Novel Heterocycloaddition Reaction of Glycals

Article abstract of DOI:10.1021/jo981047u

Diacyl thione species 1 has been generated and reacted in situ with both pyranoid and furanoid glycals to form novel [4 + 2] cycloadducts. Factors such as protecting groups and configuration of substituents in the glycals along with medium effects were varied to discover influences on face selectivity and reactivity. A qualitative correlation of reactivity with the HOMO-LUMO gap between the glycal (HOMO) and the heterodienic species (LUMO) is observed. In one example, the isolation of byproducts suggests that the cycloaddition may in fact be stepwise.

Full text of DOI:10.1021/jo981047u

J . Org. Chem. 1998, 63, 6673-6679
6673
Novel Heter ocycloa d d ition Rea ction of Glyca ls
Angeles Dios, Aloma Geer, Cecilia H. Marzabadi, and Richard W. Franck*
Department of Chemistry, Hunter College, 695 Park Avenue, New York, New York 10021
Received J une 2, 1998
Diacyl thione species 1 has been generated and reacted in situ with both pyranoid and furanoid
glycals to form novel [4 + 2] cycloadducts. Factors such as protecting groups and configuration of
substituents in the glycals along with medium effects were varied to discover influences on face
selectivity and reactivity. A qualitative correlation of reactivity with the HOMO-LUMO gap
between the glycal (HOMO) and the heterodienic species (LUMO) is observed. In one example,
the isolation of byproducts suggests that the cycloaddition may in fact be stepwise.
Sch em e 1
The attachment of ligands through heteroatom links
to C-1 of carbohydrates, known as glycosidation or
glycosyl transfer, is a field of both historic and current
interest.1 The chemistry is largely focused on creating
and controlling the electrophilicity of C-1 of the carbo-
hydrate. The concept of heterocycloaddition with glycals
as a method to functionalize carbohydrates at C-1 (and
C-2), as illustrated in Scheme 1, appeared to us as an
attractive alternative to the “classical” methods.
Since their discovery by Fischer and Zach just before
World War I,2 glycals have evolved into significant
starting materials for the simultaneous functionalization
of the anomeric carbon and carbon 2 of sugars. Most
glycal chemistry is initiated via electrophilic attack on
addition of an o-quinone methide (2 C-C bonds formed,
eq 3),7 a [4 + 2] addition of an isoquinolinium salt (2 C-C
bonds formed, eq 4),8 a [2 + 2] â-lactam synthesis (C-C
at C-2 and C-N at C-1, eq 5),9 a [4 + 2] addition of
azodicarboxylate esters (C-O at C-1 and C-N at C-2
bond, eq 6),10 a [2 + 2] addition of dichloroketene (eq 7),11
and the work to be described below, a [4 + 2] addition of
thionoketones (C-O at C-1 and C-S at C-2, eq 8).12
its very nucleophilic enol ether double bond (eq 1).3
A
The genesis of our quest for a thionoketone reagent
lay in the observation of the simplicity of carrying out
the isoquinolinium salt cycloaddition (eq 4) and of its high
yields of stereochemically homogeneous products.8 This
facile reaction was in marked contrast to our electrophilic
method for 2-deoxy-â-glycosidation which required fairly
sophisticated manipulation of corrosive reagents and
which yielded mixtures of glycosides in most cases.13
Thus, a search was undertaken for a diene with sulfur
and oxygen at its termini so that glycals might be
functionalized with heteroatoms under mild and facile
cycloaddition conditions. The key diacylthione reagent
1, generated in situ from phthalimidothiopentanedione
precursor 2, was first identified as a heterodienophile in
well-known electrophilic variation is induced by electro-
philic attack at an allylic leaving group at C-3, followed
by an SN2′-like attack of a nucleophile at C-1 (the Ferrier
rearrangement, eq 2).4 A less common and different
double bond chemistry of glycals is their participation
as the allylic component in Claisen rearrangements.5
Cycloaddition to the glycal double bond is an additional
functionalization approach.6 Examples include a [4 + 2]
(1) (a) For a complete compilation of methods for preparation of
O-disaccharides in the hexose series for1994, see: (a) Barresi, F.;
Hindsgaul, O. Modern Synthetic Methods, 1995; Ernst, B., Leumann,
C., Ed.; VCH: Basel, Switzerland, 1995; pp 281-330. (b) Barresi, F.;
Hindsgaul, O. J . Carbohydr. Chem. 1995, 14, 1043-1087. (c) For a
review of the widely used trichloroacetimidate method, see: Schmidt,
R. R.; Kinzy, W. Adv. Carbohydr. Chem. Biochem. 1994, 21-123.
(2) (a) Fischer, E.; Zach, K. Sitzungsber. K. Preuss. Akad. Wiss. 1913,
16, 311. (b) Fischer, E.; Zach, K. Ber. Bunsen-Ges 1914, 311-317.
