Is There a Generalized Reverse Anomeric Effect? Substituent and Solvent Effects on the Configurational Equilibria of Neutral and Protonated N-Arylglucopyranosylamines and N-Aryl-5-thioglucopyranosylamines

Article abstract of DOI:10.1021/jo991520j

The effects of substitution and solvent on the configurational equilibria of neutral and protonated N-(4-Y-substituted-phenyl) peracetylated 5-thioglucopyranosylamines (Y = OMe, H, CF₃, NO₂) 1-4 and N-(4-Y-substituted-phenyl) peracetylated glucopyranosylamines (Y = OMe, H, NO₂) 9-11 are described. The configurational equilibria were determined by direct integration of the resonances of the individual isomers in the ¹H NMR spectra after equilibration of both α- and β-isomers. The equilibrations of the neutral compounds 1-4 in CD₃OD, CD₃NO₂, and (CD₃)₂CO were achieved by HgCl₂ catalysis and those of the neutral compounds 9-11 in CD₂Cl₂ and CD₃OD by triflic acid catalysis. The equilibrations of the protonated compounds in both the sulfur series (solvents, CD₃OD, CD₃NO₂, (CD₃)₂CO, CDCl₃, and CD₂Cl₂) and oxygen series (solvents, CD₂Cl₂ and CD₃OD) were achieved with triflic acid. The substituent and solvent effects on the equilibria are discussed in terms of steric and electrostatic effects and orbital interactions associated with the endo-anomeric effect. A generalized reverse anomeric effect does not exist in neutral or protonated N-aryl-5-thioglucopyranosylamines and N-arylglucopyranosylamines. The anomeric effect ranges from 0.85 kcal mol^-1 in 2 to 1.54 kcal mol^-1 in 10. The compounds 1-4 and 9-11 show an enhanced endoanomeric effect upon protonation, ranging from 1.73 kcal mol^-1 in 2 to 2.57 kcal mol^-1 in 10. We estimate the increase in the anomeric effect upon protonation of 10 to be approximately 1.0 kcal mol^-1. However, this effect is offset by steric effects due to the associated counterion which we estimate to be approximately 1.2 kcal mol^-1. The values of K_eq(axial-equatorial) in protonated 1-4 increase in the order OMe < H < CF₃ < NO₂, in agreement with the dominance of steric effects (due to the counterion) over the endo-anomeric effect. The values of K_eq(axial-equatorial) in protonated 9-11 show the trend OMe > H < NO₂ that is explained by the balance of the endo-anomeric effect and steric effects in the individual compounds. The trends in the values of the C₁-H₁ coupling constants for 1-4 and the corresponding deacetylated compounds 5-8 as a function of substituent and α- or β-configuration are discussed in terms of the Perlin effect and the interplay of the endo- and exo-anomeric effects.

Full text of DOI:10.1021/jo991520j

220
J . Org. Chem. 2000, 65, 220-226
Is Th er e a Gen er a lized Rever se An om er ic Effect? Su bstitu en t a n d
Solven t Effects on th e Con figu r a tion a l Equ ilibr ia of Neu tr a l a n d
P r oton a ted N-Ar ylglu cop yr a n osyla m in es a n d
N-Ar yl-5-th ioglu cop yr a n osyla m in es†
Karla D. Randell, Blair D. J ohnston, David F. Green, and B. Mario Pinto*
Department of Chemistry, Simon Fraser University, Burnaby, British Columbia, Canada, V5A 1S6
Received September 27, 1999
The effects of substitution and solvent on the configurational equilibria of neutral and protonated
N-(4-Y-substituted-phenyl) peracetylated 5-thioglucopyranosylamines (Y ) OMe, H, CF3, NO2) 1-4
and N-(4-Y-substituted-phenyl) peracetylated glucopyranosylamines (Y ) OMe, H, NO2) 9-11 are
described. The configurational equilibria were determined by direct integration of the resonances
1
of the individual isomers in the H NMR spectra after equilibration of both R- and â-isomers. The
equilibrations of the neutral compounds 1-4 in CD3OD, CD3NO2, and (CD3)2CO were achieved by
HgCl2 catalysis and those of the neutral compounds 9-11 in CD2Cl2 and CD3OD by triflic acid
catalysis. The equilibrations of the protonated compounds in both the sulfur series (solvents, CD3OD,
CD3NO2, (CD3)2CO, CDCl3, and CD2Cl2) and oxygen series (solvents, CD2Cl2 and CD3OD) were
achieved with triflic acid. The substituent and solvent effects on the equilibria are discussed in
terms of steric and electrostatic effects and orbital interactions associated with the endo-anomeric
effect. A generalized reverse anomeric effect does not exist in neutral or protonated N-aryl-5-
thioglucopyranosylamines and N-arylglucopyranosylamines. The anomeric effect ranges from 0.85
kcal mol-1 in 2 to 1.54 kcal mol-1 in 10. The compounds 1-4 and 9-11 show an enhanced endo-
anomeric effect upon protonation, ranging from 1.73 kcal mol-1 in 2 to 2.57 kcal mol-1 in 10. We
estimate the increase in the anomeric effect upon protonation of 10 to be approximately 1.0 kcal
mol-1. However, this effect is offset by steric effects due to the associated counterion which we
estimate to be approximately 1.2 kcal mol-1. The values of Keq(axial-equatorial) in protonated
1-4 increase in the order OMe < H < CF3 < NO2, in agreement with the dominance of steric
effects (due to the counterion) over the endo-anomeric effect. The values of Keq(axial-equatorial)
in protonated 9-11 show the trend OMe > H < NO2 that is explained by the balance of the endo-
anomeric effect and steric effects in the individual compounds. The trends in the values of the
C1-H1 coupling constants for 1-4 and the corresponding deacetylated compounds 5-8 as a function
of substituent and R- or â-configuration are discussed in terms of the Perlin effect and the interplay
of the endo- and exo-anomeric effects.
