1,3-Diaxial steric effects and intramolecular hydrogen bonding in the conformational equilibria of new cis-1,3-disubstituted cyclohexanes using low temperature NMR spectra and theoretical calculations

Article abstract of DOI:10.1016/j.saa.2004.12.001

The conformational equilibria of 3-X-cyclohexanol [X = F (1), Cl (2), Br (3), I (4), Me (5), NMe₂ (6) and MeO (7)] and of 3-X- methoxycyclohexane [X = F (8), Cl (9), Br (10), I (11), Me (12), NMe₂ (13) and MeO (14)] cis isomers were

Full text of DOI:10.1016/j.saa.2004.12.001

Spectrochimica Acta Part A 62 (2005) 30–37
1,3-Diaxial steric effects and intramolecular hydrogen bonding
in the conformational equilibria of new cis-1,3-disubstituted
cyclohexanes using low temperature NMR spectra
and theoretical calculations
Paulo R. de Oliveira, Roberto Rittner∗
Physical Organic Chemistry Laboratory, Chemistry Institute, State University of Campinas, Caixa Postal 6154;
13084-971 Campinas, SP, Brazil
Received 29 October 2004; accepted 1 December 2004
Abstract
The conformational equilibria of 3-X-cyclohexanol [X = F (1), Cl (2), Br (3), I (4), Me (5), NMe2 (6) and MeO (7)] and of 3-X-
methoxycyclohexane [X = F (8), Cl (9), Br (10), I (11), Me (12), NMe2 (13) and MeO (14)] cis isomers were determined from low temperature
NMR spectra and PCMODEL calculated coupling constants. The energy differences between aa and ee conformers were obtained from these
data (ꢀGaJv and ꢀGPavC, respectively) and also by the additivity principle from data for the monosubstituted cyclohexanes (ꢀGAd). H-1 and
H-3 hydrogen vicinal coupling constants and ꢀGaJv values showed that the diequatorial conformer is predominant in the conformational equi-
av
PC
librium of the compounds studied at low temperature. However, ꢀG data show that compounds 6 and 7 constitute an exception, since they
are almost equally populated by ee and aa at room temperature, due to stabilization of their aa conformer by an intramolecular hydrogen bond.
ꢀGAd values, obtained according to the additivity principle, show a better agreement for compounds 2 and 3, since the 1,3-diaxial steric effect
is counterbalanced by the formation of an intramolecular hydrogen bond (IAHB). For the remaining compounds, ꢀGAd values underestimate
the energy differences, since the 1,3-diaxial steric effect, between X and OH or OCH3, is absent in the monosubstituted compounds used as
references. Moreover, the ꢀGaPvC, calculated from the coupling constants, obtained through the PCMODEL program, are rather smaller than
the ꢀGaJv values, since the program does not have parameters for the effect, observed in this report, of a substituent at ␥ position on coupling
constants values for the hydrogen under consideration.
© 2004 Elsevier B.V. All rights reserved.
Keywords: Conformational analysis; Low temperature NMR; 1,3-Disubstituted cyclohexanes; Theoretical calculations; Vicinal coupling constants
1. Introduction
reports the determination of conformational equilibria of new
cis-3-halocyclohexanols and the corresponding methoxy-
1
Mono- and 1,2-disubstituted cyclohexanes have been the
subject of numerous studies [1–4], but the same is not true
for 1,3-disubstituted cyclohexanes [5,6]. Recently, the occur-
rence of an intramolecular hydrogen bond (IAHB) in cis-3-
methoxy- and cis-3-N,N-dimethylamino-cyclohexanols, sta-
bilizing the diaxial conformer and suppressing the 1,3-diaxial
steric interactions, has been reported [7,8]. The present work
derivatives (Fig. 1), through low temperature H NMR and
theoretical calculations, focusing on the interplay between
steric syn-1,3-diaxial and IAHB effects in the diaxial con-
former stabilization.
2. Experimental
The 1H and 13C NMR spectra of compounds 2–5 and 8–12
were assigned through gCOSY and HSQC experiments, per-
formed with a Varian 500 spectrometer, operating at 499.88
∗
Corresponding author. Tel.: +55 19 3788 3150;
fax: +55 19 3788 3023.
E-mail address: rittner@iqm.unicamp.br (R. Rittner).
1386-1425/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.saa.2004.12.001

P.R. de Oliveira, R. Rittner / Spectrochimica Acta Part A 62 (2005) 30–37
31
1
(2)-trans: H NMR (500 MHz, CDCl3), δ 4.27 (tt, 7.95,
3.87, 1H), 4.02 (tt, 6.04, 3.13, 1H), the remaining reso-
nances could not be attributed, because the trans isomer is in
rapid equilibrium at room temperature. 13C NMR (500 MHz,
CDCl3), δ 66.4, 57.3, 43.2, 35.3, 33.6, 20.1.
Fig. 1. Conformational equilibrium of the cis isomer of 3-X-cyclohexanols
[X = F (1), Cl (2), Br (3), I (4), CH3 (5), N(CH3)2 (6) and OCH3 (7)] and
3-X-1-methoxycyclohexanes [(X = F (8), Cl (9), Br (10), I (11), CH3 (12),
N(CH3)2 (13) and OCH3 (14)].