(6) Cycloaddition Reactions in Carbohydrate Chemistry; Giuliano,
R. M., Ed.; ACS Symposium Series 494; American Chemical Society:
Washington, DC, 1992.
(7) (a) Franck, R. W.; J ohn, T. V J . Org. Chem. 1983, 48, 3269. (b)
Chew, S.; Ferrier, R. J .; Sinnwell, V. Carbohydr. Res. 1988, 174, 161-
168.
(8) Gupta, R. B.; Franck, R. W. J . Am. Chem. Soc. 1989, 111, 7668.
(9) Chmielewski, M.; KaIuza, K.; Grodner, J .; Urbanski, R. Chapter
4 in ref 5.
(10) Leblanc, Y.; Labelle, M. Chapter 6 in ref 5.
(3) For a recent review, see: Danishefsky, S. J .; Bilodeau, M. T.
Angew. Chem., Int. Ed. Engl. 1996, 35, 1380-1419.
(11) Oh, J . Tetrahedron Lett. 1997, 38, 3249-3250.
(12) For preliminary disclosures of aspects of this work, see: (a)
Capozzi, G.; Franck, R. W.; Mattioli, M.; Menichetti, S.; Nativi, C.;
Valle, G. J . Org. Chem. 1995, 60, 6416-6426. (b) Capozzi, G.; Dios,
A.; Franck, R. W.; Geer, A.; Marzabadi, C.; Menichetti, S.; Nativi, C.;
Tamarez, M. Angew. Chemie, Int. Ed. Engl. 1996, 35, 777-779.
(13) Grewal, G.; Kaila, N.; Franck, R. W.. J . Org. Chem. 1992, 57,
2084-2092.
(4) (a) Ferrier, R. J .; Prasad, N. J . Chem. Soc. C 1969, 581-586. (b)
Ferrier, R. J . Methods Carbohydr. Chem. 1972, 6, 307-11.
(5) (a) Ireland, R. E.; Wilcox, C. S.; Thaisvirongs, S.; Varnier, N. R.
Can. J . Chem. 1979, 57, 1743-1745. See the following reference for a
complete listing of Ireland’s Claisen work with carbohydrates: (b)
Curran, D. P.; Suh, Y.-G. Carbohydr. Res. 1987, 171, 161-191.
S0022-3263(98)01047-0 CCC: $15.00 © 1998 American Chemical Society
Published on Web 08/26/1998

6674 J . Org. Chem., Vol. 63, No. 19, 1998
Dios et al.
Ta ble 1. Resu lts of Cycloa d d ition s of
3-Th ion op en ta n ed ion e 1 (LUMO, - 1.62 eV)
a high-yield trapping of isoprene to form thiapyran 3 (eq
9).14 Simply by picturing diacylthione in a gauche (1a )
chloride at room temperature, adding either pyridine or
lutidine as an acid scavenger, and following the reaction
progress by NMR and/or TLC. Product formation is slow,
but the reaction mixtures are quite clean. Workup
normally consists of a basic wash to remove phthalimide
and an acid wash to removed the pyridine or lutidine,
followed by flash chromatography to separate the cy-
cloadducts from unreacted glycal and from residual
impurities carried through from the preparation of
precursor 2. Table 1 shows the glycal reactant, the
structures of the products, the approximate t1/2 for the
reaction, and the computed difference (in eV) between
the HOMO of the glycal and the LUMO of thiono species
1. The stereochemistry of the cycloadducts is easily
deduced by identifying the protons at C-1, C-2, and C-3
along with their coupling constants. X-ray crystal-
lography analysis of adducts 13, 17, and 23 (Table 2)
served to reinforce the NMR assignments.
rather than anti conformation, we were able to conceive
of it as our desired heterodiene. Its utility in cycloaddi-
tions to glycals is described below.
Resu lts
The cycloadditions are carried out by mixing glycal and
thione precursor 2 in either chloroform or methylene
(14) Capozzi, G.; Nativi, C.; Menichetti, S.; Rosi, A.; Valle, G.
Tetrahedron 1992, 48, 9023-9029.

Novel Heterocycloaddition Reaction of Glycals
J . Org. Chem., Vol. 63, No. 19, 1998 6675
Ta ble 2. Resu lts of Cycloa d d ition s of
3-Th ion op en ta n ed ion e: P r otectin g Gr ou p Va r ia bles
nonexistent. The effect of solvent on rate is small, with
the largest effect observed, about a factor of 2, with
H-bond donor solvents, t-BuOH and MeOH. Somewhat
larger solvent effects have been noted with other thiono-
carbonyl species that we have studied. It should be noted
that the effect of solvent on the Diels-Alder reaction is
a matter of intense study, with most of the focus dealing
with overall rate acceleration.15 There have also been
studies where exo-endo selectivities have been an issue.
