Experimental Demonstration of a Sizeable Nonclassical CH···G Hydrogen Bond in Cyclohexane Derivatives: Stabilization of an Axial Cyano Group

Article abstract of DOI:10.1021/acs.orglett.7b03287

Base-catalyzed equilibration of anancomeric cyanocyclohexanes demonstrates that replacement of cis-3,5-dimethyl holding groups with electron-withdrawing CF₃ groups dramatically increases the proportion of the axial cyano isomer present at equilibrium. The CF₃ groups exert an effect on the conformational energy of the cyano group worth about 0.6 kcal/mol. A nonclassical hydrogen bond between the axial CN group and the syn-axial hydrogens is a major contributor to the axial stability of the group.

Full text of DOI:10.1021/acs.orglett.7b03287

Letter
pubs.acs.org/OrgLett
Cite This: Org. Lett. XXXX, XXX, XXX-XXX
Experimental Demonstration of a Sizeable Nonclassical CH···G
Hydrogen Bond in Cyclohexane Derivatives: Stabilization of an Axial
Cyano Group
Kyle M. Lambert,† Zachary D. Stempel,† Kenneth B. Wiberg,*,‡ and William F. Bailey*,†
†Department of Chemistry, University of Connecticut, 55 North Eagleville Road, Storrs, Connecticut 06269-3060, United States
‡Department of Chemistry, Yale University, 275 Prospect Street, New Haven, Connecticut 06520-8107, United States
S
* Supporting Information
ABSTRACT: Base-catalyzed equilibration of anancomeric cyanocyclohexanes
demonstrates that replacement of cis-3,5-dimethyl holding groups with electron-
withdrawing CF3 groups dramatically increases the proportion of the axial cyano
isomer present at equilibrium. The CF3 groups exert an effect on the conformational
energy of the cyano group worth about 0.6 kcal/mol. A nonclassical hydrogen bond
between the axial CN group and the syn-axial hydrogens is a major contributor to the
axial stability of the group.
he literature of the past few decades provides many
equilibration of anancomeric cyclohexanes bearing cis-3,5-
Texamples of CH···G Coulombic interactions (nonclassical dimethyl holding groups should be different from that assessed
hydrogen bonds) affecting the conformations of cyclic and
acyclic molecules.1 Although the reported experimental and
computational results are interesting, and almost surely correct,
to our knowledge there has not been any direct assessment of
such an interaction in a simple system that also provides
information on its magnitude. We recently presented evidence of
a stabilizing CH···O interaction in the axial conformation of 5-
phenyl-1,3-dioxane between an ortho-hydrogen of the phenyl
group and an oxygen of the 1,3-dioxane: as a result, the plane of
the phenyl ring essentially bisects the dioxane ring.2 The results
of this study further demonstrated that the strength of the CH···
O bond was tunable; the interaction is enhanced when remote
electron-withdrawing substituents are present on the aromatic
ring.
We were prompted by these observations to consider a
conceptually simple experiment to probe directly a fundamental
question: the strength and consequence of such CH···G bonds in
a simple cyclohexane derivative. The requisite model systems are
illustrated in Figure 1. We reasoned that the conformational
energy (−ΔG° or A-value)3 of a group evaluated by direct
when cis-3,5-bis-trifluoromethyl holding groups are employed.
The electron-withdrawing CF3 groups would reasonably be
expected to render both C(3) and C(5), as well as the syn-axial
hydrogens at these positions, more positive than would be the
case in the dimethyl systems. To the extent that a putative CH···
G bond between an axial group at C(1) and the syn-axial
hydrogens is of any consequence, there should be a greater
proportion of the isomer having an axial substituent in the system
bearing the CF3 groups.
