Sensitivity of 1JC1-H1 magnitudes to anomeric stereochemistry in 2,3-anhydro-O-furanosides

Article abstract of DOI:10.1021/jo001747a

The magnitude of the one-bond coupling constant between C₁ and H₁ in 2,3-anhydro-O-furanosides has been shown to be sensitive to the stereochemistry at the anomeric center. A panel of 24 compounds was studied and in cases where the anomeric hydrogen is trans to the epoxide moiety, ¹J_C1-H1 = 163 - 168 Hz; and when this hydrogen is cis to the oxirane ring, ¹J_C1-H1 = 171 - 174 Hz. In contrast, for 2,3-anyhdro-S-glycosides, the size of the ¹J_C1-H1 is not sensitive to C₁ stereochemistry. Computational studies on all four methyl 2,3-anhydro-O-furanosides (5-8) demonstrated that ¹J_C1-H1 was inversely proportional to the length of the C₁-H₁ bond. A previously reported equation, which relates C₁-H₁ bond distance and atomic charges to ¹J_C1-H1 magnitudes, could be used to accurately predict the J values in the α-lyxo (5) and β-ribo (8) isomers. In contrast, with the β-lyxo (6) and α-ribo isomers (7), this equation underestimated the size of these coupling constants by 10-20 Hz.

Full text of DOI:10.1021/jo001747a

J . Org. Chem. 2001, 66, 4549-4558
4549
1
Sen sitivity of J C1-H1 Ma gn itu d es to An om er ic Ster eoch em istr y in
2,3-An h yd r o-O-fu r a n osid es
Christopher S. Callam, Rajendrakumar Reddy Gadikota, and Todd L. Lowary*
Department of Chemistry, The Ohio State University, 100 West 18th Avenue, Columbus, Ohio 43210
lowary.2@osu.edu
Received December 15, 2000
The magnitude of the one-bond coupling constant between C1 and H1 in 2,3-anhydro-O-furanosides
has been shown to be sensitive to the stereochemistry at the anomeric center. A panel of 24
compounds was studied and in cases where the anomeric hydrogen is trans to the epoxide moiety,
1
1J C1-H1 ) 163-168 Hz; and when this hydrogen is cis to the oxirane ring, J C1-H1 ) 171-174 Hz.
In contrast, for 2,3-anhydro-S-glycosides, the size of the 1J C1-H1 is not sensitive to C1 stereochemistry.
Computational studies on all four methyl 2,3-anhydro-O-furanosides (5-8) demonstrated that
1J C1-H1 was inversely proportional to the length of the C1-H1 bond. A previously reported equation,
1
which relates C1-H1 bond distance and atomic charges to J C1-H1 magnitudes, could be used to
accurately predict the J values in the R-lyxo (5) and â-ribo (8) isomers. In contrast, with the â-lyxo
(6) and R-ribo isomers (7), this equation underestimated the size of these coupling constants by
10-20 Hz.
In tr od u ction
In 1969, Perlin and Casu reported1 that the one-bond
coupling constant between the anomeric carbon and its
attached hydrogen (1J C1-H1) in R-D-glucose differed from
that of â-D-glucose by 9 Hz. Since that time, the magni-
tude of this coupling constant in pyranosyl rings has
become a reliable indicator of the stereochemistry at the
1
anomeric center. In R-pyranosides, J C1-H1 is approxi-
mately 170 Hz, and in the corresponding â-anomer, this
coupling constant is smaller, about 160 Hz.2 Although
F igu r e 1. Selected hyperconjugative effects in cyclohexane
(A) and a â-pyranoside (B).
these trends are general for all pyranosides, this method
is particularly useful for mannopyranosyl derivatives,
3
1
because J H1-H2 cannot be used to unambiguously dis-
For example, in cyclohexane the magnitude of J C-H(ax)
1
is 122 Hz, while J C-H(eq) is 126 Hz.8 Based on these J ’s,
tinguish the two anomers. More recent studies have
1
demonstrated that J C1-H1 magnitudes can also be used
it would be predicted that the axial C-H bond would be
longer than its equatorial counterpart, and indeed this
is true. The calculated (B3LYP/6-31+G**)9 lengths of the
axial and equatorial C-H bonds in cyclohexane are
1.1001 and 1.0972 Å, respectively. The increased bond
lengths for the axial C-H bond can be attributed to
σ(C-Hax) f σ*(C-Hax) hyperconjugation (Figure 1A). A
similar effect is observed at the anomeric center in
pyranosides. When the C1-H1 bond is axial, it is placed
antiperiplanar to one of the ring oxygen lone pairs. This
maximizes nO f σ*C1-H1 hyperconjugation, which in turn
lengthens the C1-H1 bond in the â-pyranosides, relative
to differentiate 1,2-orthoesters of glucose and galactose
from the isomeric glycosides.3 In contrast, 1J C1-H1 values
in furanosides have been shown to be insensitive to
anomeric stereochemistry,4 due to conformational aver-
aging of the J ’s arising from all solution conformers of
these flexible ring systems.
1
The factors that determine J C-H magnitudes in six-
membered ring carbocyclic and heterocyclic compounds
has been extensively studied.2,5-9 When ring systems
1
containing only first-row atoms are considered, J C-H
magnitudes are inversely related to the length of the
1
C-H bond and generally 1J C-H(ax) is smaller than J C-H(eq)
.
1
to the R-pyranosides (Figure 1B). Accordingly, J C1-H1 is
smaller in the â-pyranosides. This effect is often referred
to as the Perlin Effect.5 For six-membered ring hetero-
cyclic compounds containing second row atoms, reverse
Perlin effects are often seen, that is, there is not always
(1) Perlin, A. S.; Casu, B. Tetrahedron Lett. 1969, 2921.
(2) Bock, K.; Pedersen, C. J . Chem. Soc., Perkin Trans. 2 1974, 293.
(3) Ha¨llgren, C. J . Carbohydr. Chem. 1992, 11, 527.
(4) Mizutani, K.; Kasai, R.; Nakamura, M.; Tanaka, O.; Matsuura,
H. Carbohydr. Res. 1989, 185, 27.
(5) Wolfe, S.; Pinto, B. M.; Varma, V.; Leung, R. Y. N. Can. J . Chem.
1990, 68, 1051.
(6) Cuevas, G.; J uaristi, E.; Vela, A. J . Phys. Chem. A 1999, 103,
932.
(7) Tvaroska, I.; Taravel, F. R. Adv. Carbohydr. Chem. Biochem.
1995, 51, 15 and references therein.
1
an inverse relationship between J C-H and C-H bond
length. This is particularly true for sulfur-containing
molecules, and the most heavily investigated systems
have been 1,3-dithianes and 1,3-oxathianes. In these
rings it appears that these coupling constants are influ-
enced by other factors in addition to the distance between
the coupled atoms.6,8,9
(8) J uaristi, E.; Cuevas, G.; Vela, A. J . Am. Chem. Soc. 1994, 116,
5796.
(9) Alabugin, I. V. J . Org. Chem. 2000, 65, 3910.
10.1021/jo001747a CCC: $20.00 © 2001 American Chemical Society
Published on Web 05/31/2001

4550 J . Org. Chem., Vol. 66, No. 13, 2001
Callam et al.
Ch a r t 1
Ch a r t 2
Recently, we have been exploring the ability of thiogly-
coside 1 and glycosyl sulfoxide 2 (Chart 1) to serve as
glycosylating agents.10 These compounds are efficient
glycosyl donors. However, we were faced with the prob-
lem of unequivocally determining the stereochemistry at
the anomeric center in the glycoside products (e.g., 3 and
4). Unfortunately, two generally reliable4 indicators of
anomeric stereochemistry in furanosides, the chemical
shift of the anomeric carbon and the magnitude of
3J H1-H2, were very similar for both 3 and 4. Furthermore,
another possible determinant of C1 stereochemistry, the
chemical shift of the anomeric proton, was virtually
identical in both 3 and 4.
