substituted cyclohexanes ever since the
elaboration of the Perlin effect[16] (where 1J(C,Hax) <
1J(C,Heq) in cyclohexane) and which has been corroborated
by a natural bond orbital population (NBO) analysis based on
high-level ab initio MO calculations on cyclohexanol deriv-
atives.[17] Moreover, there was no evidence for 1,3-diaxial
! s*
À
C OAr
Results and Discussion
H(axial)
Experimental studies: The esters 1 11 were prepared simply
and efficiently by admixture of equimolar quantities of
cyclohexanol and the corresponding carboxylic acid in
toluene or chloroform (depending on the boiling points of
the product esters) with the removal of any accumulated
water by azeotropic distillation. Final distillation provided the
esters 1 11 in reasonable purity, confirmed by 1H and
13C NMR spectroscopy and mass spectral analysis.
À
interactions between the axial OR substituent and the axial
For the determination of the equilibrium constants K
1
(Scheme 1), the H and the 13C NMR spectra of the esters
protons in positions 3 and 5 as being a determinant factor for
the equilibrium position.[17] The same result was obtained by
Wiberg et al.[18] in a study of the conformational equilibria of
monoalkyl-substituted cyclohexanes of varying steric bulk,
and also employing high-level ab initio MO calculations. In
both cases, the agreement between calculated and experi-
mental DG8 values was found to be excellent.[17, 18] The isotope
effects of 1H versus 2D and 12C versus 13C were also in
complete agreement.[19, 20]
were recorded in CD2Cl2 at low temperatures on Bruker
NMR instruments (Avance 300 and 500); two sets of signals,
one for each of the axial and equatorial conformers, were
obtained and the set of cyclohexane carbon atoms that lay
upfield were assigned to the axial conformers in each case on
account of the steric compression effects they experience.[23]
The equilibrium constants (K ˆ [1eq]/[1ax], etc.) of the con-
formational equilibria were evaluated by careful integration
of the well-separated H1 signals in each case at 193 and 203 K
(except in the case 5 (R ˆ CF3) where measurements were
made at 183 and 193 K). These constants subsequently
provided the free energy differences (DG8 ˆ ÀRTlnK). The
13C chemical shifts of the two conformers for 1 11 are
presented in Table 1. For the 1H NMR spectra, only the
chemical shifts of H1 and the protons of the R substituent
(H10) are given because protons H2 H6 furnished subspec-
tra of higher order and severe overlap which, because of the
poor state of homogeneity at the lower temperatures, were
not amenable to simulation. The conformational energy
Because this divergence in pinpointing the underlying cause
for the equilibrium preference and the understanding of the
substituent/cyclohexane ring interaction dependency on the
substituent polarity is not yet fully understood,[21] we synthe-
sized a variety of cyclohexyl esters (1 11, Scheme 1) to study
and scrutinize the long-held belief that steric interactions in
the axial conformer prevailed in determining the position of
the conformational equilibrium (Scheme 1). The study con-
sisted of an experimental component, whereby the conforma-
tional equilibria were studied by low-temperature NMR
spectroscopy, and a theoretical component. The structures
Chem. Eur. J. 2003, 9, No. 6
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0906-1361 $ 20.00+.50/0
1361

E. Kleinpeter, F. Taddei, P. Wacker
FULL PAPER
Table 1. Experimental 1H and 13C chemical shifts and 1J(C,H) coupling constants for compounds 1 11 at 183 K (see Scheme 1 for structures and atom
numbering).
Compd.
Conf.
13Cd
C4
1Hd
1J(C1,H1)[Hz]
1J(C2,H2ax)[Hz]
1J(C2,H2eq)[Hz]
C1
C2/C6
C3/C5
C8
C10
21.2
H1
H10
1
2
ax
eq
ax
eq
ax
eq
ax
eq
ax
eq
ax
eq
ax
eq
ax
eq
ax
eq
ax
eq
ax
eq
69.2
72.2
68.9
72.4
68.8
72.4
68.7
72.3
75.5
78.3
71.9
75.1
73.8
76.9
76.2
79.2
71.8
75.1
73.5
76.7
76.1
79.0
28.9
31.2
28.9
31.1
29.0
31.2
29.1
31.2
28.6
30.6
28.9
31.1
28.8
30.9
28.7
30.5
28.8
31.0
28.7
30.7
28.5
30.3
20.0
23.8
20.0
23.8
20.2
24.0
20.3
24.1
19.7
23.8
20.0
23.9
20.0
24.0
20.0
23.9
20.0
23.9
20.1
23.9
20.0
23.7
24.9
24.4
24.8
24.4
25.0
24.6
25.1
24.7
24.5
24.1
24.8
24.3
24.6
24.4
24.8
24.5
24.8
24.3
24.9
24.4
24.7
24.2
170.5
170.3
173.8
173.5
176.5
176.5
177.5
177.5
158.8
156.2
166.5
166.3
163.4
163.4
160.6
160.6
166.3
166.1
163.9
163.9
160.7
160.7
4.98
4.62
4.98
4.63
4.96
4.61
4.88
4.51
5.29
4.91
4.99
4.62
5.15
4.78
5.21
4.84
5.08
4.72
5.12
4.76
5.19
4.83
2.10
2.10
2.31
2.31
2.50
2.50
151.2
145.8
150.4
146.7
151.9
145.9
150.3
145.4
154.7
149.2
151.3
147.0
152.8
148.0
153.7
148.8
151.7
148.7
152.5
148.0
153.0
148.1
n.a.[a]
127.7
n.a.
127.5
n.a.
130.0
126.9
127.8
128.2
130.7
n.a.
128.0
n.a.
129.4
n.a.
130.0
128.8
128.8
127.4
129.5
126.2
130.4
n.a.
130.5
n.a.
21.2
28.9
28.9
33.6
33.4
38.4
38.0
113.9
113.9
41.8
41.8
64.7
64.5
89.8
89.5
27.9
27.9
34.2
34.0
30.9
30.7
129.9
129.7
131.4
131.2
129.6
132.7
133.2
n.a.
131.1
132.5
131.8
n.a.
