Article
Cite This: J. Org. Chem. 2018, 83, 5242−5255
pubs.acs.org/joc
The Anomeric Effect: It’s Complicated
Kenneth B. Wiberg,*,†,§ William F. Bailey,*,‡ Kyle M. Lambert,‡ and Zachary D. Stempel‡
†
Department of Chemistry, Yale University, 275 Prospect Street, New Haven, Connecticut 06520-8107, United States
Department of Chemistry, University of Connecticut, 55 North Eagleville Road, Storrs, Connecticut 06269-3060, United States
‡
S Supporting Information
*
ABSTRACT: The origin of the anomeric effect has been
reexamined in a coordinated experimental and computational
investigation. The results of these studies implicate a number of
different, but correlated, interactions that in the aggregate are
responsible for the anomeric effect. No single factor is uniquely
responsible for the axial preference of a substituent that is the
hallmark of the anomeric effect. A CH···G nonbonded
attraction between a polar axial substituent (G) and the synaxial hydrogen(s) in the heterocycle has been demonstrated
experimentally. The hyperconjugation model involving electron
transfer from a ring heteroatom to an excited state of an axial
C−G bond was shown to be, at most, a minor contributor because of the very small changes in charge density at the ring
heteroatom(s): the main charge transfer is from hydrogen to G in the H−C−G unit. This appears to result from lengthening the
C−G bond to minimize repulsion with the ring atom lone pair(s) and the advantage of having a more positive hydrogen that
leads to a stabilizing Coulombic interaction with the ring heteroatom(s). In short, the anomeric effect arises mainly from two
separate CH···G nonbonded Coulombic attractions.
■
INTRODUCTION
Most six-membered heterocycles prefer to adopt chair
conformations with equatorially situated alkyl groups. To be
sure, there are qualitative differences between such heterocycles
and cyclohexane: carbon−heteroatom bond lengths are shorter
than carbon−carbon bonds and carbon−heteroatom−carbon
bond angles are generally smaller than the corresponding C−
C−C bond angles.1 Consequently, alkyl groups generally have a
greater preference for the equatorial conformation in heterocycles than in cyclohexane: for example, the conformational
energy (−ΔG° or A value) of a 2-methyl group in 1,3-dioxane
is 3.98 kcal/mol2 but only 1.76 kcal/mol in methylcyclohexane.3
Conformational studies of glycosides in the field of
carbohydrate chemistry (some of which predate similar studies
of carbocycles)4 demonstrated that, although qualitative
similarities exist between heterocyclic compounds and their
carbocyclic analogues, the presence of a heteroatom in the ring
can lead to dramatic differences in conformational behavior.
One of the most significant discoveries of the early
carbohydrate work was the observation that the ring oxygen
in a glycoside influences the conformational preference of a
polar substituent at the anomeric carbon. Pyranose derivatives
having such substituents were found to adopt a conformation
that preferentially places the aglycone in the α-position. Not
surprisingly, this contrasteric preference for the axial orientation
at the anomeric position in pyranose derivatives was termed the
“anomeric effect” in 1958 by Lemieux and Chu.5 The
phenomenon is also termed the “Edward−Lemieux effect” in
recognition of similar, contemporary studies by Edward.6,7 It
© 2018 American Chemical Society
was subsequently found that this so-called effect was not limited
to carbohydrates but was a general occurrence observable when
a heterosubstituent is located adjacent to a heteroatom in a sixmembered ring. Suffice it to note that the etiology of these
observations has been the subject of numerous studies and the
resulting voluminous literature has been extensively reviewed.8
In short, the search for a cause (or causes) for this “effect” has
captivated chemists for more than 50 years.9
A number of hypotheses have been advanced to account for
the anomeric effect.8 The four most frequently considered
explanations are summarized pictorially in Figure 1 for the case
of a tetrahydropyran. In 1955, Edward suggested that the
preference for certain polar substituents (X) at C(1) to adopt
the axial position was due to repulsive dipolar interactions in
the equatorial conformation that are relieved in the axial
conformation of the molecule.6 Indeed, the magnitude of the
anomeric effect has been found to be solvent dependent; the
more polar isomer, having an equatorial X substituent, seems to
be stabilized relative to its less polar anomer in solvents of high
dielectric constant.8 Lemieux and Chu subsequently noted that
such electrostatic dipolar interactions may be viewed as
Coulombic interactions.5,10 When the bond moments responsible for the dipole−dipole repulsions of Edward are treated as
point charges located at the center of each atom, the operation
of an attractive Coulombic interaction between an axial X
substituent at C(1) and the C(5) position of the ring can be
envisioned which enhances the preference for the substituent to
Received: March 20, 2018
Published: April 5, 2018
5242
DOI: 10.1021/acs.joc.8b00707
J. Org. Chem. 2018, 83, 5242−5255
Article
The Journal of Organic Chemistry
significance of Coulombic interactions in systems displaying an
anomeric effect.
We recently presented evidence for an attractive, nonbonded
Coulombic interaction of the type CH···G in a simple
cyclohexane derivative that serves to stabilize an axial CN
group.17 As illustrated in Figure 2, replacement of cis-3,5-
Figure 2. Effect of electron-withdrawing holding groups on the
conformational energy of a CN group.
dimethyl holding groups in anancomeric models with electronwithdrawing CF3 groups dramatically increases the proportion
of the axial cyano isomer present at equilibrium in cyclohexane
solvent. The CF3 groups render both the C(3,5) carbons as
well as the attached syn-axial hydrogens more positive vis-à-vis
the 3,5-dimethyl cyclohexane. This observation implicated an
attractive, nonclassical CH···CN hydrogen bond of the type
illustrated in Figure 1 between the axial CN and the rather
positive syn-axial hydrogens as a major contributor to the axial
stability in the system bearing CF3 groups. Moreover, it is
known from experimental studies that the hydrogen of C−H
bonds in alkanes has a small positive charge18 and a more
modest CH···CN hydrogen bond may also be involved in the
di-CH3 system.17
In an effort to explore the extent to which Coulombic
interactions of this kind contribute to an anomeric effect in
heterocyclic systems, we have investigated the axial−equatorial
energy difference in 1,3-dioxanes bearing 2-OCH3 or 2-CN
substituents. Conformational equilibria of this sort are
investigated most conveniently by equilibration of configurationally isomeric models for the conformational isomers and, as
Eliel has demonstrated,19 a cis-4,6-dimethyl-1,3-dioxane is a
good surrogate for the unbiased system.
We reasoned that the axial preference of a polar group such
as methoxy or cyano at C(2) of the 1,3-dioxane should be
enhanced when cis-4,6-di-CH3 holding groups are replaced with
strongly electron-withdrawing CF3 groups as was the case for
the cyanocyclohexanes depicted in Figure 2. The electronwithdrawing CF3 groups at C(4) and C(6) would render the
both the carbons as well as the attached syn-axial hydrogens
more positive than is the case in a 1,3-dioxane bearing di-CH3
groups at these positions. Moreover, it might reasonably be
anticipated that the CF3 holding groups would withdraw
electron density as well from the ring oxygens and lessen a
hyperconjugative interaction with the adjacent antibonding σ*orbital of the polar group at C(2).
Both r-2-methoxy-trans-4, trans-6-dimethyl-1,3-dioxane (1)
and r-2-methoxy-cis-4, cis-6-dimethyl-1,3-dioxane (2) were
prepared as described by Eliel and Nader.20 It proved
considerably more difficult to prepare the corresponding 4,-6bis-CF3 substrates (3 and 4): the syntheses are illustrated in
Scheme 1.
Transacetalization of the fluorinated diol mixture with
benzaldehyde dimethyl acetal affords the 2-phenyl-1,3-dioxane
analogue. Fortunately, r-2-phenyl-cis-4,cis-6-bis-
Figure 1. Hypotheses advanced to account for the anomeric effect.
adopt the axial position. More recently, this view was extended
by the suggestion that an attractive, nonclassical CH···X
hydrogen bond exists between the axial X group at C(1) and
the syn-axial hydrogen attached to C(5) and this has been
reviewed by Takahashi, Kohno, and Nishio.11 A currently
popular explanation for the effect, first suggested by Altona
almost 60 years ago,12 is a hyperconjugative interaction
involving electron delocalization of a 2p-type lone pair on
oxygen (often, as here, more conveniently rendered as an sp3
hybridized lone pair) into the adjacent antibonding σ*-orbital
of the C(1)−X unit; this stabilizing interaction is most effective
when the X group is axial because the sp3 lone pair on oxygen
and the axial C−X bond are antiperiplanar.13 Such an
interaction should lead to a lengthening of the axial C(1)−X
bond and a shortening of the C(1)−O bond as is often
observed in molecules that display an anomeric effect.8 It has
been argued by many that this stereoelectronic, hyperconjugative interaction is the principal factor that results in
the anomeric effect,8,9 but this view is not universally
accepted.14,15
We were prompted by our longstanding interest in the
anomeric effect16 to reexamine all four of the hypotheses
summarized in Figure 1 in a coordinated experimental and
computational investigation of the anomeric effect.
■
RESULTS AND DISCUSSION
Experimental Studies of the Role of Intramolecular
Coulombic Attraction. There is abundant experimental
evidence that electrostatic effects, arising from dipole−dipole
interactions that are affected by solvents, play a role in the
anomeric effect.8 Much less is known about the potential
5243
DOI: 10.1021/acs.joc.8b00707
J. Org. Chem. 2018, 83, 5242−5255
Article
The Journal of Organic Chemistry
Scheme 1. Preparation of 2-Methoxy-4,6-bis(trifluoromethyl)-1,3-dioxanes (3 and 4)
Table 1. Equilibria in 2-Methoxy-1,3-dioxanes 1 ⇌ 2 and 3
⇌4
(trifluoromethyl)-1,3-dioxane, a solid, is isolated as the major
product of the condensation. It is worth noting that hydrolysis
of the phenyl dioxane to afford meso-1,1,1,5,5,5-hexafluoro-2,4pentanediol is very slow as the intermediate is apparently
severely destabilized by the trifluoromethyl groups. The final
step in preparing 3 and 4 was also difficult to accomplish as
attempts to perform the condensation with typically employed
acid catalysts such as p-TsOH, camphorsulfonic acid, or
pyridinium p-toluenesulfonate resulted either in decomposition
of the diol or failed altogether to initiate the condensation.
Employing BF3·Et2O as a Lewis acid provided 3 and 4 in
modest yield, as an approximately 5/1 mixture, but only if the
reaction was conducted at room temperature; higher temperatures promoted dehydration of the diol. Careful column
chromatography on neutral alumina as described in the
Experimental Section allowed for isolation of analytically pure
3 and 4 in the low yields indicated in Scheme 1.
The 2-methoxy-4,6-dimethyl-1,3-dioxane isomers (1 and 2)
were equilibrated at room temperature (∼23 °C) in sealed
ampules under nitrogen as solutions in dry CH3CN, THF,
Et2O, cyclohexane, and n-pentane over dry Amberlyst-15 resin
or BF3·Et2O. The 2-methoxy-4,6-bis(trifluoromethyl)-1,3-dioxane isomers (3 and 4) are rather more recalcitrant and required
anhydrous AlCl3 as the catalyst for equilibration in dry THF,
Et 2O, and cyclohexane; other acids lead to substrate
decomposition. Equilibrium was approached independently
from samples of each isomer and, after the solutions were
neutralized, the area ratio of the isomeric mixtures were
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 10 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.