2.2. Cis- and trans-3-bromocyclohexanol (3)
(1H)and125.70 MHz(13C). Spectrawereofca. 0.30 mol L−1
solutions in CDCl3 with a probe temperature of 20 ◦C. 1H and
13C NMR spectra were obtained under typical conditions, as
follows: 1H NMR spectra with 128 transients, accumulated
into32 Kdatapointswithapulsewidthof45◦, sweepwidthof
ca. 5000 Hz and acquisition time of ca. 2.7 s. The FIDs were
zero-filled to 128 K data points, giving a digital resolution of
0.08 Hz/point; 13C NMR spectra with 512 transients, accu-
mulated into 32 K data points, with a pulse width of 45◦, a
sweep width of ca. 20 000 Hz and acquisition time of 1 s. The
1H NMR spectra at low temperature, in CS2/CD2Cl2 (9:1),
were recorded on a Varian 300 spectrometer. Spectra were
of ca. 0.15 mol L−1 solutions with probe temperatures of 25
and −90 ◦C, operating at 300.07 (1H), and obtained under
typical conditions, as follows: 128 transients, accumulated
into 32 K data points, with a pulse width of 37◦, sweep width
of ca. 3000 Hz and acquisition time of 2.7 s. The FIDs were
zero-filled to 128 K data points, giving a digital resolution
of 0.05 Hz/point. Most FIDs were processed with Gaussian
multiplication, typically of gf = 0.25 and gfs = 0.35 for spec-
tral resolution improvement. In all cases, SiMe4 was used as
internal reference.
Cis- and trans-3-bromocyclohexanol were prepared in a
similar way. Hydrogen bromide was generated by the action
of bromine (7.5 mL, 0.15 mol) upon tetrahydronaphthalene
(70 mL) [9]. The crude product was distilled to give cis- and
1
trans-3-bromocyclohexanol (3) in a ratio (by H NMR) of
89:11 (8.8 g, 49%). bp 77–78 ◦C/1.0 Torr.
1
(3)-cis: H NMR (500 MHz, CDCl3), δ 3.93 (tt, 11.83,
4.15, 1H), 3.60 (tt, 10.55, 4.22, 1H), 2.59 (m, 1H), 2.25 (m,
1H), 2.00 (m, 1H), 1.83 (m, 1H), 1.76 (m, 1H), 1.65 (m, 1H),
1.30 (m, 1H), 1.24 (m, 1H). 13C NMR (500 MHz, CDCl3), δ
69.9, 47.4, 46.9, 36.9, 34.0, 23.6.
1
(3)-trans: H NMR (500 MHz, CDCl3), δ 4.41 (tt, 7.95,
3.87, 1H), 4.02 (tt, 6.04, 3.13, 1H), the remaining reso-
nances could not be attributed, because the trans isomer is in
rapid equilibrium at room temperature. 13C NMR (500 MHz,
CDCl3), δ 66.9, 50.0, 43.9, 36.0, 33.6, 21.0.
2.3. Cis-3-iodocyclohexanol (4)
The preparation of cis- and trans-3-iodocyclohexanol was
similar to the above derivatives. Hydrogen iodide was pre-
pared by the reaction of a solution of two parts of iodine
(20 g) and one part of hydriodic acid (10 mL, d 1.7 and 57%)
with an excess of red phosphorus (20 g) [9]. The solvent was
removed, but the residue could not be distilled without de-
composition. However, GC–MS showed that it was a sin-
gle compound, cis-3-iodocyclohexanol (4) (17.4 g, 77%), and
pure enough for our purposes.
2.1. Cis- and trans-3-chlorocyclohexanol (2)
Dry hydrogen chloride (from 25 mL of conc. hydrochlo-
ric acid and 50 mL of conc. sulfuric acid) [9] was slowly
bubbled, for about 7 h, into 9.6 g (0.10 mol) of redistilled
2-cyclohexen-1-one, placed in a 25 mL round bottomed
flask, cooled with dry ice-ethanol bath (−30 ◦C), to give 3-
chloro-cyclohexanone. This very unstable intermediate was
added dropwise to a 250 mL 3-necked round bottomed flask
containing a lithium aluminum hydride (1.9 g, 0.05 mol)
suspension in tetrahydrofuran (60 mL), under stirring at
−10 ◦C and kept under a nitrogen atmosphere. The mixture
was allowed to warm, reaching room temperature, and stirred
for 1.5 h. Carefully addition of water destroyed the excess of
lithium aluminum hydride. The organic layer was extracted
with diethyl ether, dried over MgSO4, filtered and the solvent
was evaporated. The product was distilled to give a mixture
of cis- and trans-3-chlorocyclohexanol (2) in a ratio (by 1H
NMR) of 76:24 (2.4 g, 18%). bp 71–73 ◦C/1.0 Torr.
1
(4)-cis: H NMR (500 MHz, CDCl3), δ 4.01 (tt, 12.35,
3.99, 1H), 3.55 (tt, 10.64, 4.29, 1H), 2.70 (m, 1H), 2.32 (m,
1H), 2.07 (m, 1H), 1.92 (m, 1H), 1.80 (m, 1H), 1.69 (m, 1H),
1.29 (m, 1H), 1.27 (m, 1H). 13C NMR (500 MHz, CDCl3), δ
70.3, 49.3, 39.1, 34.2, 25.5, 23.4.
2.4. Cis- and trans-3-methylcyclohexanol (5)
A mixture of cis- and trans-3-methylcyclohexanol was
obtained commercially from Aldrich, in a ratio (by 1H
NMR) of 69:31. In this case the cis- and trans-3-
methylcyclohexanol were separated through column chro-
matography using hexane–ethyl acetate (7:1) as eluent and
silica gel, 230–400 mesh.
1
1
(2)-cis: H NMR (500 MHz, CDCl3), δ 3.82 (tt, 11.51,
(5)-cis: H NMR (500 MHz, CDCl3), δ 3.56 (tt, 10.92,
4.11, 1H), 3.61 (tt, 10.52, 4.20, 1H), 2.47 (m, 1H), 2.11 (m,
1H), 1.95 (m, 1H), 1.85 (m, 1H), 1.60 (m, 1H), 1.50 (m, 1H),
1.29 (m, 1H), 1.22 (m, 1H). 13C NMR (500 MHz, CDCl3), δ
69.4, 56.4, 45.9, 36.1, 34.1, 22.3.