The approach outlined above requires that the diastereoiso-
meric systems be easily equilibrated, and it was preferable that the
group whose conformational energies were to be evaluated be
symmetric. For these reasons, we chose to explore the cyano
group whose conformational energy has been determined by
various techniques on a number of occasions4 including base-
catalyzed equilibration of 4-tert-butylcyclohexyl cyanide.5
The diastereoisomeric cyanides (1−4) were prepared in
straightforward fashion as illustrated in Scheme 1 and detailed in
the Supporting Information (SI). The only step in the syntheses
that deserves comment involves the direct conversion of the
cyclohexylmethanols to the cis-nitriles (2 and 4) using
hexamethyldisilazane (HMDS) and Bobbitt’s salt (4-acetami-
do-2,2,6,6-tetramethyl-1-oxopiperidinium tetrafluoroborate) fol-
lowing our general procedure.6 The isomeric nitriles, 1 and 3,
were isolated from the mixtures obtained upon base-catalyzed
equilibration of the cis-isomers.
Each of the anancomeric pairs of cyanides (1 and 2; 3 and 4)
were equilibrated at room temperature (∼23 °C) in sealed
ampules under nitrogen as solutions in cyclohexane, Et2O, or
THF over dry potassium tert-butoxide. Equilibrium was
Received: October 22, 2017
Figure 1. Model systems.
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.7b03287
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Scheme 1. Preparation of the Substrates
approached independently from samples of each diastereoisomer
and, after the solutions were neutralized by shaking with
anhydrous ammonium chloride and then filtered, the area ratio of
the isomeric mixture was determined by capillary GC analysis
providing baseline separation. It was judged that equilibrium had
been reached when the same area ratios were obtained from
initially pure samples of each isomer. Area ratios for each
equilibration, which reflect the equilibrium constant for the
process, were taken as the average of 8−12 independent
determinations from each side, and the free energy difference
for the equilibrium was calculated in the normal way: ΔG° =
−RT ln K. The results of these studies are summarized in Table 1.
conformational energy of the cyano group. This effect is
strikingly dramatic: in the least polar solvents investigated,
cyclohexane and Et2O, the cyano group prefers to occupy the
axial position. Indeed, the ΔG° value for the CN group in
cyclohexane solvent (3 → 4; Table 1, entry 6) is +0.48 kcal/mol,
favoring the axial isomer. To our knowledge, this is an
unprecedented result.7 Comparison of the ΔG° value in
cyclohexane solvent of the CN group evaluated in the 3,5-
dimethyl system (−0.12 kcal/mol) with that obtained from the
3,5-bis-trifluoromethyl system (+0.48 kcal/mol) demonstrates
that the CF3 holding groups exert an effect on the conformational
energy worth some 0.60 kcal/mol. The etiology of this effect is
most certainly electrostatic.
Table 1. Equilibria in Cyanocyclohexanes
In an effort to achieve further insight into the role of these
intramolecular Coulombic interactions, we have examined some
substituted cyclohexanes at the MP2/6311+G* level.8 In
particular, cyclohexanes bearing F and CN substituents at C(1)
with cis-3,5-dimethyl or cis-3,5-bis-trifluoromethyl groups were
investigated. The fluoro compounds, we reasoned, should
provide a simple model since they involve a single atom as the
substituent; the cyano compounds, of course, correspond to
molecules 1−4 discussed above. The computed structures of the
compounds investigated may be found in the SI; the structures of
the four axially substituted compounds, along with some relevant
geometrical parameters, are shown in Figure 2.
a
entry
R
solvent
K
ΔG° (kcal/mol)
1
2
3
4
5
6
CH3 (1, 2)
THF
1.485 0.003
1.359 0.011
1.223 0.001
1.329 0.004
0.785 0.003
0.422 0.001
−0.23 0.01
−0.18 0.05
−0.12 0.01
−0.17 0.02
+0.26 0.01
+0.48 0.01
Et2O
c-C6H12
THF
CF3 (3, 4)
Et2O
The calculated axial → equatorial energy differences, corrected
for differences in zero-point energies and thermal corrections to
25 °C, are summarized in Table 2. With these compounds the
computed ΔH° and ΔG° values are fairly close, with the latter
c-C6H12
a
Determined at room temperature. Errors in K are propagated
standard deviations; the errors in ΔG° are estimated errors that are
larger than the propagated standard deviation to account for an
assumed GC response ratio of 1.0 for a given pair of isomers.