In this paper we report that the one-bond coupling
constant between C1 and H1 in 2,3-O-anhydrofuranosides
is diagnostic of the stereochemistry at the anomeric
center. The studies described here extend work previ-
ously reported by Kim, Vyas, and Szarek11 on 13C NMR
spectra of carbohydrates possessing epoxide functionality.
1
In that study, J C-H in all four methyl 2,3-anhydrofura-
nosides (5-8, Chart 2) were measured. Surprisingly,
however, no mention was made of the sensitivity of
1J C1-H1 to the stereochemistry at the anomeric center. We
1
report here the measurement of J C1-H1 for a panel of
mono- and oligosaccharides (Chart 2) and show that the
trends observed in 5-8 are general for 2,3-anhydro-O-
furanosides. Furthermore, through ab initio12 and density
functional theory (DFT) calculations,13 we have deter-
mined bond lengths of 5-8 and have shown that there
is an inverse relationship between C1-H1 bond distance
1
and J C1-H1
.
Resu lts a n d Discu ssion
Syn th esis of Su bstr a tes. Methyl glycosides 5-8 have
been previously reported.14 The other anhydrosugars
shown in Chart 2 were synthesized as described below.
2,3-An h yd r o-r-D-lyxofu r a n osid es. Compound 3 was
synthesized (Scheme 1) from octyl glycoside 2715 in 82%
(10) Gadikota, R. R.; Callam, C. S.; Lowary, T. L. Org. Lett. 2001,
3, 607.
(11) Kim, K. S.; Vyas, D. M.; Szarek, W. A. Carbohydr. Res. 1979,
72, 25.
(12) Hehre, W. J .; Radom, L.; Schleyer, P. v. R.; Pople, J . A. Ab initio
Molecular Orbital Theory; Wiley-Interscience: New York, 1986, and
references therein.
(13) (a) Becke, A. D. Phys. Rev. A 1988, 38, 3098. (b) Becke, A. D. J .
Chem. Phys. 1993, 98, 5648. (c) Lee, C.; Yang, W.; Parr, R. G. Phys.
Rev. B 1988, 37, 785.
(14) (a) 5, Martin, M. G.; Ganem, B.; Rasmussen, J . R. Carbohydr.
Res. 1983, 123, 332. (b) 6, Thome´, M. A.; Giudicelli, M. B.; Picq, D.;
Anker, D. J . Carbohydr. Chem. 1991, 10, 923. (c) 7 and 8, Callam, C.
S.; Gadikota, R. R.; Lowary, T. L. Carbohydr. Res. 2001, 330, 267.
(15) Ramneantu, O.; Lowary, T. L., unpublished results.
yield via a Mitsunobu reaction with benzoic acid, triph-
enylphosphine, and diisopropylazodicarboxylate (DIAD).
The preparation of 17, 19, 21, and 25 is outlined in
Scheme 2. Thioglycoside 2916 was reacted with alcohols
5,14 28, 30,17 and 31,18 to give 32-35 in 81-91% yield.
Deprotection with sodium methoxide proceeded in 88-

2,3-Anhydro-O-furanosides
Sch em e 1
J . Org. Chem., Vol. 66, No. 13, 2001 4551
Sch em e 3
a
BzOH, Ph3P, DIAD, THF, 0 °C, 82%. bNaOCH3, CH3OH, rt,
91%.
Sch em e 2
a
t-BuPh2SiCl, pyridine, 0 °C f rt; then BzCl, pyridine, 0 °C f
rt, 88%. b5, NIS, AgOTf, CH2CH2, 0 °C, 84%. n-Bu4NF, THF, 0
c
°C, 81%. d29, NIS, AgOTf, CH2Cl2, 0 °C, 89%. NaOCH3, CH3OH,
e
f
rt, 90%. BzOH, Ph3P, DIAD, THF, 0 °C, 89%.
Sch em e 4
a
Cyclohexanol, p-TsOH, CH2Cl2, rt, 84%. bNaOCH3, CH3OH,
c
rt, 93%. BzOH, Ph3P, DIAD, THF, 0 °C, 46% (13), 28% (14).
alcohol 44 (81%) which was then glycosylated with 29.
The resulting product, 45, was debenzoylated to give 46
(80% yield from 44). Triepoxide 23 was obtained in 89%
yield from 46, upon reaction with triphenylphosphine,
DIAD, and benzoic acid.
a
5, 28, 30 or 31, NIS, AgOTf, CH2Cl2, 0 °C, 88% (32), 81% (33),
89% (34), 91% (35). bNaOCH3, CH3OH, rt, 97% (36), 90% (37), 89%
(38), 88% (39). cBzOH, Ph3P, DIAD, THF, 0 °C, 91% (17), 83%
(19), 83% (25). dNaOCH3, CH3OH, rt, 73%, two steps from 38.
Monosaccharide 13 was synthesized from the known20
dibenzoate 47 as illustrated in Scheme 4. Treatment of
47 with cyclohexanol and p-toluenesulfonic acid afforded
a mixture of cyclohexyl glycosides 48 in 84% yield.
Deprotection with sodium methoxide afforded 49 (93%
yield) which was subsequently converted to a mixture of
13 and 14 as described above for the other epoxides.
Purification by chromatography provided pure 13 (46%)
and 14 (28%).
2,3-An h yd r o-r-D-r ibofu r a n osid es a n d 2,3-An h y-
d r o-â-D-r ibofu r a n osid es. The synthesis of 15 and 16
(Scheme 5) began from the mixture of cyclohexyl glyco-
sides 48, which was mesylated and then reacted with
sodium methoxide to afford epoxides 50 (58%) and 51
(24%). Treatment of each with benzoyl chloride in pyri-
dine provided 15 and 16 in excellent yield.
97% yield, affording disaccharides 36-39. Each of these
was then converted, in 83-91% yield, to the correspond-
ing anhydrosugars 17, 19, 25, and 40 under the Mit-
sunobu protocol used for the synthesis of 3. Reaction of
40 with sodium methoxide afforded 21.
Trisaccharide 23 was prepared as detailed in Scheme
3. The known19 thioglycoside 41 was converted in two
steps and 88% yield to 42, which was then coupled to 5,
affording disaccharide 43 (84% yield). The silyl group was
removed with tetrabutylammonium fluoride to provide
(16) Prepared from methyl 2,3,5-tri-O-benzoyl-R-D-arabinofuranoside
by reaction with p-thiocresol and boron trifluoride etherate, see
Experimental Section.
(17) Sondheimer, S. J .; Eby, R.; Schuerch, C. Carbohydr. Res. 1978,
60, 187.
2,3-An h yd r o-â-D-lyxofu r a n osid es. Compounds 4, 18,
20, 22, 24, and 26 were obtained by glycosylation of the
(18) Dasgupta, F.; Garegg, P. J . Synthesis 1994, 1121.
(19) D′Souza, F. W.; Ayers, J . D.; McCarren, P. R. Lowary, T. L. J .
Am. Chem. Soc. 2000, 122, 1251.
(20) Reist, E. J .; Hart, P. A.; Goodman, L.; Baker, B. R J . Am. Chem.
Soc. 1959, 81, 5176.

4552 J . Org. Chem., Vol. 66, No. 13, 2001
Callam et al.
1
Sch em e 5
(3-8, 13-26), the magnitude of the J C1-H1 in the R-lyxo
and R-ribo isomers (where the anomeric hydrogen is trans
to the epoxide oxygen), is 163.1-168.0 Hz. In the R-lyxo
and â-ribo isomers (anomeric hydrogen cis to the epoxide
oxygen) the size of this coupling constant is larger,
1
171.5-174.3 Hz. Therefore, the magnitude of the J C1-H1
can be used to determine the stereochemistry at the
anomeric center in 2,3-anhydrofuranosyl O-glycosides.