The free energy difference between 1 and 2 in Et2O (Table 1,
entry 3) of 0.40 kcal/mol favoring the axial isomer is identical
to the 0.41 kcal/mol value determined by Nader and Eliel in
1970.2 There is, however, a substantial solvent effect in the
axial−equatorial energy difference between 1 and 2 as would be
expected given the rather large difference in the reported dipole
moments of 1 (μ = 1.97 D) and 2 (μ = 2.93 D).20 In
acetonitrile solvent, the equatorially substituted isomer, 2, is
slightly favored by 0.04 kcal/mol (Table 1, entry 1) while in the
Determined at 23 °C; errors in K and ΔG° are propagated standard
deviations. bEquilibrated over Amberlyst-15 or BF3·Et2O. cEquilibrated over anhydrous AlCl3.
a
least polar solvents, the axial isomer, 1, is found to be more
stable by ∼0.5 kcal/mol (Table 1, entries 4 and 5). An estimate
of the free energy difference between 1 and 2 in the gas phase
may be obtained by plotting the experimental ΔG° values in
CH3CN, THF, pentane and cyclohexane versus the dielectric
constant of these solvents using the Onsager function, (ε-1)/
(2ε+1).21,22 The plot, shown in Figure 3, suggests that ΔG° in
the gas phase is ∼0.9 kcal/mol favoring the axial isomer.
Replacement of the CH3 holding groups in 1 and 2 with CF3
groups to give 3 and 4 results in a dramatic increase in the
proportion of the axial methoxy isomer (4) present at
equilibrium in all solvents studied (Table 1, entries 6−8).
The effect of solvent polarity on the energy difference between
isomers 3 and 4 is roughly twice that found for the di-CH3
isomers 1 and 2: the ΔΔG° for the more polar isomers bearing
CF3 groups in THF and cyclohexane is ∼0.59 kcal/mol (Table
1, entries 6 and 8) while that for the 4,6-diCH3 isomers in these
solvents is only ∼0.25 kcal/mol (Table 1, entries 2 and 4).
Comparison of the ΔG° value in cyclohexane solvent of the 2OCH3 group (Table 1, entry 4) evaluated in the 4,6-dimethyl
system (0.51 kcal/mol) with that found (Table 1, entry 8) in
the 4,6-bis(trifluoromethyl) system (1.57 kcal/mol) demonstrates that replacement of CH3 groups with CF3 groups effects
a 1 kcal/mol change in the proportion of axial 2-OCH3 present
at equilibrium. Remarkably, this large modification of the
conformational energy of the methoxy group is almost twice as
large as the ΔG° value of the group in 2-methoxy-4,6-dimethyl5244
DOI: 10.1021/acs.joc.8b00707
J. Org. Chem. 2018, 83, 5242−5255
Article
The Journal of Organic Chemistry
the process takes 1−2 weeks to reach equilibrium. Compounds
7 and 8 are even more reluctant to equilibrate over acid
catalysts: a sample of 8 did not form any appreciable amount of
7 (less than 2%) upon standing in anhydrous Et2O with a 10fold molar excess of AlCl3 for a period of two months.
Equilibration of compounds 7 and 8 was finally accomplished
under basic conditions (catalytic t-BuOK) in anhydrous Et2O
solvent requiring ∼2 d to reach equilibrium. Compounds 7 and
8 were equilibrated only in Et2O; neither compound is
sufficiently soluble in cyclohexane to allow for accurate
determination of an equilibrium constant and the dioxanes
decompose before reaching equilibrium in CH3CN. The
equilibration results are summarized in Figure 4.
Figure 3. Plot of ΔG° versus (ε − 1)/(2ε + 1) for 1 ⇌ 2.
1,3-dioxane. Clearly, intramolecular Coulombic interactions
involving attractive, CH···OCH3 hydrogen bonds between the
axial 2-methoxy group and the syn-axial hydrogens are at play
here. In this connection, it might be noted that Juaristi and coworkers invoked nonclassical hydrogen bonding of this sort to
account for the pronounced axial preference of the 2diphenylphosphoryl group in 1,3-dithiane23 and both Stolow24
and Kleinpeter25 have presented evidence for the operation of
attractive Coulombic interactions in cyclohexane derivatives
bearing remote, strongly electron-withdrawing substituents.
For a direct comparison with the cyanocyclohexanes
discussed above (Figure 2), 2-cyano-1,3-dioxanes with 4,6dimethyl holding groups (5 and 6) as well as those with 4,6bis(trifluoromethyl) groups (7 and 8) were investigated. The
preparation of these substrates is illustrated in Scheme 2.
The equilibration of 5 ⇌ 6 with Amberlyst-15 or p-TsOH
fails to proceed, the result, apparently, of the instability of the
cationic intermediate required for the process. Compounds 5
and 6 were successfully equilibrated over anhydrous AlCl3 in
Et2O and cyclohexane solvent in the manner described above;
Figure 4. Equilibria in 2-cyano-1,3-dioxanes 5 ⇌ 6 and 7 ⇌ 8.
Cursory inspection of the equilibration data in Figure 4
reveals that a 2-CN substituent has a greater preference for the
axial position than does a 2-OCH3 group in a common solvent
(Table 1) in both the di-CH3 and CF3 series. The 2.1 kcal/mol
difference between isomers 7 and 8 in Et2O favoring axially
substituted 7 is rather extraordinary. Unfortunately, as noted
above, it was not possible to evaluate the ΔG° for 7 ⇌ 8 in
cyclohexane where the energy difference would undoubtedly be
even larger. Nevertheless, the conformational equilibria
Scheme 2. Preparation of 2-Cyano-4,6-dimethyl-1,3-dioxanes (5 and 6) and 2-Cyano-4,6-bis(trifluoromethyl)-1,3-dioxanes (7
and 8)
5245
DOI: 10.1021/acs.joc.8b00707
J. Org. Chem. 2018, 83, 5242−5255
Article
The Journal of Organic Chemistry
summarized in Figure 4 and Table 1 implicate Coulombic
interactions as at least one major factor in the operation of an
anomeric effect.
In order to obtain further information about the role of
internal Coulombic interactions, we have carried out some
computational studies of substituted and 1,3-dioxanes at the
MP2/6-311+G* level that we found to be useful in a study of
phenyl substituted cyclohexanes.26
Computational Studies. Initially, 1,3-dioxanes with 2-F, 2CN and 2-OCH3 substituents were examined using MP2/6311+G*. The fluorine case provides an easier examination since
it involves a single atom, the CN and OCH3 molecules
correspond to the experimental studies described above. The
calculated axial−equatorial energy differences for these
molecules, corrected for the difference in zero-point energy
(ZPE) between the two isomers and the change in ΔH° on
going from 0 to 298 K, are given in Table 2.
Figure 5. Computed MP2/6-311+G* structure of r-2-methoxy-cis4,cis-6-dimethyl-1,3-dioxane (2) demonstrating close hydrogen−oxygen contacts; r (CH···O) = 2.481 Å.
Table 2. Axial ⇌ Equatorial Energy Differences in 2Substituted 1,3-Dioxanes Calculated Using the MP2/6311+G* Level in kcal/mol
a
entry
G
R
ΔH°
ΔG°
1
2
3
4
5
6
F
F
CN
CN
OMe
OMe
CH3
CF3
CH3
CF3
CH3
CF3
3.76
4.83
2.22
2.95
0.82
1.83
3.73
4.75
2.11
2.83
0.94
1.87
ΔΔH°
ΔΔG°
1.07
1.02
0.73
0.72
1.01
0.93a
above, that favors the axial isomers by 0.41 kcal/mol at 25 °C.
The ΔΔH° and ΔΔG° values in Table 2 represent the effect of
changing the holding groups from 4,6-di-CH3 to 4,6-bis-CF3.
The computed MP2/6-311+G* structures of 1 and 3, along
with relevant geometrical parameters, are shown in Figure 6;
structures of all the compounds investigated may be found in
the Supporting Information.
The calculated ΔG° of 0.93 kcal/mol for the equilibrium of 1
⇌ 2 is nicely comparable with the gas phase value of ∼0.9 kcal/
mol estimated from the experimental solution data (Figure 3).
The ΔΔH° and ΔΔG° values for the 2-fluoro-1,3-dioxanes and
the 2-cyano-1,3-dioxanes (Table 2) are similar to those
calculated previously for the analogously substituted cyclohexyl
fluorides and cyanide17 as might be expected for an attractive
Coulombic interaction between a negatively charged substituent and the positively charged syn-axial hydrogens at the
carbons bearing the holding groups. This leads to the question,
“How large are these charges?” Here, it is convenient to convert
the charge density about a nucleus to an effective atomic
charge. We believe the Hirshfeld charges that are obtained
directly from the charge density are the most useful for this
purpose (see: Appendix, Atomic Charges, in the Supporting
Information).27 We have calculated these charges, and they are
summarized in Table 3. Here, Ha is the charge of the hydrogen
at C(2) bearing the substituent, Hb is the charge of a syn-axial
hydrogen at C(4,6), X is the charge at the atom of the
substituent (G) attached to C(2), and O is the charge at the
dioxane oxygens. A complete summary of these charges may be
found in the Supporting Information.
The quantity of immediate interest is the charge, Hb, at the
syn-axial hydrogens attached to the C(4,6) carbons bearing the
CH3 or CF3 holding groups. It is seen that changing from CH3
holding groups to CF3 groups leads to a significant increase in
positive charge that will lead to increased Coulombic
interaction with a negatively charged axial substituent at
C(2). It might be noted that even with the CH3 holding
groups, Hb has a positive charge and will provide some
Coulombic stabilization of an axial substituent. This is probably
the origin of the increase in proportion of axial forms for
cyclohexanes having polar substituents whereas it is commonly
found that most nonpolar substituents strongly prefer to be
equatorial.
Includes the RTln2 (0.41 kcal/mol) symmetry correction; see text.
The 2-methoxy-1,3-dioxanes, 1 ⇌ 2 and 3 ⇌ 4, present an
additional computational complication due to the differing
rotameric arrangements of the methoxy group in the axial and
equatorial isomers. In the axially substituted isomers (1 and 3),
the methyl group adopts two energetically equivalent mirror
image rotameric arrangements in which the methyl group is
gauche to the equatorial hydrogen at C(2) and this contributes
an entropy of mixing of ΔS = R ln 2 = 1.38 eu for the axial 2methoxy isomers; the rotamer in which the CH3 group points
into the ring is excluded on steric grounds. In the equatorially
substituted isomers (2 and 4), the methoxy group may in
principle adopt three rotameric arrangements: two mirror
image arrangements in which the CH3 is gauche to the axial
hydrogen at C(2) and one in which the CH3 group is anti with
respect to that hydrogen and gauche to both ring oxygens.
However, the antiarrangement of the equatorial 2-methoxy
group, which places the rather positive methyl hydrogens quite
near the ring oxygens, is significantly more stable than the
gauche forms in both isomer 2 (by 1.7 kcal/mol) and isomer 4
(by 2.1 kcal/mol). This manifestation of what has been termed
the “exo-anomeric effect”8 is almost certainly due to Coulombic
attraction between the methyl hydrogens and the ring oxygens.
The computed structure of 2, shown in Figure 5, dramatically
reinforces this argument.
The ΔH° and ΔG° values for the 2-F and 2-CN compounds
are fairly close, with the latter being a little smaller. This
indicates that the entropy change is quite small and the values
are consistent with a ΔS° ∼ − 2 eu. For the 2-OCH3
compounds there is a significant entropy of mixing, noted
5246
DOI: 10.1021/acs.joc.8b00707
J. Org. Chem. 2018, 83, 5242−5255
Article
The Journal of Organic Chemistry
Figure 6. Computed MP2/6-311+G*structures of r-2-methoxy-trans-4,trans-6-dimethyl-1,3-dioxane (1) and r-2-methoxy-trans-4,trans-6bis(trifluoromethyl)-1,3-dioxane (3).