4.19, 1H), 1.94 (m, 2H), 1.75 (m, 1H), 1.62 (m, 1H), 1.42
(m, 1H), 1.26 (m, 1H), 1.11 (m, 1H), 0.92 (d, 6.62, 3H), 0.89
(m, 1H), 0.78 (m, 1H). 13C NMR (500 MHz, CDCl3), δ 70.7,
44.6, 35.3, 34.0, 31.4, 24.1, 22.3.

32
P.R. de Oliveira, R. Rittner / Spectrochimica Acta Part A 62 (2005) 30–37
1
(5)-trans: H NMR (500 MHz, CDCl3), δ 3.96 (tt, 4.7,
1.25 (m, 1H), 1.13 (m, 1H). 13C NMR (500 MHz, CDCl3), δ
2.4, 1H), the remaining resonances could not be attributed,
because the trans isomer is in rapid equilibrium at room tem-
perature. 13C NMR (500 MHz, CDCl3), δ 66.6, 41.8, 34.6,
33.3, 26.7, 22.2, 20.5.
78.1, 56.7, 56.0, 42.7, 36.5, 30.7, 22.6.
1
(9)-trans: H NMR (500 MHz, CDCl3), δ 4.33 (tt, 7.89,
3.86, 1H), 3.59 (tt, 6.34, 3.15, 1H), the remaining reso-
nances could not be attributed, because the trans isomer is in
rapid equilibrium at room temperature. 13C NMR (500 MHz,
CDCl3), δ 75.5, 57.6, 55.9, 40.1, 35.5, 29.6, 19.9.
2.5. Cis- and trans-3-N,N-dimethylaminocyclohexanol (6)
2.9. Cis-3-bromo-1-methoxycyclohexane (10)
Previously prepared, as recently described in Ref. [8].
This compound was prepared similarly to compound 9,
from cis-3-bromocyclohexanol. The product was distilled to
give cis-3-bromo-1-methoxycyclohexane (0.66 g, 33%); bp
62–64 ◦C/2 Torr.
2.6. Cis- and trans-3-methoxycyclohexanol (7)
Previously prepared, as recently described in Ref. [7].
1
(10)-cis: H NMR (500 MHz, CDCl3), δ 3.90 (tt, 12.09,
2.7. Cis-3-fluoro-1-methoxycyclohexane (8)
4.03, 1H), 3.35 (s, 3H), 3.10 (tt, 10.79, 4.08, 1H), 2.69
(m, 1H), 2.24 (m, 1H), 2.07 (m, 1H), 1.84 (m, 1H), 1.66
(m, 1H), 1.65 (m, 1H), 1.26 (m, 1H), 1.16 (m, 1H). 13C
NMR (500 MHz, CDCl3), δ 78.5, 56.0, 47.7, 43.7, 37.4, 30.8,
23.8.
Ten grams of 3-fluoroanisol were hydrogenated in a
100 mL autoclave, in the presence of 1.0 g of rhodium ox-
ide, Rh(Ox)Li (see below), at 60 ◦C, using a hydrogen pres-
sure of 500–700 psi. The reduction was allowed to proceed
for 6 h. The catalyst was filtered and the clear solution was
concentrated to give a complex mixture, containing 14% of
cis-3-fluoro-1-methoxycyclohexane (1.4 g), which was iso-
lated by column chromatography, using hexane as eluent and
silica gel, 230–400 mesh.
2.10. Cis-3-iodo-1-methoxycyclohexane (11)
This compound was prepared similarly to compound
9, from cis-3-iodocyclohexanol. The product was distilled
to give cis-3-iodo-1-methoxycyclohexane (1.08 g, 54%); bp
82–84 ◦C/3 Torr.
1
(8)-cis: H NMR (500 MHz, CDCl3), δ 4.46 (dtt, 48.38,
10.69, 4.12, 1H), 3.36 (s, 3H), 3.16 (m, 1H), 2.47 (m, 1H),
2.05 (m, 1H), 1.98 (m, 1H), 1.85 (m, 1H), 1.43 (m, 1H), 1.42
(m, 1H), 1.17 (m, 2H). 13C NMR (500 MHz, CDCl3), δ 90.3,
76.9, 56.0, 38.6, 32.1, 30.7, 19.2.
Catalyst. Rhodium oxide, Rh(Ox)Li, was prepared from
a lithium nitrate fusion with rhodium chloride trihydrate, as
reported by Nishimura et al. [10].
1
(11)-cis: H NMR (500 MHz, CDCl3), δ 4.00 (tt, 12.42,
3.98, 1H), 3.35 (s, 3H), 3.07 (tt, 10.72, 4.07, 1H), 2.79 (m,
1H), 2.34 (m, 1H), 2.13 (m, 1H), 1.82 (m, 1H), 1.81 (m,
1H), 1.69 (m, 1H), 1.25 (m, 1H), 1.20 (m, 1H). 13C NMR
(500 MHz, CDCl3), δ 78.9, 55.9, 45.8, 39.5, 30.9, 25.6, 23.9.
2.11. Cis- and trans-3-methyl-1-methoxycyclohexane
(12)
2.8. Cis- and trans-3-chloro-1-methoxycyclohexane (9)
These compounds were prepared similarly to compound
9. The reaction product was distilled to give cis- and trans-
3-methyl-1-methoxycyclohexane in a ratio (by 1H NMR) of
82:18 (0.88 g, 44%); bp 74–76 ◦C/17 Torr.