The conformational restraints placed on the furanose ring
by the epoxide are presumably the reason that this
coupling constant is sensitive to anomeric stereochem-
istry, whereas in conformationally unrestricted furanose
rings, there is no such relationship. The parent furanose
rings adopt multiple conformations in solution and hence
the experimentally measured couplings represent an
average of all conformers. In contrast, the epoxide moiety
locks the furanose ring into a single conformation (see
below). Interestingly, in the S-glycoside series there is
no significant difference between the two anomers (com-
pare 9 and 10, and 11 and 12).
a
MsCl, Et3N, CH2Cl2, 0 °C. bNaOCH3, CH3OH, rt, 58% (50),
c
24% (51). BzCl, pyridine, 0 °C f rt, 92% (15), 89% (16).
Sch em e 6
Ca lcu la tion of C1-H1 Bon d Len gth s in 5-8 a n d
1
1
Cor r ela tion w ith J C1-H1. To probe if the J C1-H1
magnitude in these 2,3-anhydrosugars is inversely re-
lated to the C1-H1 bond length, we were interested in
determining bond distances for some of the compounds
for which we had experimental data. To date, no crystal
structure data for any of the compounds listed in Chart
2 have been reported, and furthermore, only a few are
crystalline solids. Accordingly, we chose to determine C1-
H1 bond lengths by computational methods using both
ab initio (HF/6-31G*)12 and DFT (B3LYP/6-31G*)13 cal-
culations. In these investigations we have focused our
attention on methyl glycosides 5-8, and for each monosac-
charide, all three possible C4-C5 rotamers (gg, gt, and
tg, Figure 2) were considered.
a
b
BzOH, Ph3P, DIAD, THF, 0 °C, 73% (1), 12% (53). NaOCH3,
CH3OH, rt, 96% (9), 98% (10).
Sch em e 7
All 12 conformers were optimized at both levels of
theory described above, and then single point energies
with the 6-31+G** basis set were calculated (Table 2).
The energies of the HF- and B3LYP-optimized conform-
ers are nearly identical. For 5, 6, and 8, the lowest energy
conformer was the one in which the C4-C5 bond adopted
the tg rotamer. For 7, the gg rotamer was the lowest
energy conformer; however, the tg rotamer was of similar
energy. When comparing an R/â pair (e.g., 5 and 6, or 7
and 8) the isomer with the methoxy group on the face of
the furan ring opposite the epoxide oxygen is 1.5-4.5
kcal/mol more stable than the isomer where these groups
are cis. This would be expected given that when the
methoxy and epoxide moieties are on the same side of
the ring there is increased steric crowding and also
repulsion between the lone pair electrons on the oxygens.
a
PySH, K2CO3, acetone:toluene 1:1, rt, 88%. bNaOCH3, CH3OH,
rt, 95%. BzOH, Ph3P, DIAD, THF, 0 °C. dNaOCH3, CH3OH, rt,
56% (11), 25% (12) (two steps).
c
appropriate alcohol with thioglycoside 1 as previously
reported.10,21
Th ioglycosid es. The synthesis of 9 and 10 was carried
out as illustrated in Scheme 6. An anomeric mixture of
thioglycosides 5219 was treated with benzoic acid under
Mitsunobu conditions to afford an 85% yield of anhydro-
sugars 1 and 53 which could be easily separated by
chromatography. Deprotection with sodium methoxide
provided 9 and 10 in 96% and 98% yield, respectively.
Thioglycosides 11 and 12 were prepared as outlined
in Scheme 7. Bromide 5422 was converted to a 3:1 R: â
mixture of 2-pyridyl thioglycosides 55, which was sub-
sequently deprotected to provide 56 in 84% yield over the
two steps. Reacting 56 with benzoic acid, triphenylphos-
phine, and DIAD, followed by deprotection provided 11
(56%) and 12 (25%).
The C1-H1 bond lengths in 5-8 determined from these
calculations are shown in Table 3. The trends observed
are consistent at both levels of theory; however, the
B3LYP/6-31G* bond lengths are slightly (∼1.4%) longer
than those calculated at the HF/6-31G* level of theory.
This is to be expected given that electron-electron
correlation is neglected at the HF level of theory.12 The
difference in bond lengths between the two levels of
theory is similar to that previously reported for other
furanose rings.24 It is also interesting to note that for each
1
Mea su r em en t of Cou p lin g Con sta n ts. J C1-H1 val-
1
ues in 3-26 were measured from the H coupled 1-di-
mensional 13C NMR spectrum23 of each compound. The
results are detailed in Table 1. In the O-glycoside series
(21) Compounds 4 and 14 were also synthesized by subjecting the
corresponding â-glycoside triol (e.g., octyl â-D-arabinofuranoside) to the
sequence of reactions shown in Scheme 1.
(22) Fletcher, H. G., J r. Methods Carbohydr. Chem. 1963, 2, 228.
(23) Gansow, O. A.; Schittenhelm, W. J . Am. Chem. Soc. 1971, 93,
4294.
(24) (a) Houseknecht, J . B.; McCarren, P. R.; Lowary, T. L.; Hadad,
C. M. J . Am. Chem. Soc., Submitted. (b) Cloran, F.; Carmichael, I.;
Serianni, A. S. J . Phys. Chem. A 1999, 103, 3783.

2,3-Anhydro-O-furanosides
J . Org. Chem., Vol. 66, No. 13, 2001 4553
Ta ble 1. J C1,H1 in 2,3-An h yd r ofu r a n osid es 3-26a
ring A ring B
1
ring C
H1-Oep
H1-Oep
H1-Oep
c
c
c
compound
orientationb
1J C1-H1
orientationb
1J C1-H1
orientationb
1J C1-H1
3
4
5
6
7
cis
trans
cis
trans
trans
cis
cis
trans
cis
trans
cis
trans
trans
cis
172.4
164.1
172.8
166.5
167.8
173.5
172.3
171.5
166.6
166.3
172.5
163.1
164.5
172.5
172.7
172.8
174.0
172.3
168.4d
166.2d
173.2
173.7
163.7e
163.5e
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
8
9
-
-
-
-
-
-
-
-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
cis
cis
cis
cis
cis
trans
cis
trans
cis
trans
cis
cis
cis
trans
173.8
165.9
172.5
166.3
174.2
166.0
174.1
174.3
173.6
168.0
-
-
-
-
-
-
-
-
-
-
-
-
-
-
cis
cis
trans
-
173.8
166.9
-
cis
-
-
-
-
a
b
See Charts 1 and 2 for structures. Relative orientation of anomeric hydrogen (H1) and the epoxide oxygen (Oep). c In Hz.
R-Mannopyranosyl residue. e R-Glucopyranosyl residue.
d
lyxo-series, the C1-H1 distance is greater in the â-ano-
mer.
Oth er Con for m a tion a l F ea tu r es of 5-8. The pre-
ferred conformation of the five-membered ring in 2,3-
anhydrofuranosides has been the topic of some discus-
sion,11,25 and our calculations shed light on this issue. The
epoxide ring in 5-8 necessitates that the furan ring can
only adopt either the EO or OE conformer. In all four
compounds, the preferred conformation is a “boat”, that
is, the oxygen of the furanose ring is on the same side of
the molecule as the epoxide (Figure 3). These structures
are consistent with a previous report26 in which the
conformation of 2′,3′-anhydronucleoside derivatives was
probed through X-ray crystallography, 1H NMR spec-
troscopy, and molecular mechanics calculations.
F igu r e 2. Definition of gauche-gauche (gg), gauche-trans
(gt), and trans-gauche (tg) rotamers about the C4-C5 bond.
