Conformational study of cyclodecane and substituted cyclodecanes by dynamic NMR spectroscopy and computational methods

Article abstract of DOI:10.1021/ja973116c

Low-temperature ¹³C NMR spectra of cyclodecane (1) showed the presence of a minor conformation, assigned to the twist-boat-chair-chair (TBCC), in addition to the expected boat-chair-boat (BCB). If only the TBCC and BCB conformations were assumed to be appreciably populated, then a free-energy difference between the two conformations of 0.73 ± 0.3 kcal/mol could be obtained from the five area measurements over a temperature range of -148.6 to -131.0 °C, with populations of 5.2 and 94.8% for the TBCC and BCB conformations at -146.1 °C. However, an alternative description of the conformations of 1 was suggested by the ab initio calculations, which predicted that the twist-boat-chair (TBC) and TBCC conformations have comparable free energies and populations. Equal amounts of TBCC and TBC would give populations of 5.2, 5.2, and 89.6% and relative free energies of 0.72, 0.72, and 0.00 kcal/mol for the TBCC, TBC, and BCB conformations at -146.1 °C, based on the experimental areas at this temperature. The experimental spectra could neither confirm nor disprove the presence of the TBC. Saunders' calculations of the strain energies of 1 using Allinger's MM3 program were reproduced to obtain a complete set of these parameters and drawings of the conformations, and free energies and populations were obtained at +25 and - 171.1 °C. Free energies were also calculated at the HF/6-31G* and HF/6- 311G* levels, and chemical shifts were obtained for three conformations at the HF/6-311G* level by the GIAO method. Chlorocyclodecane (2) was shown by ¹³C and ¹H NMR spectroscopy to have three conformations at -165.5 °C. To aid in conformational assignments, the ¹³C chemical shifts were calculated for all of the BCB and TBCC conformations of 2 using the GIAO method at the HF/6-311G* level. The free energies for each of the possible BCB, TBCC, and TBC conformations were also calculated using Allinger's MM3 program. From the line shape changes in the experimental ¹³C NMR spectra, the free-energy barriers, a consideration of the X-ray structures of substituted cyclodecanes, and these calculated chemical shifts and free energies, the three conformations of 2 at -165.5 °C were suggested to be 2e BCB (31.2%), 2a BCB (14.9%), and a TBCC conformation (53.9%) (numbering as in Figure 1); the 2e and 2a BCB assignments could be reversed. Free-energy barriers for interconversion of BCB conformations of 2 at -159.8 °C were 5.4 ± 0.2 and 5.5 ± 0.2 kcal/mol, and the free-energy barriers at -120.9 °C for equilibration of the TBCC conformation with the rapidly interconverting BCB conformations were 7.07 ± 0.2 and 7.08 ± 0.2 kcal/mol. The ¹³C NMR spectrum of cyclodecyl acetate (3) at -160.0 °C showed a similar pattern of chemical shifts and intensities for the substituted ring carbon.

Full text of DOI:10.1021/ja973116c

J. Am. Chem. Soc. 1998, 120, 10715-10720
10715
Conformational Study of Cyclodecane and Substituted Cyclodecanes
by Dynamic NMR Spectroscopy and Computational Methods
Diwakar M. Pawar, Sumona V. Smith, Hugh L. Mark, Rhonda M. Odom, and
Eric A. Noe*
Department of Chemistry, Jackson State UniVersity Jackson, Mississippi 39217-0510
ReceiVed September 4, 1997
Abstract: Low-temperature 13C NMR spectra of cyclodecane (1) showed the presence of a minor conformation,
assigned to the twist-boat-chair-chair (TBCC), in addition to the expected boat-chair-boat (BCB). If only the
TBCC and BCB conformations were assumed to be appreciably populated, then a free-energy difference between
the two conformations of 0.73 ( 0.3 kcal/mol could be obtained from the five area measurements over a
temperature range of -148.6 to -131.0 °C, with populations of 5.2 and 94.8% for the TBCC and BCB
conformations at -146.1 °C. However, an alternative description of the conformations of 1 was suggested by
the ab initio calculations, which predicted that the twist-boat-chair (TBC) and TBCC conformations have
comparable free energies and populations. Equal amounts of TBCC and TBC would give populations of 5.2,
5
.2, and 89.6% and relative free energies of 0.72, 0.72, and 0.00 kcal/mol for the TBCC, TBC, and BCB
conformations at -146.1 °C, based on the experimental areas at this temperature. The experimental spectra
could neither confirm nor disprove the presence of the TBC. Saunders’ calculations of the strain energies of
1
using Allinger’s MM3 program were reproduced to obtain a complete set of these parameters and drawings
of the conformations, and free energies and populations were obtained at +25 and -171.1 °C. Free energies
were also calculated at the HF/6-31G* and HF/6-311G* levels, and chemical shifts were obtained for three
1
3
conformations at the HF/6-311G* level by the GIAO method. Chlorocyclodecane (2) was shown by C and
1
H NMR spectroscopy to have three conformations at -165.5 °C. To aid in conformational assignments, the
1
3
C chemical shifts were calculated for all of the BCB and TBCC conformations of 2 using the GIAO method
at the HF/6-311G* level. The free energies for each of the possible BCB, TBCC, and TBC conformations
1
3
were also calculated using Allinger’s MM3 program. From the line shape changes in the experimental
C
NMR spectra, the free-energy barriers, a consideration of the X-ray structures of substituted cyclodecanes,
and these calculated chemical shifts and free energies, the three conformations of 2 at -165.5 °C were suggested
to be 2e BCB (31.2%), 2a BCB (14.9%), and a TBCC conformation (53.9%) (numbering as in Figure 1); the
2e and 2a BCB assignments could be reversed. Free-energy barriers for interconversion of BCB conformations
of 2 at -159.8 °C were 5.4 ( 0.2 and 5.5 ( 0.2 kcal/mol, and the free-energy barriers at -120.9 °C for
equilibration of the TBCC conformation with the rapidly interconverting BCB conformations were 7.07 ( 0.2
and 7.08 ( 0.2 kcal/mol. The 13C NMR spectrum of cyclodecyl acetate (3) at -160.0 °C showed a similar
pattern of chemical shifts and intensities for the substituted ring carbon.