However, the effect of solvent on diastereofacial selectiv-
ity has not been addressed. The number of data points,
from this and related work, and the magnitude of the
effect are too limited for us to offer any rationale. As a
control, it has been ascertained that the isomer ratios
are not due to reversibility of cycloaddition. We simply
record that, for yield optimization purposes, solvent and
blocking group effects should not be overlooked. It is also
worth noting that carbohydrate hydroxyls need not be
blocked at all.
isolated
entry
1
yield (%)a
83
R3,R4 ) Bn,
95 (11) 05 (23)
68 (25) 32 (26)
83 (28) 17 (29)
86 (31) 14 (32)
88 (13) 12 (33)
75 (35) 25 (36)
R6 ) BnO (10)
R3 ) H, R4 ) Bn,
R6 ) BnO (24)
R3 ) Bn, R4 ) H,
R6 ) BnO (27)
R3,R4 ) Bn,
2
3
4
5
6
87
41
48
58
75
R6 ) HO (30)
R3 ) H, R4,R6
isopropylidene (12)
R3 ) SEM, R4 ) Bn,
R6 ) H (34)
)
The rate of cycloaddition appears to be a function of
the HOMO-LUMO gap between the glycal and the
diacylthione. In the one case (entry 5 Table 1) where the
gap exceeded 8 eV, no reaction product was detected. On
the other hand, entry 3 (Table 1), with reactants having
the smallest gap, produces the fastest rate. Sustman has
recently demonstrated an excellent correlation of rates
versus the reciprocal of the HOMO-diene/LUMO-di-
enophile difference in a Diels-Alder reaction with normal
polarity.16 The energy gaps computed and the rates
observed for our system are in qualitative agreement
with those reported by Sustman. Our LUMO value for
diacylthione is taken for its minimum energy conforma-
tion. This conformation has a dihedral angle between
the thiocarbonyl and the carbonyl of 50°. One would
guess that the optimum dihedral angle for cycloaddition
would be 0°, where the LUMO is quite a bit lower.
However, the conformational energy of this 0° conforma-
tion is so high that it would not be populated to a
significant extent.
An attempt was made to model the transition state of
the cycloaddition.17 For three model reactions shown in
Figure 1, we were able to locate structures, whose geom-
etries are plotted in the figure, with single imaginary
frequencies; structures we thus assign as transition
states. It can be seen in the simplest case of the addition
of ethylene to thionoglyoxal, the transition state has the
appearance of a normal cycloaddition, with roughly equal
lengths for the forming C-O and C-S bonds. In cases
where the dienophile or the diene are substituted to
resemble the experimental examples, the transition
states are still cycloaddition-like, but the forming C-S
bond is much shorter, really approaching the final bond
length of the product. However, since starting materials
are quite polarized, this unequal bond length is not
unexpected. In the case where we examined the combi-
nation of both a polarized diene and a polarized dieno-
phile, we were unable to locate a cycloaddition transition
state. Our inital geometries usually collapsed to struc-
tures where the C-S bond was essentially fully
a
The remaining organic material is unreacted glycal.
Ta ble 3. Resu lts of Cycloa d d ition s of
3-Th ion op en ta n ed ion e: Solven t Va r ia bles
isolated
yield
entry
1
R3,R4 ) Bn,
R6 ) BnO (10)
(10)
(10)
(10)
CHCl3
95 (11) 05 (23) 83a
2
3
4
5
6
7
DMSO
DMF
THF
72 (11) 28 (23) 59a
68 (11) 32 (23)
86 (11) 14 (23)
b
b
b
(10)
(10)
t-BuOH 77 (11) 23 (23)
MeOH
65 (11) 35 (23) 28c
68 (25) 32 (26) 87a
R3 ) H, R4 ) Bn, CHCl3
R6 ) BnO (24)
8
(24)
DMF
67 (25) 33 (26) 60a
a
b
The remaining material is unreacted glycal. Not isolated,
ratios determined by NMR. c 50% yield of methyl glycosides, see
text.
Table 2 records the effect of protecting group variation
on face selectivity of the cycloaddition of 1 and glucals
where all substituents are equatorial. Table 3 records
the face selectivity outcomes as a function of solvent. In
both tables, only cycloadduct product ratios are recorded.
With one exception, the other components in the reaction
mixture are starting glycal and materials derived from
the thionopentanedione. Notably, the reaction described
in entry 6, Table 3, with methanol as solvent, afforded
two other products, namely, methyl glycosides, the
significance of which will be discussed below.
Discu ssion
In general, substituent and solvent effects on stereo-
selectivity are small. The results in Table 1 suggest that
the obvious influence on the face selectivity of the
cycloaddition is the steric effect of axial substituents on
the glycal. In the ribal series, where the 3-substituent
is not truly axial, the size of the blocking group on the
oxygen has an effect. In the all-equatorial glucal series,
below-plane selectivity is observed. This is the general
rule for attacks on glucal. An apparent small effect of
size variation of the C-3 group from benzyloxy to hydroxy
might be invoked (entry 2, Table 2), but the data in
entries 5 and 6 of Table 2 are not supportive. In Table
3, the solvent effects on face selectivity are small or
(15) van der Wel, G. K.; Wijnen, J . W.; Engberts, J . B. F. N. J . Org.