Ta ble 2. Rela tive En er gies of 5-8a
B3LYP/6-31+G**
B3LYP/6-31+G**
//B3LYP/6-31G*
compound
//HF/6-31G*
5-gg
5-gt
5-tg
6-gg
6-gt
6-tg
7-gg
7-gt
7-tg
8-gg
8-gt
8-tg
4.5
1.7
0.0
9.0
5.7
3.9
4.5
6.9
4.7
2.9
2.5
0.6
4.5
1.7
0.0
9.0
5.7
4.0
4.5
6.9
4.7
3.0
2.5
0.6
In the case of the lyxo isomers, 5 and 6, the five-
membered ring adopts an OE conformation, which would
be expected to be favorable because the sterically de-
manding hydroxymethyl group at C4 is oriented in a
pseudoequatorial orientation. On the other hand, the
furanoid ring in the ribo isomers, 7 and 8, is in the EO
conformation. In this conformation, the C4 hydroxymethyl
group is placed in a pseudoaxial orientation, which, a
priori, would be expected to be unfavorable. However, the
furanose ring in 7 and 8 is less highly puckered than in
5 and 6. It appears, therefore, that in the methyl 2,3-
anhydro-O-ribofuranosides, the ring flattens appreciably
to minimize unfavorable transannular steric interactions
involving the hydroxymethyl group at C4. All four ring
systems are appreciably less puckered than conforma-
tionally free furanose rings.27 In addition to the expected
flattening induced by the epoxide moiety, electron repul-
sion between the lone pairs of the furan, glycosidic, and
a
Bottom of the well energies in kcal/mol at the specified level
of theory.
ring system, the C1-H1 bond length calculated for the
gt and tg rotamers are very similar, if not identical, while
that for the gg conformer is significantly different. As is
clear from the data in Table 3, for a given anomeric pair
of 2,3-anhydrosugar methyl glycosides (e.g., 5 and 6), the
C1-H1 bond is longer in the anomer in which H1 is trans
to the epoxide, which is also the isomer that has the
1
smaller J C1-H1 (Table 1). Thus, there is an inverse
1
relationship between C1-H1 bond length and J C1-H1 in
these compounds. In contrast to pyranose systems, the
anomer with the longer bond length is not the same in
both families of compounds. In the ribo-family, the
R-isomer has the longest C1-H1 bond, whereas in the
(25) (a) Hall, L. D. Chem. Ind. 1963, 950. (b) Hiraoka, T.; Iwashige,
T.; Iwai, I. Chem. Pharm. Bull. 1965, 13, 285.
(26) Koole, L. H.; Neidle, S.; Crawford, M. D.; Krayevski, A. A.;
Gurskaya, G. V.; Sandstro¨m, A.; Wu, J .-C.; Tong, W.; Chattopadhyaya.
J . Org. Chem. 1991, 56, 6884.

4554 J . Org. Chem., Vol. 66, No. 13, 2001
Ta ble 3. Ca lcu la ted C1-H1 Bon d Len gth s a n d J C1,H1 in Meth yl 2,3-An h yd r ofu r a n osid es 5-8
Callam et al.
1
1
d
compounda
H1-Oep orientationb
level of theory
C1-H1 distancec
calculated J C1,H1
experimentale 1J C1,H1
5-gg
5-gt
5-tg
6-gg
6-gt
6-tg
7-gg
7-gt
7-tg
8-gg
8-gt
8-tg
5-gg
5-gt
5-tg
6-gg
6-gt
6-tg
7-gg
7-gt
7-tg
8-gg
8-gt
8-tg
cis
cis
cis
trans
trans
trans
trans
trans
trans
cis
HF/6-31G*
HF/6-31G*
HF/6-31G*
HF/6-31G*
HF/6-31G*
HF/6-31G*
HF/6-31G*
HF/6-31G*
HF/6-31G*
HF/6-31G*
HF/6-31G*
HF/6-31G*
B3LYP/6-31G*
B3LYP/6-31G*
B3LYP/6-31G*
B3LYP/6-31G*
B3LYP/6-31G*
B3LYP/6-31G*
B3LYP/6-31G*
B3LYP/6-31G*
B3LYP/6-31G*
B3LYP/6-31G*
B3LYP/6-31G*
B3LYP/6-31G*
1.0836
1.0841
1.0841
1.0920
1.0910
1.0907
1.0874
1.0906
1.0901
1.0852
1.0839
1.0838
1.0989
1.0994
1.0994
1.1082
1.1070
1.1065
1.1038
1.1061
1.1055
1.1007
1.0991
1.0989
173.5
171.4
171.4
142.9
146.6
147.6
158.3
147.6
149.3
167.7
172.2
172.5
116.1
114.2
114.2
84.7
172.8
166.5
167.8
173.5
172.8
166.5
167.8
173.5
cis
cis
cis
cis
cis
trans
trans
trans
trans
trans
trans
cis
88.8
90.4
99.9
91.5
93.4
110.0
114.9
115.6
cis
cis
a
b
See Figure 2 for definitions of the gg, gt, and tg rotamers about the C4-C5 bond. Relative orientation of anomeric hydrogen (H1) and
the epoxide. c In angstroms. In hertz, coupling constants calculated using eq 1. e From Table 1.
d
In the original report,5 the bond lengths were determined
by HF/6-31G* optimization of model compounds, and the
atomic charges were obtained by Mulliken population
analysis.28 The use of model compounds was necessary
given that the computational resources available at the
time made such calculations on monosaccharides a
formidable challenge.
1J C1-H1 ) -3432 + 182.2qCqH + 3889/r
(1)
1
Using eq 1, we have calculated J C1-H1 for all three
possible C4-C5 rotamers of 5-8 (Table 3).29 The most
noticeable observation is that when the B3LYP/6-31G*
bond lengths are used, the coupling constants are dra-
matically underestimated. This is to be expected because
eq 1 was developed by correlating experimental 1J C,H data
with bond lengths calculated by the HF/6-31G* method.
Accordingly, when the “longer” B3LYP/6-31G* bond
distances are used, it would be expected that the J values
would be underestimated.
When the HF/6-31G* bond lengths are used, better
agreement between the calculated (Table 3) and experi-
mental (Table 1) coupling constant magnitudes is seen.
In particular, for those isomers in which the anomeric
hydrogen is cis to the epoxide (5 and 8) the agreement is
excellent. For example, in 5, the experimental value is
172.8 Hz, whereas the calculated values are 171.4 or
173.4 Hz, depending upon the C4-C5 rotamer. Although
the correlation between calculated and experimental J
values for 5 and 8 is excellent, for 6 and 7, there is a
10-20 Hz discrepancy. In an effort to improve the
accuracy of these predictions, we also carried out a series
of calculations in which atomic charges measured from
NBO analysis30 were used in place of those from the
Mulliken population analysis. This resulted in only
F igu r e 3. Representative examples of HF/6-31G* optimized
conformers of 5-8; note the relatively flat furan ring.
epoxide oxygens would also be expected to further force
the ring into a more planar conformation. The pseudoro-
tational phase angles and puckering amplitudes for all
of the conformers listed in Table 3 can be found in the
Supporting Information (Table S1).
P r ed iction of 1J C1-H1 Ma gn itu d es by a n Em p ir ica l
Rela tion sh ip . Armed with the bond length data pre-
1
sented in Table 3, we next explored how well the J C1-H1
magnitudes could be predicted using a previously re-
1
ported5 relationship (eq 1) that correlates J C1-H1 with
C1-H1 bond length (r, in Å) and the atomic charge on
the carbon and hydrogen atoms, qC and qH, respectively.
(28) (a) Mulliken, R. S. J . Chem. Phys. 1955, 23, 1833. (b) Mulliken,
R. S. J . Chem. Phys. 1955, 23, 1841. (c) Mulliken, R. S. J . Chem. Phys.
1955, 23, 2338. (d) Mulliken, R. S. J . Chem. Phys. 1955, 23, 2343.
(29) The required bond lengths were taken from the optimized
geometries of 5-8 at both levels of theory. Atomic charges were
determined by Mulliken population analysis.