In tr od u ction
ments. The effect has been classified further in terms of
the endo- and exo-anomeric effect. The endo-anomeric
effect4 refers to the preference of electronegative groups
attached to the anomeric carbon for the axial orientation.
The preference is dictated partly by stabilizing nXfσ*C-Y
orbital interactions (Chart 1, a).5 The exo-anomeric effect6
is the preference for the gauche conformation around the
C1-aglyconic carbon bond of glycopyranosides that per-
mits expression of an nYfσ*C-X stabilizing orbital inter-
action (Chart 1, b and c).5
In recent years, another phenomenological conforma-
tional effect, the reverse anomeric effect (RAE), has also
gained notoriety and has provoked some debate wherein
even its very existence has been questioned. The latter
effect was defined by Lemieux and Morgan7 as the
tendency of an aglycon bearing a positive charge in a
The preference for the axial orientation of electro-
negative substituents at the anomeric carbon of a pyra-
nose ring is in marked contrast to expectations based
solely on the consideration of steric interactions. This
anomalous behavior was first noted by Edward1 in his
investigation of the relative stability of methyl R- and
â-glycopyranosides to acid hydrolysis and was clearly
defined as the anomeric effect by Lemieux and Chu¨2 as
a result of their investigations of the anomeric equilibria
of peracetylated pento- and hexopyranoses. Since that
time, the anomeric effect seems to have taken on a life
of its own and has engaged the imagination and efforts
of both experimental and theoretical chemists in their
attempts to define its nature and assign its origin.3 The
effect is stereoelectronic in nature and has since been
shown to be a general effect operating in X-A-Y seg-
(3) For leading references: (a) Thatcher, G. R. J ., Ed. The Anomeric
Effect and Associated Stereoelectronic Effects; ACS Symposium Series
539; American Chemical Society: Washington, DC, 1993. (b) J uaristi,
E.; Cuevas, G., Eds. The Anomeric Effect; CRC Press: Boca Raton,
FL, 1995.
† This work is dedicated, with respect, to the memory of J . T.
Edward.
* To whom correspondence may be addressed. Tel: 604-291-4327.
Fax: 604-291-3765. E-mail: bpinto@sfu.ca.
(1) Edward, J . T. Chem. Ind. (London) 1955, 1102.
(2) Lemieux, R. U.; Chu¨, N. J . 133rd Abstr. Papers Am. Chem. Soc.
Meeting 1958, 31N.
(4) Praly, J .-P.; Lemieux, R. U. Can. J . Chem. 1987, 65, 213.
(5) Pinto, B. M.; Leung, R. Y. N. The Anomeric Effect and Associated
Stereoelectronic Effects; Thatcher, G. R. J ., Ed.; ACS Symposium Series
539; American Chemical Society: Washington, DC, 1993.
(6) Lemieux, R. U.; Koto, S. Tetrahedron 1974, 30, 1933.
10.1021/jo991520j CCC: $19.00 © 2000 American Chemical Society
Published on Web 12/17/1999

Is There a Generalized Reverse Anomeric Effect?
J . Org. Chem., Vol. 65, No. 1, 2000 221
Ch a r t 1
Ch a r t 2
Sch em e 1
Sch em e 2
method, the Queen’s group12 claimed the existence of such
an effect in protonated N-(xylopyranosyl)imidazoles
(Scheme 2) by examination of conformational equilibria
using an approximate average coupling-constant method,
which they attributed to intramolecular electrostatic
attraction in the equatorial isomers. Such subtle differ-
ences indicate that the magnitude of the measured effect
is small. The matter has been resolved very recently by
use of an NMR titration method to estimate the RAE in
protonated N-(xylopyranosyl)imidazoles.11g The authors
conclude that no RAE exists in these compounds and that
the previous conclusions8,9,12b resulted because of the
approximations inherent in the average coupling-con-
stant method.11g
We chose to address the question of the existence of a
generalized RAE by systematic examination of substitu-
ent and solvent effects on the configurational equilibria
of N-aryl-5-thioglucopyranosylamines 1-4 and N-aryl-
glucopyranosylamines 9-11 and the corresponding pro-
tonated species (Chart 2). The 5-thio compounds 1-4 are
more stable than their oxygen congeners 9-11 and are
readily amenable to analysis. Both series also display
equilibria that are not highly biased and differ from
systems studied to date; they constitute, therefore, a
valuable test system. The question of a generalized RAE
is of interest for an understanding of the enzyme inhibi-
tory activity of glucopyranosylamines and their 5-thio
analogues.13
sugar ring to adopt the equatorial orientation; the
systems under study contained quaternary nitrogen
aromatic substituents such as pyridinium and imidazo-
lium (e.g. Scheme 1). The work was later extended by
Paulsen et al.8 to the study of the conformational
equilibria in peracetylated pentopyranosyl imidazoles
and the corresponding protonated species (e.g. Scheme
2) and by Finch and Nagpurkar9 to the neutral and
positively charged N-(hexopyranosyl)imidazoles and their
tetraacetates. It has also been suggested that neutral and
protonated amino and alkylamino substituents would
show a RAE,10 but this has been questioned.5,11 The
greater equatorial preferences have now been attributed
to accentuated steric effects, although the conclusions are
based on data from highly biased equilibria.5,11a A recent
study with a more sterically balanced, 2,2′-substituted-
1,3-dioxane system confirmed these conclusions.11f The
general picture that is emerging from recent studies is
that protonation of either an alkylamino substituent or
an imidazole substituent results in a stronger anomeric
effect11a-g and that the equatorial preference has its
origin in favorable electrostatic interactions.12 Neverthe-
less, the conclusions were not universal and appeared to
be system-dependent. Thus, whereas Perrin and co-
workers11a,b claimed an absence of a RAE with protonated
N-(glucopyranosyl)imidazoles (Scheme 1) by examination
of configurational equilibria using an NMR titration
Resu lts a n d Discu ssion
(7) (a) Lemieux, R. U.; Morgan, A. R. Can. J . Chem. 1965, 43, 2205.