2.0 g (13.5 mmol) of a cis and trans mixture of 3-
chlorocyclohexanol and 50 mL of dry THF were placed in
a 2-necked 125 mL round bottomed flask, fitted with a cal-
cium chloride protected reflux condenser, a dropping funnel
and a magnetic stirrer. 0.65 g (27.0 mmol) of sodium hydride
were added and the reaction mixture was stirred at room tem-
perature for 1 h. The reaction mixture was cooled to 0 ◦C and
3.8 g (27.0 mmol) of methyl iodide in 15 mL of dry THF
were gradually added. The ice bath was removed and stirring
continued for a further 1.5 h, under reflux, and then the so-
lution was cooled to 20 ◦C and water was gradually added
to the reaction flask. The organic layer was separated, dried
over MgSO4 and filtered and the solvent was evaporated.
The product was distilled to give cis- and trans-3-chloro-1-
methoxycyclohexane in a ratio (by 1H NMR) of 86:14 (1.3 g,
59%); bp 62–64 ◦C/2 Torr.
(12)-cis: 1H NMR (500 MHz, CDCl3), δ 3.35 (s, 3H), 3.10
(tt, 10.87, 4.11, 1H), 2.02 (m, 2H), 1.76 (m, 1H), 1.60 (m,
1H), 1.41 (m, 1H), 1.24 (m, 1H), 1.04 (m, 1H), 0.87 (d, 6.65,
3H), 0.81 (m, 2H). 13C NMR (500 MHz, CDCl3), δ 79.4,
55.5, 40.8, 34.4, 31.7, 31.4, 24.1, 22.4.
1
(12)-trans: H NMR (500 MHz, CDCl3), δ 3.48 (tt, 4.4,
2.3, 1H), the remaining resonances could not be attributed,
because the trans isomer is in rapid equilibrium at room tem-
perature. 13C NMR (500 MHz, CDCl3), δ 75.7, 55.6, 38.4,
29.4, 26.7, 22.3, 20.3.
2.12. Cis- and trans-3-N,N-dimethylamino-1-
methoxycyclohexanol (13)
1
(9)-cis: H NMR (500 MHz, CDCl3), δ 3.78 (tt, 11.89,
4.14, 1H), 3.35 (s, 3H), 3.12 (tt, 10.96, 4.11, 1H), 2.58 (m,
1H), 2.15 (m, 1H), 2.05 (m, 1H), 1.86 (m, 1H), 1.50 (m, 2H),
Previously prepared, as recently described in Ref. [8].

P.R. de Oliveira, R. Rittner / Spectrochimica Acta Part A 62 (2005) 30–37
33
and low temperature (3JH1e/H2e and 3JH1a/H2a) were used for
the determination of the molar fraction of ee conformer (Xee)
and to estimate the ꢀGJ values through Eqs. (1) and (2),
respectively [12]. As it was not possible to determine the
experimental value of 3JH1e,H2e for the aa conformer, an ap-
proximation, suggested by other authors [13], was made by
2.13. Cis- and trans-1,3-dimethoxycyclohexanol (14)
Previously prepared, as recently described in Ref. [7].
3. Results and discussion
3
considering that 3JH1a/H2e ≈ JH1e/H2e
3.1. Temperature effects
3
3
Xee = (3Jobs − JH1e/H2e/3JH1a/H2a − JH1e/H2e
)
(1)
Conformational energies can usually be obtained by mea-
surements of integral intensities in the NMR spectra at
low temperatures [11]. However, it was not possible to de-
termine the axial–axial conformer population for the cis-
1,3-disubstituted cyclohexanes, described here, through this
method, since it was usually at too low concentration in the
equilibrium, due to 1,3-diaxial steric effects. It is noteworthy
that even for compounds 6 and 7, which show a very large
proportion of the aa conformer at room temperature, due to
an IAHB [7,8], the equilibrium is fully shifted to the ee con-
former at low temperatures, indicating that the IAHB of these
compounds is lost under these conditions.
as Xee + Xaa = 1, the free energy difference (ꢀG◦) is, thus,
readily obtained from Eq. (2), where R = 1.99 cal mol K−1
T = 298 K and K1 = Xee/Xaa, where X is the conformer molar
,
fraction
ꢀG◦ = −RT ln K1
(2)
In the second method [14,15], the experimental coupling con-
stants values (3Jobs) at room temperature, together with the
calculated values for aa and ee conformers obtained through
the PCMODEL program [16] using Haasnoot–Altona equa-
tions [17], were used to estimate the conformer populations.