The conformations of cyclodecane have been of interest for
many years.1 X-ray diffraction studies of a number of
crystalline derivatives generally showed the conformation to be
the BCB (1a), and this conformation has also been assigned to
1a
1
3
2a
cyclodecane in the solid state by C NMR spectroscopy and
2
b
3-7
X-ray diffraction. The early force-field calculations
on
cyclodecane included this conformation along with others, and
these calculations also indicated the BCB to be of lowest energy.
The initial studies by X-ray diffraction and molecular mechanics
(1) Summaries of work in this area include (a) Dunitz, J. D. In
PerspectiVes in Structural Chemistry; Dunitz, J. D., Ibers, J. A., Ed.; Wiley:
New York, 1968; Vol. II. (b) Burkert, V.; Allinger, N. L. Molecular
Mechanics; American Chemical Society: Washington, DC, 1982; pp 105,
1
06. (c) Dale, J. Stereochemistry and Conformational Analysis; Verlag
Chemie: New York, 1978, pp 50, 207-213.
2) (a) Drotloff, H.-O., Doctoral Thesis, Universitat Freiburg, West
Germany, 1987. (b) Shenhav, H.; Schaeffer, R. Crystallogr. Struct. Commun.
981, 10, 1181.
(
1
(3) Hendrickson, J. B. J. Am. Chem. Soc. 1964, 86, 4854.
(4) Wiberg, K. B. J. Am. Chem. Soc. 1965, 87, 1070.
(5) Bixon, M.; Lifson, S. Tetrahedron 1966, 23, 769.
(6) Hendrickson, J. B. J. Am. Chem. Soc. 1967, 89, 7036.
(7) Hendrickson, J. B. J. Am. Chem. Soc. 1967, 89, 7047.
(8) Reference 1b, p 106.
8
led to the perception that the preference for the BCB in
cyclodecane is comparable to the situation for the six-membered
1
0.1021/ja973116c CCC: $15.00 © 1998 American Chemical Society
Published on Web 10/06/1998

10716 J. Am. Chem. Soc., Vol. 120, No. 41, 1998
Pawar et al.
most stable of these conformations. Populations were estimated
by weighting the four theoretical curves to obtain the best match
with the experimental data. In this way, the following composi-
tion was predicted: BCB, 49%; TBC, 35%; TBCC, 8%; and
BCC (1d), 8%.
1
,1-Difluorocyclodecane was studied in solution by 19F NMR
spectroscopy, and a single AB quartet was observed at slow
1
3
exchange (-164 °C). This result was interpreted in terms of
the BCB conformation, with the fluorines at positions 2a and
2
e, and the free-energy barrier was 5.7 kcal/mol. Proton NMR
spectra for a 0.4% solution of cyclodecane in CF2Cl2 have been
14
reported for temperatures to -162 °C. Three broad, overlap-
ping peaks were found at the lowest temperature. No discussion
1
4
of the number or identity of possible conformations was made,
but a barrier of approximately 6 kcal/mol was estimated. 13C
NMR spectra for cyclodecane were obtained below room
temperature, but a limiting spectrum was not recorded because
of the low solubility of the compound. A successful low-
temperature C NMR study of cyclodecane does not appear to
have been reported in the primary literature, although the slow-
exchange C NMR spectrum of this compound has been
14
1
3
Figure 1. Strain energies of cyclodecane conformations, calculated
1
6
by Saunders using Allinger’s MM2 and MM3 force fields. The
structures are shown in order of increasing strain energy according to
MM3 (second number in parentheses). For the TBC conformation,
carbons 5 and 6 are in the front.
1
3
1
5
described as consisting of only 3 peaks for the BCB in a ratio
of 4:4:2.
Later molecular mechanics calculations of strain energies by
Saunders, using both the MM2 and MM3 force fields of
ring, where the chair conformation is found except in unusual
circumstances.9
16
Allinger, appear to support the possibility that more than one
conformation of cyclodecane could be populated. We have
reproduced these calculations, and results for the first six
conformations are shown in Figure 1. Although the steric
energies are not directly related to populations, the low values
for the TBCC, the TBC, and a slightly distorted BCC conforma-
tion indicated that the BCB may not be favored by as large a
margin as previously believed. Barriers for interconversion of
cyclodecane conformations also have been calculated recently
Two examples of substituted cyclodecanes with conforma-
tions other than BCB have been found by X-ray crystal-
lography,1 and in one case the departure from the usual ring
conformation can be explained in terms of steric interactions.