(b) Lemieux, R. U. Pure Appl. Chem. 1971, 25, 527.
(8) (a) Paulsen H.; Gyorgydeak, Z.; Friedmann, M. Chem. Ber. 1974,
107, 1590. (b) Paulsen, H. Die Starke 1975, 27, 397.
Con figu r a tion a l An a lysis. The equilibrium popula-
tions of the 5-thio compounds 1-4 and their protonated
derivatives were assessed by 1H NMR spectroscopy at 294
K. Equilibration of the neutral species (12, Scheme 3)
was achieved by the HgCl2 catalysis of the individual
isomers in the polar solvents CD3OD, CD3NO2, and
(CD3)2CO only, owing to the limited solubility of HgCl2
in nonpolar solvents. The equilibria were approached
from both directions, i.e., starting from pure R- or
â-anomers, to ensure that a true equilibrium had been
(9) Finch, P.; Nagpurkar, A. G. Carbohydr. Res. 1976, 49, 275.
(10) (a) Wolfe, S.; Whangbo, M.-H.; Mitchell, D. J . Carbohydr. Res.
1979, 69, 1. (b) Krol, M. C.; Huige, C. H. J .; Altona, C. J . Comput.
Chem. 1990, 11, 765.
(11) (a) Perrin, C. L.; Armstrong, K. B. J . Am. Chem. Soc. 1993,
115, 6825. (b) Fabian, M. A.; Perrin, C. L.; Sinnott, M. L. J . Am. Chem.
Soc. 1994, 116, 8398. (c) Perrin. C. L.; Fabian, M. A.; Armstrong, K.
B. J . Org. Chem. 1994, 59, 5246. (d) Perrin, C. L. Pure Appl. Chem.
1995, 67, 719. (e) Perrin, C. L. Tetrahedron 1995, 51, 11901. (f) J ones,
P. G.; Kirby, A. J .; Komarov, I. V.; Wothers, P. D. J . Chem. Soc., Chem.
Commun. 1998, 1695. (g) Perrin, C. L.; Fabian, M. A.; Brunckova, J .;
Ohta, B. K. J . Am. Chem. Soc. 1999, 121, 6911.
(13) (a) Andrews, J . S.; Weimar, T.; Frandsen, T. P.; Svensson, B.;
Pinto, B. M. J . Am. Chem. Soc. 1995, 117, 10799. (b) Randell, K. D.;
Frandsen, T. P.; Stoffer, B.; J ohnson, M. A.; Svensson, B.; Pinto, B.
M. Carbohydr. Res. 1999, 321, 143.
(12) (a) Chan, S. S. C.; Szarek, W. A.; Thatcher, G. R. J . J . Chem.
Soc., Perkin Trans. 2 1995, 45. (b) Vaino, A. R.; Chan, S. S. C.; Szarek,
W. A.; Thatcher, G. R. J . J . Org. Chem. 1996, 61, 4514.

222 J . Org. Chem., Vol. 65, No. 1, 2000
Randell et al.
Sch em e 3
When Y ) CF3 (3) or NO2 (4), there is no significant
change in Keq upon protonation. The positive charge on
nitrogen must enhance the endo-anomeric effect as above.