Table 3. Calculated Hirshfeld Charges, q (e)
entry
isomer
R
G
Ha
Hb
X
1
2
3
4
5
6
7
8
9
10
11
12
A
B
A
B
A
B
A
B
A
B
A
B
CH3
CH3
CF3
CF3
CH3
CH3
CF3
CF3
CH3
CH3
CF3
CF3
F
F
F
F
CN
CN
CN
CN
OMe
OMe
OMe
Me
+0.0622
+0.0368
+0.0771
+0.0476
+0.0741
+0.0468
+0.0871
+0.0563
+0.0475
+0.0228
+0.0611
+0.0318
+0.0426
+0.0453
+0.0549
+0.0514
+0.0352
+0.0341
+0.0517
+0.0521
+0.0326
+0.0301
+0.0490
+0.0481
−0.1493
−0.1136
−0.1286
−0.0898
+0.0695
+0.0844
+0.0730
+0.0906
−0.1814
−0.1744
−0.1739
−0.1628
N
O
−0.2048
−0.1945
−0.1809
−0.1709
−0.1759
−0.1760
−0.1659
−0.1697
−0.1680
−0.1681
−0.1574
−0.1583
−0.1804
−0.1763
−0.1720
−0.1681
carried out in solution, and as seen in the experimental results
presented above, solvent effects can be quite large in these
systems, and this makes it very difficult to evaluate comparisons
of results reported by different groups. The effect of solvent is
an important electrostatic effect.
A spherical ion in the gas phase has a large electrostatic effect
that is given by ΔG = q2/2r, where q is the charge on the ion
and r is its radius. If this ion is placed in a medium with a
dielectric constant of ε, the energy is only 1/ε as great. Clearly,
solvation of ions is important. The same is true with molecules
having a dipole moment. An early approximate treatment by
Onsager found that the energy of a dipolar molecule is reduced
in solution by a factor of (ε − 1)/(2ε + 1).21 This concept has
been extended and incorporated into modern ab initio codes so
geometry optimizations can be carried out using this polarizable
continuum model (PCM).28 It is an approximation of course,
but it proves to be useful. More accurate solvent effects can be
obtained using molecular dynamics methods,29 but they are
computationally much more difficult.
The dipole moment of 1,3-dioxane is 2.293 D, and it lies
along the equatorial direction.30 As a result, polar equatorial
substituents that are aligned along this axis will lead to a large
dipole moment. The dipole moment of equatorial fluorocyclohexane is 2.428 D, for cyanocyclohexane it is 4.415 D, and for
methoxycyclohexane it is 1.362 D. The calculated dipole
An unexpected result of this analysis of effective atomic
charge is the discovery that the charge at the C(2) hydrogen
(Ha in Table 3) in compounds with an axial C(2) substituent
(A in Table 3) is appreciably more positive than in the
corresponding, but less stable, equatorial isomers (B in Table
3). We believe this feature is a fundamental but unrecognized
factor that contributes significantly to the anomeric effect. We
defer discussion of this important topic to the sequel.
Electrostatic Effects. As noted in the Introduction, the
importance of repulsion involving the ring and substituent
dipoles was first suggested by Edward as an explanation of the
anomeric effect and this sort of interaction was more recently
reinforced by Perrin, Armstrong, and Fabian15 in their study of
2-substituted N,N-dimethyltetrahydropyrimidines. The idea is a
simple one. Taking the 1,3-dioxanes as an example, the ring
dipole lies along the equatorial direction so that the dipole
associated with the bond to an equatorial polar substituent will
lead to a repulsive interaction, whereas if the substituent were
in the axial position, the repulsion would be smaller. As a result,
the compound with an axial substituent should have the lower
energy.
Attempts have been made to make this qualitative argument
a more quantitative one, but they are not definitive.12 However,
the consensus is that this interaction accounts for only a part of
the anomeric effect.8 The measurements have always been
5247
DOI: 10.1021/acs.joc.8b00707
J. Org. Chem. 2018, 83, 5242−5255
Article
The Journal of Organic Chemistry
because nitrogen is less electronegative than oxygen and its
lone pair has a higher energy than that in dioxane, and this
should lead to greater hyperconjugative interaction with the
C(2) substituent. This system is a bit more complicated than
the 1,3-dioxanes discussed above. The N-CH3 groups in 2methoxy-N,N-dimethylhexahydropyrimidines may adopt aa, ae,
or ee conformations. We found that the methoxy group in
lower energy isomer, having an axial 2-OCH3, adopts a
rotameric arrangement in which the methyl group is gauche to
the equatorial hydrogen at C(2) and has ae-N-CH3 groups; the
equatorial 2-OCH3 isomer has ee-N-CH3 groups and the
OCH3 moiety adopts a rotameric arrangement in which the
methyl group is anti with respect to the axial hydrogen at C(2).
This later arrangement is similar to that of equatorial 2methoxy-1,3-dioxane and, here again, is likely a consequence of
attractive Coulombic attraction between the methyl hydrogens
of the OCH3 and the ring nitrogens.
In an effort to obtain further information, we also examined
the corresponding 2-methoxy-1,3-dithianes. We focus at the
outset on the rotameric conformations of the 2-methoxy group:
they are gauche with respect to the C(2)-hydrogen in both the
axially and equatorially substituted 1,3-dithiane as found with
the corresponding cyclohexanes. Indeed, for equatorial 2methoxy-1,3-dithiane, the anticonformation of the OCH3 group
was found to be a transition state. The difference with respect
to the favored anticonformation of equatorial 2-methoxy-1,3dioxane discussed above is probably due to the much smaller
charge at sulfur vis-à-vis oxygen. The calculated axial ⇌
equatorial energy differences for substituted cyclohexane, 1,3dioxane, 1,3-dithiane and N,N-dimethylhexahydropyrimidine
are given in Table 5; a more detailed table may be found in the
Supporting Information.
moments for the 2-substituted 1,3-dioxanes of interest and
estimates of the moments for the equatorially substituted
compounds from a simple sum of the dipole moment of 1,3dioxane and dipole moments of the appropriate substituted
cyclohexane are given in Table 4.
Table 4. Calculated and Estimated Dipole Moments, μ, in D
entry
isomer
G
calcd μ
1
2
3
4
5
6
A
B
A
B
A
B
F
F
CN
CN
OMe
OMe
2.872
4.812
3.774
6.152
1.979
3.168
estd μ
4.668
6.654
3.601
It can be seen that the estimate for the equatorial forms are
fairly good and the dipole moments for the axial forms must be
considerably smaller since there is a large angle between the
dioxane and C−G dipoles. Clearly, electrostatic destabilization
for the equatorial forms will be considerably greater than that
for the axial forms.
How large might this effect be? It may be estimated using
PCM. The range of the Onsager function is from 0 for the gas
phase to 0.5 for a very large dielectric constant. The value for
cyclohexane is 0.202 and that for acetonitrile is 0.480, so these
solvents span the range of possible solvent effects. To take but
two examples: the axial ⇌ equatorial free energy difference for
2-fluoro-1,3-dioxane decreases to 3.19 kcal/mol in c-C6H12 and
2.05 kcal/mol in CH3CN and that of 2-cyano-1,3-dioxane
decreases to 1.69 and 0.80 kcal/mol in these solvents.
It is clear that the electrostatic energy of these polar
molecules is large and it favors the conformation in which the
polar substituent is in the axial position. Solvents, which reduce
the electrostatic energy, have a profound effect on the
conformational free energy difference between such isomers,
and this term is a significant component of any experimentally
determined anomeric effect that is determined in solution. In
short, any solvent will decrease the actual magnitude of the
effect.
Hyperconjugation or Coulombic Interactions? Current
conventional wisdom posits that hyperconjugation, the
interaction of a heteroatom’s lone pair p-orbital with an axial
substituent’s antibonding orbital, is the major cause of the
anomeric effect. This hypothesis leads to the expectation that
there should be a significant shift in charge from the ring
heteroatom to an axial substituent. In light of the unexpected
charge shifts discussed below, all compounds were reexamined
at MP2aug-cc-pVTZ using the MP2/6-311+G* geometries and
vibrational frequencies. In two cases, optimization and
frequency calculations were carried out using MP2/aug-ccpVTZ, but the energy changes were the same as for the singlepoint calculations. The use of the larger basis set led to only
small changes in calculated energies and charges.
In the studies described above we focused on substituted
dioxanes that would exhibit large anomeric effects. Some time
ago, Perrin15 suggested that a comparison with 2-methoxy-N,Ndimethylhexahydropyrimidines could be useful in this context
Table 5. Axial ⇌ Equatorial Energy Differences Calculated
at the MP2/aug-cc-pVTZ Level in kcal/mol
entry
1
2
3
4
5
6
7
8
9
10
11
12
ring
cyclohexane
1,3-dioxane
1,3-dithiane
hexahydropyrimidine
2-substituent
ΔH°
ΔG°
F
CN
OMe
F
CN
OMe
F
CN
OMe
F
CN
OMe
0.09
0.53
0.22
3.17
2.00
0.75
4.52
3.09
3.71
5.31
3.40
−0.07
0.05
0.53
−0.04
3.14
1.89
0.78
4.54
2.82
3.57
5.40
3.59
0.26
Hirshfeld charges and relevant bond lengths for the
substituted cyclohexanes, 1,3-dioxanes, 1,3-dithianes, and
hexahydropyrimidines were also calculated at the MP2/augcc-pVTZ level. The results are summarized in an unavoidably
long Table 6. The quantities of foremost interest are the
changes in these quantities (axial−equatorial) on going from an
axial to an equatorial conformation and these are highlighted in
red. Among these charges and lengths, the larger variations are
further highlighted in blue.
An examination of the data in Table 6 demonstrates that an
axial substituent (G) in the dioxane, the dithiane and the N,Ndimethylhexahydropyrimidines systems gains significant electron density with respect to one in the equatorial position. No
5248
DOI: 10.1021/acs.joc.8b00707
J. Org. Chem. 2018, 83, 5242−5255
Article
The Journal of Organic Chemistry
Table 6. MP2/aug-cc-pVTZ Calculated Hirshfeld Charges, q (e), and Relevant Bond Lenghts, r (Å)
a
Distance from the C−G carbon to the adjacent ring atom. bFor the molecules with gauche rotamers of the methoxy groups, there are two different
carbon−ring atom distances. cAxial−equatorial values.
5249
DOI: 10.1021/acs.joc.8b00707
J. Org. Chem. 2018, 83, 5242−5255
Article
The Journal of Organic Chemistry
such change is found with the cyclohexanes, and consequently,
the increase in charge at the axial group must be due to the
presence of a heteroatom in the ring. However, this increase in
charge density at an axial G is not derived from the ring
heteroatoms; the variation in the charge density of the ring
heteroatoms is fairly small. Rather, there is a much larger
reduction in charge density at the hydrogen attached to the C−
G carbon. The bond lengths change accordingly: the axial C−G
bond is lengthened, the adjacent equatorial C−H bond is
shortened, and the carbon to ring heteroatom bond is also
shortened in the isomer having an axial substituent.
If one’s focus is directed exclusively to the bond length
changes in the heteroatom−carbon−G unit, the hyperconjugation model would appear to be validated by the data
in Table 6. Indeed, on average, over the nine systems
investigated, the axially substituted isomers have longer C−G
bonds (∼0.031 Å longer) and shorter C−heteroatom bonds
(∼0.009 Å shorter) than their equatorial counterparts.
However, the unanticipated and rather significant reduction
in the charge density at the equatorial hydrogen in the axially
substituted heterocycles (on average ∼0.018 e), coupled with
only modest change in the electron density of the ring
heteroatom (on average ∼0.004 e), suggests that more than
simple hyperconjugation is at play here.