The 1a and 3a hydrogens of 1a form two sets of three hydrogens
with substantial nonbonded interactions, as shown in the diagram
above, and these positions will be avoided if possible by
substituents larger than hydrogen. Substitution of the ring by
four methyl groups in a 1,1,5,5 relationship would place a
methyl group in a highly strained intraanular position (1a or
0,11
1
7
using MM2. For example, a barrier of 5.43 kcal/mol was
found for conversion of TBCC to the distorted BCC by K3-
1
7
17
3
a), and the crystal structure of 4,4,8,8-tetramethylcyclodecane-
kayaking. BCB was reported to be connected only to TBC,
with a barrier of 6.38 kcal/mol for this K3-kayaking.
Allinger18 has carried out a conformational search for
10
carboxylic acid was determined to examine the effects of this
type of substitution. Initial results for the compound did not
correspond to a reasonable structure, but the use of the apparent
ring geometry as a starting point for force-field calculations of
cyclodecane resulted in two conformations with relative strain
energies of 2.1 kcal/mol (TBC, 1c of Figure 1) and 3.1 kcal/
mol (TBCC, 1b). It was found that the X-ray diffraction data
could be interpreted in terms of a 4:1 mixture of TBC and TBCC
conformations randomly distributed in the crystal. More
cyclodecane using the MM4 force field, and a total of 10
conformations were found to have free energies within 3.2 kcal/
mol of the lowest (TBCC) conformation. The BCB conforma-
tion was calculated to have a lower enthalpy than the TBCC
conformation by 0.52 kcal/mol at room temperature, but the
calculations indicated a free energy slightly higher for the BCB
than for the TBCC at this temperature (by 0.14 kcal/mol). Most
of the difference between the enthalpy and the free energy for
the two conformations was attributed to the fact that the TBCC
exists as a mixture of D and L forms, while unsubstituted BCB
11
recently, the TBCC and BCB conformations have been found
in the crystals of cis-1,6-cyclodecanediol.
Cyclodecane has been studied in the gas phase at +130 °C
by a combination of electron diffraction and molecular mechan-
18
cyclodecane does not. MM4 calculations at 298 K gave
1
2
ics calculations. Excellent agreement between the experi-
mental intensity curve with a curve calculated for C2h BCB
cyclodecane was obtained, but significant differences between
this structure and the X-ray structure were observed, leading
the authors to suggest that other conformations may actually
be populated. In an alternative interpretation12 molecular
mechanics was used to calculate the strain energies for the four
populations of 30% BCB, 37% TBCC, 12% of a C1 conforma-
12
tion not considered by Hilderbrandt, 10% TBC, and 11% of
6
additional conformations.
Cyclodecane possesses the common types of strain, including
the transannular H‚‚‚H repulsions noted above for 1a, and the
(
(
13) Noe, E. A.; Roberts, J. D. J. Am. Chem. Soc. 1972, 94, 2020.
14) Anet, F. A. L.; Cheng, A. K.; Wagner, J. J. J. Am. Chem. Soc. 1972,
(
9) Conformational BehaVior of Six-Membered Rings, Juaristi, E., Ed;
94, 9250.
(15) Reference 1c, pp 50 and 208. On p 50, the ratio is given as 4:2:2.
VCH: New York, 1995.
(
10) Dunitz, J. D.; Eser, H.; Bixon, M.; Lifson, S. HelV. Chim. Acta
967, 50, 1572.
11) Ermer, O.; Vincent, B. R.; Dunitz, J. D. Isr. J. Chem. 1989, 29,
37.
12) Hilderbrandt, R. L.; Wieser, J. D.; Montgomery, L. K. J. Am. Chem.
Soc. 1973, 95, 8598.
(16) Saunders: M. J. Comput. Chem. 1991, 12, 645. Free energies for
several conformations have been calculated recently: Senderowitz, H.;
Guarnieri, F.; Still, W. C. J. Am. Chem. Soc. 1995, 117, 8211.
(17) Kolossvary, I.; Guida, W. C. J. Am. Chem. Soc. 1993, 115, 2107.
(18) Allinger, N. L.; Chen, K.; Lii, J.-H. J. Comput. Chem. 1996, 17,
642.