However, since the nitrogen atom bears a greater positive
charge because of the inductive and field effects of these
substituents, we suggest that the counterion is bound
more tightly and that there is an increased steric effect
that offsets the increased anomeric effect (see Scheme
7). Corroboration of this hypothesis derives from the
effects of substituents and solvent on the equilibria of
the protonated species. Thus, when Y ) OMe (1) or H
(2), the axial isomer (13R, Scheme 4) dominates for all
solvents and the endo-anomeric effect is pronounced. One
predicts that the p-OMe substituent will cause the
nitrogen to be less positive, and therefore, the endo-
anomeric effect will not be as strong as in the parent
aniline derivative 2.16 One also predicts that the coun-
terion will not be as tightly bound and that steric effects
will be less pronounced. The observation of a greater
proportion of the axial isomer for 1 (Y ) OMe) than 2 (Y
) H) suggests that the steric effects dominate. When Y
) CF3 (3) or NO2 (4), the â-isomer (13â, Scheme 4) is
favored more than in the compounds in which Y ) OMe
(1) or H (2). The observation supports our argument that
the electron-withdrawing substituents cause the nitrogen
atom to have a greater positive charge, a tightly bound
counterion, and, therefore, a greater steric effect (see
Scheme 7); these increased steric effects favor the equa-
torial isomer (13â, Scheme 4). In nonpolar solvents, the
steric effect of the associated counterion should be even
more significant because the ion-pair is not solvated
effectively. In accord with this hypothesis, when Y ) CF3
(3) or NO2 (4), there is a greater proportion of the
â-isomer in nonpolar solvents than in polar solvents; this
effect is more pronounced than for compounds in which
Y ) OMe (1) or H (2). In 3 and 4, the counterion is bound
more tightly and these equilibria are more sensitive to
the effects of solvent. The values of Keq increase in the
order OMe < H < CF3 < NO2, in agreement with the
dominance of steric effects (due to the counterion) over
the endo-anomeric effect.
Equ ilibr a tion of th e Oxygen Con gen er s 9-11. The
percentages of the â-isomers in compounds 9-11 are
greater than in the corresponding sulfur analogues, in
accord with the greater steric effects in the former series.
As was the case with the 5-thio compounds 1-4, there
is no marked substituent effect on the equilibria of the
neutral species. However, in this series we were able to
examine the equilibria in both a nonpolar and polar
solvent and there is a notable solvent effect: the propor-
tion of the R-anomer is greater in CD2Cl2 compared to
CD3OD. We attribute this effect to the lesser ability of
the nonpolar solvent CD2Cl2 to solvate the dipole-dipole
interactions in the â-anomers that are normally associ-
ated with the endo-anomeric effect (see Scheme 8). There
was also a greater proportion of the â-isomer for 9-11
in CD3NO2 than in CD2Cl2 (data not shown). The greater
effect of dipolar interactions in the oxygen series as
compared to the sulfur series is expected on the basis of
the greater electronegativity of oxygen than sulfur.
The effects of protonation and the effects of solvent on
the equilibria of the protonated species reveal some
interesting trends. Unfortunately, the comparison is only
Sch em e 4
Sch em e 5
reached. The corresponding equilibrations of the proto-
nated species (13, Scheme 4) were studied in the presence
of 1.5 equiv of triflic acid, in polar and nonpolar solvents.
The addition of 1.5 equiv of triflic acid would ensure
complete protonation of the amines since the pKa of triflic
acid is -5.914 while the pKas of the isolated aglycons are
5.31 for p-anisidine, 4.60 for aniline, 2.45 for p-tri-
fluoromethylaniline, and 1.00 for the weakest base,
p-nitroaniline.15 The equilibration of the oxygen conge-
ners 9-11 was performed at 230 K only in CD2Cl2 and
CD3OD. In this series of compounds, we were restricted
in our choice of solvents because of the instability of the
compounds or line broadening effects in the spectra which
did not permit unambiguous assignment of signals or
their accurate integration. The equilibrations of the
neutral species (14, Scheme 5) were achieved by addition
of catalytic amounts of triflic acid to solutions of the
individual anomers, and those of the protonated species
(13, Scheme 4) were achieved by addition of excess triflic
acid, as described above. The equilibrium constants and
free energy values for 1-4 are listed in Table 1 and those
for 9-11 in Table 2.
Equ ilibr a tion of th e 5-Th io Com p ou n d s 1-4.
Examination of the equilibria of the neutral species 1-4
in polar solvents indicates that there is no marked
substituent or solvent effect. Interestingly, the effects of
protonation on the equilibria are a function of the
substituent. When Y ) OMe (1) or H (2), the proportion
of the axial isomer (13R, Scheme 4) increases upon
protonation. This can be attributed to an increase in the
endo-anomeric effect because the strength of the n-σ*
interactions increases due to a smaller energy gap
between interacting fragment orbitals (see Scheme 6).10
(14) Guthrie, J . P. Can. J . Chem. 1978, 56, 2342.
(15) Perrin, D. D. Dissociation Constants of Organic Bases in
Aqueous Solution; Butterworth: London, 1965.
(16) Pinto, B. M.; J ohnston, B. D.; Sandoval-Ramirez, J .; Sharma,
R. D. J . Org. Chem. 1988, 53, 3766.

Is There a Generalized Reverse Anomeric Effect?