A Unified Paradigm for the Anomeric Effect. A noted
above, were hyperconjugation the major factor responsible for
the anomeric effect, one would expect a shift in charge density
from the ring heteroatom to the axial substituent, a lengthening
of the bond from the axial substituent to the carbon to which it
is attached, and compression of the bond from that carbon to
the ring heteroatom. The shorter C−H bond in the axially
substituted isomers is likely attributable to the enhanced pcharacter of the axial C−G bond resulting from hyperconjugative interaction with the ring heteroatom. As a result,
the bond from the C−G carbon to the equatorial hydrogen in
the these isomers would be expected to have a higher scharacter.31
While the anticipated effects from a hyperconjugative
interaction are confirmed by the data summarized in Table 6,
these computational observations do not, ipso facto, speak to
the etiology of the anomeric effect.9 In particular: (1) why does
the equatorial hydrogen in a heterocycle with an axial polar
substituent become substantially more positive, and (2) is
hyperconjugation connected to the nonclassical CH···G
hydrogen bonds between an axial group and the syn-axial
hydrogens? We believe the answer to these questions lies in the
enhancement of attractive Coulombic interactions engendered
by the sizable charge density shifts.
The shift of electron density from an equatorial hydrogen to
an axial substituent renders the hydrogen more positive, and
the substituent more negative, than is the situation in which the
substituent is in the equatorial orientation. In short, the charge
density shifts conspire to maximize attractive Coulombic
interactions, and minimize repulsive ones, in the axially
substituted isomers.
Consider, as an example, axial 2-fluoro-1,3-dioxane (Figure
7). The oxygen p-orbitals are roughly parallel to the C−F bond,
leading to a repulsive interaction. This repulsion would be
lessened by an increase the C−F bond length, and such bond
lengthening is observed both computationally and experimentally.13 The longer C−F bond requires a carbon orbital with
high p-character.31 This then leads to a higher s-character in the
bond from that carbon to the hydrogen, resulting in a shorter
Figure 7. Interactions in axial 2-fluoro-1,3-dioxane.
bond and a larger positive charge at the hydrogen. Thus, the
longer C−F bond allows a shift in charge from the hydrogen to
the fluorine providing an increase in the attractive Coulombic
interaction of the equatorial hydrogen with the neighboring
ring oxygens. The axial isomer further benefits energetically
from a more negative fluorine capable of participating in a
stronger CH···F hydrogen bond. An analogous analysis appears
to be true for all of the substituents in the three
heterosubstituted ring systems (9 examples) that were studied.
Applying Occam’s razor,32 it appears that the anomeric effect is
mainly a result of Coulombic interactions. As a check on these
conclusions, we also studied 2-fluorotetrahydropyran and again
found the major charge shift was from H to F in the H−C−F
group.
As a corollary to this study we suggest that proposals of
hyperconjugative interactions in neutral closed shell systems
should be checked by calculating the atomic charges to see if
the proposed charge transfer is found. We are currently
reassessing some of these cases.
■
CONCLUSIONS
The anomeric effect, defined as the difference in conformational free energy between a substituent on a heterocyclic ring
and that on cyclohexane, is the result of a composite of a
number of different, but interrelated, nonbonded intramolecular interactions. There is a CH···G attraction between
the axial substituent (G) and the syn-axial hydrogen(s) in the
heterocycle of the sort advocated by Nishio and co-workers.11
This factor was demonstrated experimentally by replacing
dimethyl holding groups at the carbons bearing the syn-axial
hydrogens with electron-withdrawing CF3 groups and finding
that the axial preference is significantly increased. The
electrostatic effect, due to the much larger dipole moment of
an equatorially substituted isomer vis-à-vis its axially substituted
analog, attenuates the magnitude of the anomeric effect in
solution. The resultant solvent effects, which are often quite
large, make it difficult to compare results from different studies
of substituted heterocycles conducted in disparate solvent
systems.
The hyperconjugation model, frequently proposed as the
origin of the anomeric effect, may contribute to bond length
changes in the heteroatom−carbon−G unit, but it is likely not
5250
DOI: 10.1021/acs.joc.8b00707
J. Org. Chem. 2018, 83, 5242−5255
Article
The Journal of Organic Chemistry
3.50 (s, 3H), 3.85 (dqd, J = 11.1 Hz, J = 6.1 Hz, J = 2.4 Hz, 2H), 5.16
(s, 1H); 13C NMR (100 MHz, CDCl3) δ 21.3, 39.6, 53.2, 71.2, 112.1.
d,l- and meso-1,1,1,5,5,5-Hexafluoro-2,4-pentanediol.36 A
solution of 400 mL of MeOH and 100 mL of deionized water was
prepared in a 2 L, two-neck, round-bottomed flask fitted with a
mechanical stirrer. The solution was made basic by addition of 0.550 g
(13.8 mmol) of solid NaOH, cooled to −78 °C, and then, with
stirring, 9.309 g (246 mmol) of solid NaBH4 was added in portions. A
solution of 41.612 g (200 mmol) of 1,1,1,5,5,5-hexafluoropentanedione in 100 mL of MeOH was cooled to −10 °C and added to a 200
mL addition funnel. The addition funnel was then fitted to the second
neck of the reaction flask, and the dione was added at a rate of 50 mL/
h to the stirred reaction mixture. The reaction mixture was held at −78
°C for 12 h, and then allowed to warm to room temperature over 18 h.
The addition funnel was removed, 400 mL of a 10% aqueous HCl
solution was slowly added to the reaction mixture over a period of 20
min to quench the excess NaBH4, and then allowed to stir for an
additional 1 h to hydrolyze the borate ester. The resulting solution was
transferred to a round-bottomed flask and the MeOH was removed
using a rotary evaporator operating at 35 °C and water aspirator
pressure. The remaining solution was transferred to a separatory
funnel and extracted with five 100 mL portions of Et2O. The organic
layers were combined, 75 mL of hexanes was added and the solvent
was removed under reduced pressure; this process was repeated twice
to azeotropically remove any excess water or MeOH. The resulting oil
was distilled (Kugelrohr, bath temperature = 65−100 °C, 10 mm) to
afford 29.560 g (70%) of a mixture (55:45; d,l: meso) of the title
compounds as a clear oil that partially crystallizes upon standing: 1H
NMR (400 MHz, CD2Cl2) δ 1.96−2.05 (m, 3H), 2.16 (dt, J = 14.8
Hz, J = 3.1 Hz, 1H), 2.75 (d, J = 5.8 Hz, 2H), 3.24 (d, J = 3.9 Hz, 2H),
4.25−4.40 (m, 4H); 13C NMR (100 MHz, CD2Cl2) δ 29.3, 29.4, 66.6
(quartet, JC−F = 32.1 Hz), 69.4 (quartet, JC−F = 32.1 Hz), 124.4
(quartet, JC−F = 281.3 Hz), 125.1 (quartet, JC−F = 281.3 Hz); 19F NMR
(376 MHz, CD2Cl2) δ − 79.7, − 80.1; HRMS (DART-TOF) m/z [M
+ H]+ calcd for C5H7F6O2 213.0345, found 213.0343.
r-2-Phenyl-cis-4,cis-6-bis(trifluoromethyl)-1,3-dioxane. A
mixture of 29.506 g (139.1 mmol) of 1,1,1,5,5,5-hexafluoro-2,4pentanediol, consisting of ∼55% racemic and ∼45% meso isomers,
9.045 g (59.5 mmol) of benzaldehyde dimethyl acetal, and 120 mg
(0.63 mmol) of p-TsOH in 150 mL of hexanes was heated at reflux
under a Dean−Stark trap37 for 12 h, at which point the reaction
mixture was allowed to cool to room temperature. Upon cooling to
room temperature, off-white, needle-like crystals formed upon the
walls of the reaction flask, and after 3 h of standing, the flask was
further cooled to 5 °C in a cold water bath resulting in further
crystallization. The contents of the flask were filtered through a
Büchner funnel containing a coarse filter paper, and the crystals were
rinsed with 50 mL of cold hexanes (the filtrate contained the excess
racemic 1,1,1,5,5,5-hexafluoro-2,4-pentanediol), and allowed to air-dry,
affording 17.805 g (95%) of an off-white, crystalline solid (mp 97.0−
99.0 °C; hexanes). 1H NMR analysis revealed that the solid was
contaminated with ∼10% by mass of racemic 1,1,1,5,5,5-hexafluoro2,4-pentanediol. Further purification was achieved as follows: the solid
was dissolved in CH2Cl2, the solution was extracted with three 100 mL
portions of a saturated, aqueous NaHCO3 solution,38 dried over
anhydrous Na2SO4, and solvent removed by rotary evaporation to
afford 9.737 g (52%) of the title compound as a white, crystalline solid:
mp 100.5−101.5 °C (CH2Cl2); 1H NMR (400 MHz, CDCl3) δ 2.00
(dt, J = 12.9 Hz, J = 2.8 Hz, 1H), 2.11 (dt, J = 13.0 Hz, J = 11.6, 1H),
4.35 (dqd, J = 11.6 Hz, J = 5.7 Hz, J = 3.1 Hz, 2H), 5.64 (s, 1H),
7.38−7.43 (m, 3H) 7.48−7.54 (m, 2H); 13C NMR (100 MHz,
CDCl3) δ 22.9, 73.6 (quartet, JC−F = 34.0 Hz), 101.3, 123.1 (quartet,
JC−F = 278.9 Hz), 126.5, 128.6, 130.0, 135.8; 19F NMR (376 MHz,
CDCl3) δ − 79.4; HRMS (DART-TOF) m/z [M + H]+ calcd for
C12H11F6O2 301.0658, found 301.0672.
meso-1,1,1,5,5,5-Hexafluoro-2,4-pentanediol. A solution of
9.506 g (31.7 mmol) of r-2-phenyl-cis,cis-4,6-bis(trifluoromethyl)-1,3dioxane, 5 mL of deionized water, 10 mL of a 10% aqueous HCl
solution, and 150 mL of MeOH were heated at reflux for 48 h. The
solution was then allowed to cool to room temperature, 100 mL of
uniquely responsible for the axial preference of a substituent
that is the hallmark of the anomeric effect. Indeed, in the
systems that have been investigated (1,3-dioxanes, 1,3dithianes, and hexahydropyrimidines) there was no evidence
for a large role of such an interaction: the atomic charges on the
ring heteroatoms are only slightly altered when the orientation
of the substituent is changed from equatorial to axial. Rather,
there is a significant charge shift from hydrogen to G in the H−
C−G unit that is caused by a combination of repulsion between
the ring heteroatom p-orbital and the substituent, leading to a
longer C−G bond, coupled with a decreased C−H bond length
and an increased positive charge at H. These charge shifts lead
to enhanced Coulombic attraction between the equatorial
hydrogen and the negatively charged ring heteroatoms. In the
aggregate, the anomeric effect appears to arise mainly from two
separate CH···heteroatom nonbonded Coulombic attractions;
in a sense, nonclassical hydrogen bonds.
■
EXPERIMENTAL SECTION
General Methods. Spectroscopic and chromatographic procedures and methods used for the purification of reagents and solvents
have been previously described.33 Pure meso-2,4-pentanediol was
obtained in two steps. A 70:30 mixture of meso and d,l-2,4-pentanediol
was obtained from the sodium borohydride reduction of 2,4pentanedione as described by Pritchard and Vollmer.34 This mesoenriched mixture was then separated following the efficient process
described by Zhang and co-workers35 to afford pure meso-2,4pentanediol.
r-2-Methoxy-trans-4,trans-6-dimethyl-1,3-dioxane (1) and r2-Methoxy-cis-4,cis-6-dimethyl-1,3-dioxane (2). Following a
procedure developed by Eliel and Nader20 with minor modifications,
a mixture of 2.621 g of meso-2,4-pentanediol, 2.991 g (27.5 mmol) of
trimethyl orthoformate, 116 mg (0.61 mmol) of p-TsOH, and 100 mL
of cyclohexane was heated at reflux for 1 h under a 10 mL Dean−Stark
trap. After 1 h of reflux, the contents of the trap were discarded, and
the trap was allowed to refill; this process was repeated twice more,
allowing for ca. 30 mL of a MeOH/solvent mixture to be discarded.