Chem. 1996, 61, 9001-9005.
(16) Sustman, R.; Siangouri-Feulner, I. Chem. Ber. 1993, 126, 1241-
1245.
(17) The AM1 semiempirical molecular orbital method implemented
by the Spartan software package, version 3.1.3, was used for all
calculations.

6676 J . Org. Chem., Vol. 63, No. 19, 1998
Dios et al.
or be intercepted by methanol. The related intermediate
top-face episulfonium species (not illustrated) and mech-
anism would hold for derived materials 23 and 40. The
slight difference in top/bottom ratios of cycloadducts and
glycosides suggests slight differences between the dia-
stereomers in rates of rearrangement versus methanol
trapping. It is tempting to extrapolate from the methanol
result that our cycloaddition is in fact a two-step process
in every solvent; but at this time, there is no experiment
which can support this generalization.
In summary, a novel cycloadditive functionalization of
glycals by 3-thionopentanedione has been discovered. The
virtue of the new reaction is its mildness and the fact
that it is in a reactivity manifold orthogonal to most
carbohydrate chemistry. The reaction is normally highly
stereoselective. The ease of preparation of the novel
heterodiene has led to the preparation of other thiono
species with similar behavior.19 The heterocycle formed
from the cycloaddition also displays some interesting
chemistry in its own right.20 These newer developments
coupled with the results described here serve to vali-
date the original cycloaddition concept as presented in
Scheme 1.
F igu r e 1. Chem3D+ representations of computed transitions
states for concerted cycloadditions for the reactants (shown
in 2D) with TS bond lengths and product bond lengths (in
parentheses).
Sch em e 2
Exp er im en ta l Section
Gen er a l In for m a tion . 1H and 13C spectra were recorded
in CDCl3 solutions on a at 300 and 75 MHz, respectively. In
some cases, 500 MHz 1H spectra were obtained. The 1H
chemical shifts are reported in ppm downfield from Me4Si. The
13C chemical shifts are reported in ppm relative to the center
line of CDCl3 (77.0 ppm). Optical rotations were recorded on
an automatic polarimeter under standard conditions. Infrared
spectra were recorded on an FTIR. Melting points are uncor-
rected. Medium resolution mass spectra were obtained using
a direct exposure probe and CI (NH3) ionization methods on a
quadropole instrument.
All reactions were performed under an inert atmosphere.
Solvents were dried and distilled prior to use: THF from
sodium/benzophenone ketyl, CH2Cl2 and CHCl3 from P2O5,
CH3CN from CaH2, and t-BuOH from Na. Anhydrous metha-
nol, DMF, DMSO, 2,4-pentanedione, and 2,6-lutidine were
purchased from Aldrich Chemical Co. (Milwaukee, WI). 3,4,6-
Tri-O-benzyl-D-galactal,21 3,4,6-tri-O-benzyl-D-allal,22 and 4,6-
O-isopropylidene-D-glucal23 were prepared using literature
procedures. 3,4-Di-O-benzyl-D-glucal was prepared as de-
scribed by Blackburne.24 Thiophthalimide 2 was prepared as
reported elsewhere.25 Molecular sieves (3A) were purchased
from Aldrich and were powdered and activated prior to use.
Crude products were purified by flash column chromatography
on silica gel (Merck, 230-400 mesh).
formed and the computed solution did not have a single
imaginary frequency. Alternately, the two components
of the cycloaddition simply moved away from each other
to produce very unreasonable structures that were not
transition states. This computational “discontinuity”
from the simpler cases suggested to us that the mecha-
nism might not be a true cycloaddition but might involve
a charged intermediate.
(18) It must be noted that semiempirical MO calculations suggest
that in the pyranose series, the episulfonium species is somewhat less
stable than the ring-opened thio-oxonium form: (a) J ones, D. K.; Liotta,
D. C. Tetrahedron Lett. 1993, 34, 7209-7212. (b) J ones, D. K.; Liotta,
D. Adv. Mol. Model 1995, 3, 67.
(19) (a) Capozzi, G.; Falciani, C.; Menichetti, S.; Nativi, C.; Franck,
R. W. Tetrahedron Lett. 1995, 36, 6755-6758. (b) Win, W. W.; Franck,
R. W. J . Org. Chem. 1997, 62, 4510-4512. (c) Capozzi, G.; Falciani,
C.; Franck, R. W.; Menichetti, S.; Nativi, C.; Raffaelli, B. Chem. Eur.
J ., in press.