J . Org. Chem., Vol. 65, No. 1, 2000 223
Ta ble 1. Effects of P r oton a tion , Solven t, a n d Su bstitu en t on Keq a n d ∆G for th e Neu tr a l Com p ou n d s in Wh ich X ) S
neutral protonated
∆Ga (error)
solvent
compd
Y
K
eq (error)
∆Ga (error)
Keq (error)
CD3NO2
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
OMe
H
CF3
NO2
OMe
H
CF3
NO2
OMe
H
CF3
NO2
OMe
H
CF3
NO2
OMe
H
CF3
NO2
1.31 (0.15)
1.26 (0.08)
1.15 (0.13)
1.51 (0.10)
1.37 (0.10)
1.22 (0.08)
1.24 (0.19)
1.34 (0.16)
1.15 (0.08)
1.15 (0.06)
0.98 (0.07)
0.98 (0.08)
d
d
d
d
d
d
d
d
-0.16 (0.07)
0.32 (0.07)
0.45 (0.15)
1.05 (0.08)
1.27 (0.07)
b
0.67 (0.10)
0.47 (0.20)
-0.03 (0.04)
-0.14 (0.03)
b
-0.14 (0.04)
-0.08 (0.07)
-0.24 (0.04)
CD3OD
(CD3)2CO
CDCl3
-0.18 (0.04)
-0.12 (0.03)
b
b
-0.13 (0.09)
0.97 (0.09)c
b
0.02 (0.05)c
b
-0.17 (0.01)
-0.08 (0.04)
0.38 (0.11)
0.81 (0.13)
1.07 (0.08)
b
0.48 (0.13)
0.81 (0.08)
1.69 (0.09)
1.93 (0.14)
0.32 (0.03)
0.65 (0.03)
1.40 (0.12)
1.71 (0.07)
0.56 (0.17)
0.12 (0.09)
-0.04 (0.04)
b
0.43 (0.10)
0.13 (0.06)
-0.31 (0.02)
-0.39 (0.04)
0.68 (0.05)
0.26 (0.03)
-0.20 (0.05)
-0.31 (0.02)
-0.08 (0.03)
0.01 (0.04)
0.01 (0.05)
d
d
d
d
d
d
d
d
CD2Cl2
a
b
In kcal mol-1 at 294 K. Decomposition of samples did not permit accurate determination of Keq
.
c Only equilibrated from Rfâ (3
d
days). Equilibration of the neutral species in these solvents was not possible owing to the insolubility of HgCl2.
Ta ble 2. Effects of P r oton a tion , Solven t, a n d Su bstitu en t on Keq a n d ∆G for th e Neu tr a l Com p ou n d s in Wh ich X ) O
neutral
protonated
∆Ga (error)
solvent
CD2Cl2
compd
Y
Keq (error)
∆Ga (error)
Keq (error)
9
10
11
9
10
11
OMe
H
NO2
OMe
H
2.65 (0.03)
2.74 (0.07)
3.18 (0.03)
4.04 (0.11)
3.95 (0.06)
b
-0.45 (0.01)
-0.46 (0.01)
-0.53 (0.01)
-0.64 (0.02)
-0.63 (0.01)
b
6.00 (0.09)
4.24 (0.07)
6.58 (0.03)
6.87 (0.25)
3.87 (0.06)
10.9 (0.20)
-0.82 (0.01)
-0.66 (0.01)
-0.82 (0.01)
-0.88 (0.03)
-0.62 (0.01)
-1.09 (0.03)
CD3OD
NO2
a
b
In kcal mol-1 at 230 K. No observable R-isomer.
Sch em e 6
monopole interaction between the resultant dipole of the
lone pairs on the ring oxygen and the positive monopole
on the protonated nitrogen atom in the â-isomer. When
Y ) OMe (9), there is an increase in the proportion of
the â-isomer upon protonation in both solvents. When Y
) NO2 (11), the R-isomer is barely detectable in the
neutral compound, but in CD3OD, the R-isomer is present
in the protonated form. The positive charge must enhance
the endo-anomeric effect, as no R-isomer is present in the
unprotonated species. The more pronounced R-preference
in CD3OD relative to CD2Cl2 likely results from better
solvation of the dipole-monopole interactions in the
â-isomer. Since the nitrogen atom is more positive in this
compound because of the substituent, we propose that
the counterion is more tightly bound and that there is
an increased steric effect that offsets the increased endo-
anomeric effect (see Scheme 7).
Sch em e 7
Sch em e 8
It is instructive to compare the effects of substituents
on the equilibria of the protonated species, as with the
sulfur series of compounds. When Y ) H (10), there is a
greater proportion of the R-isomer present as compared
to the compounds in which Y ) OMe (9) or NO2 (11). With
regard to the p-OMe compound 9, one predicts again that
the substituent will cause the nitrogen to be less positive,
and therefore, the anomeric effect will not be as strong
as in the parent aniline derivative 10.16 One also predicts
that the counterion will not be as tightly bound and that
steric effects will be less pronounced. The observation of
a greater proportion of the R-isomer for 10 (Y ) H) than
9 (Y ) OMe) suggests that the endo-anomeric effect
dominates, in contrast to the situation with the sulfur
possible for the equilibrations in CD3OD and CD2Cl2 since
attempts to extend these studies to other solvents, e.g.
CD3NO2, were frustrated by problems of line broadening
that made assignment of the signals ambiguous and
integrations inprecise. When Y ) H (10), there is no effect
upon protonation in CD3OD, but in CD2Cl2 more of the
â-isomer is observed. We attribute this to the solvation
or damping by the polar solvent of the favorable dipole-

224 J . Org. Chem., Vol. 65, No. 1, 2000
Randell et al.
Ta ble 4. Ster ic Com p on en t of ∆G in 17-22a
Sch em e 9
compd
-∆G
17
18
19
20
21
22
1.26 (0.10)
2.00 (0.14)
0.99 (0.05)
2.03 (0.02)
3.23 (0.23)
1.60 (0.08)
a
In kcal mol-1
.