The coupling constants values for the H-1 and H-3 hydrogens
ofthecisisomerofcompounds1–14, atroomandlowtemper-
Therefore, two different methods were used to determine
the aa and ee conformer populations. In the first method, the
vicinal coupling constants values obtained at room (3Jobs
)
Table 1
Calculateda and experimentalb,c coupling constants (3JH,H
)
d
Conformer
3JH1/H2a or 3JH1/H6a
3JH1/H2e or JH1/H6e
3JH3/H2a or 3JH3/H4a
3JH3/H2e or JH3/H4e
1aa
1ee
2.44
11.14
3.69
4.82
1.52
11.23
4.07
4.84
2aa
2ee
3aa
3ee
4aa
4ee
5aa
5ee
6aa
6ee
7aa
7ee
8aa
8ee
9aa
9ee
2.62
3.59
3.40
3.09
11.14, 10.47b, 10.85c
4.79, 4.08b, 3.77c
3.56
11.56, 11.38b, 11.72c
4.23, 4.10b, 3.95c
2.68
3.75
2.86
11.14, 10.48b, 10.74c
2.50
4.78, 4.16b, 4.07c
3.75
11.79, 11.77b, 12.12c
4.07, 4.11b, 4.09c
4.48
2.25
11.14, 10.52b
2.60
4.79, 4.06b
3.61
12.24, 12.22b, 12.50c
3.69, 3.95b, 3.95c
4.74
12.34
2.92
11.81
1.98
11.21, 7.25b
2.03
3.31
3.71
3.11
4.23
4.71, 3.61b
11.13, 10.85b, 10.95c
2.05
4.79, 4.25b, 4.18c
4.41
11.15, 7.88b
1.91
4.73, 3.81b
4.40
11.14, 7.35b
2.38
4.79, 3.67b
3.84
1.64
4.08
11.21, 11.38c
2.56
4.69, 3.82c
3.72
11.26, 10.63b, 11.19c
4.75, 4.41b, 4.02c
3.45
3.15
11.22, 10.79b, 11.05c
2.65
4.66, 4.09b, 4.03c
3.65
11.58, 11.78b, 11.97c
4.16, 4.17b, 4.04c
10aa
10ee
11aa
11ee
12aa
12ee
13aa
13ee
14aa
14ee
a
3.76
2.94
11.21, 10.72b, 10.90c
2.65
4.65, 4.08b, 4.03c
3.64
11.81, 12.05b, 12.18c
4.00, 4.11b, 4.04c
4.45
2.36
11.21, 10.63b, 10.76c
2.53
4.67, 4.08b, 4.01c
3.76
12.23, 12.34b, 12.48c
3.69, 3.94b, 3.88c
4.80
12.35
3.19
11.81
2.01
3.23
3.50
3.05
11.21, 10.84b, 10.92c
2.44
4.66, 4.12b, 4.07c
4.04
11.22, 10.83b, 10.86c
2.38
4.60, 4.07b, 4.04c
3.98
2.48
3.76
11.23, 10.62b, 10.98c
4.62, 3.98b, 3.98c
11.21, 10.62b, 10.98c
4.67, 3.98b, 3.98c
Calculated through the PCMODEL program.
b
c
d
From 1H NMR spectra at 20 ◦C in CS2/CD2Cl2 (9:1).
From 1H NMR spectra at −90 ◦C in CS2/CD2Cl2 (9:1).
In Hz.

34
P.R. de Oliveira, R. Rittner / Spectrochimica Acta Part A 62 (2005) 30–37
ature, togetherwithvaluescalculatedthroughthePCMODEL
program are shown in Table 1. For cis-3-fluorocyclohexanol
(1), only theoretical data are reported since it was not possible
to synthesize this compound.
into account the substituent electronegativity ␣ effect on vic-
inal coupling constants.
Compounds 12–14 display a similar behavior to 8–11,
due to the lack of parameters for the substituent [CH3, OCH3
and N(CH3)2] at the ␥ position. The calculated (PCMODEL)
coupling values for H-1 hydrogen (Table 1) are constant
(∼11.22 Hz), while the low temperature values show small
changes (10.92, 10.86 and 10.98 Hz).
Table 1 shows that the coupling constants (3JH1/H2a
3
or JH1/H6a) decrease with the increase in halogen size
(compounds 2, 3 and 8–11), at low temperature, since
they decrease from 11.38 to 10.76 Hz for compounds
8–11, respectively. The same vicinal coupling constants,
at room temperature, also decrease from 10.79 to 10.63 Hz
for compounds 9–11. However, the PCMODEL program
gives the same value for compounds 1–4 (11.14 Hz)
and also for compounds 8–11 (11.21 Hz), which is in
disagreement with experimental data. The program takes
into account the presence of a fixed substituent at C-1,
for compound 1–4 (hydroxyl) and 8–11 (methoxyl), but
not the halogen atom at C-3, and, thus, the experimental
results suggest the need of establishing parameters for a
␥-halogen atom for use in the Haasnoot–Altona equations
[17].
Moreover, the experimental couplings for 6 and 7 cannot
be compared with those from PCMODEL, since it is known
they are largely dependent on concentration and solvent, ex-
hibiting values of 6.82 and 6.65 Hz in C6D12 and 10.43 and
10.92 in DMSO, for 6 and 7, respectively [7,8].
Table 2 presents ꢀGaJv and ꢀG values, which are av-
av
PC
eraged from the corresponding ꢀG for H-1 and H-3 hy-
drogens, from experimental and calculated (PCMODEL) J
values, respectively. They show that the ee conformer pop-
ulation (≥92%) is always much larger than that of aa con-
former, according to their ꢀGaJv values (≥1.46 kcal mol−1),
except for compounds 6 and 7. Data from Table 2 also
show that these equilibria are very sensitive to substituent
steric effects, since ꢀGaJv increase on increasing the sub-
stituent size at the C-3 carbon. Thus, the largest ꢀGJav
value corresponds to compound 13, which presents the
bulkiest substituent, the dimethylamino group. Moreover,
ꢀGaJv for 14 (1.73) lies between 8 (1.46) and 9 (2.07),
as expected, since OCH3 presents a steric effect inter-
mediate between F and Cl, according to Charton’s [20]
steric effect (υef 0.36, 0.27 and 0.55, for OCH3, F and Cl,
respectively).
Low temperature coupling constants for H-3 hydrogen
(3JH3/H2a or 3JH3/H4a) increase with the increase in the halo-
gen atom size, since they increase from 11.19 to 12.48 Hz
for compounds 8–11. This shows that the substituent X ␣ ef-
fect, in the vicinal coupling constants, is opposite to ␥ effect
of same substituent. The electronegative substituent ␣ effect
on vicinal coupling constants, was also reported by Karplus
[18,19], who showed that 3JH,H decreases as the substituent
electronegativity increase.