The reaction mixture was allowed to cool to room temperature, 1.50 g
of solid, anhydrous K2CO3 was added, and the mixture was stirred for
1.5 h. The solvent was carefully removed by rotary evaporation
operating at room temperature and water aspirator pressure to afford
2.578 g (70%) of a clear oil containing a mixture of compounds 1 and
2. The compounds are extremely acid sensitive, and to prevent
decomposition due to any adventitious acid, a spatula tip of anhydrous,
solid K2CO3 was added to all glassware to which compounds were
transferred. Separation of 1 and 2 was achieved by column
chromatography on 200 g neutral alumina (5 → 20% Et2O/hexanes):
50 mL fractions were collected; 863 mg (24%) of pure trans-isomer
(1) was collected in fractions 5−9; 101 mg (3%) of pure cis-isomer (2)
was collected in fraction 16. Progress of the chromatography was
monitored by TLC on neutral alumina (15% Et2O in hexanes), and
visualization of 1 and 2 was achieved either by an I2 chamber or
staining with a solution of 20% phosphomolybdic acid in EtOH
followed by heating. The trans-isomer (1), which streaks on alumina,
had an Rf = 0.41 and the cis-isomer (6) had an Rf = 0.15. The NMR
spectra of 1 and 2, detailed below, were identical to those reported for
these compounds.20
r-2-Methoxy-trans-4,trans-6-dimethyl-1,3-dioxane (1): 863 mg
(24%), clear, colorless oil; 1H NMR (400 MHz, CDCl3) δ 1.15 (d, J =
6.3 Hz, 6H), 1.32 (dt, J = 13.1 Hz, J = 11.5 Hz, 1H), 1.54 (apparent
dtd, J = 13.1, Hz, J = 2.5 Hz, J = 1.4 Hz, 1H), 3.30 (s,12 h 3H), 4.18
(dqd, J = 11.7 Hz, J = 6.1 Hz, J = 2.4 Hz, 2H), 5.37 (s, 1H); 13C NMR
(100 MHz, CDCl3) δ 21.6, 40.2, 53.0, 64.2, 109.4.
r-2-Methoxy-cis-4,cis-6-dimethyl-1,3-dioxane (2): 101 mg (3%),
white, crystalline solid; mp 38.3−38.8 °C (sublimation) (lit.20 mp 38−
39 °C); 1H NMR (400 MHz, CDCl3) δ 1.25 (dt, J = 13.0 Hz, J = 11.1
Hz, 1H), 1.26 (d, J = 6.1 Hz, 6H), 1.54 (dt, J = 13.1, Hz, J = 2.4, 1H),
5251
DOI: 10.1021/acs.joc.8b00707
J. Org. Chem. 2018, 83, 5242−5255
Article
The Journal of Organic Chemistry
278.8 Hz); 19F NMR (376 MHz, CDCl3) δ − 79.2; HRMS (DARTTOF) m/z [M + H]+ calcd for C7H9F6O3 255.0450, found 255.0430.
N-Phthalimidyl-2-aminoacetaldehyde. A solution of 22.960 g
(87.2 mmol) of phthalimidoacetaldehyde diethyl acetal,42 22 mL (132
mmol) of a 6 M aqueous HCl solution, and 175 mL of acetone was
heated at reflux for 5 h. The solution was cooled to room temperature,
transferred to a separatory funnel, and diluted with 200 mL of water.
The solution was extracted with two 200 mL portions of EtOAc, the
combined organic layers were dried (MgSO4), and the solvent
removed under reduced pressure to yield a brown solid that was
recrystallized from CH2Cl2/pentane to afford 12.351 g (75%) of the
title compound as an off-white, crystalline solid: mp 112−113 °C
(sublimation) (lit.43 mp 114 °C); 1H NMR (400 MHz, CDCl3) δ 4.56
(s, 2H), 7.76 (dd, J = 5.5 Hz, J = 3.1 Hz, 2H), 7.89 (dd, J = 5.4 Hz, J =
3.0 Hz, 2H), 9.66 (s, 1H); 13C NMR (100 MHz, CDCl3) δ 47.6,
123.9, 132.2, 134.5, 167.7, 193.7.
N-((cis-4,cis-6-Dimethyl-1,3-dioxan-2-yl)-r-methyl)phthalimide. A mixture of 5.730 g (55.0 mmol) of meso-2,4pentanediol, 10.000 g (52.8 mmol) of N-phthalimidyl-2-aminoacetaldehyde, 254 mg (1.33 mmol) of p-TsOH, and 250 mL of
benzene was heated at reflux under a Dean−Stark trap for 8 h. The
reaction mixture was allowed to cool to room temperature, transferred
to a separatory funnel, washed successively with two 100 mL portions
of saturated, aqueous Na2CO3 and one 150 mL portion of brine, and
dried (Na2SO4), and solvent was removed under reduced pressure to
give a brown solid. Recrystallization (CH2Cl2/cyclohexane) yielded
12.667 g (88%) of the title compound as a white, crystalline solid: mp
173.8−174.2 °C (CH2Cl2/cyclohexane); 1H NMR (400 MHz,
CDCl3) δ 1.18 (d, J = 6.2 Hz, 6H), 1.29 (dt, J = 13.1 Hz, J = 11.2
Hz, 1H), 1.49 (dt, J = 13.1 Hz, J = 2.4 Hz, 1H), 3.68 (dqd, J = 11.1 Hz,
J = 6.1 Hz, J = 2.3 Hz, 2H), 3.85 (d, J = 5.4 Hz, 2H), 4.88 (t, J = 5.4
Hz, 1H), 7.71 (dd, J = 5.4 Hz, J = 3.1 Hz, 2H), 7.86 (dd, J = 5.4 Hz, J
= 3.1 Hz, 2H) ; 13C NMR (100 MHz, CDCl3) δ 21.6, 40.4, 41.5, 72.7,
97.2, 123.5, 132.4, 134.1, 168.2; HRMS (DART-TOF) m/z [M + H]+
calcd for C15H18NO4 276.1230, found 276.1232.
(cis-4,cis-6-Dimethyl-1,3-dioxan-2-yl)-r-methanamine. A mixture of 12.667 g (46.4 mmol) of N-((cis-4,cis-6-dimethyl-1,3-dioxan-2yl)-r-methyl)phthalimide, 3.60 mL (2.375 g, 47.4 mmol) of hydrazine
hydrate (64 wt % hydrazine), and 170 mL of EtOH was heated at
reflux for 5 h and then allowed to cool to room temperature, at which
point a white precipitate of 2,3-dihydro-1,4-phthalazine formed. The
mixture was filtered through a bed of Celite, rinsing with a 100 mL
portion of CH2Cl2 to remove the white solid, and the filtrate was then
concentrated under reduced pressure to give a brown oil. The oil was
triturated with 100 mL of cold CH2Cl2 to further precipitate the
byproduct and filtered a second time through a bed of Celite rinsing
with cold CH2Cl2, and the solvent was removed under reduced
pressure to afford a light orange oil. The crude oil was distilled
(Kugelrohr, bath temperature = 85−90 °C, 40 mm) to yield 4.915 g
(74%) of the title compound as a clear, colorless oil: nD22 = 1.4445; 1H
NMR (400 MHz, CDCl3) δ 1.22 (d, J = 6.2 Hz, 6H), 1.24 (dt, J = 13.0
Hz, J = 11.1 Hz, 1H), 1.30 (br s, 2H), 1.52 (dt, J = 13.2 Hz, J = 2.4 Hz,
1H), 2.80 (d, J = 4.5 Hz, 2H), 3.75 (dqd, J = 11.1 Hz, J = 6.2 Hz, J =
2.3 Hz, 2H), 4.52 (t, J = 4.5 Hz, 1H); 13C NMR (100 MHz, CDCl3) δ
21.7, 40.7, 46.0, 72.5, 101.7; HRMS (DART-TOF) m/z [M + H]+
calcd for C7H16NO2 146.1176, found 146.1180.
r-2-Cyano-cis-4,cis-6-dimethyl-1,3-dioxane (6). Following the
conditions of our previously reported procedure,44 a slurry of 15.995 g
(26.0 mmol) of Oxone, 72 mg (0.45 mmol) of pyridinium bromide,
and 100 mL of CH2Cl2 was stirred for 15 min until the solution
developed a light yellow color. Then, 4.764 g (60.2 mmol) of dry
pyridine was added, the mixture was stirred for 10 min, and 110 mg
(0.52 mmol) of 4-acetamido-2,2,6,6-tetramethylpiperidine-1-oxyl
catalyst was added. The reaction mixture was stirred for an additional
15 min until the solution returned to a light yellow color. A solution of
1.453 g (10.1 mmol) of (cis-4,cis-6-dimethyl-1,3-dioxan-2-yl)-rmethanamine in 20 mL of CH2Cl2 was then added via syringe
pump at a rate of 10 mL/h, and the reaction mixture was allowed to
stir for 12 h at room temperature. The reaction mixture was filtered
through a short plug of silica gel (∼50 g) while rinsing with 150 mL of
hexanes was added, and the solvent was removed under reduced
pressure; this process was repeated twice to azeotropically remove any
excess water or MeOH. The mixture was diluted with 200 mL of
water, transferred to a separatory funnel, and the aqueous layer was
extracted with four 75 mL portions of hexanes.39 The organic extracts
were discarded, and the remaining aqueous solution was extracted with
three 100 mL portions of Et2O, the organic layers were combined and
dried (Na2SO4), 100 mL of hexanes was added, and the solvent was
removed under reduced pressure; this process was repeated twice to
azeotropically remove any excess water. The resulting clear oil
solidified on cooling to −78 °C, and the last traces of solvent were
removed under high vacuum while the flask was allowed to warm to
∼20 °C affording 6.47 g (96%) of the title compound as a white,
crystalline solid: mp 24.3−25.5 °C (Et2O); 1H NMR (400 MHz,
CD2Cl2) δ 2.00 (dt, J = 14.8 Hz, J = 9.9 Hz, 1H), 2.15 (dt, J = 14.8 Hz,
J = 3.1, 1H), 3.25 (d, J = 3.9 Hz, 2H), 4.26−4.35 (m, 2H); 13C NMR
(100 MHz, CD2Cl2) δ 29.4, 69.4 (quartet, JC−F = 32.1 Hz), 101.3,
124.4 (quartet, JC−F = 281.3 Hz); 19F NMR (376 MHz, CD2Cl2) δ −
80.2; HRMS (DART-TOF) m/z [M + H]+ calcd for C5H7F6O2
213.0345, found 213.0349.