Ta ble 5. An om er ic Effect in Com p ou n d s 2 a n d 10a
compd
solvent
neutral
protonated
2
2
2
CD3NO2
CD3OD
(CD3)2CO
CDCl3
0.85 (0.06)
0.87 (0.06)
0.91 (0.06)
c
b
b
b
2
1.73 (0.08)
1.86 (0.08)
b
2
10
10
CD2Cl2
CD3OD
CD2Cl2
c
1.37 (0.14)
1.54 (0.14)
2.57 (0.23)
a
b
In kcal mol-1
.
Not appropriate (see text). c Experiments not
performed because of limited solubility of HgCl2.
K)17 was then used to give an estimate of the conforma-
tional free energy for N-phenylcyclohexylamine 17 and
its protonated counterpart 20. These data (Table 4) give
an estimate of approximately 0.8 kcal mol-1 for the steric
effect upon protonation that can be used as a rough
measure of the steric effect of the associated counterion.
The steric components of the conformational free energies
for N-phenyltetrahydropyranylamine 18 and its proto-
nated counterpart 21, and N-phenyltetrahydrothiapy-
ranylamine 19 and its protonated counterpart 22 were
then calculated (see Table 4) by taking into account the
ratios of the respective conformational free energies for
the axial-equatorial equilibria in 2-Me-tetrahydropyran
(2.86 ( 0.20 kcal mol-1),18 2-Me-tetrahydrothiapyran
(1.42 ( 0.07 kcal mol-1),19 and methylcyclohexane (1.80
( 0.02 kcal mol-1).17 Finally, estimates of the anomeric
effects operating in the neutral and protonated com-
pounds 2 and 10 were obtained by adding the steric
components to the conformational free energies (Tables
1 and 2) for neutral and protonated 2 and 10 (see Table
5). We realize that different temperatures were used to
obtain the different ∆G values described above, but feel
that the treatment will provide a rough estimate of the
magnitude of the anomeric effect. For the neutral com-
pounds, the anomeric effect ranges from 0.85 kcal mol-1
in 2 to 1.54 kcal mol-1 in 10. We have shown that in the
equilibria of the protonated species, the associated coun-
terion is critical in determining the steric effect and that
the solvent and substituent play a crucial role in the
solvation of ion pairs and hence the steric effect. There-
fore, since we have only assessed the steric component
for the protonated, unsubstituted compound 16 in a
nonpolar solvent, we have only estimated the magnitudes
of the anomeric effect for the protonated, unsubstituted
compounds 2 and 10 in the nonpolar solvents CDCl3 and
CD2Cl2. The anomeric effect in the protonated derivatives
ranges from 1.73 kcal mol-1 in 2 to 2.57 kcal mol-1 in
10. In the case of 10, the anomeric effect increases on
protonation by approximately 1.0 kcal mol-1. This is
offset by a steric effect of the associated counterion of
Ta ble 3. Equ ilibr iu m Da ta for 15 a n d 16a
Keq (error)
∆G (error)b
15
16
0.18 (0.01)c
2.03 (0.01)d
0.54 (0.01)
-0.23 (0.01)
a
b
In CD2Cl2:CFCl3 (85:15). In kcal mol-1 at 160 K. c Derived
from integration of signals for H1axial (δ 3.20), NH (δ 3.60) in the
minor conformer and H1equatorial (δ 3.70), NH (δ 3.94) in the major
d
conformer. Derived from integration of signals for H1axial (δ 3.17)
in the major conformer and H1equatorial (δ 3.68) in the minor
conformer.
series. With regard to the p-NO2 compound 11, the
â-isomer (13â, Scheme 4) is favored more than in the
compounds in which Y ) OMe (9) or H (10). As with the
sulfur series, the observation supports our argument that
the electron-withdrawing substituents cause the nitrogen
atom to have a greater positive charge, a tightly bound
counterion, and therefore, a greater steric effect (see
Scheme 7); these increased steric effects favor the equa-
torial isomer (13â, Scheme 4).
The results presented in the foregoing sections suggest
that there is no generalized reverse anomeric effect
operating in X-C-N or X-C-N+ fragments. In fact,
protonation leads to a greater endo-anomeric effect than
in the corresponding neutral X-C-N fragments. The
effects of substitution on the equilibria of the protonated
species can be interpreted in terms of the dominance of
steric effects (due to the counterion) over the endo-
anomeric effect. For compounds 1-4, in nonpolar sol-
vents, ion pair separation is not as effective and the
accentuated steric effects are more pronounced when X
) CF3 or NO2 than when X ) H or OMe.
Th e An om er ic Effect. To estimate the magnitudes
of the anomeric effects in 1-4 and 9-11, cis-4-methyl-
1-N-phenylcyclohexylamine 15 was used as a model
compound in order to approximate the steric component.
The methyl group was used as a counterpoise group to
give a more balanced conformational equilibrium, and the
conformational equilibria of 15 and the corresponding
protonated species 16 (Scheme 9) were analyzed at 160
K in a mixture of CD2Cl2:CFCl3 (85:15) (see Table 3). The
A-value of the methyl group (1.80 ( 0.02 kcal mol-1, 157
(17) Wiberg, K. B.; Hammer, J . D.; Castejon, H.; Bailey, W. F.;
DeLeon, E. L.; J arret, R. M. J . Org. Chem. 1999, 64, 2085.
(18) Eliel, E. L.; Hargrave, K. D.; Pietrusiewicz, K. M.; Manoharan,
M. J . Am. Chem. Soc. 1982, 104, 3635.
(19) Willer, R. L.; Eliel, E. L. J . Am. Chem. Soc. 1977, 99, 1925.