The experimental free energy difference (1 → 2; Table 1, entry
1) of −0.23 0.01 kcal/mol in THF at room temperature is in
good agreement with the CN group’s A-value of 0.25 kcal/mol
determined by Allinger from the equilibration of 4-tert-
butylcyclohexyl cyanide with potassium tert-butoxide in THF
solvent at 66 °C;5 demonstrating, not unexpectedly, that the
entropy contribution to ΔG° is negligible in this system. The
equilibration data for the 3,5-dimethyl system (Table 1, entries
1−3) reveals a small solvent effect at play. As might be expected,
in the most polar solvent, THF, the molecule with the higher
dipole moment (2) is stabilized relative to the axial isomer (1). In
the least polar solvent studied, cyclohexane, the free energy
difference between 1 and 2 decreases by half to −0.12 kcal/mol.
In short, the results gleaned from an examination of the 3,5-
dimethyl system are unremarkable.
Examination of the influence of 3,5-bis-trifluoromethyl
holding groups on the conformational energy of the cyano
group (3 → 4; Table 1, entries 4−6) vis-a-vis found in the 3,5-
̀
dimethyl system (1 → 2; Table 1, entries 1−3) demonstrates that
the electron-withdrawing nature of the CF3 groups affects the
Figure 2. Computed MP2/6-311+G* structures.8,10
B
DOI: 10.1021/acs.orglett.7b03287
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
The quantity of immediate interest is the charge at H(3,5), the
equivalent axial hydrogens at the carbons bearing the holding
groups. It is clear, as expected, that replacement of the methyl
holding groups with CF3 groups results in a substantial increase
(∼0.02 e) in positive charge at these axial hydrogens, and this
would lead to an increased Coulombic interaction with a
negatively charged axial substituent at C(1). It should be noted
that H(3,5) is rather positive even in the cis-3,5-dimethyl
systems, and this will provide some Coulombic stabilization of an
axial C(1) group; this likely contributes to the origin of the rather
low A-values of F and CN groups in cyclohexane.
A rough “back of the envelope” estimate of the Coulombic
energy (E = q1q2/r12) associated with the CH···F interaction
between a H(3,5) hydrogen and an axial F at C(1) is easily
obtained from the Hirshfeld charges (Table 3, entries 1 and 3)
and the computed distances from Figure 2 in the di-CH3 and the
bis-CF3 systems of 2.65 and 2.60 Å, respectively.13 For the system
with di-CH3 holding groups, the Coulombic energy is attractive
by 0.67 kcal/mol per hydrogen; for the corresponding bis-CF3
compound, the attractive energy is 0.94 kcal/mol. An analogous
but slightly more tedious computation for the CN compounds,
involving separate calculations of the Coulombic energy of the
interactions of carbon and the nitrogen of the group with a
H(3,5) hydrogen, gave the following results: the overall CH···
CN interaction in the di-CH3 molecule (1) is attractive by 0.54
kcal/mol per hydrogen; the corresponding value for the bis-CF3
system is attractive by 0.69 kcal/mol.
In this connection we must note, as we have noted elsewhere,
that simple summation of Coulombic energies between
nonbonded atoms is an approximation, at best, of the actual
situation.2 Indeed, as Kirkwood and Westheimer demonstrated
some 80 years ago, simple two-center Coulombic energies will be
attenuated by the electric fields attendant with the bonds in
molecules.14 Nonetheless, this modest exercise provides an order
of magnitude estimate of the effect of replacing dimethyl holding
groups with bis-trifluoromethyl groups that comports well with
both the experimental data for the CN isomers and the
computational results for both the CN and F molecules.
The MP2/6-311+G* structures (Figure 2) reinforce the
conclusion that there is a significant attractive interaction
between an axial group at C(1) and the positive 3,5-positions
of the bis-CF3 molecules. Cursory inspection of the structures
indicates that an axial group is considerably closer to the syn-axial
hydrogens in the systems bearing CF3 groups at C(3,5) than in
the corresponding CH3 molecules. This is most noticeable in the
Table 2. Energy Differences Calculated Using the MP2/6-
311+G* Level in kcal/mol
a
a
entry
R
G
ΔH°
ΔG°
ΔΔH°
ΔΔG°
1
2
3
4
CH3
CF3
CH3
CF3
F
0.16
1.61
0.55
1.50
0.12
1.54
0.44
1.35
1.45
1.42
F
CN
CN
0.95
0.91
a
(ΔH° or ΔG° in the R = CF3 series) − (ΔH° or ΔG° in the R = CH3
series).