Hence, the outcome of our cycloaddition reaction in
methanol as solvent (Table 3, entry 6) is telling. Along
with the bottom- (11) and top-face (23) cycloadducts
shown in the tables, we also obtained methyl glycosides
39 and 40. The ratio of adducts to glycosides before
chromatography is 1/2.5 (Scheme 2). The ratio of bottom/
top cycloadducts is 1.5/1 whereas the ratio of derived
glycosides is 1.25/1. This result suggests that the cy-
cloaddition (for the bottom-face) might proceed via the
epithiosulfonium species 37 or its open oxonium ion
isomer 38.18 Either could then rearrange to cycloadduct
(20) (a) Marzabadi, C. H.; Franck, R. W. J . Chem. Soc., Chem.
Commun. 1996, 2651-2652. (b) Marzabadi, C. H.; Franck, R. W. J .
Org. Chem. 1998, 63, 2197-2208. (c) Capozzi, G.; Mannocci, F.;
Menichetti, S.; Nativi, C.; Paoletti, S. J . Chem. Soc., Chem. Commun.
1997, 2291-2292.
(21) Kozikowski, A. P.; Lee, J . J . Org. Chem. 1990, 55, 863.
(22) Witman, M. D.; Halcomb, R. L.; Danishefsky, S. J .; Golik, J .;
Vyas, D. J . Org. Chem. 1990, 55, 1979.
(23) Fraser-Reid, B.; Walker, D. L.; Tam, S. Y.; Holder, N. L. Can.
J . Chem. 1973, 51, 3950.
(24) Blackburne, I. D.; Fredericks, D. M.; Guthrie, R. D. Aust. J .
Chem. 1976, 29, 381.
(25) (a) Capozzi, G.; Nativi, C.; Menichetti, S.; Rosi, A.; Valle, G.
Tetrahedron 1992, 48, 9023. (b) Reference 19b.

Novel Heterocycloaddition Reaction of Glycals
J . Org. Chem., Vol. 63, No. 19, 1998 6677
Gen er a l P r oced u r e for Cycloa d d ition Rea ction s. 2,6-
Lutidine (1.0 equiv) was added to a suspension of glycal (1.0
equiv), thiophthalimide 2 (1-2 equiv), and powdered, acti-
vated, 3A molecular sieves (typically, a portion equal in weight
to the weight of the combined reactants) in dry solvent (0.5-
1.0 mM). For ribal derivatives, it was necessary to add lutidine
prior to addition of the thiophthalimide 2 in order to prevent
formation of undesired furan derivatives. The mixture was
stirred at room temperature, and the reaction progress was
monitored at intervals by taking an aliquot from the suspen-
O-benzyl-D-galactal (14, 0.48 g, 1.2 mmol), 2 (0.64 g, 2.3 mmol),
and 2,6-lutidine (0.14 mL, 1.2 mmol) in CHCl3 (4.7 mL)
afforded 1-O,2-S-(2-acetyl-1-methyl-1,2-ethenediyl)-3,4,6-tri-
O-benzyl-2-deoxy-2-thio-R-D-galactopyranoside (15) as an oil
(0.46 g, 0.8 mmol, 73%, 4 days) following column chromatog-
raphy (SiO2, 20% EtOAc/petroleum ether): [R]D + 30.3° (c )
0.8; CHCl3); IR (thin film) 1674 cm-1; 1H NMR (CDCl3) δ 7.36-
7.26 (m, 15H), 5.62 (d, 1H, J ) 3.0 Hz), 4.89 (d, 1H, J ) 13.8
Hz), 4.71-4.39 (m, 5H), 4.13-4.07 (m, 1H), 4.00 (d, 1 H, J )
1.2 Hz), 3.73 (dd, 1H, J ) 10.8, 3.0 Hz), 3.60 (dd, 2 H, J ) 6.9,
1.8 Hz), 3.53-3.44 (m, 1H), 2.29 (s, 3H), 2.25 (s, 3H); 13C NMR
(CDCl3) δ 195.4, 159.3, 138.2, 137.7, 137.5, 128.5, 128.3, 128.2,
128.1, 128.0, 127.9, 127.8, 127.7, 127.5, 101.9, 96.7, 75.0, 73.7,
73.6, 73.3, 72.9, 71.8, 68.4, 38.0, 30.1, 29.7; MS m/z 564 (M +
NH4). Anal. Calcd for C32H34O6S: C, 70.31; H, 6.52; S, 5.87.
Found: C, 69.94; H, 6.64; S, 5.87.
1
sion for analysis by H NMR (CDCl3). When the composition
of the reaction mixture showed no additional change (3-7
days), the reaction mixture was quenched with saturated
aqueous NH4Cl, and the mixture was extracted with CH2Cl2
(3×). The organics were dried (Na2SO4), filtered, and concen-
trated in vacuo. The crude mixtures were chromatographed
(SiO2) using gradients of ethyl acetate in petroleum ether.
Removal of residual phthalimide impurities from the chro-
matographed fractions was accomplished by stirring with 0.1
M NaOH, followed by extraction with ethyl acetate.