1
(
1
(

Conformational Study of Cyclodecanes
J. Am. Chem. Soc., Vol. 120, No. 41, 1998 10717
compound has been useful in refining the force fields used in
molecular mechanics calculations.18,19 The level of accuracy
and Northrup model 8690-2 millivolt potentiometer. The uncertainty
in the temperatures was estimated to be (2 °C. The free-energy barriers
obtained by dynamic NMR spectroscopy are not highly sensitive to
reached by the MM3 and MM4 programs is high enough that
the results from these calculations suggested that further
experimental studies of the conformations of 1 by NMR could
be useful. No details (e.g., temperatures, solvent, chemical
22
the rate constant or the temperature. For a process with a rate constant
-
1
of 220 s at -120.9 °C, a change in temperature of (2 °C raises or
lowers the barrier of 7.07 kcal/mol by 0.10 kcal/mol. Temperatures
were not rounded off to the nearest degree because this would have
15
shifts, or signal-to-noise ratio) were described for the early
resulted in temperature changes of up to 0.5 °C, which would have
1
3
C NMR study of cyclodecane, and it seemed possible that
1
had about /
4
the effect on the barrier of a 2 °C change.
one or more additional conformations could be present and
detected with the greater sensitivity of NMR spectrometers now
available. The conformations of 1 have also been studied in
this work using ab initio molecular orbital calculations. In
addition, chlorocyclodecane and cyclodecyl acetate have been
examined by dynamic NMR spectroscopy and computational
The strain energies for 1, calculated by Saunders,16 were reproduced
23
using Allinger’s MM2 and MM3 programs built into Spartan. An
24
external MM3 program, updated on January 1, 1994, and purchased
from the Quantum Chemistry Program Exchange at Indiana University
was used to obtain free energies for the conformations of 1 and 2. The
geometries determined for the first five conformations of 1 according
to MM3 strain energies were used as starting points for the ab initio
calculations, using the Gaussian 94 series of programs.2 For all five
conformers, full geometry optimizations were done at the HF/6-31G*
level; smaller basis sets were used initially. Results for these
conformations were used as input for calculations at the HF/6-311G*
level. Energy-minimized structures were characterized by calculation
of harmonic vibrational frequencies at the same levels, and free energies
were obtained. For the frequency calculations, the default scaling factor
of 0.89 was used.
methods. A low-temperature ESR study of the cyclodecyl-
5
methyl radical concluded20 that two or three conformations were
present, with the CH2 group in an axial and one or two equatorial
positions, but no dynamic NMR studies of substituted cyclo-
decanes have been described.
Experimental Section
Cyclodecane, listed as 98% pure, was purchased from Aldrich
Chemical Co. and used as received. The actual level of purity was
The gauge-including atomic orbitals (GIAO) method was employed
to calculate isotropic absolute shielding constants (σ, in ppm) for the
carbons of TMS, three conformations of 1, and all of the possible BCB
and TBCC conformations of 2. Subtraction gave the calculated
chemical shifts, relative to TMS. These calculations were done at the
13
indicated to be greater than 99.5% by the room-temperature C and
1
H NMR spectra. A 0.4% solution of 1 in CF
2 2
Cl was prepared in a
5
-mm thin-walled screw-capped NMR tube. A small amount of
tetramethylsilane was added for an internal reference. Caution: high
pressure. The sample tube was stored and handled below 0 °C most
of the time.
2
5
HF/6-311G* level using the Gaussian 94 series of programs.
Spectra were recorded on a General Electric model GN-300 wide-
Results and Discussion
1
3
bore NMR spectrometer operating at 75.58 MHz for C and 300.52
1
13
MHz for H. The C NMR spectra for 1 were obtained from 18.2 to
171.1 °C with a 5-mm dual probe. A pulse width of 12.5 µs was
The 13C NMR spectrum of 1 at +18.2 °C shows a single
-
sharp peak at δ 26.10 (Figure 2). The spectrum is exchange-
broadened by -120.1 °C, and at -131.0 °C two overlapping
peaks are observed at δ 25.0 and 23.7, with line widths of 47
and 546 Hz, respectively. As the temperature is lowered further,
the major signal decoalesces into three peaks at δ 28.4, 22.4,
and 19.2 in a ratio of 2:2:1. From the number and relative
intensities of the peaks, they are assigned to the BCB conforma-
tion (1a). Calculated chemical shifts (Table 1) permit assign-
ment of these signals to carbons of types 2, 3, and 1,
respectively; the assignment for C-1 is unambiguously confirmed
used, corresponding to a tip angle of 45°, and the pulse repetition period
was 1 s. The delay time was shorter than optimal for integration,21
but this was necessary to obtain an adequate signal-to-noise ratio. From
2
800 to 3200 acquisitions were used, with a sweep width of (12 500
Hz, data size of 64 K, and 3.0 Hz line broadening to increase the signal-
to-noise ratio. Spinning was discontinued below about -120 °C. Area
measurements in the temperature region -148.6 to -131.0 °C were
made by a cut-and-weigh procedure.
Chlorocyclodecane (2) was purchased from Columbia Organics and
1
3
used as received after checking for purity by C NMR spectroscopy.
It was studied as a 7% solution in a 2:1:1 mixture of CHClF , CHCl F,
C NMR spectra were taken with 500 ( 100 acquisitions,
2
2
13
by the intensity ratios. These C assignments are in agreement
13
2 2
and CF Cl .
2
a
with those made for 1a in the solid state. The minor peak at
131.0 °C sharpens at first as the temperature is lowered and
a sweep width of ( 11 600 Hz, data size of 32 K, pulse width of 9.0
µs (80° tip angle), and pulse repetition period of 8 s, over a temperature
range of -0.1 to -165.5 °C.