131.8
131.9
131.6
132.3
131.1
128.4
132.5
3
4
5
6
4.15
4.10
6.16
6.01
7
8
9
4.03
3.99
6.00
5.92
10
11
[a] n.a. ˆ Not available because the coupling pattern was too complex and because of an insufficient signal-to-noise ratio.
differences between the axial and equatorial conformers of
11 are summarized in Table 2. These values differ appreci-
both in the gas phase and in solution. In addition, DFT
calculations at the B3LYP/6-31G** level of theory were also
performed. The Onsager reaction field theory was applied to
calculate the solvent effect by self-consistent reaction field
theory (SCRF). A self-consistent isodensity polarized con-
tinuum model (SCIPCM) was employed with the 6-311G**
basis set in a solvent with a dielectric constant e ˆ 2.6. The
solvent effect was estimated with the routines of the
GAUSSIAN98 program package.[24]
1
ably from those reported previously[11, 12] as they are depend-
ent both on the method used and the solvent employed.
However, the sequence order with respect to the polarity of
the R substituent remains intact.
The 1J(C,H) coupling constants were determined from
standard HMQC spectra (without proton decoupling).
À
À
By means of internal rotation about the C1 O7 and O7 C8
Table 2. Axial equatorial conformational equilibria (K ˆ [eq]/[ax]) and
DG8exp (DG8exp ˆ ÀRTlnK) for compounds 1 11.
À
bonds (rotation about the C8 C10 bond was not taken into
consideration), a number of nondegenerate, but stable, con-
formers were assessed for each of the orientations of the ester
group. This set of conformers was whittled down to the stage
where only one conformer was found to be significant for each
of the equatorial and axial orientations (Scheme 2). For
example, for the equatorial conformation, a second con-
Compd
K193
À DG8
K203
À DG8
[kcalmolÀ1
]
[kcalmolÀ1
]
1
2
3
4
5
6
7
8
9
7.4704
9.5275
6.1489
2.9079
5.7954[a]
6.1818
4.1900
2.7106
5.9925
3.8176
2.3456
0.771
0.864
0.696
0.409
0.639
0.698
0.549
0.384
0.686
0.513
0.327
6.7145
7.9572
5.9011
2.6448
5.3600[b]
5.6146
3.8675
2.6056
5.2484
3.4987
2.1963
0.678
0.836
0.716
0.392
0.644
0.696
0.545
0.386
0.668
0.505
0.317
10
11
[a] At 183 K. [b] At 193 K.
Computational studies: Ab initio MO calculations were
performed with the GAUSSIAN98 program package.[24]
Different levels of theory were previously tested on three of
the compounds (1, 3, and 6) and a number of sulfur
analogues.[22] The results at the HF/6-311G**//HF/6-31G**,
and MP2/6-311G**//HF/6-311G** theoretical levels proved
to be the most reliable and were therefore used to calculate
the energies of the axial and equatorial conformers of 1 11,
À
À
Scheme 2. Preferred rotamers about C1 O7 and O7 C8 bonds for the
axial and equatorial conformers of compounds 1 11.
1362
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0906-1362 $ 20.00+.50/0
Chem. Eur. J. 2003, 9, No. 6

Conformational Equilibria of Cyclohexyl Esters
1360 1368
former, identified as a local minimum, was more than
3 kcalmolÀ1 higher in energy and was therefore not consid-
ered to contribute significantly to the population of the
equatorial form. The two conformers, axial and equatorial,
both adopted staggered orientations for the groups of C1 and
O7 with, in each case, a lone pair of electrons on the oxygen
atom orientated towards the cyclohexane ring (Scheme 2).[22]
In both cases, the ester group was found to be in the Z
configuration.[25]
The structural parameters of the molecular-relaxed geo-
metries of the two conformations of 1 11 are summarized in
Tables 3 and 4; the theoretical free energies of activation,
DG8calcd, obtained by the different methods, both in the gas
phase and in solution, are reported in Table 5. The electron
Table 3. Selected geometrical parameters for compounds 1 11 calculated at the B3LYP/6-31G** level of theory.
Bond lengths[ä][a]
Angles[8]
O7-C1-C4
À
C1 C2
À
C1 C6
À
À
C2 C3
À
À
À
C2 H2ax
À
C6 H6eq
À
C6 H6ax
Compd
C1 O7
C5 C6
C2 H2eq
q[b]
equatorial conformers
1
2
3
4
5
6
7
8
9
1.5301
1.5269
1.5301
1.5269
1.5284
1.5259
1.5286
1.5281
1.5260
1.5286
1.5283
1.5268
1.5302
1.5267
1.5300
1.5251
1.5293
1.5253
1.5249
1.5293
1.5254
1.5250
1.4544
1.4542
1.4542
1.4539
1.4658
1.4600
1.4640
1.4675
1.4588
1.4623
1.4653
1.5380
1.5377
1.5380
1.5376
1.5383
1.5378
1.5384
1.5385
1.5376
1.5383
1.5385
1.5377
1.5379
1.5379
1.5378
1.5380
1.5385
1.5383
1.5383
1.5380
1.5380
1.5380
1.0943
1.0956
1.0943
1.0957
1.0945
1.0952
1.0944
1.0944
1.0952
1.0944
1.0944
1.0983
1.0980
1.0984
1.0981
1.0977
1.0977
1.0977
1.0976
1.0979
1.0977
1.0975
1.0956
1.0943
1.0956
1.0943
1.0952
1.0945
1.0952
1.0951
1.0945
1.0952
1.0952
1.0980
1.0983
1.0980
1.0983
1.0976
1.0982
1.0974
1.0974
1.0981
1.0976
1.0974
147.56
148.80
148.86
148.82
148.60
148.57
148.51
148.53
148.75
148.44
148.43
157.660
157.648
157.699
157.658
157.464
157.450
157.391
157.410
157.593
157.324
157.323
10
11
axial conformers
1
2
3
4
5
6
7
8
9
1.5331
1.5330
1.5329
1.5293
1.5312
1.5285
1.5318
1.5307
1.5285
1.5319
1.5310
1.5294
1.5294
1.5296
1.5334
1.5277
1.5321
1.5277
1.5275
1.5326
1.5281
1.5275
1.4608
1.4611
1.4607
1.4610
1.4728
1.4677
1.4716
1.4753
1.4652
1.4684
1.4720
1.5362
1.5363
1.5364
1.5372
1.5362
1.5371
1.5363
1.5365
1.5377
1.5367
1.5367
1.5371
1.5370
1.5371
1.5361
1.5373
1.5364
1.5369
1.5368
1.5365
1.5377
1.5376
1.0943
1.0943
1.0943
1.0956
1.0946
1.0955
1.0944
1.0946
1.0956
1.0946
1.0945
1.0987
1.0956
1.0956
1.0956
1.0938
1.0953
1.0945
1.0952
1.0951
1.0942
1.0953
1.0952
1.0983
1.0983
1.0983
1.0987
1.0980
1.0986
1.0982
1.0981
1.0985
1.0980
1.0980
102.58
102.54
102.38
103.14
101.71
102.43
102.39
102.59
100.88
100.84
100.83
93.377
93.326
93.190
93.894
92.460
93.125
93.087
93.236
91.767
91.669
91.643
1.0987
1.0987
1.0984
1.0984
1.0982
1.0985
1.0985
1.0980
1.0984
1.0984
10
11
À
À
À
À
À
À
[a] No changes in the following bond lengths for 1 11: C1 H1, C3 H3ax, C3 H3eq, C5 H5ax, and C5 H5eq. [b] Angle between the axis of the bond C1 O7
and the plane defined by C3-C1-C5.