Ad d u ct 11 fr om 3,4,6-Tr i-O-ben zyl-D-glu ca l (10) w ith
3-(P h th a lim id osu lfen yl)-2,4-p en ta n ed ion e (2). 3-(Phthal-
imidosulfenyl)-2,4-pentanedione (1.32 g, 4.8 mmol, 2.0 equiv)
and tri-O-benzyl-D-glucal (1.01 g, 2.4 mmol, 1 equiv) in 10 mL
of dry CHCl3 under N2 atmosphere with powdered 3A molec-
ular sieves (activated, 0.5 g) plus dry 2,6 lutidine (0.28 mL,
2.4 mmol, 1 equiv) gave 1.10 g (83%) of 1-O,2-S-(2-acetyl-1-
methyl-1,2-ethenediyl)-3,4,6-tri-O-benzyl-2-deoxy-2-thio-R-D-
glucopyranoside (11) plus recovered starting material (tri-
benzyl-D-glucal 0.1555 g, 15.4%): mp 115-116 °C, identical
to that reported in ref 12a; 1H NMR (300 MHz, CDCl3) δ 2.30
(s, 6H), 3.23 (dd, J ) 3.1, and 10.5 Hz, 1H (H2)), 3.59-3.97
(m, 5H), 4.51-4.92 (m, 6H), 5.62 (d, J ) 3 Hz, 1H (H1)), 7.10-
7.40 (m, 15H); 13C NMR (CDCl3) δ 22.5, 31.0, 42.5, 69.0, 73.9,
74.9, 76.5, 77.5, 78.6, 79.5, 97.3, 128-130 (18 C), 138.6, 138.7,
160.5; IR(CDCl3, cm-1) 1678.
Isola tion of Ad d u ct 23 fr om 3,4,6-Tr i-O-ben zyl-D-glu ca l
(10) w ith 3-(P h th a lim id osu lfen yl)-2,4-p en ta n ed ion e (2)-
DMSO a s Solven t. 3-(Phthalimidosulfenyl)-2,4-pentanedione
(2.02 g, 5.3 mmol, 2.0 equiv) and tri-O-benzyl-D-glucal (1.11
g, 2.7 mmol, 1 equiv) in 15 mL of dry DMSO under N2
atmosphere with powdered 3A molecular sieves and dry 2,6-
lutidine (0.32 mL, 2.7 mmol, 1 equiv) over a 7 day period
afforded 241 mg of 1-O,2-S-(2-acetyl-1-methyl-1,2-ethenediyl)-
3,4,6-tri-O-benzyl-2-deoxy-2-thio-â-D-mannopyranoside (23, 17%
of total yield) and 611 mg of 11 (42% of total yield). The total
yield, including top and bottom cycloadduct, was 59%. The
amount of starting material (tri-benzyl-D-glucal) recovered was
0.279 g (25%). For 23: 1H NMR (300 MHz, CDCl3) δ 2.20 (s,
3H), 2.30 (s, 3H), 3.56 (m, J ) 9.3, 3, 1H), 3.65 (dd, J ) 1.2,
4.2, 1H (H2 )), 3.68 (m, J ) 2.4, 1.8, 4.2, 2H), 3.89 (dd, J ) 4.5,
8.4, 1H), 3.96 (dd, J ) 9, 8.7, 1H), 4.4-4.7 (m, 4H), 4.7 (AB
quartet, J ) 10.8, 11.4, 2H), 5.1 (d, J ) 0.9, 1H), 7.1-7.40 (m,
15H); 13C NMR δ 21.4, 29.3, 41.2, 68.9, 71.0,73.5, 73.7, 75.2,
80.2, 92.3, 105.6, 127.5-128.5 (15 C), 137.2-138.2 (3 C), 155.6,
195.5; IR (cm-1) 1679; M/S m/z 564.
Cycloa d d u ct of 3,4,6-Tr i-O-b en zyl-D-a lla l a n d 3-
(P h th a lim id osu lfen yl)-2,4-p en ta n ed ion e (17). 3,4,6-Tri-
O-benzyl-D-allal (16, 0.38 g, 0.9 mmol), 2 (0.5 g, 1.8 mmol),
and 2,6-lutidine (0.1 mL, 0.9 mmol) in CHCl3 (3.8 mL) gave
1-O,2-S-(2-acetyl-1-methyl-1,2-ethenediyl)-3,4,6-tri-O-benzyl-
2-deoxy-2-thio-â-D-altropyranoside 17 (0.34 g, 0.6 mmol, 68%,
3 days) as a white solid following column chromatography
(SiO2, 33-50% EtOAc/petroleum ether): mp 68-71 °C (Et2O/
hexanes); [R]D + 4.0° (c 0.02, CHCl3); IR (thin film) 1678 cm-1
;
1H NMR (CDCl3) δ 7.30-7.25 (m, 15H), 5.65 (d, 1H, J ) 1.8
Hz), 4.65-4.50 (m, 6H), 4.28 (m, 1H), 4.09 (dd, 1H, J ) 7.2,
3.0 Hz), 3.82 (dd, 1H, J ) 5.4, 3.0 Hz), 3.75-3.60 (m, 2H), 3.53
(dd, 1H, J ) 5.7, 1.8 Hz), 2.30 (s, 3H), 2.25 (s, 3H); 13C NMR
(CDCl3) δ 195.1, 157.2, 138.1, 137.7, 137.5, 128.4, 127.9, 127.6,
93.3, 93.2, 76.1, 73.5, 72.8, 72.3, 72.2, 71.8, 69.5, 39.1, 29.6,
21.9; MS m/z 564 (M + NH4). Anal. Calcd for C32H34O6S: C,
70.31; H, 6.52; S, 5.87. Found: C, 70.23; H, 6.44; S, 6.11.