(27) (a) Gordon, M. T.; Lowary, T. L.; Hadad, C. M. J . Org. Chem.
2000, 65, 4954. (b) Gordon, M. T.; Lowary, T. L.; Hadad, C. M. J . Am.
Chem. Soc. 1999, 121, 9682. (c) Church, T. J .; Carmichael, I.; Serianni,
A. S. J . Am. Chem. Soc. 1997, 119, 8946. (d) de Leeuw, H. P. M.;
Haasnoot, C. A. G.; Altona, C. Isr. J . Chem. 1980, 20, 108.

2,3-Anhydro-O-furanosides
J . Org. Chem., Vol. 66, No. 13, 2001 4555
1
Ta ble 4. Ca lcu la ted C1-H1 Bon d Len gth s a n d J C1,H1 in
Meth yl P yr a n osid es 58-65
Ch a r t 3
calculated
experimentald
c
compounda
C1-H1 distanceb
1J C1,H1
1J C1,H1
58
59
60
61
1.0869
1.0923
1.0915
1.0866
1.0868
1.0873
1.0869
1.0927
1.0917
1.0921
1.0866
1.0868
1.0869
1.0920
1.0915
1.0911
158.7
141.5
144.1
159.7
159.4
157.6
159.2
140.0
143.8
141.5
160.0
159.3
159.2
142.3
144.5
144.7
170
158
161
179
62-gg
62-gt
62-tg
63-gg
63-gt
63-tg
64-gg
64-gt
64-tg
65-gg
65-gt
65-tg
170
159
170
159e
a
See Figure 2 for definitions of the gg, gt, and tg rotamers about
b
the C4-C5 bond. In angstroms; measured from HF/6-31G*
optimized geometries. c In hertz, calculated using eq 1. From ref
d
2. e Value taken from ref 31.
marginal improvement (see Table S2 in the Supporting
Information).
We were curious as to whether this discrepancy
between the predicted and experimentally measured
coupling constants was unique to these rings systems.
Consequently, bond lengths obtained from the HF/6-31G*
optimization of methyl pyranosides 58-65 were used in
the difference is smaller, 0.0045-0.006 Å in the O-
glycosides vs 0.0055-0.008 Å in the S-glycosides. The
lack of correlation between C1-H1 bond length and
1J C1-H1 in these thioglycosides is consistent with previous
studies8,9 on 1,2-oxathiane (72, Figure 4). In 72, the C2-
1
conjunction with eq 1 to predict J C1-H1 magnitudes in
1
H2ax bond is longer than the C2-H2eq bond, but J C2-H2eq
these glycosides. The results of these calculations are
shown in Table 4, together with the experimentally
measured couplings.2,31 From these data it is clear that
for 58-65 (and likely other pyranose glycosides) that eq
)
1J C2-H2zx. As in the case of the sulfur-containing
heterocycles studied to date (1,3-oxathiane, 1,2-dithiane),
it appears that in thioglycosides 9-12, the factors that
influence one bond C-H coupling constants is an inter-
play of a number of factors including, but not limited to
the C-H bond length (e.g, orbital hybridization, atom
electronegativities.)
1
1 underestimates the J C1-H1 magnitudes by approxi-
mately 10-20 Hz relative to the experimental values.
In the initial report describing eq 1, two different
conformers of acetaldehyde hydrate, 66 and 67, were used
as model compounds for an R- and â-glycoside, respec-
tively. The calculated C-H bond lengths were 1.0837 Å
(66) and 1.08906 Å (67), which are shorter than those
determined for 58-65 (Table 4). In light of this informa-
tion, it is not surprising that this equation underesti-
mates these 1J C1-H1 magnitudes. It can also be concluded
that (1) the underestimation of these couplings for 6 and
7 is not unique to these anhydrosugars; (2) the good
correlation between the calculated and experimental
values observed for 5 and 8 is fortuitous; and (3) eq 1
should be updated through the calculation of C1-H1
bond lengths in a range of carbohydrate derivatives.
F igu r e 4. C2-H2ax and C2-H2eq bond lengths and the associ-
1
ated J C-H in 1,3-oxathiane. B3LYP/6-31+G** bond lengths
1
were taken from ref 9 and J C-H from ref 8.
Con clu sion s
1
Rela tion sh ip betw een J C1-H1 a n d C1-H1 Bon d
1
In conclusion we have shown that the J C1-H1 in 2,3-
Len gth in Th ioglycosid es. As mentioned above, the
anhydro-O-furanosides is sensitive to the stereochemistry
at the anomeric center and that the magnitude of the
coupling constant is inversely related to the C1-H1 bond
length. We have also shown that a previously reported
1
magnitude of J C1-H1 in thioglycosides 9-12 is not
sensitive to the stereochemistry at the anomeric center.
This prompted us to determine the C1-H1 bond lengths
in 68-71 (Chart 3) at the B3LYP/6-31G* level of theory.
These calculations demonstrated that the trend observed
in 5-8, i.e., that the C1-H1 bond was longer in the â-lyxo
and R-ribo isomers, was also present in these thioglyco-
sides (see Supporting Information, Table S3). However,
1
equation (eq 1) relating C1-H1 bond length to J C1-H1
magnitudes underestimates these coupling in many
cases. Circumventing this problem requires that this
relationship be reparametrized, and such investigations
are currently in progress.
(30) (a) Reed, A. E.; Weinhold, F. J . Chem. Phys. 1985, 83, 1736.
(b) Glendening, E. D.; Reed, A. E.; Carpenter, J . E.; Weinhold, F. NBO
3.1; Theoretical Chemistry Institute, University of Wisconsin: Madi-
son, WI, 1996.
(31) Podlasek, C. A.; Wu, J .; Stripe, W. A.; Bondo, P. B.; Serianni,
A. S. J . Am. Chem. Soc. 1995, 117, 8625.
Exp er im en ta l Section
Gen er a l. Solvents were distilled from the appropriate
drying agents before use. Unless stated otherwise, all reactions
were carried out under a positive pressure of argon and were

4556 J . Org. Chem., Vol. 66, No. 13, 2001
Callam et al.
monitored by TLC on silica gel 60 F254 (0.25 mm, E. Merck).
Spots were detected under UV light or by charring with 10%
H2SO4 in ethanol. Solvents were evaporated under reduced
pressure and below 40 °C (bath). Organic solutions of crude
products were dried over anhydrous Na2SO4. Column chro-
matography was performed on silica gel 60 (40-60 µM). The
ratio between silica gel and crude product ranged from 100 to
50:1 (w/w). Optical rotations were measured at 21 ( 2 °C.
Melting points are uncorrected. 1H NMR spectra were recorded
at 400 or 500 MHz, and chemical shifts are referenced to either
TMS (0.0, CDCl3), CD3OD (3.31, CD3OD), or external dioxane
(3.75, D2O). 13C NMR spectra were recorded at 100 or 125
MHz, and 13C chemical shifts are referenced to CDCl3 (77.00,
CDCl3), CD3OD (49.15, CD3OD), or external dioxane (68.11,
D2O). Elemental analyses were performed by Atlantic Microlab
Inc., Norcross, GA. Electrospray mass spectra were recorded
on samples suspended in THF or CH3OH. Physical data (1H
NMR, 13C NMR, MS, optical rotation, and elemental analysis)
for all new compounds is included in the Supporting Informa-
tion.
n -Oct yl 2,3-An h yd r o-5-O-b en zoyl-r-D-lyxofu r a n osid e
(3). To a solution of 2715 (1.0 g, 3.8 mmol), triphenylphosphine
(2.5 g, 9.5 mmol), and benzoic acid (0.7 g, 5.7 mmol) in THF
was added DIAD (1.9 mL, 9.5 mmol) at 0 °C. The reaction
mixture was allowed to stir at room temperature for 30 min.