Table 4. Selected geometrical parameters (interatomic distances) for compounds 1 11 calculated at the B3LYP/6-31G** level of theory.
Interatomic distances[ä]
À
À
À
À
À
À
À
À
À
À
À
À
À
À
À
O7 H2eq O7 H2ax O7 C2 O7 H6eq O7 H6ax O7 C6 O7 H3eq O7 H3ax O7 H5eq O7 H5ax O7 C1 O7 H1 O9 H2eq O9 H6eq C1 C10
equatorial conformers
1
2
3
4
5
6
7
8
9
2.638
2.638
2.719
2.640
2.721
2.637
2.721
2.722
2.637
2.720
2.722
2.584
2.585
2.636
2,587
2.631
2.582
2.629
2.631
2.584
2.627
2.627
2.394
2.395
2.449
2.395
2.450
2.395
2.449
2.450
2.396
2.448
2.449
2.719
2.719
2.638
2.716
2.643
2.722
2.641
2.642
2.720
2.638
2.638
2.634
2.635
2.586
2.633
2.585
2.632
2.582
2.583
2.634
2.580
2.581
2.449
2.449
2.395
2.447
2.399
2.450
2.397
2.398
2.449
2.395
2.396
4.556
4.557
4.607
4.557
4.605
4.556
4.604
4.605
4.557
4.603
4.604
4.125
4.125
4.159
4.126
4.164
4.129
4.163
4.164
4.127
4.162
4.163
4.606
4.606
4.557
4.605
4.559
4.606
4.557
4.558
4.605
4.555
4.556
4.160
4.159
4.125
4.158
4.133
4.161
4.131
4.131
4.161
4.129
4.129
1.454
1.454
1.454
1.454
1.466
1.460
1.464
1.468
1.459
1.462
1.465
2.070
2.070
2.068
2.068
2.073
2.071
2.072
2.074
2.070
2.072
2.073
4.269
4.271
2.793
4.263
2.847
4.279
2.826
2.818
4.269
2.820
2.824
2.801
2.799
4.271
2.790
4.297
2.821
4.294
4.266
2.835
4.293
4.284
3.720
3.726
3.735
3.752
3.735
3.741
3.752
3.756
3.726
3.741
3.746
10
11
axial conformers
1
2
3
4
5
6
7
8
9
2.678
2.677
2.677
2.584
2.680
2.588
2.682
2.676
2.593
2.681
2.687
3.378
3.378
3.377
3.333
3.377
3.334
3.379
3.378
3.332
3.374
3.377
2.451
2.450
2.449
2.394
2.448
2.394
2.454
2.452
2.388
2.443
2.447
2.588
2.590
2.591
2.680
2.594
2.679
2.585
2.588
2.689
2.592
2.587
3.334
3.335
3.335
3.380
3.337
3.378
3.335
3.337
3.376
3.333
3.332
2.394
2.394
2.394
2.456
2.396
2.452
2.395
2.398
2.447
2.390
2.388
4.016
4.014
4.012
3.970
3.997
3.959
4.019
4.019
3.937
3.980
3.985
2.746
2.743
2.741
2.712
2.718
2.695
2.754
2.756
2.665
2.692
2.699
3.962
2.699
2.699
2.693
2.766
2.692
2.748
2.693
2.707
2.694
2.668
2.664
1.461
1.461
1.461
1.461
1.473
1.468
1.472
1.475
1.465
1.468
1.472
2.068
2.069
2.069
2.067
2.073
2.071
2.071
2.072
2.073
2.074
2.074
2.768
2.765
2.773
4.304
2.801
4.282
2.775
2.822
4.342
2.781
2.752
4.273
4.274
4.270
2.683
4.312
2.789
4.305
4.290
2.718
4.317
4.325
3.726
3.733
3.741
3.758
3.740
3.754
3.758
3.763
3.729
3.743
3.750
3.962
3.959
4.029
3.956
4.016
3.957
3.966
3.983
3.939
3.936
10
11
Chem. Eur. J. 2003, 9, No. 6
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0906-1363 $ 20.00+.50/0
1363

E. Kleinpeter, F. Taddei, P. Wacker
FULL PAPER
Table 5. Comparison of experimental free energy differences, DG8exp, with
calculated[a] thermodynamic values, DG8calcd or DE8calcd, at various levels of
theory for compounds 1 11.
group, the more preferred is the axial conformer (Table 2).