P r ep a r a tion of 3,6- a n d 4,6-Di-O-ben zyl-D-glu ca l. To
a suspension of NaH (1.73 g, 4.3 mmol) in DMF (50 mL) was
added D-glucal (3.16 g, 22 mmol) portionwise as a solution in
DMF (25 mL). After 40 min of stirring at room temperature,
the mixture was cooled to 5 °C and tetra-N-butylammonium
iodide (0.31 g) and benzyl bromide (5.2 mL) were added. The
mixture was slowly warmed to room temperature (2 h) and
was stirred an additional 48 h. The reaction was quenched
with water (10 mL), and the mixture was extracted with
CH2Cl2 (3×, 75 mL). The combined organic fractions were
extracted with additional water (5×, 100 mL) and with brine
(100 mL) to afford after concentration in vacuo a mixture of
glucals. Column chromatography (SiO2; 15-40% ethyl acetate
in petroleum ether) gave as the major products 3,4,6-tri-O-
benzyl-D-glucal (10, 1.65 g) and 4,6-di-O-benzyl-D-glucal (24,
0.96 g): IR (thin film) 3409, 1647 cm-1 1H NMR (CDCl3) δ
;
7.29-7.19 (m, 10H), 6.33 (dd, 1H, J ) 6.0, 0.9 Hz), 4.74-4.48
(m, 5H), 4.28-4.26 (m, 1H), 3.93-3.89 (m, 1H), 3.75-3.74 (m,
1H), 3.61 (dd, 1H, J ) 8.9, 6.3 Hz); 13C NMR δ 144.6, 128.5,
128.4, 127.9, 127.8, 127.7, 102.7, 97.4, 77.3, 76.8, 73.7, 73.6,
69.1, 68.9; MS m/z 344 (M + NH4). Anal. Calcd for C20H22O4:
C, 73.60; H, 6.79. Found: C, 73.52; H, 6.94. 3,6-Di-O-benzyl-
D-glucal26 (27, 0.08 g) was also obtained.
Cycloa d d u ct of 4,6-O-Isop r op ylid en e-D-glu ca l a n d
3-(P h th a lim id osu lfen yl)-2,4-p en ta n ed ion e (13). 4,6-Iso-
propylidene-D-glucal (12, 0.17 g, 0.9 mmol), thiophthalimide
(2, 0.5 g, 1.8 mmol), and 2,6-lutidine (0.1 mL, 0.9 mmol) in
CHCl3 (3.8 mL) gave 1-O,2-S-(2-acetyl-1-methyl-1,2-ethene-
diyl)-4,6-O-isopropylidene-2-deoxy-2-thio-R-D-glucopyrano-
side (13) as a solid (0.15 g, 0.5 mmol, 54%, 5 days) following
column chromatography (SiO2, 33-50% EtOAc in petroleum
ether): mp 142-144 °C (Et2O/hexanes); [R]D + 49.45° (c 0.64,
CHCl3); IR (thin film) 3446, 2994, 1683, 1558, 1374, 1234,
1130, 1038 cm-1; 1H NMR (CDCl3) δ 5.60 (d, 1H, J ) 3.3 Hz),
3.96-3.68 (m, 5H), 3.17-3.12 (m, 1H), 2.63 (s, 1H), 2.35 (s,
6H), 1.55 (s, 3H), 1.44 (s, 3H); 13C NMR δ 195.3, 159.9, 101.8,
100.1, 95.9, 74.4, 67.0, 65.5, 61.8, 42.6, 30.3, 28.9, 19.1, 19.0;