Is There a Generalized Reverse Anomeric Effect?
J . Org. Chem., Vol. 65, No. 1, 2000 225
Ta ble 6. Cou p lin g Con sta n ts J C1-H1 (Hz) for Com p ou n d s
1-8
R-isomers may be attributed to the counterbalance of the
endo- and exo-anomeric effects. For example, the greater
endo-anomeric effect in 8 relative to that in 6 is offset by
the decreased exo-anomeric effect.14 The two effects thus
partially cancel in 8, giving a C-H bond strength that is
similar to that in 6.
J C1-H1
compd (Y; R)
R
â
1 (OMe; Ac)
2 (H; Ac)
3 (CF3; Ac)
4 (NO2; Ac)
5 (OMe; H)
6 (H; H)
153.9
154.5
154.4
154.9
152.8
152.9
151.0
152.0
156.2
151.7
155.9
154.5
156.4
149.5
151.2
152.1
Con clu sion s
There is no evidence to support the existence of a
generalized reverse anomeric effect in neutral or proto-
nated N-aryl-5-thioglucopyranosylamines and N-arylglu-
copyranosylamines. For the neutral compounds, the
anomeric effect ranges from 0.85 kcal mol-1 in 2 to 1.54
kcal mol-1 in 10. The compounds 1-4 and 9-11 show
an enhanced endo-anomeric effect upon protonation. The
anomeric effect in the protonated derivatives ranges from
1.73 kcal mol-1 in 2 to 2.57 kcal mol-1 in 10. We estimate
the increase in the anomeric effect upon protonation of
10 to be approximately 1.0 kcal mol-1. However, this
effect is offset by steric effects due to the associated
counterion which we estimate to be approximately 1.2
kcal mol-1. The values of Keq in protonated 1-4 increase
in the order OMe < H < CF3 < NO2, in agreement with
the dominance of steric effects (due to the counterion)
over the endo-anomeric effect. The values of Keq in
protonated 9-11 show the trend OMe > H < NO2 that
is explained by the balance of the endo-anomeric effect
and steric effects in the individual compounds.
7 (CF3; H)
8 (NO2; H)
about 1.2 kcal mol-1. Unfortunately, in the case of 2, the
equilibrium data for the neutral species in CD2Cl2 are
not available, and an estimate of the magnitude of the
differential anomeric effect on protonation is not possible.
Nevertheless, the results indicate that the reverse ano-
meric effect does not exist in both neutral and protonated
N-phenylglucopyranosylamine 10 and in protonated
N-phenyl-5-thioglucopyranosylamine 2; rather, the ano-
meric effect exists.
P er lin Effect. As a final point of interest, we comment
here on the J C1-H1 coupling constants in compounds 1-8
and their relationship to the Perlin effect, which cor-
relates larger J C1-H1 values with greater C-H bond
strengths.20 The data are summarized in Table 6. The
trends are the same for the acetylated and free sugars.
The p-CF3 and p-NO2 derivatives (3, 4, 7, 8) have similar
coupling constants for both the R- and the â-isomers. The
p-OMe derivatives (1, 5) have larger coupling constants
for the â-isomer, whereas the aniline derivatives (2, 6)
have larger coupling constants for the R-isomer. Com-
pounds 2 and 6 follow the normal trend in that an
nSfσ*C1-H1 orbital interaction in the â-isomers results
in weaker C-H bonds and smaller coupling constants.20
We propose that in the case of the p-OMe derivatives
Exp er im en ta l Section
The synthesis of compounds 1-8 are described elsewhere.13b
Compounds 9-11 were synthesized by the method of Honey-
man.22 The R- and â-isomers of 10 and 11 have been charac-
terized previously.23 The synthesis of R- and â-9 is described
below. Compound 15 was synthesized as described.24
1H NMR spectra were recorded on a Bruker AMX-400 NMR
spectrometer at 400.13 MHz. Chemical shifts are given in ppm
downfield from TMS. Chemical shifts and coupling constants
were obtained from a first-order analysis of the spectra.
Chemical shifts of the neutral and protonated species 1-4 and
9-11 are listed in Tables 7-10 of the Supporting Information.
N-(4-Meth oxyp h en yl)-2,3,4,6-tetr a -O-a cetyl-r- a n d â-D-
glu cop yr a n osyla m in e 9r a n d 9â. A mixture of D-glucose
(9.01 g, 50.0 mmol), p-anisidine (6.16 g, 50.0 mmol) in
methanol (100 mL), and glacial acetic acid (0.2 g) was refluxed
for 45 min. The reaction mixture was concentrated to give a
â-1 and â-5, a greater exo-anomeric effect (nNfσ*C-S
)
relative to that in the unsubstituted compounds 2 and
6, respectively,16 leads to greater s-character in the C-H
bond and to a reversal in C1-H1 bond strengths. This
effect outweighs the opposing bond lengthening effect
caused by the Perlin effect.
Within the series of the acetylated R-isomers the J C1-H1
values are very similar, in contrast to those for the
â-isomers which increase in 1, 3, and 4 relative to the
value for the unsubstituted compound 2, the greatest
difference being observed for the p-OMe derivative 1.