For example, with an increasing number of chlorine atoms the
differences in energy between the two conformers are: CH3
(À0.771 kcalmolÀ1)) < CH2Cl (À0.699 kcalmolÀ1)) < CHCl2
(À0.549 kcalmolÀ1)) < CCl3 (À0.382 kcalmolÀ1)). Similarly
with an increasing number of bromine atoms, the population
of the axial conformer in the conformational equilibria
increases, with the larger bromine atoms not limiting a
greater shift towards the axial conformer. This result is in
complete agreement with previously obtained measure-
DG8exp
[kcalmolÀ1
DG8calcd or DE8calcd[b, c] [kcalmolÀ1
]
]
A
B
C
D
E
F
1
2
3
4
5
6
7
8
9
À 0.771
À 0.865
À 0.695
À 0.409
À 0.639
À 0.699
À 0.549
À 0.382
À 0.687
À 0.514
À 0.327
À 0.547 0.070
À 0.548 0.096
À 0.579 0.061
À 0.739 0.121
À 0.286 0.370
À 0.542 0.421
À 0.348 0.436
À 0.323 0.317
À 0.479 0.420
À 0.451 0.450
À 0.511 0.492
À 0.626
À 0.699
À 0.664
À 0.719
À 0.347
À 0.542
À 0.453
À 0.435
0.001 0.294
À 0.514
À 0.507
À 0.523
À 0.654
À 0.407
À 0.403
À 0.442
À 0.429
0.350
À 0.025 0.221
0.093 0.178
0.141 0.277
0.328 0.514
0.257 0.579
0.332 0.568
0.438 0.614
0.577
13]
ments.[11
substituents is also noticeable; however, the differences
in DG8 are not that remarkable: CH2CH3
On the other hand, the effect of the alkyl
10
11
0.584
0.626
0.231
0.455
(À0.865 kcalmolÀ1)) < CH3 (À0.771 kcalmolÀ1)) < CH(CH3)2
(À0.695 kcalmolÀ1)) < C(CH3)3 (À0.409 kcalmolÀ1)).
Clearly, since polarity differences in the series 1 4 can be
neglected, the increasing volume of the substituent also
increases the population of the axial conformer, or, to state it
alternatively, destabilizes the equatorial counterpart. This
latter result, in particular, does not bode well for the
interpretation of substituent influences on the cyclohexane
ring in terms of destabilizing steric 1,3-nonbonding interac-
tions.
[a] At 173.15 K. [b] A: HF/6-311G**//HF/6-311G**, GP; B: MP2/6-311G**//
HF/6-311G**, GP; C: HF/6-311G**//HF/6-31G**, S; D: MP2/6-311G**//HF/6-
311G**, S; E: MP2/6-311G**//HF/6-31G**, DE8; F: B3LYP/6-31G**, DE8.
(Note: GP ˆ gas phase, S ˆ solution state). [c] Solution-state values were not
calculated for the brominated compounds 9 11.
populations of the atoms and lone pairs of the axial and
equatorial conformers of the cyclohexane ring in compounds
1
11 were obtained by NBO analysis[26] and refer to the
B3LYP/6-31G** molecular geometries. Both the hyperconju-
The total energy of the preferred conformers of the
equatorial and axial conformers of compounds 1 11 was
calculated with the 6-311G** basis set at the HF level of
theory; for the DFT calculations, the 6-31G** basis set was
used. The results, together with the free-energy differences,
DG8, for the two conformers of 1 11 are presented in Table 5.
Correction for the solvent effect, calculated according to the
methods described in the Experimental Section, was applied
to HF and MP2 calculations (data sets C and D in Table 5).
The correlation between the experimental DG8 values for
À
À
gative interactions of the C H and C C bonds of the
cyclohexane moiety with the antibonding orbital of the
À
C1 O7 bond and the electron population of the oxygen O7
lone pairs, which are both generally of interest with respect to
the effects dominating the conformational equilibria of 1 11,
are presented in Table 6.
Relative energies of the equatorial and axial conformers: The
free-energy differences of the monosubstituted cyclohexanes
1
11 exhibit a distinct trend: the more polar (ÀI effect) the R
1
11 with the theoretical values is only reasonable if the
Table 6. Hyperconjugative interactions [kcalmolÀ1] of C H and C C bonds with the antibonding s*(C1 O7) and occupancy by electrons of this orbital and
of the O7 lone pairs[a] in compounds 1 11 from NBO analysis.
À
À
À
Conf.
Donor orbital
S of all s Sax À Seq Lone pair
s*(C1 O7)
occupancy (O7) occupancy
s(C2 H2eq) s(C2 H2ax) s(C6 H6eq) s(C6 H6ax) s(C2 C3) s(C5 C6)
1
2
3
4
5
6
7
8
9
eq
ax
eq
ax
eq
ax
eq
ax
eq
ax
eq
ax
eq
ax
eq
ax
eq
ax
0.345
0.692
0.446
0.681
0.343
0.692
0.439
0.828
0.362
0.718
0.452
0.842
0.359
0.453
0.460
0.767
0.454
0.827
0.352
0.697
0.360
0.692
0.699
5.387
0.707
5.402
0.687
5.366
0.697
5.015
0.736
5.663
0.733
5.153
0.736
0.745
0.749
5.784
0.705
5.282
0.717
5.832
0.742
5.676
0.445
0.817
0.347
0.801
0.443
0.807
0.345
0.695
0.452
0.845
0.349
0.719
0.453
0.359
0.367
0.878
0.352
0.651
0.459
0.856
0.464
0.863
0.709
4.978
0.697
5.049
0.700
4.988
0.694
5.436
0.744
5.283
0.724
5.592
0.745
5.263
0.737
5.388
0.697
5.802
0.724
5.405
0.749
5.146
3.677
0.162
3.429
0.141
3.683
0.169
3.421
0.161
3.916
0.196
3.519
0.189
3.897
0.171
3.982
0.169
3.599
0.204
3.934
0.192
3.945
0.193
3.384
0.178
3.717
0.159
3.404
0.182
3.672
0.145
3.647
0.205
3.796
0.169
3.622
0.194
3.689
0.187
3.839
0.181
3.660
0.210
3.638
0.218
9.259
12.214
9.344
12.233
9.260
12.204
9.268
12.280
9.857
12.910
9.573
12.664
9.812
12.934
9.984
13.173
9.646
2.955
2.889
2.944
3.012
3.053
3.091
3.122
3.189
3.301
3.346
3.060
1.8768
1.8770
1.8769
1.8768
1.8762
1.8760
1.8762
1.8760
1.8624
1.8627
1.8680
1.8663
1.8635
1.8636
1.8642
1.8642
1.9056
1.9057
1.9016
1.9022
1.8659
1.8672
0.0431
0.0489
0.0431
0.0364
0.0432
0.0489
0.0430
0.0490
0.0470
0.0533
0.0444
0.0509
0.0449
0.0513
0.0457
0.0525
0.0313
0.0376
0.0316
0.0379
0.0451
0.0518
12.947
9.846
13.192
9.898
10 eq
ax
11 eq
ax
12.788
[a] Averaged value of the two lone pairs.