3
3
Low temperature JH3/H2a or JH3/H4a agree with val-
ues calculated through PCMODEL, since, in this case, the
Karplus [18,19] and Haasnoot–Altona [17] equations take
av
PC
ꢀG underestimates the conformer energy differences
for compounds 8–14, showing again the effect of lacking a
Table 2
Energy differencesa obtained from J values calculated through the PCMODEL program (ꢀGPC), from experimental coupling constants (ꢀGJ), the corresponding
av
averaged values (ꢀG and ꢀGaJv), and obtained through additivity of data from monosubstituted compounds (ꢀGAd
)
PC
ꢀGaJv
ꢀGAd
b
c
b
c
av
PC
d
Compound
ꢀGPC
ꢀGPC
ꢀGJ
ꢀGJ
ꢀG
1
2
3
4
5
6
7
8
9
10
11
12
13
14
–
–
2.27
3.62
3.69
–
–
–
–
–
1.53 0.20
1.75 0.15
1.78 0.09
1.72 0.15
2.91 0.04
2.64 0.06e
1.66 0.04e
0.97 0.20
1.19 0.15
1.22 0.09
1.16 0.15
2.35 0.02
2.08 0.06
1.10 0.02
1.38
1.39
1.42
1.93
0.04
−0.15
1.39
1.66
1.58
1.48
1.75
1.69
1.42
1.70
1.90
–
2.49
–
1.83
1.83
2.01
–
–
–
1.83 0.45
2.51 1.12
2.55 1.14
1.93 0.04
0.04 0.04
−0.16 0.01
1.39 0.04
1.66 0.04
1.58 0.04
1.48 0.04
1.75 0.04
1.69 0.04
1.42 0.04
1.77 0.07
1.87 0.04
2.01 0.04
2.49 0.04
–
–
−0.16
–
–
–
1.46
1.93
2.14
2.33
2.63
3.22
1.73
–
1.46 0.04
2.07 0.14
2.29 0.15
2.38 0.05
2.63 0.04
3.22 0.04
1.73 0.04
f
–
–
–
2.20
2.44
2.43
–
–
–
f
f
–
–
–
a
In kcal mol−1
.
b
c
d
e
f
From H-1 coupling constants.
From H-3 coupling constants.
See text.
These figures have no physical meaning (see text).
Jobs is out of the range Jaa–Jee (calculated through the PCMODEL program).

P.R. de Oliveira, R. Rittner / Spectrochimica Acta Part A 62 (2005) 30–37
35
parameter to take into account the ␥ substituent effect for the
estimation of hydrogen coupling constants.
3.2. Additivity of conformational energies
The additivity principle considers that the conforma-
tional equilibrium constants for 3-X-cyclohexanols and 3-X-
1-methoxycyclohexanes [X = F, Cl, Br, I, CH3, N(CH3)2 and
OCH3] can be obtained from the three series of monosub-
stituted cyclohexanes: X-cyclohexanes, cyclohexanols and
methoxycyclohexanes [21]. Clearly, this leads to the further
implication that the conformational energies are additive,
that is, for ꢀGAd = ꢀGX + ꢀGOH or ꢀGOR, where ꢀGX is
for X-cyclohexanes, ꢀGOH is for cyclohexanol and ꢀGOR
for methoxycyclohexane. Table 2 lists the additivity energies
(ꢀGAd) calculated from ꢀG◦ values for the monosubstituted
cyclohexanes. We have chosen the following values of ꢀG◦
(kcal mol−1) for F (0.42 0.20) [22], Cl (0.64 0.15) [22],
Br (0.67 0.09) [22], I (0.61 0.15) [22], CH3 (1.80 0.02)
[23], N(CH3)2 (1.53 0.06) [24], OCH3 (0.55 0.02) [25]
and OH (1.11 0.04) [26]. Error limits have been chosen
from the data in the cited references.
ꢀGAd is in good agreement with experimental values
(ꢀGaJv) for compounds 2–5 and 12, while for compounds
8–11, 13 and 14 ꢀGAd is significantly smaller than ꢀGJav
(Table 2) and the differences can be attributed to the
1,3-diaxial steric effect between the substituent X and the
OCH3 group. The excellent agreement for compounds 2 and
3 may be due to an IAHB, which stabilizes the aa conformer
(see next section), overcoming the steric effect between the
substituents.
Fig. 2. Possible conformers for the cis isomer of compounds 1–5 (R = H)
and 8–12 (R = CH3).
ues of 0.04 and −0.16 for compounds 6 and 7, respectively.
Therefore the ꢀGAd values for these compounds, presented
in parenthesis in Table 2, have no physical meaning, since
they are in complete disagreement with experimental data.
3.3. Theoretical calculations
Theoretical calculations were performed for the cis
isomers of compounds 1–5 and 8–12, since the remaining
compounds have already been studied [7,8]. The possible ee
and aa conformers are shown in Fig. 2, and their energies
were minimized using the Gaussian98 program [27] with
6-311 + g** basis set at the B3LYP [28] level of theory.
The resulting relative energies (ꢀEaa–ee) are given in
Table 3.
The ꢀEaa–ee values for the cis isomer of compound
1 show that the 1aa1 conformer is slightly more stable
(0.06 kcal mol−1) than 1ee1, which means that the IAHB ef-
fect is slightly larger than the 1,3-diaxial steric effect. More-
over, the geometry of 1aa2 conformer does not allow the
formation of an IAHB and turns this conformer much less
stable (2.84 kcal mol−1) than 1aa1 and less stable than the
other 1ee conformers. The 1aa3 conformer is not stable at all
since, during optimization, it assumes the 1aa1 geometry.