Similar trends are observed for the R- and â-isomers in
the deprotected compounds 5-8. The larger coupling
constants in the p-OMe derivatives â-1 or â-5 can be
readily rationalized in terms of a greater exo-anomeric
effect, which leads to greater C-H bond strengths (see
above). The origins of the increases in J C1-H1 values for
the p-CF3 and p-NO2 â-compounds are not as readily
obvious. In these cases, the exo-anomeric effect is re-
duced.16 We propose that these substituents cause field
effects and inductive effects that cause a contraction of
the C-H bond. The effect can be viewed as an electro-
static effect, as proposed by Pross and Radom21 for the
case of acetals. The invariance of the J C1-H1 values in the
1
red oil which was shown by H NMR spectroscopy to contain
a mixture of R- and â-isomers. The oil was dissolved in pyridine
(80 mL) and acetic anhydride (60 mL) containing a catalytic
amount of DMAP and the mixture was stirred for 1 h at room
temperature and then for 45 min at 45 °C. The reaction
mixture was poured into ice/water (800 mL) and stirred to give
a reddish-brown precipitate that was recrystallized from ether
to give a 2:1 R:â mixture of 9 (13.5 g, 64%). Selective
recrystallization from ethanol gave pure 9R (mp 131-132 °C)
and pure 9â (mp 140-141 °C). Anal. Calcd for C21H27NO10
:
C, 55.63; H, 6.00; N, 3.09. Found (R:â mixture): C, 55.66; H,
6.02; N, 3.08.
The equilibrations of the neutral compounds 1-4 were
carried out at 294 K in CD3NO2, CD3OD, and (CD3)2CO. To a
solution of either the R- or â-isomer (1.5 mg) in 0.6 mL of the
appropriate deuterated solvent was added 10% HgCl2 from a
100 mM stock solution in deuterated solvent. The equilibra-
(20) (a) Wolfe, S.; Pinto, B. M.; Varma, V.; Leung, R. Y. N. Can. J .
Chem. 1990, 68, 1051. (b) J uaristi, E.; Cuevas, G. Tetrahedron Lett.
1992, 33, 1847. (c) J uaristi, E.; Cuevas, G.; Flores-Vela, A. Tetrahedron
Lett. 1992, 33, 6927. (d) Cuevas, G.; J uaristi, E.; Vela, A. J . Phys. Chem.
1999, 103, 932.
(22) Honeyman, J . Methods Carbohydr. Chem. 1963, 2, 95.
(23) (a) Capon, B.; Connett, B. E. J . Chem. Soc. 1965, 4492. (b)
Capon, B.; Connett, B. E. J . Chem. Soc. 1965, 4497.
(24) Hutchins, R. O.; Su, W.-Y.; Sivakumar, R.; Cistone, F.; Stercho,
Y. P. J . Org. Chem. 1983, 48, 3412.
(21) Pross, A.; Radom, L. J . Comput. Chem. 1980, 1, 295.

226 J . Org. Chem., Vol. 65, No. 1, 2000
Randell et al.
tions of the protonated compounds 1-4 were carried out at
294 K in CDCl3, CD2Cl2, CD3NO2, CD3OD, and (CD3)2CO. To
a solution of either the R- or â-isomer (1.5 mg) in 0.6 mL of
the appropriate deuterated solvent was added 1.5 equiv of
triflic acid from a 0.565 M stock solution in deuterated solvent.
integrated to ensure accuracy. The average values of several
integrations were taken into account for the final calculation
of K. The errors in K are the standard deviations of the
measurements. The errors in ∆G were derived from errors in
K and in temperature ((2 K). Calculations of the magnitude
of the anomeric effect were performed using the formula: AE
) ∆G2 - R(∆G1), where R ) ∆G(2-methyltetrahydropyran or
2-methyltetrahydrothiapyran)/∆G(methylcyclohexane); ∆G1 )
N-(phenyl)cyclohexylamine; ∆G2 ) N-arylglycopyranosylamine
(X ) S, O). The errors in the values of ∆G in Tables 4 and 5
were derived by standard treatment for the propagation of
errors.
1
The H NMR spectra were recorded periodically until equilib-
rium had been reached. Equilibration usually took place in
less than a few hours. Because both isomers were used in the
equilibrations, it was deemed that equilbrium had been
reached when the spectra starting from both the R- and the
â-isomers were identical.
The equilibrations of the neutral and protonated compounds
9-11 were carried out at 230 K in CD2Cl2 and CD3OD.
Attempts to extend these studies to other solvents, e.g. CD3-
NO2, were frustrated by line broadening effects that did not
permit unambiguous assignment of the signals or their ac-
curate integration. To a solution of either the R- or â-isomer
(10 mg) in 0.6 mL of the appropriate deuterated solvent at
230 K was added 5% triflic acid for the equilibrations of the
neutral species or 2.5 equiv of triflic acid for the equilibriations
of the protonated species. The equilibrations were immediate,
and the proton spectra did not change over time.
Ack n ow led gm en t. We are grateful to the Natural
Sciences and Engineering Research Council of Canada
for financial support and to J .-C. Brodovitch for helpful
discussions regarding the error analysis.
1
Su p p or tin g In for m a tion Ava ila ble: Tables of H NMR
chemical shifts for neutral and protonated 1-4 and 9-11 in
different solvents. This material is available free of charge via
the Internet at http://pubs.acs.org.
The equilibrium constants were derived from the integration
of different pairs of peaks of both isomers. Isolated signals were
J O991520J