1364
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0906-1364 $ 20.00+.50/0
Chem. Eur. J. 2003, 9, No. 6

Conformational Equilibria of Cyclohexyl Esters
1360 1368
calculations take account of both electron correlation and the
solvent effect. The best results were obtained by the use of the
MP2/6-311G**//HF/6-311G** calculations (data set D in
Table 5); the corresponding diagram is depicted in Figure 1. In
Molecular geometries: The structural parameters of the
molecular-relaxed geometries of the axial and equatorial
conformers of 1 11 by means of the B3LYP/6-31G** method
are reported in detail in Tables 3 and 4 including bond lengths
(Table 3), interatomic distances (Table 4), and the inclination
angle q (Table 3). These theoretical structural parameters,
together with the experimental NMR parameters (13C and 1H
chemical shifts and 1J(C,H), see Table 1), are now discussed in
light of the long-range substituent influence of the R group on
the conformational equilibria of 1 11.
First, the trend of the bond lengths and bond angles of
monosubstituted cyclohexanes reported previously[1, 21] are
also found for these compounds 1 11, for example, the
À
C1 O7 bonds are longer in the axial conformer than in the
respective equatorial conformer. The same holds for the
À
C1 H1 bond lengths, with the axial bonds being longer than
those in the equatorial conformer and in agreement with the
corresponding observed coupling constants, 1J(C,H), (Ta-
ble 1) and the well-described Perlin effect.[16]
Figure 1. Plot of DG8calcd (MP2/6-311G**//HF/6-311G**,S) versus DG8exp
for compounds 1 8.
Second, the electron densities at O7 and along the axial
À
C H bond fragments in the 3- and 5-positions in the axial
conformers have been examined with respect to the steric
bond polarization model, which is widely accepted for the
conformational analysis of substituted cyclohexanes and
heterocyclic analogues.[1, 2] For the calculated electron density
at O7, the results were consistent with expectations (Table 7);
electron-donating substituents (compounds 1 4) increase the
electron density at O7 whilst electron-withdrawing substitu-
ents (compounds 5 11) reduce it. For these parameters, a fine
correlation of q(O7) versus the inductive effect of the R group
was obtained.
this case, the correct order of the conformational energies as a
function of the R substituent was obtained. In addition, there
seem to be two linear dependencies, one for the more polar
substituents R and another one (of lower ascent) for the alkyl
substituents R, further corroborating the former view of a
combined dependence of the long-range influence of the R
substituent on the conformational equilibria of 1 11 based on
both electronic and steric substituent effects. For this reason,
the substituent cyclohexane interaction has been studied in
more detail on the basis of the experimental NMR spectra and
the calculated results.
Finally, the electron densities at H3ax and H5ax and at C3
and C5 correlate exceptionally well (Figure 2). As the
electron density at C3 and C5 increases, the electron density
Table 7. Electronic properties (dipole moments and selected atomic charge densities) for compounds 1 11 calculated at the B3LYP/6-31G** level of theory.
m
Atomic charge densities, q[e]
[Debye]
C1
C3
C5
H2eq
H2ax
H6eq
H6ax
H3eq
H3ax
H5eq
H5ax
O7
O9
equatorial conformers
1
2
3
4
5
6
7
8
9
1.929
1.801
1.861
1.897
3.486
2.253
2.418
3.499
2.381
2.249
3.172
0.1697
0.1703
0.1702
0.1701
0.1557
0.1603
0.1546
0.1547
0.1611
0.1562
0.1555
À 0.1884
À 0.1884
À 0.1886
À 0.1882
À 0.1908
À 0.1904
À 0.1903
À 0.1911
À 0.1899
À 0.1905
À 0.1910
À 0.1887
À 0.1886
À 0.1883
À 0.1904
À 0.1911
À 0.1894
À 0.1912
À 0.1917
À 0.1894
À 0.1908
À 0.1915
0.0988
0.0988
0.1140
0.0993
0.1176
0.1081
0.1165
0.1189
0.1082
0.1165
0.1184
0.1002
0.1002
0.0955
0.0995
0.1065
0.1052
0.1078
0.1085
0.1035
0.1071
0.1079
0.1142
0.1140
0.0993
0.1143
0.1076
0.1142
0.1082
0.1094
0.1138
0.1078
0.1087
0.0956
0.0956
0.0997
0.0957
0.1090
0.0998
0.1095
0.1104
0.0986
0.1089
0.1097
0.0950
0.0950
0.0950
0.0951
0.1017
0.0988
0.1005
0.1011
0.0982
0.0999
0.1013
0.0953
0.0952
0.0984
0.0954
0.1010
0.0966
0.0998
0.1011
0.0968
0.0996
0.1007
0.0951
0.0950
0.0950
0.0951
0.1015
0.0978
0.1006
0.1019
0.0975
0.1000
0.1011
0.0984
0.0983
0.0953
0.0985
0.0982
0.0997
0.0972
0.0984
0.0998
0.0969
0.0979
À 0.4816
À 0.4947
À 0.4982
À 0.5008
À 0.4745
À 0.4605
À 0.4498
À 0.4537
À 0.4694
À 0.4586
À 0.4644
À 0.4732
À 0.4819
À 0.4860
À 0.4916
À 0.4519
À 0.4627
À 0.4522
À 0.4269
À 0.4623
À 0.4579
À 0.4391
10
11
axial conformers
1
2
3
4
5
6
7
8
9
1.983
1.851
1.892
1.904
3.229
1.633
2.069
3.213
1.912
1.866
2.821
0.1520
0.1523
0.1508
0.1554
0.1383
0.1424
0.1385
0.1359
0.1466
0.1389
0.1399
À 0.1781
À 0.1780
À 0.1786
À 0.1804
À 0.1820
À 0.1835
À 0.1841
À 0.1846
À 0.1817
À 0.1859
À 0.1863
À 0.1796
À 0.1797
À 0.1792
À 0.1807
À 0.1834
À 0.1836
À 0.1842
À 0.1846
À 0.1862
À 0.1840
À 0.1847
0.1149
0.1146
0.1148
0.1008
0.1185
0.1028
0.1179
0.1189
0.1016
0.1171
0.1199
0.0948
0.0946
0.0942
0.0961
0.1031
0.0989
0.1014
0.1037
0.0988
0.1002
0.1021
0.1008
0.1005
0.0999
0.1175
0.1096
0.1162
0.1104
0.1115
0.1190
0.1092
0.1105
0.0958
0.0956
0.0957
0.0942
0.1027
0.0983
0.1008
0.1030
0.0969
0.1000
0.1018
0.0907
0.0905
0.0905
0.0898
0.0965
0.0913
0.0946
0.0964
0.0906
0.0929
0.0945
0.0995
0.0998
0.1006
0.1057
0.1085
0.1170
0.1135
0.1133
0.1124
0.1160
0.1148
0.0895
0.0894
0.0892
0.0908
0.0954
0.0925
0.0939
0.0955
0.0913
0.0927
0.0940
0.1064
0.1066
0.1063
0.1017
0.1137
0.1145
0.1166
0.1159
0.1182
0.1169
0.1169
À 0.4744
À 0.4876
À 0.4905
À 04956
À 0.4679
À 0.4530
À 0.4441
À 0.4477
À 0.4630
À 0.4521
À 0.4585
À 0.4730
À 0.4817
À 0.4862
À 0.4918
À 0.4523
À 0.4656
À 0.4523
À 0.4270
À 0.4621
À 0.4576
À 0.4393
10
11
Chem. Eur. J. 2003, 9, No. 6
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0906-1365 $ 20.00+.50/0
1365

E. Kleinpeter, F. Taddei, P. Wacker
FULL PAPER
Figure 2. Plot of the H3ax (H5ax) electron density, q, against the electron
density of C3 (C5), respectively, for the axial conformers of compounds 1
11.