Cyclodecyl acetate (3) was synthesized by boiling a mixture of
cyclodecanol (5.0 g) and acetic anhydride (13.1 g). Ether was added
to the cooled mixture, and aqueous sodium hydroxide solution was
used to remove excess anhydride and acetic acid. Cyclodecyl acetate
was isolated by drying over Drierite, removing the ether, and distilling
-
as the rate of exchange with the BCB conformation decreases
and then broadens at still lower temperatures as interconversion
of carbon positions within the conformation is slowed suf-
ficiently on the NMR time scale. Slow exchange was not seen
for the latter process. The total concentration of cyclodecane
is 0.4%, and only 5.2% of this amount at -146.1 °C is due to
2 2
under reduced pressure. An 8% solution in 2:1:1 CHClF , CHCl F,
13
2 2
CF Cl was studied by C NMR from +0.1 to -160.0 °C. The sweep
(22) For a discussion of sources and magnitudes of errors in dynamic
width, number of acquisitions, data size, pulse width, tip angle, pulse
NMR spectroscopy, see: Raban, M.; Kost, D.; Carlson, E. H. J. Chem.
Soc. D 1971, 656.
repetition period, and line broadening were ( 10 500 Hz, 250 ( 50,
2 K, 4.2 µs, 45°, 6.0 s, and 3.0 Hz.
Because of the difficulty in ejecting the sample at the lower
(
23) Spartan version 3.0 from Wavefunction, Inc., Irvine, CA.
3
(24) Version MM3 (94) was used. The latest version of the MM3
program, which is referred to as MM3 (96), is available to academic users
from the Quantum Chemistry Program Exchange and to commercial users
from Tripos Associates, 1699 South Hanley St., St. Louis, MO 63144.
(25) Gaussian 94: Rev. E. 2, Frisch M. J.; Trucks, G. W.; Gill, P. M.
W.; Johnson, B. G.; Schlegel, H. B.; Robb, M. A.; Cheesman, J. R.;
Gomperts, R.; Keith, T.; Petersson, G. A.; Montgomary, J. A.; Raghavachari,
K.; Al-Laham, M. A.; Zakrzewski, J. V. O.; Foresman, J. B.; Cioslowski,
J.; Stefanov, B. B.; Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala,
P. Y.; Chen, W.; Wong, M. W.; Andres, J. L.; Replogle, E. S.; Fox, D. J.;
Head-Gordon, M.; Binkley, J. S.; Gonzales, C.; Martin, R. L.; Defrees, D.
J.; Baker, J.; Stewart, J. J.; Pople, J. A.; Gaussian, Inc.: Pittsburgh, PA,
1995.
temperatures, the temperature calibrations were performed separately,
using a copper-constantan thermocouple immersed in the same solvents
contained in a nonspinning dummy sample tube and under conditions
as nearly identical as possible. The emf’s were measured using a Leeds
(
19) Allinger, N. L.; Yuh, Y.; Lii, J.-H. J. Am. Chem. Soc. 1989, 111,
8
551.
(
(
20) Ingold, K. U.; Walton, J. C. Acc. Chem. Res. 1989, 22, 8.
21) Ideal conditions are a tip angle of 83° and a pulse-repetition period
of 4.5 T1: Traficante, D. D. Concepts Magn. Reson. 1992, 4, 153. See also:
Traficante, D. D.; Steward, L. R. Concepts Magn. Reson. 1994, 6, 131.

10718 J. Am. Chem. Soc., Vol. 120, No. 41, 1998
Pawar et al.
a
Table 2. Relative Strain Energies, Relative Free Energies, and
Populations of Different Conformations of Cyclodecane, Calculated
Using Allinger’s MM3 Programb
2
5 °C
-171.1 °C
relative
free
relative
strain
energies
relative
free
con-
energies
popula-
tions
energies
(kcal/mol)
popula-
tions
former (kcal/mol) (kcal/mol)
1
1
1
1
1
1
1
1
1
1
1
1
a
b
c
d
e
f
g
h
i
0.000
0.930
1.389
2.079
2.194
2.890
3.013
3.997
4.333
4.382
4.963
5.089
0.828
0.000
1.303
1.007
1.308
1.196
2.683
2.511
4.044
2.971
3.291
2.920
0.136
0.546
0.061
0.100
0.060
0.073
0.006
0.008
0.001
0.004
0.002
0.004
0.427
0.000
1.231
1.315
1.138
1.585
2.245
2.851
4.159
3.262
3.440
3.583
0.108
0.885
0.004
0.002
0.004
0.000
j
k
l
a
Previously calculated in ref 16. b Conformations are numbered
according to increasing strain energy.