The solution was concentrated to an oil which was purified by
column chromatography (hexanes/EtOAc, 6:1) to yield 3 as a
colorless oil (1.08 g, 82%).
n -Octyl 2,3-An h yd r o-r-D-lyxofu r a n osid e (28). To a solu-
tion of 3 (1.08 g, 3.1 mmol) in CH3OH (25 mL) was added 0.1
M NaOCH3 in CH3OH (3 mL). After stirring for 4 h at room
temperature, the reaction mixture was neutralized with pre-
washed Amberlite IR-120 H+ resin, filtered, and concentrated.
The resulting syrup was chromatographed (hexanes/EtOAc,
2:1) to obtain 28 (0.68 g, 91%) as a colorless liquid.
p-Tolyl 2,3,5-Tr i-O-ben zoyl-1-th io-r-D-a r a bin ofu r a n o-
sid e (29). To a solution of methyl 2,3,5 tri-O-benzoyl-R-D-
arabinofuranoside22 (10 g, 21.0 mmol) in CH2Cl2 (50 mL) was
added BF3‚Et2O (7.9 mL, 63.0 mmol) at 0 °C under an argon
atmosphere. The solution was allowed to stir for 10 min, and
then p-thiocresol (2.60 g, 21.0 mmol) was added. The reaction
mixture was stirred for 5 h, diluted with CH2Cl2, and then
poured into a cold saturated NaHCO3 solution. The layers were
separated, and organic layer was washed successively with a
saturated NaHCO3 solution and water and then dried and
concentrated. Upon addition of methanol to the resulting oil,
the product crystallized. Recrystallization from methanol gave
29 (9.6 g, 81%) as a white solid.
Syn th esis of 32-35. All alcohols and donors were predried
under vacuum in the presence of P2O5 for 4 h prior to use. To
a solution of the alcohol (5,14 28, 30,17 3118) (0.8 mmol) and
donor 2916 (1.0 mmol) in CH2Cl2 (10 mL) at 0 °C was added
powdered 4 Å molecular sieves (0.5 g). The reaction mixture
was stirred for 20 min at 0 °C, and then N-iodosuccinimide
(1.0 mmol) and silver triflate (0.25 mmol) were added. After
stirring for 30 min, triethylamine was added. The reaction
mixture was diluted with CH2Cl2 and filtered through Celite.
The filtrate was washed with a saturated solution of sodium
thiosulfate, water, and brine. The organic layer was subse-
quently dried, filtered, and concentrated. The resulting crude
oil was purified by chromatography to give disaccharides 32-
35.
Chromatography (hexanes/EtOAc, 4:1) provided 35 (1.21 g,
91%) as a white solid.
Syn th esis of 36-39. A solution of the each disaccharide
32-35 (0.6 mmol) in CH2Cl2 (10 mL) and CH3OH (10 mL) was
treated with a catalytic amount of 1 M NaOCH3 in CH3OH (2
mL) solution at room temperature. After stirring for 6 h the
reaction mixture was neutralized with dry ice and concen-
trated. The crude product was purified by chromatography to
yield compounds 36-39.
Met h yl 5-O-(r-D-Ar a b in ofu r a n osyl)-2,3-a n h yd r o-r-D-
lyxofu r a n osid e (36). Chromatography (CHCl3:CH3OH, 10:
1) provided 36 (686 mg, 97%) as a colorless syrup.
n -Octyl 5-O-(r-D-Ar a bin ofu r a n osyl)-2,3-a n h yd r o-r-D-
lyxofu r a n osid e (37). Chromatography (hexanes/EtOAc, 1:1)
provided 37 (492 mg, 90%) as a colorless syrup.
Meth yl 6-O-(r-D-Ar a bin ofu r a n osyl)-2,3,4-tr i-O-ben zyl-
r-D-m a n n op yr a n osid e (38). Chromatography (hexanes/
EtOAc, 1:1) provided 38 (771 mg, 89%) as a colorless syrup.
Meth yl 3-O-(r-D-Ar a bin ofu r a n osyl)-4,6-O-ben zylid en e-
2-O-b en zyl-r-D-glu cop yr a n osid e (39). Chromatography
(CHCl3/CH3OH, 15:1) provided 39 (541 mg, 88%) as a white
solid.
Syn th esis of 17, 19, 21, a n d 25. Each disaccharide 36-
39 (0.4 mmol), triphenylphosphine (1.0 mmol), and benzoic acid
(0.6 mmol) was dissolved in THF (20 mL), and the solution
was cooled to 0 °C. DIAD (1.0 mmol) was added dropwise to
the solution over a 10 min period, and the reaction was then
stirred for an additional 1 h while allowing to warm to room
temperature. The solution was then concentrated, and the
residue was purified by chromatography to give the products
in 73-91% yield.
Meth yl 5-O-(2,3-An h yd r o-5-O-ben zoyl-r-D-lyxofu r a n o-
syl)-2,3-a n h yd r o-r-D-lyxofu r a n osid e (17). Chromatography
(hexanes/EtOAc, 7:1) provided 17 (771 mg, 91%) as a white
solid.
n -Octyl 5-O-(2,3-An h yd r o-5-O-ben zoyl-r-D-lyxofu r a n o-
syl)-2,3-a n h yd r o-r-D-lyxofu r a n osid e (19). Chromatography
(hexanes/EtOAc, 5:1) provided 19 (412 mg, 83%) as a white
solid.
Meth yl 6-O-(2,3-An h yd r o-r-D-lyxofu r a n osyl)-2,3,4-tr i-
O-ben zyl-r-D-m a n n op yr a n osid e (21). It was impossible to
completely purify 40 due to the presence of DIAD-based
impurities which were inseparable by chromatography. There-
fore, the crude product obtained after an initial chromato-
graphic purification (hexanes/EtOAc, 4:1) was deprotected with
0.1 M NaOCH3 in CH3OH as described for the synthesis of
36. Chromatography (hexanes/EtOAc, 6:1) yielded 21 (585 mg,
73%) as a clear syrup.
Meth yl 3-O-(2,3-An h yd r o-r-D-lyxofu r a n osyl)-4,6-O-ben -
zylid en e-3-O-ben zyl-r-D-glu cop yr a n osid e (25). Chroma-
tography (hexanes/EtOAc, 6:1) afforded 25 (391 mg, 83%) as
a white solid.
p-Cr esyl 5-O-ter t-Bu tyld ip h en ylsilyl-2,3-d i-O-ben zoyl-
1-th io-r-D-a r a bin ofu r a n osid e (42). To a stirred solution of
4119 (1.05 g, 4.1 mmol) in pyridine (10 mL) was added
tertbutylchlorodiphenylsilane (1.35 g, 4.9 mmol) at 0 °C. The
reaction mixture was allowed to stir for 5 h while warming to
room temperature. The solution was recooled to 0 °C, and
benzoyl chloride (1.2 mL, 10.2 mmol) was added. The reaction
mixture was allowed to stir overnight while warming to room
temperature, and then CH3OH was added. The mixture was
diluted with CH2Cl2 and washed successively with a saturated
solution of NaHCO3 and water. The organic layer was dried
and concentrated, and the resulting residue was purified by
chromatography (hexanes/EtOAc, 8:1) to give 42 (2.52 g, 88%)
as a colorless syrup.
Meth yl 5-O-[5-O-ter t-Bu tyld ip h en ylsilyl-2,3-d i-O-ben -
zoyl-r-D-a r a bin ofu r a n osyl]-2,3-a n h yd r o-r-D-lyxofu r a n o-
sid e (43). To a solution of alcohol 514 (0.70 g, 4.8 mmol) and
donor 42 (3.36 g, 4.8 mmol) in CH2Cl2 (30 mL) was added
powdered 4 Å molecular sieves (1.0 g). The reaction mixture
was stirred for 20 min at 0 °C, and then N-iodosuccinimide
(1.07 g, 4.80 mmol) and silver triflate (0.38 g, 1.4 mmol) were
added. After 15 min, triethylamine was added. The reaction
mixture was diluted with CH2Cl2 and filtered through Celite.