Scheme 3. Lewis bond/nonbonding structures of monosubstituted cyclo-
hexane.
at the axial protons, H3ax and H5ax, abates. Interestingly, the
increased electron density at O7 does not polarize the axial
To see if this effect also has an influence on the position of
the conformational equilibria of 1 11, subject to the different
polarity of the R substituents, the characteristic bond lengths,
À
C H bonds in the 3- and 5-positions in the expected manner
by decreasing the H3ax and H5ax electron density whilst
concomitantly increasing the C3 and C5 electron densities and
thereby destabilizing the axial conformer overall. However,
the opposite is in fact happening: an increasing O7 electron
À
À
À
À
axial C2(6) H2(6), C2 C3, C2 C1, and C1 O7 in the two
conformers have been appropriately compared. In addition,
the electron populations of the atoms involved in the hyper-
conjugation mechanism have been calculated by NBO
analysis[6] and are presented in Table 6. Both types of
measurements provide promising indications for the presence
of hyperconjugation in the compounds 1 11 for the following
reasons:
À
density increases the polarization of the axial C3 H3
À
(C5 H5) bond and shifts electron density from C3 and C5
to the H3ax and H5ax protons, respectively. The correlations
are without doubt fine and the bond polarization model (used
also for the quantification of the g effect in 13C NMR
spectroscopy[23]) is not active in the accepted manner in the
conformational analysis of saturated six-membered ring
1) There is a clear tendency that, with increasing polarity of
À
the R substituent, the C1 C2 bond length shortens and the
À
À
systems. In addition, the axial C3(5) H3(5) bond length in
corresponding C1 O7 bond lengthens in both conformers.
the axial conformers does not change in 1 11; the same is also
true for the inclination angle q (see Table 4).
The corresponding effect on the lengths of the axial
À
À
C2(6) H2(6) bonds in the axial conformer and C2 C3 in
the equatorial conformer is less sensitive. However, there is a
clear tendency in the corresponding direct C,H coupling
Hyperconjugation: According to the Perlin effect,[16] but also
based on detailed ab initio quantum-chemical calculations,
hyperconjugation should be active in cyclohexane and its
monosubstituted derivatives.[27] This electronic interaction
within the saturated six-membered ring skeleton was origi-
nally purported to be responsible for the anomeric effect of
polar substituents in the 2-position of six-membered saturated
heterocyclic compounds (e.g., in the 2-OH oxanes as donation
of the ring-oxygen lone pair into the antibonding C O orbital
À
constants of the C(2,6) H(2,6-axial) fragments (Table 1);
with increasing population of the axial conformer also the
1J(C-2,6)(H-2,6-axial) values decrease indicating less effective
À
coupling on account of the increasing C H bond length. In
terms of hyperconjugation, this implies the presence of this
electronic interaction in the two conformers and because the
À
À
C1 O7 and C1 H1 bond lengths in the axial conformers were
found to be longer than in the equatorial conformers (the
1J(C1,H1) coupling constants also change appropriately), a
stronger contribution by hyperconjugation occurs in the case
of the axial conformer. In short, the higher the polarity of R,
the higher the population of the axial conformer. This is
complete agreement with the experimental results (Table 2).
2) Similar conclusions can be drawn from the occupation
numbers of lone pair/antibonding orbitals involved in hyper-
conjugation: the total S of occupancies of all s orbitals
À
of the exocyclic C O bond, n ! s* O). In light of the studied
À
O
C
compounds 1 11, hyperconjugation can be best represented
by the Lewis bond/nonbonded structures depicted in
Scheme 3. Consequently, for the axial conformers, extended
À
À
axial C2(6) H2(6) and C1 O7 bond lengths and concom-
À
itantly reduced C1 C2 bond lengths are anticipated if hyper-
conjugation is present or increasing. In the corresponding
À
À
equatorial conformer, the C2 C3 and C1 O7 bond lengths
À
À
are expected to appropriately increase whilst the C1 C2 bond
donating into the antibonding orbital of the C1 O7 bond
(s* O7) by way of hyperconjugation proved to be remarkably
À
C1
should shorten. In a comparison of the axial and equatorial
À
À
conformers of compounds 1 11, the C1 O7 and the C1 H1
bond lengths, and also the 1J(C1,H1) coupling constants in the
axial conformers were found to be larger than in the
equatorial conformers. In the context of hyperconjugation,
this means stronger hyperconjugation in the axial conformer
in comparison to its equatorial counterpart.
higher in the axial conformer, indicating more extensive
hyperconjugation in this conformer and corroborating also
the former experimental results. In addition, the stronger
hyperconjugation in the axial conformer correlates with the
experimental free energy differences. This is depicted in
Figure 3 which shows that increasing hyperconjugation sta-
1366
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0906-1366 $ 20.00+.50/0
Chem. Eur. J. 2003, 9, No. 6

Conformational Equilibria of Cyclohexyl Esters
1360 1368
Figure 3. Plot of the calculated, stronger donation of the s orbitals, Dd, for
the axial conformers of 1 11 against DG8exp
.
bilizes the axial conformer with respect to its equatorial
counterpart.