Table 3. Relative Free Energies of Cyclodecane Conformations
from ab Initio Calculations at the HF/6-31G* and HF/6-311G*
Levelsa
HF/6-31G*
HF/6-311G*
Figure 2. Low-temperature 13C NMR spectra of cyclodecane.
relative free relative free relative free relative free
Table 1. 13C Chemical Shifts in PPM (δ) for the BCB, TBCC,
and TBC Conformations of Cyclodecane. Calculated Chemical
Shifts Are from ab Initio Calculations at the HF/6-311G* Level,
Using the GIAO Method
energiesb
at 25 °C
energies
at -171.1 °C
energies
at 25 °C
energies at
-171.1 °C
conformer
1
1
1
1
1
a
b
c
d
e
0.000
0.678
0.767
2.020
2.120
0.000
0.810
0.819
2.181
2.319
0.000
0.781
0.815
2.088
2.204
0.000
0.915
0.868
2.243
2.395
TBCC
TBC
BCB conformation
experimentalb calculated
carbon
no.
conformation conformation
calculated
a
calculated
1
2
3
4
5
6
19.2
28.4
22.4
20.3
27.6
21.4
19.3
26.5
25.4
20.8
27.9
31.2
22.6
29.0
20.0
19.6
25.0
a
Conformations are numbered as in Figure 1. b Kcal/mol.
positions within each of these conformations, then the TBC
could be present in small amounts and not be detected in the
low-temperature C NMR spectra. At slow exchange, the TBC
13
a
Numbering of the ring carbons is shown in Figure 1. b At -171.1
absorption would be divided among six peaks, each too low in
intensity to be observed, and at higher temperatures, the line
shapes of the broad signals would probably not be influenced
enough by the presence of the TBC to make detection possible.
Equal amounts of the TBCC and the TBC would give popula-
tions of 5.2, 5.2, and 89.6% and relative free energies of 0.72,
0.72, and 0.00 kcal/mol for the TBCC, TBC and BCB
conformations at -146.1 °C, respectively, based on the experi-
mental areas at this temperature. The experimental spectra
neither confirm nor disprove the presence of the TBC confor-
mation. Populations at -171.1 °C can be estimated as 2.7 and
97.3% by assuming that only the TBCC and the BCB confor-
mations are present, separated by a constant free-energy
difference of 0.73 kcal/mol. If equal amounts of the TBC and
the TBCC are assumed, each separated from the BCB by a
constant free-energy difference of 0.72 kcal/mol, then popula-
tions of 2.6, 2.6, and 94.7% would be predicted for -171.1 °C.
On the basis of the contribution of Rln2 to the entropy of 1b
and 1c, as a consequence of the presence of D and L forms, the
free energy of either of these conformations would be expected
to rise, relative to 1a, as the temperature is lowered, and this
effect is seen in the MM3 calculations of Table 2. However,
the Gaussian 94 calculations (Table 3) do not include the Rln2
entropy term for 1b or 1c, resulting in calculated changes in
relative free energies from +25 to -171.1 °C that are smaller
than expected for 1a and 1b or for 1a and 1c. The free-energy
difference between 1a and 1b would actually be expected to
°
C.
the minor conformation. In addition, the peaks of the second
conformation may still be exchange-broadened at -171.1 °C.
Calculated free energies for conformations of 1 at -171.1
°
C by MM3 and ab initio methods (Tables 2 and 3) show that
the observed minor conformation could be the TBCC (1b) or
the TBC (1c). The line shape changes for 1 in Figure 2 suggest
that the barriers to interconversion of 1a with the second
conformation are larger than the barrier to exchange of the BCB
carbon positions. According to the pruned network of confor-
mational interconversions of Kolossvary and Guida, 1a can
interconvert only with 1c. The low-temperature 13C line shapes
of 1 indicate that exchange of carbon positions in 1a does not
occur through conversion to the minor conformation, and the
second conformation is then assigned to the TBCC. If only
the TBCC and the BCB conformations are assumed to be
appreciably populated, then a free-energy difference between
the two conformations of 0.73 ( 0.3 kcal/mol could be obtained
from the five area measurements over a temperature range of
1
7
-
148.6 to -131.0 °C. However, an alternative description of
the conformations of 1 is suggested by the ab initio calculations
of Table 3, which predict that the TBC and the TBCC
conformations have comparable free energies and populations.
If interconversion of the BCB and the TBC conformations
provides the lowest-energy pathway for exchange of carbon

Conformational Study of Cyclodecanes
J. Am. Chem. Soc., Vol. 120, No. 41, 1998 10719
Table 4. Relative Free Energies of the BCB, TBCC, and TBC
Conformations of Chlorocyclodecane at -165.5 °C from MM3
relative free
energies
BCB
substituent
position
TBCC
substituent
position
TBC
substituent
position
(kcal/mole)
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
2
2
2
6
8
8
8
9
9
.000
.018
.316
.396
.443
.512
.562
.040
.146
.159
.307
.576
.727
.965
.985
.038
.073
.319
.436
.016
.219
.987
.722
.842
2a
2e
3e
2e
3e
2a
1e
1e
5e
6
4e
2e
3e
4e
5e
1
3a
Figure 3. Low-temperature 13C NMR spectra of the substituted ring
carbon of chlorocyclodecane.
5a
1a
3a
1a
4a
2a
5a
4a
3a
1
1
1.358
2.127
Table 5. 13C NMR Chemical Shifts of the Substituted Ring
Figure 4. 1H NMR spectrum of the methine proton of chlorocyclo-
decane at -165.5 °C, with decoupling of the adjacent protons.