being a little smaller. This result indicates that the entropy
change is quite small, as expected.9 It will be noted that, in each
instance, the computed energies show a distinct preference for
the axially substituted isomer that is significantly larger than the
experimentally determined ΔG° values in cyclohexane solvent
(Table 1). This discrepancy is likely due to the large difference in
dipole moment between the axial and equatorial forms and the
resultant large solvent effect.11 However, the change from axial to
equatorial of a C(1) substituent represents a small perturbation,
and for this reason, the energy differences between isomers in the
3,5-di-CH3 systems and that in the corresponding 3,5-bis-CF3
systems (ΔΔH° and ΔΔG°) should be more accurate than the
total energies. Indeed, a comparison between the computed
ΔΔG° = 0.91 kcal/mol for the cyano compounds in each series
(Table 2, entry 4) with the experimentally determined ΔΔG° =
0.60 kcal/mol in cyclohexane solution (Table 1) lends some
credence to this suggestion.
How large are the charge-induced Coulombic interactions that
are apparently responsible for the difference in conformational
energies of a polar substituent when methyl holding groups are
replaced by trifluoromethyl holding groups? For an answer to
this seemingly simple question, it is convenient to convert the
charge density about a nucleus to an effective atomic charge. We
believe that the Hirshfeld charges,12 obtained directly from the
charge density, are the most useful for this purpose. We have
calculated these charges for all atoms in the molecules depicted in
Table 2, and they may be found in the SI. Selected charges,
germane to the present question, are summarized in Table 3.
Table 3. Calculated Hirshfeld Charges, q (e)
entry
isomer
G
R
H(1)
C(1)
H(3,5)
F
CN
CN
1
2
3
4
5
6
7
8
A
B
A
B
A
B
A
B
F
CH3
CH3
CF3
CF3
CH3
CH3
CF3
CF3
0.0425
0.0399
0.0534
0.0454
0.0596
0.0561
0.0689
0.0606
0.0585
0.0557
0.0655
0.0618
0.0015
0.0007
0.0078
0.0063
0.0367
0.0338
0.0544
0.0521
0.0364
0.0347
0.0531
0.0528
−0.1460
−0.1522
−0.1348
−0.1356
F
F
F
CN
CN
CN
CN
0.0799
0.0764
0.0807
0.0807
−0.2325
−0.2349
−0.2144
−0.2163
C
DOI: 10.1021/acs.orglett.7b03287
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
110, 6049. (b) Alkorta, I.; Elguero, J. Chem. Soc. Rev. 1998, 27, 163.
(c) Desiraju, G. R.; Steiner, T. The Weak Hydrogen Bond in Structural
Chemistry and Biology; Oxford Univ. Press: Oxford, 1999.
(2) (a) Bailey, W. F.; Lambert, K. M.; Stempel, Z. D.; Wiberg, K. B.;
Mercado, B. Q. J. Org. Chem. 2016, 81, 12116. (b) Wiberg, K. B.;
Lambert, K. M.; Bailey, W. F. J. Org. Chem. 2015, 80, 7884.
(3) Winstein, S.; Holness, N. J. J. Am. Chem. Soc. 1955, 77, 5562.
(4) Eliel, E. L.; Wilen, S. H.; Mander, L. N. Stereochemistry of Organic
Compounds; Wiley: New York, 1994; pp 696−697.
(5) Allinger, N. L.; Szkrybalo, W. J. Org. Chem. 1962, 27, 4601.
(6) Kelly, C. B.; Lambert, K. M.; Mercadante, M. A.; Ovian, J. M.;
Bailey, W. F.; Leadbeater, N. E. Angew. Chem., Int. Ed. 2015, 54, 4241.
Angew. Chem. 2015, 127, 4315.
smaller C(1)−C(2)−C(3) bond angle and a larger H(1)−G
bond angle when 3,5-CF3 groups are present. More subtle is the
nonlinear C(1)−C−N angle of 178° in compound 3 in
comparison to an almost linear (179.5°) angle in compound 1.