Figure 5. Plot of DG8exp against DG8calcd after taking account of the volume
V of the CR3 substituents and the hyperconjugation Dd, for compounds 1
11 against DG8calcd ˆ 0.125Dd ‡ 0.024VÀ 1.499. Linear regression analysis
(r2 ˆ 0.80) provides an equation, y ˆ ax ‡ b with, values of 0.899 (Æ0.154)
for a and 0.047 (Æ0.113) for b.
Steric substituent effect: From the correlation of DG8exp
versus DG8calcd given in Figure 1, it is clear that, besides the
electronic long-range substituent effect of R in 1 11, there
should be another substituent effect present, probably of
steric origin on account of the two dependencies indicated in
Figure 1. Although the correlation coefficient is not very
good, the tendency that there really is a steric effect is,
nonetheless, evident. Thus, the volumes of the R groups in 1
11 have been tabulated (Table 8) and correlated with the
employing the two variables of R volume and hyperconjuga-
tion (Figure 5). The correlation coefficient improves consid-
erably (r2 ˆ 0.89); the slope is close to 1 and the intercept close
to 0, both of which are evidence for a close dependence
[Eq. (1)].
DG8exp ˆ 0.8986DG8calcd ‡ 0.0474
(1)
Table 8. Spatial volumes[a] of the R substituents in compounds 1 11.
R
V[cm3 molÀ1
]
R
V[cm3 molÀ1
]
The stabilization of the axial conformer by the hyper-
conjugative effect of the OC(O)CR3 substituent is evident, yet
it only has a partial influence on the relative stability DGo as a
function of R. The dependence of DGo on the volume of the
CR3 group is present and should be caused by steric
interactions involving the cyclohexane ring and the whole
OC(O)CR3 group. Nevertheless, how this effect intervenes to
destabilize the equatorial conformer is a point that needs
further insight. Although bond polarization via nonbonding
1,3-diaxial interactions as a cause can be discarded (vide
supra), conversely, the reversed-substituent effect has to be
1
2
3
4
5
6
Me
Et[b]
iPr[b]
tBu[b]
CF3
11.76
16.88
22.00
27.12
15.93
16.98
7
8
9
10
11
CHCl2
CCl3
CH2Br
CHBr2
CBr3
21.85
26.36
19.43
26.26
32.37
CH2Cl
[a] Reference [29]. [b] Calculated by employing van der Waals atomic radii
and assuming spherical shapes for the atoms.
conformational equilibria of 1 11 (Figure 4). It provides a
rather good correlation (r2 ˆ 0.78), even better than the
corresponding correlation of DG8exp versus hyperconjugation
given in Figure 3. The best correlation, however, was obtained
if the least-squares treatment of a linear fit is conducted
À
considered. Thus, the characteristic bond lengths C1 O7 and
spatial distances (internuclear distances between substituent
atoms and the protons and carbon atoms of the cyclohexane
ring, see Table 4) have been examined to illuminate the steric
effects present in the two conformers of 1 11. First, there is
no clear trend amongst any of the distances examined, except
À
À
the C1 O7 bond lengths and the spatial distances H1 O7 and
À
H1 O9, and this occurs in both conformers. Furthermore, the
increase of the bond lengths and spatial distances is very
similar in the two conformers with the same R substituent.
The inclination angle q in the equatorial and axial conformers
was found to be rather independent of the volume of CR3
(Table 4). This should mean that the steric interaction of the
OC(O)CR3 substituent does not modify the geometry of the
cyclohexane ring, while it is influential in changing the relative
energy of axial and equatorial conformers. The fact that the
modifications of the geometrical features of the cyclohexane
ring in the series of compounds with different R are rather
small should be assumed as further evidence that hyper-
conjugation only partly changes the relative stability of axial/
Figure 4. Plot of the volumes of the CR3 substituents in compounds 1 11
against DG8exp. Linear regression analysis (r2 ˆ 0.78) provides an equation,
y ˆ ax ‡ b with values of 31.17 (Æ5.53) for a and 40.06 (Æ3.41) for b.
Chem. Eur. J. 2003, 9, No. 6
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0906-1367 $ 20.00+.50/0
1367

E. Kleinpeter, F. Taddei, P. Wacker
FULL PAPER
[1] Conformational Behaviour of Six-Membered Rings (Ed.: E. Juaristi),
VCH, New York, 1995 .
equatorial conformers as a funcion of R, since more significant
geometric distortions should have otherwise been expected.[28]
When the size of the CR3 group increases as a function of R,
more probably, it is able to provoke deformations within the
[2] a) F. R. Jensen, C. H. Bushweller, Adv. Alicycl. Chem. 1971, 3, 140;
b) C. H. Bushweller, Stereodynamics ofCyclohexane and Substituted
Cyclohexanes. Substituent A Values, in Conformational Behaviour of
Six-Membered Rings (Ed.: E. Juaristi), VCH, New York, 1995, p. 25.
[3] W. Steiner, W. S‰nger, J. Chem. Soc. Perkin Trans. 2, 1998, 371.
[4] J. Kirby, The Anomeric Effect and Related Stereoelectronic Effects at
Oxygen, Springer, Berlin, 1983.
À
ester group itself. In fact, the calculated C10 C1 distances
(Table 4) always have a greater value in the axial conformer
than in the equatorial analogue and this interatomic distance
increases with increasing bulk of CR3. This is particularly
evident within the series of the alkyl substituents, since the
[5] P. A. Petillo, L. E. Lerner, A.C.S. Symp. 539, 1993, p. 156, and
references therein.
À
[6] J. A. Bohmann, F. Weinhold, T. C. Farrar, J. Chem. Phys. 1997, 107,
1173; NBO 5.0, E. D. Glendenig, J. K. Badenhoop, A. E. Reed, J. E.
Carpenter, J. A. Bohmann, C. M. Morales, F. Weinhold, Theoretical
Chemistry Institute, University of Wisconsin (USA), 2001.
[7] J. T. Edward, Chem. Ind. 1959, 1102.
[8] a) E. Kleinpeter, Conformational Analysis of Six-Membered Sulfur-
Containing Heterocycles, in Conformational Behaviour of Six-Mem-
bered Rings (Ed.: E. Juaristi), VCH, New York, 1995, p. 201; b) E.
Kleinpeter, Adv. Heterocycl. Chem. 1998, 69, 217.