r-2-Methoxy-trans-4,trans-6-bis(trifluoromethyl)-1,3-dioxane (3) and r-2-Methoxy-cis-4,cis-6-bis(trifluoromethyl)-1,3dioxane (4). A flame-dried, argon-flushed, round-bottomed flask
containing a magnetic stir bar was charged with 1.552 g (7.31 mmol)
of meso-1,1,1,5,5,5-hexafluoro-2,4-pentanediol and 3.753 g (35.4
mmol) of trimethyl orthoformate. The neat mixture was stirred for
5 min to dissolve the diol, 0.1 mL (ca. 43.5 mg, 0.31 mmol) of a 49
wt/wt % solution of BF3·Et2O in Et2O was added via syringe, and the
mixture was stirred for 2 h at room temperature. Under a cone of
argon gas, ∼0.50 g of solid, anhydrous K2CO3 was added, and the
mixture was stirred under a gentle flow of argon until bubbling of the
mixture ceased. The mixture was then diluted with 15 mL of pentane,
and the contents of the flask were transferred to a separatory funnel,
diluted with an additional 150 mL of pentane, and rinsed with three
portions of a 10% aqueous K2CO3 solution. The pentane layer was
checked by GC−MS to ensure no excess diol remained,40 rinsed with
25 mL of brine, and dried (Na2SO4). The solvent was carefully
removed via use of a rotary evaporator operating at ∼15 °C and water
aspirator pressure to afford ∼1.2 g of a light yellow oil containing a
mixture of trimethyl orthoformate, and a mixture of the trans (3) and
cis (4) title compounds in a ∼5:1 ratio.41 Separation of compounds 3
and 4 was achieved by column chromatography (75 g neutral alumina,
0 → 10% Et2O/pentane); 20 mL fractions were collected: 286 mg
(16%) of pure trans-isomer (3) was collected in fractions 7−8; 68 mg
(4%) of pure cis-isomer (4) was collected in fractions 26−42. Progress
of the chromatography was monitored by GC−MS as TLC
visualization methods were insufficient. GC−MS analysis on a 25 m
× 0.20 mm × 0.33 μm DB-5 5% phenyl/95% dimethyl polysiloxane
column using the following temperature program: 5 min at an initial
temperature of 40 °C then a gradual increase in temperature at a rate
of 15 °C/min to a final temperature of 240 °C which is held constant
for 20 min. A solvent delay of 2 min was used and compounds of
interest eluted in the following order: trimethyl orthoformate, 2.60
min; 3, 5.95 min; 4, 6.20 min; meso-1,1,1,5,5,5-hexafluoro-2,4pentanediol, 7.30 min.
r-2-Methoxy-trans-4,trans-6-bis(trifluoromethyl)-1,3-dioxane (3):
clear, colorless oil; 1H NMR (400 MHz, CDCl3) δ 1.91 (dtd, J = 12.9
Hz, J = 2.2 Hz, J = 1.0 Hz, 1H), 2.03 (dt, J = 12.5 Hz, J = 12.1, 1H),
3.42 (s, 3H), 4.56 (dqd, J = 11.8 Hz, J = 5.8 Hz, J = 2.9 Hz, 2H), 5.61
(s, 1H); 13C NMR (100 MHz, CDCl3) δ 22.2, 53.8, 65.5 (quartet, JC−F
= 34.4 Hz), 108.8, 123.4 (quartet, JC−F = 278.3 Hz); 19F NMR (376
MHz, CDCl3) δ − 79.9; HRMS (DART-TOF) m/z [M + H]+ calcd
for C7H9F6O3 255.0450, found 255.0438.
r-2-Methoxy-cis-4,cis-6-bis(trifluoromethyl)-1,3-dioxane (4):
white, crystalline solid; mp 27.7−28.4 °C (pentane); 1H NMR (400
MHz, CDCl3) δ 1.90 (dt, J = 13.0 Hz, J = 3.1 Hz, 1H), 1.98 (dt, J =
12.8 Hz, J = 11.7, 1H), 3.55 (s, 3H), 4.26 (dqd, J = 11.4 Hz, J = 5.6 Hz,
J = 3.1 Hz, 2H), 5.38 (s, 1H); 13C NMR (100 MHz, CDCl3) δ 22.1,
52.5, 71,8 (quartet, JC−F = 34.6 Hz), 110.6, 122.8 (quartet, JC−F =
5252
DOI: 10.1021/acs.joc.8b00707
J. Org. Chem. 2018, 83, 5242−5255
Article
The Journal of Organic Chemistry
The mixture was filtered through a bed of Celite, the filtrate was
concentrated under reduced pressure resulting in further precipitation
of the byproduct and filtered a second time through a bed of Celite
rinsing with a cold 2:1 mixture of pentane/CH2Cl2, and the solvent
was removed under reduced pressure to afford a light yellow oil. The
oil was distilled (Kugelrohr, bath temperature = 120−130 °C, 40 mm)
to yield 1.402 g (77%) of the title compound as a crystalline, white
solid: mp 36.4−37.2 °C (sublimation); 1H NMR (400 MHz, CDCl3)
δ 1.30 (br s, 2H), 1.85−2.00 (m, 2H), 2.94 (d, J = 4.1 Hz, 2H), 4.11−
4.20 (m, 2H), 4.65 (t, J = 4.1 Hz, 2H); 13C NMR (100 MHz, CDCl3)
δ 22.9, 45.2, 73.0 (quartet, JC−F = 34.0 Hz), 102.0, 123.0 (quartet, JC−F
= 279.0 Hz); 19F NMR (376 MHz, CDCl3) δ − 79.5; HRMS (DARTTOF) m/z [M + H]+ calcd for C7H10F6NO2 254.0610, found
254.0613.
r-2-Cyano-cis-4,cis-6-bis(trifluoromethyl)-1,3-dioxane (8).
Following the conditions of our previously reported procedure,44 a
slurry of 16.010 g (26.1 mmol) of Oxone, 71 mg (0.44 mmol) of
pyridinium bromide, and 100 mL of CH2Cl2 was stirred for 15 min
until the solution developed a light yellow color. Then 4.754 g (60.1
mmol) of dry pyridine was added, and the mixture was stirred for 10
min before the addition of 108 mg (0.51 mmol) of the 4-acetamido2,2,6,6-tetramethylpiperidine-1-oxyl catalyst and then stirred for an
additional 15 min until the solution returned to a light yellow color. A
solution of 582 mg (2.30 mmol) of (cis-4,cis-6-bis(trifluoromethyl)1,3-dioxan-2-yl)-r-methanamine in 20 mL of CH2Cl2 was added to the
slurry via syringe pump at a rate of 10 mL/h, and following the
addition, the reaction mixture was stirred for 12 h at room
temperature. The reaction mixture was filtered through a short plug
of SiO2 (∼50 g) while rinsing with 150 mL of CH2Cl2. The filtrate was
concentrated under reduced pressure to yield a clear oil, which was
triturated with pentane. Upon removal of the solvent, 550 mg (96%)
of the title compound was isolated as a white solid: mp 30.1−30.8 °C
(pentane); 1H NMR (400 MHz, CDCl3) δ 2.00 (dt, J = 13.3 Hz, J =
2.7 Hz, 1H), 1.98 (dt, J = 13.3 Hz, J = 11.8, 1H), 4.26 (dqd, J = 11.6
Hz, J = 5.4 Hz, J = 2.7 Hz, 2H), 5.47 (s, 1H); 13C NMR (100 MHz,
CDCl3) δ 22.5, 73.5 (quartet, JC−F = 35.5 Hz), 87.9, 112.0, 122.1
(quartet, JC−F = 278.6 Hz); 19F NMR (376 MHz, CDCl3) δ − 79.1;
HRMS (DART-TOF) m/z [M − CN]+ calcd for C6H5F6O2 223.0188,
found 223.0217.
r-2-Cyano-trans-4,trans-6-bis(trifluoromethyl)-1,3-dioxane
(7). To an oven-dried, argon-flushed 2 mL ampule was added 68 mg
(0.273 mmol) of r-2-cyano-cis-4,cis-6-bis(trifluoromethyl)-1,3-dioxane
(8), 1.5 mL of Et2O, 11 mg (0.10 mmol) of potassium tert-butoxide,
and a small magnetic stir bar. The ampule was flame-sealed and stirred
for 1 d, at which point it was opened and the contents were transferred
by pipet into a 3 mL vial containing 1 mL of a saturated, aqueous
NH4Cl and shaken. The organic layer was carefully separated from the
aqueous layer by pipet and transferred into a separate vial, and the
solution was concentrated to ∼0.3 mL with a gentle stream of N2 gas.
Compound 7 was isolated from the concentrated solution by
preparative TLC on a 1000 μm SiO2 glass-backed plate using a 10%
solution of Et2O in pentane as eluent to afford 32 mg (47%) of the
title compound (Rf = 0.58−0.78) as a crystalline, white solid: mp
33.1−33.7 °C (Et2O); 1H NMR (400 MHz, CDCl3) δ 2.04−2.19 (m,
2H), 4.50−4.60 (m, 2H), 5.97 (s, 1H); 13C NMR (100 MHz, CDCl3)
δ 22.5, 69.9 (quartet, JC−F = 35.5 Hz), 86.4, 111.8, 122.4 (quartet, JC−F
= 279.1 Hz); 19F NMR (376 MHz, CDCl3) δ − 79.2; HRMS (DARTTOF) m/z [M − CN]+ calcd for C6H5F6O2 223.0188, found
223.0187.
Equilibrations. For each pair of anancomeric diastereomers,
equilibrium was approached independently from pure samples of
each isomer. The diastereomers were equilibrated at room temperature (∼23 °C) in sealed vials or ampules as solutions in dry solvent
over the specified catalyst (Table 1 and Figure 4) and neutralized prior
to GC analysis. The area ratio of the isomers was determined by GC
analysis using one of the following columns: a 30 m × 0.25 mm × 0.25
μm Optima-225 50% cyanopropyl/50% phenylmethyl polysiloxane
column or a 30 m × 0.25 mm × 0.25 μm EC-1 100% dimethyl
polysiloxane column. The analysis parameters are detailed in the
Supporting Information. When the same area ratios were obtained
CH2Cl2. The filtrate was concentrated under reduced pressure to give
1.307 g of a light yellow oil, which was distilled (Kugelrohr, bath
temperature = 80−100 °C, 10 mm) to yield 1.159 g (82%) of the title
compound as a clear, colorless oil: nD23 = 1.4285; 1H NMR (400 MHz,
CDCl3) δ 1.27 (d, J = 6.2 Hz, 6H), 1.41 (dt, J = 13.5 Hz, J = 11.1 Hz,
1H), 1.58 (dt, J = 13.5 Hz, J = 2.4 Hz, 1H), 3.81 (dqd, J = 11.1 Hz, J =
6.2 Hz, J = 2.4 Hz, 2H), 5.33 (s, 1H); 13C NMR (100 MHz, CDCl3) δ
21.3, 39.9, 74.3, 88.6, 114.7; HRMS (DART-TOF) m/z [M − H]+
calcd for C7H10NO2 140.0717, found 140.0712.
r-2-Cyano-trans-4,trans-6-dimethyl-1,3-dioxane (5). To an
oven-dried, argon-flushed, 5 mL vial was added 417 mg (2.96
mmol) of r-2-cyano-cis-4,cis-6-dimethyl-1,3-dioxane (6), 4.5 mL of
pentane, 61 mg (0.54 mmol) of potassium tert-butoxide, and a small
magnetic stir bar. The vial was tightly sealed with a Teflon-coated cap
and to stirred for 24 h, at which point, it was opened and the contents
were transferred by pipet and filtered through a bed of Celite into a 5
mL vial. The solution was concentrated to ∼0.3 mL with a gentle
stream of N2 gas. Compound 7 was isolated from the concentrated
solution by preparative TLC on a 1000 μm SiO2 glass-backed plate
using a 10% solution of Et2O in hexanes as eluent to give a clear oil,
which crystallized upon cooling to 0 °C to afford 100 mg (24%) of the
title compound (Rf = 0.47−0.54) as a crystalline, white solid: mp
27.8−28.9 °C (Et2O/pentane); 1H NMR (400 mHz, CDCl3) δ 1.26
(d, J = 6.2 Hz, 6H), 1.40 (dt, J = 13.5 Hz, J = 11.3 Hz, 1H), 1.66 (dt, J
= 13.6 Hz, J = 2.0 Hz, 1H), 4.18 (dqd, J = 11.3 Hz, J = 6.1 Hz, J = 2.3
Hz, 2H), 5.74 (s, 1H); 13C NMR (100 mHz, CDCl3) δ 21.3, 39.9,
69.2, 87.4, 114.7; HRMS (DART-TOF) m/z [M + NH4]+ calcd for
C7H15N2O2 159.1128, found 159.1130.