Carbon of Chlorocyclodecane from ab Initio Calculations at the
HF/6-311G* Level, Using the GIAO Methoda
BCB
substituent
positions
TBCC
substituent
positions
increase slightly as the temperature is lowered to -171.1 °C
from -146.1 °C.
chemical shifts
chemical shifts
(ppm, δ)
(ppm, δ)
Calculated chemical shifts for the five types of carbons of
b and six kinds of carbons for 1c are given in Table 1. The
average shifts for 1b and 1c are 24.0 and 23.9 ppm, respectively,
1
2
2e
3e
e
a
58.20
62.57
64.06
57.70
1a
1e
2e
2a
3e
60.50
56.81
61.64
61.87
62.83
61.53
57.97
61.01
62.04
62.25
1
and the experimental averaged shift of the minor conformation
1
3
is 25.1 ppm at -146.1 °C. The calculated C shifts of 1b
suggest that it should be possible to see at least some of its five
peaks in the presence of 1a if the signal-to-noise ratio is high
enough and if slow exchange can be reached. These conditions
3
4
4
a
e
a
5e
5a
were probably not met in an earlier 1 C NMR study of
cyclodecane, resulting in the conclusion15 that only the BCB
conformation is populated for 1.
3
15
a
Ring carbons are numbered as in Figure 1.
The 13C NMR spectrum of chlorocyclodecane at -60.1 °C
shows a single peak for the substituted carbon at δ 63.16, and
by -120.9 °C the signal has decoalesced into two broad peaks
at δ 60.95 and 66.12, with relative areas of 47.4 and 52.6%,
respectively (Figure 3). By -165.5 °C, the signal at higher
frequency has split into two peaks. The three chemical shifts
at this temperature are δ 67.67, 66.14, and 61.63, and the
program are summarized in Table 4. The TBC conformations
are of higher energy and are not considered further here.
Chemical shifts calculated for the substituted carbon of the BCB
and the TBCC conformations of 2 (Table 5) provide support
for the assignment of BCB conformations to peaks 1 and 2,
since the calculated TBCC chemical shifts are all at substantially
lower frequency than those found experimentally for peak no.
1
populations are 31.2, 14.9, and 53.9%, respectively. The H
1
. Assignment of peaks 1 and 2 to the 2e and 2a BCB
NMR spectrum of the methine proton, with decoupling of the
adjacent hydrogens, also shows three peaks (Figure 4).
For cyclodecane, exchange between the BCB and the TBCC
conformations first slowed as the temperature was lowered,
followed by slowing of the exchange of the BCB carbon
positions and then the TBCC carbon positons. The results for
conformations, respectively, gives the best match of calculated
and experimental chemical shifts. However, axial hydrogens
generally absorb at lower frequency than equatorial hydrogens,
26
and on this basis the chemical shifts of Figure 4 would predict
the reverse assignments for peaks 1 and 2 of Figure 3. Of the
two crystalline monosubstituted cyclodecanes studied by X-ray
2
show some similarity to 1 and are best interpreted in terms
27
crystallography, cyclodecanol was found to have the OH group
of a mixture of two BCB conformations, assigned to peaks 1
and 2 at -165.5 °C, and a TBCC conformation absorbing at
the lowest frequency (peak no. 3).
Free-energy calculations for the possible BCB, TBCC, and
TBC conformations of chlorocyclodecane using the MM3
+
in the 2e position of the BCB ring, and the NH3 group of
(
26) Jensen, F. R.; Bushweller, C. H.; Beck, B. H. J. Am. Chem. Soc.
969, 91, 344.
27) Valente, E. J.; Pawar, D. M.; Noe, E. A. Acta Crystallogr., Section
C, in press.
1
(

10720 J. Am. Chem. Soc., Vol. 120, No. 41, 1998
Pawar et al.
are probably exchange-broadened, and three conformations are
also populated for 3.
Conclusions
The only observable 13C peaks for cyclodecane at -171.1
C were for the BCB conformation (1a), which can be identified
°
by the 2:2:1 ratio of the areas for the three peaks. At
temperatures between -131.0 and -154.0 °C, a peak for a
minor conformation was visible and was assigned to the TBCC
conformation on the basis of free energies estimated from MM3
and ab initio calculations, in addition to a consideration of the
likely pathway for exchange of carbon positions in the BCB
conformation. Because this exchange probably occurs by way
of the TBC conformation, a small amount of the TBC, if present,
would not be detected in the low-temperature C NMR spectra.
The free energy of 1b relative to that of 1a was found
experimentally to be 0.73 kcal/mol at -146.1 °C, with the
assumption that only these two conformations are present. If
we assume equal amounts of 1b and 1c, the free-energy
difference between 1a and 1b changes only slightly, to 0.72
kcal/mol.
Figure 5. 13C NMR spectra of cyclodecyl acetate.
13
cyclodecylamine hydrochloride sesquihydrate28 was located at
position 3e of the BCB. trans-1,6-Disubstituted cyclodecanes
with both groups in the 3e,29 2e,
30,31
or 2a
29,32
positions of the
BCB conformation are also known, but no examples of
substitution at position 1e have been reported. The possibility
that peak 1 or 2 of Figure 3 is associated with the 3e BCB is
not excluded, but the calculated chemical shifts of Table 5 are
in better agreement with the proposed assignments for this
compound. The small calculated differences in free energies
for several conformations of BCB and TBCC cyclodecane
suggest the possibility that the positions occupied could change,
depending on the group attached to the ring.