The ee conformer stability increases with the increasing
size of the halogen atom for compounds 2–4, since 2ee1, 3ee1
and 4ee2 conformers are 1.59, 1.83 and 3.10 kcal mol−1 more
stable than 2aa1, 3aa1 and 4aa3, respectively (Table 3). The
significant differences between 2 and 3 in relation to 4 must
It was not possible to obtain the vicinal coupling con-
stants for H-1 and H-3 hydrogens, in the low temperature
1H NMR spectra for compounds 6 and 7 since the multiplets
were not clearly resolved, but it could be concluded from
the H-1 chemical shift and from the 13C NMR spectrum that
just the ee conformer was present at low temperature. It had
been observed that the aa and ee conformer populations are
very similar at room temperature [7,8], showing that this the-
ory cannot be applied when an strong IAHB is formed, since
av
ꢀGaJv values are near to zero, as is shown by ꢀG val-
PC
Table 3
Relative energies (ꢀEaa–ee)a of possible conformers for the cis isomer of compounds 1–5 and 8–12, at B3LYP/6-311 + g** level
Conformer
ꢀEaa–ee
1
2
3
4b
5
8
9
10
11b
12
ee1
ee2
ee3
aa1
aa2
aa3
−0.06
0.10
0.20
0.00
2.84
0.00
0.10
0.20
1.59
0.00
0.15
0.18
1.83
0.03
0.00
0.13
3.89
–
3.10
0.00
0.01
0.14
–
3.66
3.65
0.00
0.00
2.44
–
2.55
2.00
0.12
0.00
2.55
–
3.87
3.15
0.10
0.00
2.55
0.00
0.03
2.27
0.00
0.01
2.55
–
3.41
3.58
c
c
c
c
c
c
–
–
–
–
c
c
c
c
c
–
–
d
d
d
–
–
–
3.42
3.72
a
In kcal mol−1
Basis set 3–21 g.
ꢀE > 4.0 kcal mol−1
.
b
c
d
.
Not stable, changing to aa1 conformer in the optimization process.

36
P.R. de Oliveira, R. Rittner / Spectrochimica Acta Part A 62 (2005) 30–37
be due to presence of an IAHB in the 2aa1 (Cl HO) and
3aa1 (Br HO) conformers.
Acknowledgments
Hydrogen bond energies are usually calculated as differ-
ences between the energies of a bonded and a non-bonded
species (ꢀEHB = Eref − Ebonded). There are two ways to esti-
mate the strength of an IAHB: through optimization of a ref-
erence structure [29–31] or without its optimization [32,33].
The energy of IAHB (ꢀEHB) for the diaxial conformers of
compounds 1–3, with optimization of reference structure
(aa2 conformer), was performed at the B3LYP level with
6-311 + g** basis set, since it was found to be more appro-
priate for similar compounds recently studied [7,8]. It was
observed that the ꢀEHB value (3.12 kcal mol−1) for com-
The authors thank FAPESP for financial support of this re-
search and for a fellowship (to PRO), CNPq for a fellowship
(to RR) and CENAPAD-SP for computer facilities (Gaus-
sian 98). Professor C.H. Collins’ assistance in revising this
manuscript is also gratefully acknowledged.
References
[1] E.L. Eliel, S.H. Wilen, L.N. Mander, Stereochemistry of Organic
Compounds, Wiley, New York, 1994, p. 690 and references therein.
[2] K.B. Wiberg, J.D. Hammer, H. Castejon, W.F. Bailey, E.L. Deleron,
R.M. Jarret, J. Org. Chem. 64 (1999) 2085–2095.
[3] D. Kost, N. Peor, G.S. Moriah, Y. Sharabi, D.T. Durocher, M. Raban,
J. Org. Chem. 67 (2002) 6938–6943.
pound 1 is larger than for compounds 2 (2.67 kcal mol−1
)
and 3 (2.70 kcal mol−1), showing that the IAHB energy is
larger for F HO than for Cl HO and Br HO, since F is
more electronegative than Cl and Br.
[4] M.P. Freitas, C.F. Tormena, P.R. Oliveira, R. Rittner, J. Mol. Struct.
(Theochem.) 589 (2002) 147–151.
Table 3 also shows that 8ee2, 9ee2, 10ee2 and 11ee1
are 2.00, 3.15, 3.42 and 3.72 kcal mol−1 more stable than
8aa3, 9aa3, 10aa3 and 11aa3, respectively. A comparison of
ꢀEaa–ee values of compounds 8–11 with those of compounds
1–4, shows that the energy differences for the former com-
pounds are much larger because the 1,3-diaxial steric effect
for the aa conformer of those compounds is also larger, since
they can not form an IAHB. ꢀEaa–ee values for 1–4 and 8–11
show that the increase in halogen size leads to the expected
increase in the 1,3-diaxial steric effect.
It was also observed that for X = CH3 (compounds 5 and
12), the ee conformer (5ee1 and 12ee1) is far more stable than
the corresponding aa conformer (5aa3 and 12aa2) by 3.65
and 3.41 kcal mol−1, respectively, due to the steric effect of
the methyl group.
[5] E.J. Corey, N.F. Feiner, J. Org. Chem. 45 (1980) 765–780.
[6] B. Franzus, B.E. Hudson, J. Org. Chem. 28 (1963) 2238–2244.
[7] P.R. Oliveira, R. Rittner, Spectrochim. Acta, Part A 61 (2005)
1737–1745.
[8] P.R. Oliveira, D.S. Ribeiro, R. Rittner, J. Phys. Org. Chem. 17
(2004).