N-((cis-4,cis-6-Bis(trifluoromethyl)-1,3-dioxan-2-yl)-rmethyl)phthalimide. A mixture of 3.679 g (17.3 mmol) of meso1,1,1,5,5,5-hexafluoro-2,4-pentanediol, 4.572 g (17.3 mmol) of
phthalimidoacetaldehyde diethyl acetal,42 239 mg (1.26 mmol) of pTsOH, and 150 mL of hexanes was heated at reflux under a Dean−
Stark trap for 24 h, and the reaction mixture was then allowed to cool
to room temperature. Upon cooling, a light brown solid precipitated.
The reaction mixture was then transferred to a separatory funnel and
diluted with 200 mL of CH2Cl2, and the organic layer was rinsed with
two 100 mL portions of a saturated, aqueous Na2CO3 and dried
(Na2SO4), and solvent was removed under reduced pressure to afford
a brown solid. The crude material was dissolved in 100 mL of EtOAc
and adsorbed onto 10 g of SiO2 for flash chromatography. Separation
of the title compound was achieved by “double” flash column
chromatography (200 g SiO2, 30 → 50% EtOAc/hexanes); 50 mL
fractions were collected: 3.469 g (52%) of the title compound of
∼90% purity was collected in fractions 7−16. This almost pure
material was dissolved in 100 mL of EtOAc and was adsorbed onto 8.5
g of SiO2 for additional chromatography (150 g SiO2, 30 → 70%
Et2O/hexanes); 50 mL fractions were collected: 3.036 g (46%) of the
title compound was collected in fractions 16−26. Progress of the
chromatography was monitored by TLC (SiO2, eluent: 50% Et2O in
hexanes) and UV-light allowed for visualization of the title compound.
The title compound had an Rf = 0.43: white, crystalline solid; mp
196.1−197.0 °C (Et2O/hexanes); 1H NMR (400 MHz, CDCl3) δ 1.88
(dt, J = 13.0 Hz, J = 2.6 Hz, 1H), 2.00 (dt, J = 12.8 Hz, J = 11.8, 1H),
4.00 (d, J = 5.3 Hz, 2H), 4.12 (dqd, J = 11.6 Hz, J = 5.6 Hz, J = 2.9 Hz,
2H), 5.08 (t, J = 5.4 Hz, 2H), 7.74 (dd, J = 5.4 Hz, J = 3.0, 2H), 7.87
(dd, J = 5.4 Hz, J = 3.0, 2H); 13C NMR (100 MHz, CDCl3) δ 22.8,
40.4, 73.1 (quartet, JC−F = 34.1 Hz), 97.4, 122.8 (quartet, JC−F = 278.9
Hz), 123.8, 132.0, 134.4, 167.9; 19F NMR (376 MHz, CDCl3) δ −
79.4; HRMS (DART-TOF) m/z [M + H]+ calcd for C15H12F6NO4
384.0665, found 384.0645.
(cis-4,cis-6-Bis(trifluoromethyl)-1,3-dioxan-2-yl)-r-methanamine. A mixture of 2.756 g (7.19 mmol) of N-((cis-4,cis-6bis(trifluoromethyl)-1,3-dioxan-2-yl)-r-methyl)phthalimide, 0.850 mL
(528 mg, 10.5 mmol) of hydrazine hydrate (64 wt % hydrazine), and
100 mL of EtOH was heated at reflux for 5 h and then allowed to cool
to room temperature, at which point a white precipitate of 2,3dihydro-1,4-phthalazine formed. An additional 100 mL of a 2:1
mixture of pentane/CH2Cl2 was added to the flask, which was then
cooled to 0 °C in an ice/water bath to further precipitate the solid.
5253
DOI: 10.1021/acs.joc.8b00707
J. Org. Chem. 2018, 83, 5242−5255
Article
The Journal of Organic Chemistry
from initially pure samples of each epimer, it was deemed that
equilibrium had been attained. Area ratios for each equilibrium were
taken as the average of 10 independent determinations from each side,
the isomers were assumed to have identical GC response ratios, and
the equilibrium constant for each equilibrium was determined from
these area ratios.
Calculations. All calculations were carried out using Gaussian-16.45
Hirshfeld charges were obtained using the keywords pop = Hirshfeld
and density = current. The structures were drawn using CYLview
1.0b.46
■
(8) For a short, curated list of reviews, see: Juaristi, E.; Bandala, Y.
Anomeric Effect in Saturated Heterocyclic Ring Systems. Adv.
Heterocycl. Chem. 2012, 105, 189−222 and references cited therein.
Juaristi, E.; Cuevas, G. The Anomeric Effect; CRC Press: Boca Raton,
1995. Kirby, A. J. The Anomeric Effect and Related Stereoelectronic Effects
at Oxygen; Springer-Verlag: New York, 1983. Anomeric Effect: Origin
and Consequences; Szarek, W. A., Horton, D., Eds.; ACS Symposium
Series; American Chemical Society: Washington, DC, 1979; Vol. 87.
(9) For an intriguing discussion advocating “a clear semantic
distinction between cause and effect” when discussing the “anomeric
effect”, see: Filloux, C. M. The Problem of Origins and Origin of the
Problem: Influence of Language on Studies Concerning the Anomeric
Effect. Angew. Chem., Int. Ed. 2015, 54, 8880−8894.
(10) Lemieux, R. U. Effects of Unshared Pairs of Electrons and Their
Solvation on Conformational Equilibria. Pure Appl. Chem. 1971, 25,
527−548.
(11) (a) Takahashi, O.; Yamasaki, K.; Kohno, Y.; Ueda, K.; Suezawa,
H.; Nishio, M. The Origin of the Generalized Anomeric Effect:
Possibility of CH/n and CH/pi Hydrogen Bonds. Carbohydr. Res.
2009, 344, 1225−1229. (b) Takahashi, O.; Yamasaki, K.; Kohno, Y.;
Ueda, K.; Suezawa, H.; Nishio, M. The Origin of the Relative Stability
of Axial Conformers of Cyclohexane and Cyclohexanone Derivatives:
Importance of CH/n and CH/pi Hydrogen Bonds. Bull. Chem. Soc.
Jpn. 2009, 82, 272−276. (c) For a comprehensive review of nonclassical hydrogen bonds, see: Takahashi, O.; Kohno, Y.; Nishio, M.
Relevance of Weak Hydrogen Bonds in the Conformation of Organic
Compounds and Bioconjugates: Evidence from Recent Experimental
Data and High-Level ab Initio MO Calculations. Chem. Rev. 2010, 110,
6049−6076.
(12) (a) Altona, C.; Romers, C.; Havinga, E. Molecular Structure and
Conformation of Some Dihalogenodioxanes. Tetrahedron Lett. 1959, 1,
16−20. (b) Romers, C.; Altona, C.; Buys, H. R.; Havinga, E. In Topics
in Stereochemistry; Allinger, N. L., Eliel, E. L., Eds.; Wiley-Interscience:
New York, 1969; Vol. 4, pp 39−97.
(13) Alabugin, I. V. Stereoelectronic Effects; Wiley: Hoboken, NJ, 2016.
(14) (a) Mo, Y. Computational Evidence that Hyperconjugation
Interactions are not Responsible for the Anomeric Effect. Nat. Chem.
2010, 2, 666−671. (b) Wang, C.; Chen, Z.; Wu, W.; Mo, Y. How the
Generalized Anomeric Effect Influences the Conformational Preference. Chem. - Eur. J. 2013, 19, 1436−1444. (c) Wang, C.; Ying, F.; Wu,
W.; Mo, Y. How Solvent Influences the Anomeric Effect: Roles of
Hyperconjugation verses Steric Interactions on the Conformational
Preference. J. Org. Chem. 2014, 79, 1571−1581.
(15) Perrin, C. L.; Armstrong, K. B.; Fabian, M. A. The Origin of the
Anomeric Effect. Conformational Analysis of 2-Methoxy-1,3-dimethylhexahydropyrimidine. J. Am. Chem. Soc. 1994, 116, 715−722.
(16) (a) Bailey, W. F.; Eliel, E. L. Conformational Analysis. 29. 2Substituted and 2,2-Disubstituted 1,3-Dioxanes. Generalized and
Reversed Anomeric Effects. J. Am. Chem. Soc. 1974, 96, 1798−1806.
(b) Wiberg, K. B.; Murcko, M. A. Rotational Barriers. 4. Dimethoxymethane − the Anomeric Effect Revisited. J. Am. Chem. Soc. 1989, 111,
4821−4828. (c) Wiberg, K. B.; Marquez, M. The Energy Components
of the Anomeric Effect for 2-Methoxytetrahydropyran − An
Experimental Comparison of the Gas Phase and Solution. J. Am.
Chem. Soc. 1994, 116, 2197−2198. (d) Bailey, W. F.; Rivera, A. D.
Conformational Change Occasioned by Complexation: ContraAnomeric Effect Epimerization of 2-Methoxy-1,3-dioxanes in the
Presence of Magnesium Bromide. J. Org. Chem. 1987, 52, 1559−1562.
(e) Bailey, W. F.; Lambert, K. M.; Wiberg, K. B.; Mercado, B. Q. Effect
of Remote Substituents on the Conformational Equilibria of 2,2Diaryl-1,3-dioxanes: Importance of Electrostatic Interactions. J. Org.
Chem. 2015, 80, 4108−4115.
(17) Lambert, K. M.; Stempel, Z. D.; Wiberg, K. B.; Bailey, W. F.
Experimental Demonstration of Sizable Nonclassical CH···G Hydrogen Bond in Cyclohexane Derivatives: Stabilization of an Axial Cyano
Group. Org. Lett. 2017, 19, 6408−6411.
(18) Signorell, R.; Marquardt, R.; Quack, M.; Suhm, M. A. The
Permanent Dipole Moment of CH2D2: FIR Spectroscopy, Centrifugal
Distortion Effects and Quantum Monte Carlo Calculations with 9-
ASSOCIATED CONTENT
* Supporting Information
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.joc.8b00707.
Appendix entitled Atomic Charges; NMR spectra of all
products; equilibration data; analytical GC data; a
summary of the calculations, including computed
energies and coordinates (PDF)
■
AUTHOR INFORMATION
Corresponding Authors
*E-mail: kenneth.wiberg@yale.edu.
*E-mail: william.bailey@uconn.edu.
ORCID
Kenneth B. Wiberg: 0000-0001-8588-9854
William F. Bailey: 0000-0001-9159-0218
Kyle M. Lambert: 0000-0002-8230-2840
Present Address
§
(K.B.W.) 865 Central Avenue, Apt 404, Needham, MA 02492.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
We are grateful to Prof. Steven R. Kass for providing
suggestions on the preparation of meso-1,1,1,5,5,5-hexafluoro2,4-pentanediol. The work at the University of Connecticut was
supported by a grant from Procter & Gamble Pharmaceuticals.
■
REFERENCES
(1) Eliel, E. L.; Wilen, S. H.; Mander, L. N. Stereochemistry of Organic
Compounds; Wiley: New York, 1994; pp 686−754.
(2) Nader, F. W.; Eliel, E. L. Conformational Analysis 22.
Conformational Equilibria in 2-Substituted 1,3-Dioxanes. J. Am.
Chem. Soc. 1970, 92, 3050−3055.