Meth yl 5-O-(2,3,5-Tr i-O-ben zoyl-r-D-ar abin ofu r an osyl)-
2,3-an h ydr o-r-D-lyxofu r an oside (32). Chromatography (hex-
anes/EtOAc, 5:1) provided 32 (1.79 g, 88%) as a colorless syrup.
n -Octyl 5-O-(2,3,5-Tr i-O-ben zoyl-r-D-ar abin ofu r an osyl)-
2,3-an h ydr o-r-D-lyxofu r an oside (33). Chromatography (hex-
anes/EtOAc, 6:1) provided 33 (1.19 g, 81%) as a colorless syrup.
Meth yl 6-O-(2,3,5-Tr i-O-ben zoyl-r-D-ar abin ofu r an osyl)-
2,3,4-tr i-O-ben zyl-r-D-m a n n op yr a n osid e (34). Chromatog-
raphy (hexanes/EtOAc, 5:1) provided 34 (1.35 g, 89%) as a
colorless syrup.
Meth yl 3-O-(2,3,5-Tr i-O-ben zoyl-r-D-ar abin ofu r an osyl)-
4,6-O-ben zylid en e-3-O-ben zyl-r-D-glu cop yr a n osid e (35).

2,3-Anhydro-O-furanosides
J . Org. Chem., Vol. 66, No. 13, 2001 4557
The filtrate was concentrated, and the resulting syrup was
chromatographed (hexanes/EtOAc, 8:1) to give 43 (2.91 g, 84%)
as a colorless syrup.
Cycloh exyl 2,3-An h yd r o-5-O-ben zoyl-â-D-r ibofu r a n o-
sid e (16). Benzoate 16 was synthesized from 51 (100 mg, 0.46
mmol) as described for the preparation of 15. Chromatography
(hexanes/EtOAc, 6:1) provided 16 (132 mg, 89%) as a clear
syrup.
Met h yl 5-O-[2,3-Di-O-b en zoyl-r-D-a r a b in ofu r a n osyl]-
2,3-a n h yd r o-r-D-lyxofu r a n osid e (44). To a solution of dis-
accharide 43 (2.85 g, 3.9 mmol) in THF (30 mL) was added
TBAF (1.13 g, 4.3 mmol) at 0 °C under argon atmosphere. The
reaction mixture was allowed to stir for 1 h and then
concentrated to a syrup which was purified by chromatography
(hexanes/EtOAc, 4:1) providing 44 (1.55 g, 81%) as a clear
syrup.
Meth yl 5-O-[5-O-(2,3,5-Tr i-O-ben zoyl-r-D-a r a bin ofu r a -
n osyl)-2,3-d i-O-ben zoyl-r-D-a r a bin ofu r a n osyl]-2,3-a n h y-
d r o-r-D-lyxofu r a n osid e (45). Reaction of alcohol 44 (1.5 g,
3.0 mmol) and donor 29 (2.27 g, 4.0 mmol) as described for
the synthesis of 32, followed by chromatography (hexanes/
EtOAc, 4:1) provided 45 (2.56 g, 89%) as a white solid.
Meth yl 5-O-[5-O-(r-D-Ar a bin ofu r a n osyl)-r-D-a r a bin o-
fu r a n osyl]-2,3-a n h yd r o-r-D-lyxofu r a n osid e (46). Trisac-
charide 46 was prepared from 45 (2.5 g, 2.68 mmol) as
described for the prepar ation of 36. Chromatography (CHCl3/
CH3OH, 10:1) provided 46 (990 mg, 90%) as a colorless syrup.
Met h yl 5-O-[5-O-(2,3-An h yd r o-5-O-b en zoyl-r-D-lyxo-
fu r a n osyl)-2,3-a n h yd r o-r-D-lyxofu r a n osyl)-2,3-a n h yd r o-
r-D-lyxofu r a n osid e (23). Trisaccharide 23 was synthesized
from 46 (700 mg, 1.7 mmol) as described for the preparation
of 17. Chromatography (hexanes/EtOAc, 3:1) afforded 23 (720
mg, 89%) as a white solid.
p-Tolyl 2,3-An h yd r o-5-O-ben zoyl-1-th io-r-D-lyxofu r a -
n osid e (1) a n d p-Tolyl 2,3-An h yd r o-5-O-ben zoyl-1-th io-
â-D-lyxofu r a n osid e (53). Compound 5219 (2.0 g, 7.81 mmol),
triphenylphosphine (5.2 g, 19.8 mmol), and benzoic acid (1.42
g, 11.6 mmol) were dissolved in THF (50 mL), and the solution
was cooled to 0 °C. DIAD (3.86 mL, 19.5 mmol) was added
dropwise over 10 min. After complete addition, the reaction
mixture was allowed to warm to room temperature and was
stirred for 45 min. The solution was subsequently evaporated
to yield an which, upon trituration with cold diethyl ether,
precipitated triphenylphosphine oxide. The solid was filtered
off, and the filtrate was concentrated to an oil which was
purified by chromatography (hexanes/EtOAc, 5:1), providing
1 (1.95 g, 73%) and 53 (310 mg, 12%) as white crystalline
solids.
p-Tolyl 2,3-An h yd r o-1-th io-r-D-lyxofu r a n osid e (9). To
a solution of 1 (1.0 g, 2.92 mmol) in CH3OH (20 mL) and CH2-
Cl2 (5 mL) was added 0.1 M NaOCH3 in CH3OH (4 mL). After
stirring for 8 h at room temperature, the reaction mixture was
neutralized with Amberlite IR-120 H+ resin, filtered, and
concentrated to an oil which was purified by chromatography
(hexanes/EtOAc, 2:1) to yield 9 (670 mg, 96%) as a white solid.
Cycloh exyl 3,5-Di-O-ben zoyl-r/â-D-a r a bin ofu r a n osid e
(48). To a solution of diol 4720 (1.0 g, 2.8 mmol) and cyclohex-
anol (1.1 g, 12 mmol) in CH2Cl2 (20 mL) was added p-TsOH
(50 mg, 0.23 mmol). After stirring for 6 h, the solution was
diluted with CH2Cl2 (30 mL), washed with saturated NaHCO3
solution and water, and dried. The solution was filtered and
concentrated to an oil which was purified by chromatography
(hexanes/EtOAc, 3:1) to yield 48 (1.02 g, 84%) as a white solid.
Cycloh exyl r/â-D-Ar a bin ofu r a n osid e (49). To a solution
of 48 (1.0 g, 2.27 mmol) in CH3OH (20 mL) was added 0.1 M
NaOCH3 in CH3OH (0.5 mL). After stirring for 6 h at room
temperature, the reaction mixture was neutralized with Am-
berlite IR-120 H+ resin, filtered, and concentrated. The residue
was purified by chromatography (10:1, CHCl3, CH3OH) to give
49 (484 mg, 93%) as a colorless syrup.
Cycloh exyl 2,3-An h yd r o-5-O-ben zoyl-r-D-lyxofu r a n o-
sid e (13) a n d Cycloh exyl 2,3-An h yd r o-5-O-ben zoyl-â-D-
lyxofu r a n osid e (14). Epoxides 13 and 14 was synthesized
from 49 (468 mg, 2.0 mmol) as described for the preparation
of 17. Chromatography (hexanes/EtOAc, 8:1 f hexanes/EtOAc,
6:1) provided 13 (280 mg, 46%) and 14 (180 mg, 28%) as white
crystalline solids.
p -Tolyl 2,3-An h yd r o-1-t h io-â-D-lyxofu r a n osid e (10).
Thioglycoside 10 was synthesized from 53 (300 mg, 0.88 mmol)
as described for the preparation of 9. Chromatography (hex-
anes/EtOAc, 2:1) provided 10 (204 mg, 98%) as a colorless
syrup.