[9] B.S. Furniss, A.J. Hannford, V. Rogers, P.W.G. Smith, A.R. Tatchell,
Vogel’s Textbook of Practical Organic Chemistry, Longman, London,
1978.
[10] S. Nishimura, H. Uchino, H. Yoshino, Bull. Chem. Soc. Jpn. 41
(1968) 2194–2195.
[11] E.A. Basso, P.R. Oliveira, J. Caetano, I.T.A. Schuquel, J. Braz.
Chem. Soc. 12 (2001) 215–222.
[12] N.S. Zefirov, L.G. Gurvich, A.S. Shashkov, M.Z. Krimer, E.A.
Vorob’eva, Tetrahedron 32 (1976) 1211–1219.
[13] R.J. Abraham, N. Aboitiz, J. Merrett, B. Sherborne, I. Whitcombe,
J. Chem. Soc., Perkin Trans. 2 (2000) 2382–2392.
[14] R.J. Abraham, E.J. Chambers, W.A. Thomas, J. Chem. Soc., Perkin
Trans. 2 (1993) 1061–1066.
[15] R.J. Abraham, T.A.D. Smith, W.A. Thomas, J. Chem. Soc., Perkin
Trans. 2 (1996) 1949–1955.
ꢀEaa–ee values are in very good agreement with the ex-
perimental values for 2 and 3, but overestimate the energy
differences for 4, 5 and 8–12.
[16] PCModel, Version 7, Serena Software, Bloomington, IN, USA,
2001.
[17] C.A.G. Haasnoot, F.A.A.M. Leeuw, C. Altona, Tetrahedron 36
(1980) 2783–2792.
4. Conclusion
[18] M. Karplus, J. Chem. Phys. (1959) 30–31.
H-1 and H-3 hydrogen vicinal coupling constants and
ꢀGaJv values show that the diequatorial conformer is al-
ways predominant in the conformational equilibrium of 1,3-
disubstituted cyclohexanes at low temperature. ꢀGAd val-
ues, obtained according to the additivity principle, show a
better agreement for compounds 2 and 3, since the 1,3-
diaxial steric effect is counterbalanced by the formation of
an intramolecular hydrogen bond (IAHB). For the remaining
compounds, ꢀGAd values underestimate the energy differ-
ences, since the 1,3-diaxial steric effect, between X and OH
or OR, is absent in the monosubstituted compounds used
[19] M. Karplus, J. Am. Chem. Soc. 85 (1963) 2870–2871.
[20] M. Charton, Top. Curr. Chem. 114 (1983) 57–91.
[21] E.L. Eliel, S.H. Schroeter, T.J. Brett, F.J. Biros, J.C. Richer, J. Am.
Chem. Soc. 88 (1966) 3327–3334.
[22] D.E. Bugay, C.H. Bushweller, C.T. Danehey, S. Hoogasian,
J.A. Blersch, W.R. Leenstra, J. Phys. Chem. 93 (1989) 3908–
3911.
[23] K.B. Wiberg, J.D. Hammer, H. Castejon, W.F. Bailey, E.L. Deleron,
R.M. Jarret, J. Org. Chem. 64 (1999) 2085–2095.
[24] H. Booth, M.L. Jozefowicz, J. Chem. Soc., Perkin Trans. 2 (1976)
895–901.
[25] F.R. Jensen, C.H. Bushweller, B.H. Beck, J. Am. Chem. Soc. 91
(1969) 344–351.
[26] H.J. Schneider, V. Hoppen, J. Org. Chem. 43 (1978) 3866–3873.
[27] M.J. Frisch, G.W. Trucks, H.B. Schlegel, J.R. Cheeseman, P.M.W.
Gill, M.W. Wong, J.B. Foresman, B.G. Johnson, M.A. Robb, E.S.
Replogle, R. Gomperts, J.L. Andres, K. Raghavachavi, J.S. Blink-
ley, C. Gonzalez, R.L. Martin, D.J. Fox, D.J. Defrees, J. Baker,
J.P. Stewart, T. Keith, G.A. Peterson, J.A. Montgomery, M.A. Al-
Laham, V.G. Zakrzewski, J.V. Ortiz, J. Cioslowski, B.B. Stefanov,
av
as references. Moreover, the ꢀG calculated through the
PC
PCMODEL program are lower than the ꢀGaJv values, since
the program does not have parameters for a substituent at
the ␥ position, in relation to the hydrogen under consid-
eration, when used for the estimation of vicinal coupling
constants.

P.R. de Oliveira, R. Rittner / Spectrochimica Acta Part A 62 (2005) 30–37
37
A. Nanayakkara, M. Challacombe, C.Y. Peng, P.Y. Ayala, W. Chen,
M. Head-Gordon, J.A. Pople, Gaussian 98, Revision A.7, Gaussian,
Inc., Pittsburgh, PA, 1998.
[30] J.S. Craw, G.B. Bacskay, J. Chem. Soc., Faraday Trans. 88 (1992)
2315–2321.
[31] T. Schaefer, R. Sebastian, D.M. Mckinnon, P.W. Spevack, K.J. Cox,
C.S. Takeucki, Can. J. Chem. 71 (1993) 960–967.
[32] A. Simperler, W. Mikenda, Monatsh. Chem. 128 (1997) 969–980.
[33] H. Lampert, W. Mikenda, A. Karpfen, J. Phys. Chem. 100 (1996)
7418–7425.
[28] J.B. Foresman, A. Frish, Exploring Chemistry with Electronic Struc-
ture Methods, Gaussian, Inc., Pittsburgh, PA, 1996.
[29] P. George, C.W. Bock, M. Trachtman, J. Mol. Struct. (Theochem.)
133 (1985) 11–24.