(3) Wiberg, K. B.; Hammer, J. D.; Castejon, H.; Bailey, W. F.;
DeLeon, E. L.; Jarret, R. M. Conformational Studies in the
Cyclohexane Series. 1. Experimental and Computational Investigations
of Methyl, Ethyl, Isopropyl and tert-Butylcyclohexanes. J. Org. Chem.
1999, 64, 2085−2095.
(4) For example, see: (a) Hassel, O.; Ottar, B. The Structure of
Molecules Containing Cyclohexane or Pyranose Rings. Acta Chem.
Scand. 1947, 1, 929−943. (b) Capon, B. Mechanism in Carbohydrate
Chemistry. Chem. Rev. 1969, 69, 407−498.
(5) (a) Lemieux, R. U.; Chu, N. J. Abstracts of Papers; 133rd National
Meeting of the American Chemical Society; San Francisco, CA, April
1958; American Chemical Society: Washington, DC, 1958; p 31N.
(b) Chu, N. J. Ph.D. Dissertation, University of Ottawa, 1959.
(6) Edward, J. T. Stability of Glycosides to Acid Hydrolysis. A
Conformational Analysis. Chem. Ind. (London) 1955, 1102−1104.
(7) Edward, J. T. Anomeric Effect. How it Came to be Postulated. In
ACS Symposium Series; Thatcher, G. R. J., Ed.; American Chemical
Society: Washington, DC, 1993; Vol. 539, pp 1−5.
5254
DOI: 10.1021/acs.joc.8b00707
J. Org. Chem. 2018, 83, 5242−5255
Article
The Journal of Organic Chemistry
Dimensioanl Analytic Dipole Moment and Potential Function for
Methane. Mol. Phys. 1996, 89, 297−313.
(19) Eliel, E. L.; Knoeber, M. C. Conformational Analysis. XVI. 1,3Dioxanes. J. Am. Chem. Soc. 1968, 90, 3444−3458.
(20) Eliel, E. L.; Nader, F. W. Conformational Analysis. XX.
Stereochemistry of Reaction of Grignard Reagents with Ortho Esters.
Synthesis of 1,3-Dioxanes with Axial Substituents at C-2. J. Am. Chem.
Soc. 1970, 92, 584−590.
(21) Onsager, L. Electric Moments of Molecules in Liquids. J. Am.
Chem. Soc. 1936, 58, 1486−1493.
(22) We have found that plots of ΔG° vs the Onsager function, (ε −
1)/(2ε + 1), usually give a fairly good linear relationship when the
solvents were cyclohexane, di-n-butyl ether, THF, acetone, and
MeCN, and in one case where the experimentally determined gas
phase value was available, it fell on the line. The correlation was
unsuccessful with Et2O, aromatic solvents, and halogenated solvents.
See: (a) Wiberg, K. B.; Rush, D. J. Solvent Effects on the Thioamide
Rotational Barrier: An Experimental and Theoretical Study. J. Am.
Chem. Soc. 2001, 123, 2038−2046. (b) Castejon, H.; Wiberg, K. B.
Solvent Effects on Methyl Transfer Reactions. 1. The Menschutkin
Reaction. J. Am. Chem. Soc. 1999, 121, 2139−2146.
(23) Juaristi, E.; Valle, L.; Valenzuela, B. A.; Aguilar, M. A. S-C-P
Anomeric Interactions. 4. Conformational Analysis of 2-(Diphenylphosphinoyl)-1,3-dithiane. J. Am. Chem. Soc. 1986, 108, 2000−2005.
(24) (a) Stolow, R. D.; Giants, T. W.; Roberts, J. D. Fluorine and
Proton NMR Studies of the Conformation of 4,4-Difluorocyclohexanol. Tetrahedron Lett. 1968, 9, 5777−5780. (b) Stolow, R. D.; Lewis,
D. I.; D’Angelo, P. A. Conformational Studies. XV. Transannular
Electrostatic Interaction in 4-Substituted 1,1-Dimethylpiperidinnium
Cations. Tetrahedron 1970, 26, 5831−5844.
(25) Kleinpeter, E.; Heydenreich, M.; Koch, A.; Linker, T. Synthesis
and NMR Spectroscopic Conformational Analysis of Esters of 4Hydroxycyclohexanone: the More Polar the Molecule the More Stable
the Axial Conformer. Tetrahedron 2012, 68, 2363−2373 and
references cited therein.
(26) Wiberg, K. B.; Castejon, H.; Bailey, W. F.; Ochterski, J.
Conformational Studies in the Cyclohexane Series. 2. Phenylcyclohexane and 1-Methyl-1-phenylcyclohexane. J. Org. Chem. 2000,
65, 1181−1187.
(27) (a) Hirshfeld, F. L. Bonded Atom Fragments for Describing
Molecular Charge Densities. Theor. Chim. Acc. 1977, 44, 129−138.
(b) Marenich, A. V.; Jerome, S. V.; Cramer, C. J.; Truhlar, D. G.
Charge Model 5. An Extension of Hirshfeld Population Analysis for
the Accurate Description of Molecular Interactions in Gaseous and
Condensed Phases. J. Chem. Theory Comput. 2012, 8, 527−541.
(28) Tomasi, J.; Mennucci, M.; Cammi, R. Quantum Chemical
Continuum Solvation Models. Chem. Rev. 2005, 105, 2999−3094.
(29) Jorgensen, W. L.; Thomas, L. L. Perspective on Free-Energy
Perturbation Calculations for Chemical Equilibria. J. Chem. Theory
Comput. 2008, 4, 869−876.
(30) The orientation of the dipole is readily found as follows. After a
geometry optimization that also gives the components of the dipole
moment, one requests a new .com file with just the energy requested.
This file is edited to add two new entries after the coordinates. The
first has H 0.0 0.0 00 and the second has H followed by the
components of the dipole moment. The connection table is edited to
add a bond between the two entries above. Now, if this file is opened
in Gaussview, the dipole vector can be seen, and it will rotate with the
molecule.
(31) Bent, H. A. An Appraisal of Valence Bond Structures and
Hybridization in Compounds of the 1st Row Elements. Chem. Rev.
1961, 61, 275−311.
(32) “Entities are not to be multiplied without necessity” (Non sunt
multiplicanda entia sine necessitate).
(33) Bailey, W. F.; Lambert, K. M.; Stempel, Z. D.; Wiberg, K. B.;
Mercado, B. Q. C. Controlling the Conformational Energy of a Phenyl
Group by Tuning the Strength of a Nonclassical CH···O Hydrogen
Bond. The Case of 5-Phenyl-1,3-Dioxane. J. Org. Chem. 2016, 81,
12116−12127.
(34) Pritchard, J. G.; Vollmer, R. L. Conversion of Acetates into
Alcohols in 1,3-Diol Systems. J. Org. Chem. 1963, 28, 1545−1549.
(35) Pan, H.; Tu, S.; Zhang, C.; Young, A.; Fontaine, P. P. Efficient
Separation of Diastereomeric Mixtures of syn- and anti-Pentanediol.
Org. Process Res. Dev. 2015, 19, 463−469.
(36) (a) Shokri, A.; Kass, S. R. Solvent Effects on the Molecular
Recognition of Anions. Chem. Commun. 2013, 49, 11674−11676.
(b) Shokri, A.; Wang, X. B.; Kass, S. R. Electron Withdrawing
Trifluoromethyl Groups in Combination with Hydrogen Bonds in
Polyols: Bronsted Acids, Hydrogen Bond Catalysts, and Anion
Receptors. J. Am. Chem. Soc. 2013, 135, 9525−9530.
(37) An azeotrope of unknown composition sometimes arises leading
to removal of the diol from the reaction mixture. This can be detected
by the appearance of a milky, turbid lower layer in the Dean−Stark
trap within 1 h of reflux. This is easily dealt with by cooling the
reaction mixture to room temperature, emptying the contents of the
Dean−Stark trap into the reaction flask, and adding 2−3 mL of
deionized water to the reaction mixture before continuing.
(38) The aqueous layers were extracted with two 100-mL portions of
Et2O, dried over Na2SO4, and solvent removed to afford the racemic
1,1,1,5,5,5-hexafluoro-2,4-pentanediol as a white crystalline solid. This
diol was not of interest in the present study, however, it had the
following properties: mp 75.1−76.0 °C (Et2O); 1H NMR (400 MHz,
CDCl3) δ 1.99 (dd, J = 7.5, Hz, J = 5.7, 2H), 2.69 (d, J = 5.8 Hz, 2H),
4.30−4.40 (m, 2H); 13C NMR (100 MHz, CD2Cl3) δ 29.3, 66.6
(quartet, JC−F = 32.1 Hz), 125.1 (quartet, JC−F = 281.3 Hz); 19F NMR
(376 MHz, CD2Cl2) δ −79.7; HRMS (DART-TOF) m/z [M + H]+
calcd for C5H7F6O2 213.0345, found 213.0366.
(39) The hexane extraction facilitates the removal of benzaldehyde;
the meso diol is not significantly soluble in hexanes and remains in the
aqueous layer.
(40) It is essential to ensure no excess diol remains as its separation
from the title compounds is unsuccessful through column chromatography. If excess diol remained, additional aqueous K2CO3 washings are
necessary.
(41) The boiling point of the title compounds appears to be
extremely low, and care has to be taken during solvent removal to
prevent loss of the product.
(42) Anderson, M. A.; Shim, H.; Raushel, F. M.; Cleland, W. W.
Hydrolysis of Phosphotriesters: Determination of Transition States in
Parallel Reactions by Heavy-Atom Isotope Effects. J. Am. Chem. Soc.
2001, 123, 9246−9253.
(43) Alvarez, M.; Fernandez, D.; Joule, J. A. Synthesis of
Deoxyvariolin B. Tetrahedron Lett. 2001, 42, 315−317.
(44) Lambert, K. M.; Bobbitt, J. M.; Eldirany, S. A.; Kissane, L. E.;
Sheridan, R. K.; Stempel, R. D.; Sternberg, F. H.; Bailey, W. F. MetalFree Oxidation of Primary Amines to Nitriles through Coupled
Catalytic Cycles. Chem. - Eur. J. 2016, 22, 5156−5159.
(45) Gaussian 16, Revision A.03: Frisch, M. J.; Trucks, G. W.;
Schlegel, H. B.; Scuseria, G. E.; Robb, M. A; Cheeseman, J. R.;
Scalmani, G.; Barone, V.; Peterson, G. A.; Nakatsuji, H.; Li, X.;
Caricto, M.; Marenich, A. V.; Bloino, J.; Janesko, B. G.; Gomperts, R.;
Mennucci, B.; Hratchian, H. F.; Ortiz, J. V.; Izmayiov, A. F.;
Sonnenberg, J. L.; Williams-Young, D.; Ding, F.; Lipparini, F.; Egidi,
F.; Goings, I.; Peng, B.; Petrone, A.; Henderson, T.; Ranasinghe, D.;
Zakrzewski, V. G.; Gao, J.; Rega, N.; Zheng, G.; Liang, W.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Throssell, E.;
Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M. J.; Heyd,
J. J.; Brothers, E. N.; Kudin, K. N.; Staroverov, V. N.; Keith, T. A.;
Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A. P.; Burant, J.
C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Millam, J. M.; Klene, M.;
Adamo, C.; Cammi, R.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.;
Farkas, O.; Foresman, J. B.; Fox, D. J. Gaussian, Inc., Wallingford, CT,
2016.
(46) Legault, C. Y. CYLview 1.0b; Université de Sherbrook,
Sherbrook, QC, Canada, 2009; http://www.cylview.org.
5255
DOI: 10.1021/acs.joc.8b00707
J. Org. Chem. 2018, 83, 5242−5255