2-P yr id yl 2,3,5-Tr i-O-ben zoyl-1-th io-r/â-D-a r a bin ofu r a -
n osid e (55). To a mixture of 2-mercaptopyridine (1.29 g, 11.7
mmol) and potassium carbonate (2.01 g, 14.6 mmol) in acetone
(30 mL) was added 5422 (5.11 g, 9.75 mmol) in toluene (30 mL).
The reaction mixture was stirred overnight. The resulting salts
were filtered through Celite, and the filtrate was concentrated.
Chromatography (hexanes/EtOAc, 4:1) afforded 55 (4.74 g,
88%) as a white foam.
2-P yr id yl 1-Th io-r/â-D-a r a bin ofu r a n osid e (56). To a
solution of 55 (4.32 g, 8.3 mmol) in CH3OH (40 mL) and CH2-
Cl2 (20 mL) was added 0.1 M NaOCH3 in CH3OH (4 mL). After
stirring for 6 h at room temperature, the reaction mixture was
neutralized with Amberlite IR-120 H+ resin, filtered, and
concentrated. The residue was purified by column chromatog-
raphy (CHCl3/CH3OH, 10:1) to give 56 (1.91 g, 95%) as a
colorless syrup.
Cycloh exyl 2,3-An h yd r o-r-D-r ibofu r a n osid e (50) a n d
Cycloh exyl 2,3-An h yd r o-â-D-r ibofu r a n osid e (51). To a
solution of 48 (1.20 g, 2.7 mmol) in CH2Cl2 (25 mL) and
triethylamine (1.51 mL, 11 mmol) at 0 °C was added meth-
anesulfonyl chloride (0.32 mL, 2.9 mmol). The reaction mixture
was stirred for 30 min and then cold water added. The reaction
mixture was diluted with CH2Cl2, and the organic layer was
washed with a saturated solution of NaHCO3 and then water.
After drying the organic layer was filtered and concentrated
to a syrup. The resulting residue was redissolved in CH3OH,
and 1 M NaOCH3 in CH3OH (2 mL) was added. The solution
was stirred overnight, neutralized with Amberlite IR-120 H+
resin, filtered, and concentrated. The residue was purified by
column chromatography (hexanes/EtOAc, 5:1) to give 50 (311
mg, 58%) as a white solid and 51 (132 mg, 24%) as a colorless
syrup.
Cycloh exyl 2,3-An h yd r o-5-O-ben zoyl-r-D-r ibofu r a n o-
sid e (15). To a solution of 50 (250 mg, 1.2 mmol) in pyridine
(2 mL) at 0 °C was added benzoyl chloride (0.2 mL, 1.8 mmol)
dropwise, and the reaction mixture was stirred for 30 min at
room temperature. The reaction was then diluted with CH2-
Cl2 and then washed with chilled 5% HCl, a saturated solution
of NaHCO3, and water. The organic layer was dried, filtered,
and concentrated. Chromatography (hexanes/EtOAc, 6:1) pro-
vided 15 (342 mg, 92%) as a white solid.
2-P yr id yl 2,3-An h yd r o-1-th io-r-D-lyxofu r a n osid e (11)
an d 2-P yr idyl 2,3-An h ydr o-1-th io-â-D-lyxofu r an oside (12).
Compound 56 (1.5 g, 6.7 mmol), triphenylphosphine (4.0 g, 15.4
mmol), and benzoic acid (1.12 g, 9.2 mmol) were dissolved in
THF (50 mL), and the solution was cooled to 0 °C. DIAD (3.86
mL, 19.5 mmol) was added dropwise over 10 min. After
complete addition, the reaction mixture was allowed to warm
to room temperature and was stirred for 45 min. The solution
was subsequently evaporated to yield a crude oil which, upon
trituration with cold diethyl ether, precipitated triphenylphos-
phine oxide. The solid was filtered off, and the filtrate was
concentrated. The resulting oil was dissolved in CH3OH (40
mL) and dichloromethane (20 mL), and then 0.1 M NaOCH3
in CH3OH (4 mL) was added. After stirring for 6 h at room
temperature, the reaction mixture was neutralized with Am-
berlite IR-120 H+ resin, filtered, and concentrated. The residue
was purified by chromatography (EtOAc/hexanes, 2:1) to give
11 (770 mg, 56%) and 12 (350 mg, 25%) as colorless syrups.
Mea su r em en t of J C1-H1. The 1J C1-H1 were measured from
1
the 1H-coupled 13C NMR spectra23 of 3-26, that were recorded
at 125 MHz on samples dissolved in CDCl3.
Com p u ta tion a l In vestiga tion s. Ab initio molecular or-
bital and DFT calculations were conducted using Gaussian

4558 J . Org. Chem., Vol. 66, No. 13, 2001
Callam et al.
98.32 For each of compounds 5-8, all three possible staggered
C4-C5 rotamers (Figure 2) were considered: gauche-gauche
(gg), gauche-trans (gt), or trans-gauche (tg). Optimizations
were carried out using both Hartree-Fock (HF)12 and density
functional theory (B3LYP)13 calculations with the 6-31G* basis
set. In all optimizations, the orientation about the C1-O1 bond
was initially chosen to maximize the exo-anomeric effect33 (the
methyl group was placed trans to the C1-C2 bond). The
orientation about the C5-O5 bond was initially set with the
hydrogen of the hydroxyl group oriented trans to the C4-C5
bond. During the optimizations, no significant changes were
seen in the orientations about these two bonds, or about the
C4-C5 bond. In all calculations, the initial furan ring geometry
was OE. For the conformers of 7 and 8, this ring minimized to
EO, but for 5 and 6 it remained OE. To ensure that the OE
conformer was not a local minima for 5 and 6, we also began
these calculations with furan ring in the EO conformation;
however, upon optimization, these conformers minimized to
OE. An identical approach was used in the optimization of
thioglycosides 68-71. The Cartesian coordinates for the
optimized conformers are given in the Supporting Information.
Single-point energy calculations were carried out at the B3LYP
level of theory with the 6-31+G** basis set; these energies can
be found in the Supporting Information. Atomic charges were
determined by Mulliken population28 and NBO30 analysis.
Ack n ow led gm en t. This work was supported by the
National Institutes of Health (AI44045), the National
Science Foundation (CHE-9875163), and the OSU Spec-
troscopy Institute. We also acknowledge support from
the Ohio Supercomputer Center where some of the
calculations were performed. We thank Justin B. House-
knecht for technical assistance and Dr. Christopher M.
Hadad (The Ohio State University) for help with the
computational work and for helpful discussions.
(32) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J . R.; Zakrzewski, V. G.; Montgomery, J . A.,
J r.; Stratmann, R. E.; Burant, J . C.; Dapprich, S.; Millam, J . M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J .;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,
C.; Clifford, S.; Ochterski, J .; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J . B.; Cioslowski, J .; Ortiz, J . V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J .; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; J ohnson, B.; Chen, W.;
Wong, M. W.; Andres, J . L.; Gonzalez, C.; Head-Gordon, M.; Replogle,
E. S.; Pople, J . A. Gaussian 98, Revision A.7, Gaussian, Inc.; Pitts-
burgh, PA, 1998.
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
spectra for 3, 4, and 9-26. Characterization data for all new
compounds. Cartesian coordinates, pseudorotational phase
angles and puckering amplitudes for the HF/6-31G* and
B3LYP/6-31G*-optimized geometries of 5-8, tables of relevant
structural parameters in 5-8 and 68-71. This material is
available free of charge via the Internet at http://pubs.acs.org.
(33) (a) Lemieux, R. U. Pure Appl. Chem. 1971, 25, 527. (b) Lemieux,
R. U.; Koto, S. Tetrahedron 1974, 30, 1933.
J O001747A