In conclusion, direct base-catalyzed equilibration of ananco-
meric cyanocyclohexanes demonstrates that the conformational
energy of the CN group is substantially affected by the nature of
the holding groups at the 3,5-positions. When di-CH3 groups are
used, the conformational energy is identical to that obtained
using a classical 4-t-butyl holding group. However, the
conformational energy is considerably reduced when bis-CF3
groups are present. Remarkably, in cyclohexane, the least polar
solvent system investigated, the cyano group favors the axial
orientation to the extent of almost 0.5 kcal/mol. The
computational results described above indicate that the axial-
stabilizing effect of the electron-withdrawing CF3 groups is
plainly electrostatic in origin; a manifestation of a moderately
strong, nonclassical CH···CN bond between the syn-axial
hydrogens and the CN group worth at least 0.6 kcal/mol.
MP2/6-311+G* calculations suggest that the effect of 3,5-bis-
CF3 groups on the conformational energy of a fluorine is even
more dramatic; the attractive CH···F interaction is computed to
be worth more than 1 kcal/mol. In the aggregate, the results of
the investigation suggest that nonclassical CH···G attractive
interactions merit further consideration when evaluating the
origin of conformational phenomena in more complex systems
involving polar substituents such as the anomeric effect.
(7) The only substituents previously known to adopt an axial
orientation are HgX moieties as a consequence of their long Hg−X
bonds and the polarizable nature of the Hg atom. See ref 4.
(8) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.;
Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.;
Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, Jr., J. A.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.;
Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.;
Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich,
̈
S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.;
Fox, D. J. Gaussian 09, Revision C.01; Gaussian, Inc.: Wallingford, CT,
2009.
ASSOCIATED CONTENT
■
(9) Hofner, D.; Lesko, S. A.; Binsch, G. Org. Magn. Reson. 1978, 11,
179.
(10) Legault, C. Y. CYLview; Universite de Sherbrook, 2009. http://
́
www.cylview.org.
S
* Supporting Information
̅
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.orglett.7b03287.
(11) The computed energies refer to the gas phase, and it might be
noted that the polarizable continuum model suggests that experimental
data obtained in cyclohexane should represent values that are
approximately 40% of the maximum effect obtained in the absence of
a solvent.
(12) (a) Hirshfeld, F. L. Theor. Chim. Acta 1977, 44, 129. (b) Parr, R.
G.; Ayers, P. N.; Nalewajski, R. F. J. Phys. Chem. A 2005, 109, 3957.
(13) The Coulombic energy, E = q1q2/r12, is conveniently calculated in
atomic units (hartrees), using the charge in e, r in Bohr (1 Bohr = 0.529
Å), and the conversion factor 1 H = 627.51 kcal/mol.
Detailed experimental procedures for all synthesized
compounds; 1H, 13C, and 19F NMR spectra of all products;
1H NOESY spectra and analyses; details of the analytical
GC data; a summary of the calculations, including
computed energies, coordinates, and Hirshfeld charges
(PDF)
AUTHOR INFORMATION
Corresponding Authors
■
(14) (a) Kirkwood, J. G.; Westheimer, F. H. J. Chem. Phys. 1938, 6, 506.
(b) Kirkwood, J. G.; Westheimer, F. H. J. Chem. Phys. 1938, 6, 513.
*E-mail: william.bailey@uconn.edu.
*E-mail: kenneth.wiberg@yale.edu.
ORCID
Kyle M. Lambert: 0000-0002-8230-2840
Kenneth B. Wiberg: 0000-0001-8588-9854
William F. Bailey: 0000-0001-9159-0218
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported in part by the facilities and staff of the
Yale University Faculty of Arts and Sciences High Performance
Computing Center. The work at the University of Connecticut
was supported by grants from Procter & Gamble Pharmaceut-
icals, Mason, OH.
REFERENCES
■
(1) For comprehensive reviews of nonclassical CH···G hydrogen
bonds, see: (a) Takahashi, O.; Kohno, Y.; Nishio, M. Chem. Rev. 2010,
D
DOI: 10.1021/acs.orglett.7b03287
Org. Lett. XXXX, XXX, XXX−XXX