Chlorocyclodecane (2) was found to have three conformations
at -165.1 °C, with populations of 31.2, 14.9, and 53.9%. The
major conformation (peak number 3 for the methine carbon of
13
the slow-exchange C NMR spectrum of Figure 3) was assigned
to the TBCC ring conformation, and peaks 1 and 2 were
assigned to BCB conformations, based on several considerations
including the following: (1) The free-energy barriers of 5.4 and
5.5 kcal/mol separating the conformations associated with peaks
1 and 2 were close to the free-energy barrier of 5.7 kcal/mol
for 1,1-difluorocyclodecane, which was assigned a BCB con-
From the free energies of Table 4, the 2a and all of the
conformations with the chlorine equatorial are possible for the
TBCC conformation of 2. The calculated chemical shifts for
the 1e- and 4e-substituted conformations make these conforma-
tions less likely, but there is no firm basis to decide among the
remaining four possibilities.
13
formation. If peaks 1 and 2 represented TBCC conformations,
then lower barriers would have been expected, based on the
results for cyclodecane. (2) The TBC conformations of 2 were
calculated by MM3 to have free energies higher than those for
the favorable TBCC or BCB conformations. (3) The chemical
shifts for peaks 1 and 2 of 2 were at high frequency (δ 67.67
and 66.14). The chemical shift at the highest frequency
calculated for this carbon was δ 64.06 for the 2e BCB
conformation, which was then assigned to peak no. 1. The
closest calculated chemical shift remaining for peak no. 2 of
BCB 2 was at δ 62.57 for the 2a BCB conformation. However,
At -159.8 °C, exchange of the TBCC conformation of 2 with
the BCB conformations is slow on the NMR time scale, and
free-energy barriers of 5.4 ( 0.2 and 5.5 ( 0.2 kcal/mol for
interconversion of 2a and 2e BCB chlorocyclodecane were
calculated from the rate constants of 85.0 and 59.1 s , which
were obtained from complete line shape matching of this part
of the spectrum. The similarity of these barriers to the barrier
for exchange of 2a and 2e positions of 1,1-difluorocyclodecane
-
1
(5.7 kcal/mol) provides some support for assignment of peaks
1
as noted above, the experimental H shifts of Figure 4 suggest
1
and 2 of the chloride to BCB conformations. By -120.9 °C,
the reverse assignments for peaks 1 and 2 of Figure 3.
the BCB conformations of 2 are exchanging rapidly, and free-
energy barriers of 7.07 ( 0.2 and 7.08 ( 0.2 kcal/mol were
The MM3 calculations indicate that a large number of TBCC
and BCB conformations of 2 are of low energy, and other
substituted cyclodecanes may have different conformational
preferences, although the low-temperature spectra of cyclodecyl
acetate (3) suggested an equilibrium for this compound similar
to that of the chloride.
-
1
calculated from the rate constants of 220 and 211 s for
1
3
interconversion of BCB and TBCC conformations. The
C
NMR spectrum (Figure 5) of the substituted ring carbon of
cyclodecyl acetate at -160.0 °C shows a pattern of chemical
shifts and intensities similar to that of 2, although the signals
Acknowledgment. We thank the National Science Founda-
tion (Grant No. HRD-9450455), the National Institutes of Health
Grant No. SO6 GM08047), the Department of Energy, Col-
(
28) (a) Nowacki, W.; Mladeck, M. H. Chimia 1961, 15, 531. (b)
Mladeck, M. H.; Nowacki, W. HelV. Chim. Acta 1964, 47, 1280.
29) trans-1,6-Diaminocyclodecane dihydrochloride was found to exist
(
(
in two crystal modifications, which corresponded to different locations on
the BCB ring for the NH3+ groups (3e or 2a): (a) Huber-Buser, E.; Dunitz,
J. D. HelV. Chim. Acta 1960, 43, 760. (b) Huber-Buser, E.; Dunitz, J. D.
HelV. Chim. Acta 1961, 44, 2027. (c) Huber-Buser, E.; Dunitz, J. D. HelV.
Chim. Acta 1966, 49, 1821.
laborative Research and Manpower Development Program
between JSU and AGMUS (Grant No. DE-FG05-86ER75274)
for support of this work, and the Mississippi Center for
Supercomputing Research for a generous amount of time on
the Cray Y-MP supercomputer. We thank Professor Norman
L. Allinger of the University of Georgia for helpful discussions
and for pointing out that Gaussian 94 does not include the Rln2
contribution to entropy for 1b and 1c.
(
30) trans-1,6-Dibromocyclodecane: Dunitz, J. D.; Weber, H. P. HelV.
Chim. Acta 1964, 47, 951.
31) The OH groups of trans-1,6-cyclodecanediol in a 2:1 molecular
compound with cis-1,6-cyclodecanediol are in the 2e positions: ref 11.
32) trans-1,6-Cyclodecanediol: (a) Ermer, O.; Dunitz, J. D. Chem.
(
(
Commun. 1971, 178. (b) Ermer, O.; Dunitz, J. D.; Bernal, I. Acta
Crystallogr. 1973, B29, 2278.
JA973116C