C10 C1 distance increases in the order Me < Et < iPr< tBu.
Furthermore, we performed the natural steric analysis within
the framework of the NBO model with the facilities of the
NBO5.0 program.[6] From the pairwise additive estimate of
steric exchange energy and referring to the difference
between the two conformers of each R-substituted molecule,
it turns out that the steric effect is always greater in the
equatorial conformer. For example, for the alkyl groups, the
contribution DE to the relative destabilization of the two
conformers are: Me 0.47, Et 0.73, iPr 0.77, tBu 1.12 kcalmolÀ1
(B3LYP/6-31G** wavefunctions). The steric effect should
thus also influence the molecular stabilization of the two
conformers as a result of intramolecular hydrogen bonding
since the relative orientation of O9 and the hydrogen atoms
H1, H2eq/H6eq changes as a function of the spatial require-
ments of the CR3 group.
[9] C. Tschierske, H. Kˆhler, Zaschke, E. Kleinpeter, Tetrahedron 1989,
45, 6987.
[10] K. Pihlaja, E. Kleinpeter, Carbon-13 NMR Chemical Shifts in
Structural and Stereochemical Analysis; Methods in Stereochemical
Analysis (Ed.: A. P. Marchand), VCH, New York, 1994, p. 207.
[11] R. Borsdorf, R. M¸ller, R. Tenner, E. Kleinpeter, Z. Chem. 1976, 16,
106.
[12] R. Borsdorf, M. Arnold, E. Kleinpeter, Z. Chem. 1977, 17, 378.
[13] R. Borsdorf, E. Kleinpeter, C. Meinel, D. Lenk, Z. Chem. 1978, 18,
185.
[14] M. J. Cook, K. Nasri, S. M. Vather, Tetrahedron Lett. 1986, 27, 3853.
[15] A. J. Kirby, N. H. Williams, J. Chem. Soc. Chem. Commun. 1992, 1285.
[16] W. F. Bailey, A. D. Rivera, K. Rossi, Tetrahedron Lett. 1988, 29, 5621.
[17] U. Salzner, P. von R. Schleyer, J. Org. Chem. 1994, 59, 2138.
[18] K. B. Wiberg, J. D. Hammer, H. Castejon, W. F. Bailey, E. L. DeLeon,
R. M. Jarrett, J. Org. Chem. 1999, 64, 2085.
Conclusion
[19] C. A. Carr, M. J. T. Robinson, D. J. Young, Tetrahedron Lett. 1989, 30,
4593.
The long-range substituent influences on the conformational
equilibria of a number of monosubstituted cyclohexanes are
partly based on electronic factors–hyperconjugation by way
[20] F. A. L. Anet, D. J. O×Leary, Tetrahedron Lett. 1989, 30, 1059.
[21] D. Cremer, K. J. Szabo, Ab-initio Studies ofSix-Membered Rings;
Present Status and Future Developments, in Conformational Behav-
iour ofSix-Membered Rings (Ed.: E. Juaristi), VCH, New York, 1995,
p. 59.
[22] E. Kleinpeter, F. Taddei, J. Molec. Struct. 2002, 585, 223.
[23] Conformational Behaviour of Six-Membered Rings (Ed.: E. Juaristi),
VCH, New York, 1995, p. 51.
[24] Gaussian98 (Revision A.12), M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski,
J. A. Montgomery, Jr., R. E. Stratmann, J. C. Burant, S. Dapprich,
J. M. Millam, A. D. Daniels, K. N. Kudin, M. Strain, C. O. Farkas, J.
Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C.
Adamo, C. S. Clifford, J. Ochterski, G. A. Petersson, P. Ayala, Y. Q.
Cui, K. D. Morokuma, K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. Cioslowski, J. V. Ortiz, A. G. Baboul, B. B. Stefanov,
G. Liu, A. Liashenko, P. Piskorz, L. Komaromi, R. Gomperts, L. R.
Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A.
Nanayakkara, C. Gonzalez, M. Challacombe, P. M. W. Gill, B.
Johnson, W. Chen, M. W. Wong, J. L. Andres, C. Gonzalez, M.
Head-Gordon, E. S. Replogle, J. A. Pople, Gaussian Inc., Pittsburgh
PA, 2002.
of sC2
! s*
and sC2 ! s* O7–but also partly on
À
H2ax
À
À
C3
À
C1
C1 O7
steric influences–ironically, by destabilization of the equato-
rial conformer with increasing volume of the R substituent.
The hyperconjugation sC2
! s* O7 in the axial conformer
À
H2ax
À
C1
is more effective than hyperconjugation sC2 ! s* O7 in the
À
C3
À
C1
equatorial counterpart, thereby leading to a preference of the
former conformer with increasingly stronger electronic inter-
À
actions between the O2CR substituent and the cyclohexane
skeleton. The linear fit obtained employing the two variables
of R, hyperconjugation and R volume, is excellent. Signifi-
cantly, there is no hint of the existence of sterically destabiliz-
ing 1,3-diaxial nonbonding interactions in the axial con-
formers. Steric interactions, occurring in the whole molecule,
estimated also from the NBO method, destabilize the
equatorial conformer and the increasing bulk of the CR3
group is one relevant component of the global effect.
[25] K. B. Wiberg, K. E. Laidig, J. Am. Chem. Soc. 1987, 109, 5935.
[26] A. E. Reed, R. B. Weinstock, F. Weinhold, J. Chem. Phys. 1985, 83,
735.
Because this electronic interaction found here is similarly
responsible for the stabilization of the axial conformer in the
case of 2-substituted, six-membered saturated heterocyclic
compounds (i.e. the anomeric effect), the question therefore
arises: can the anomeric effect really be construed as
anomalous or rather, is it a general physical-organic phenom-
enon? The mass of previous theoretical work and the results
of this study clearly indicate that the answer is ™No∫.
[27] I. V. Alabugin, J. Org. Chem. 2000, 65, 3910.
[28] V. Pophristic, L. Goodman, Nature, 2001, 411, 565.
[29] A. Bondi, Physical Properties ofMolecular Crystals , Liquids and
Glasses, Wiley, New York 1968, pp. 453 468.
Received: July 29, 2002
Revised: November 18, 2002 [F4293]
1368
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0947-6539/03/0906-1368 $ 20.00+.50/0
Chem. Eur. J. 2003, 9, No. 6