MS (DP/PCI) m/z (rel intensity) 334 (M + NH4) (100), 317 (M
+ H) (82). Anal. Calcd for C14H20O6S: C, 53.15; H, 6.37; S,
10.14. Found: C, 53.09; H, 6.57; S, 10.15.
Rea ction of 4,6-Di-O-ben zyl-D-glu ca l w ith 3-(P h th a l-
im id osu lfen yl)-2,4-p en ta n ed ion e. CHCl3 a s solven t: Glu-
cal 24 (0.087 g, 0.27 mmol), thiophthalimide (2, 0.35 mmol),
and 2,6-lutidine (0.03 mL, 0.27 mmol) in CHCl3 (1.8 mL) after
6 days at room temperature afforded following chromatogra-
phy (SiO2; 25-50% EtOAc in petroleum ether) the R-gluco
cycloadduct 1-O,2-S-(2-acetyl-1-methyl-1,2-ethenediyl)-4,6-tri-
O-benzyl-2-deoxy-2-thio-R-D-glucopyranoside (25) as an oil
(0.08 g, 63%): 1H NMR (CDCl3) δ 7.54-7.21 (m, 10H), 5.21
(d, 1H, J ) 3.3 Hz), 4.83 (d, 1H, J ) 11.1 Hz), 4.67 (d, 1H, J
) 12.3 Hz), 4.58 (d, 1H, J ) 10.8 Hz), 4.55 (d, 1H, J ) 12.3
Hz), 3.95 (dd, 1H, J ) 6.9, 2.1 Hz), 3.81 (dd, 1H, J ) 11.1, 3.3
Hz), 3.74-3.70 (m, 2H), 3.13 (dd, 1H, J ) 10.2, 3.0 Hz), 2.69
(s, 1H), 2.27 (s, 3H), 2.05 (s, 3H); MS (DEP/CI) m/z 474 (M +
NH4) 344. The â-manno cycloadduct 1-O,2-S-(2-acetyl-1-
Cycloa d d u ct of 3,4,6-Tr i-O-b en zyl-D-ga la ct a l a n d 3-
(P h th a lim id osu lfen yl)-2,4-p en ta n ed ion e (15). 3,4,6-Tri-
(26) Danishefsky, S. J .; Behar, V.; Randolph, J . T.; Lloyd, K. O. J .
Am. Chem. Soc. 1995, 117, 5701.

6678 J . Org. Chem., Vol. 63, No. 19, 1998
Dios et al.
methyl-1,2-ethenediyl)-4,6-tri-O-benzyl-2-deoxy-2-thio-â-D-man-
nopyranoside (26) was also obtained as a white powder (0.03
g, 24%): 1H NMR (CDCl3) δ 7.29-7.16 (m, 10H), 5.17 (d, 1H,
J ) 0.9 Hz), 4.66-4.49 (m, 4H), 3.99-3.97 (m, 2H), 3.78-3.75
(m, 2H), 3.56-3.53 (m, 2H), 2.27 (s, 3H), 2.24 (s, 3H); MS m/z
474 (M + NH4).
DMF a s solven t : Glucal 24 (0.05 g, 0.15 mmol), thio-
phthalimide (2, 0.15 mmol), and 2,6-lutidine (0.02 mL, 0.15
mmol) in DMF (0.9 mL) were stirred at room temperature for
5 days. Column chromatography on the crude reaction
mixture using conditions described above afforded recovered
glycal 24 (0.01 g), R-gluco cycloadduct 25 (0.03 g, 44%), and
â-manno cycloadduct 26 (0.01 g, 16%).
((((trimethylsilyl)ethyl)oxy)methyl)-4-O-benzyl-2,6-dideoxy-2-
thio-R-D-glucopyranoside (35) and 43.6 mg (19%) of 1-O,2-S-
(2-acetyl-1-methyl-1,2-ethenediyl)-3-((((trimethylsilyl)ethyl)oxy)-
methyl)-4-O-benzyl-2,6-dideoxy-2-thio-â-D-mannopyranoside (36).
For 35 (bottom-face diastereomer): 1H NMR (300 MHz,
CDCl3) δ 0.00 (s, 3H), 0.02 (s, 3H), 0.03 (s, 3H), 0.92 (m, 2H),
1.34 (d, J ) 6.3 Hz, 3H), 2.32 (s, 3H), 2.35 (s, 3H), 3.14 (dd, J
) 3, 10.5 Hz, 1H) 3.23 (dd, J ) 9.3, 8.1 Hz, 1H), 3.65 (dd, J )
10.8, 10.5 Hz, 1H), 3.66-3.80 (m, 3H), 3.95 (q, J ) 6.4 Hz,
1H), 4.7, 4.9 (AB quartet, J ) 10.8, 11.1 Hz, 2H), 4.97 (s, 2H),
5.57 (dd, J ) 3.3 Hz, 1H), 7.3-7.4 (m, 5H); 13C NMR δ -1.2,
0.2, 18.1, 18.2, 21.8, 30.4, 42.3, 66.8, 69.7, 75.2, 75.6, 76.6,
77.95, 84.5, 96.0, 96.6, 102.5, 128.0, 128.2, 128.7, 137.95, 159.7,
195.4; IR (cm-1) 3036, 2966, 2907, 1666, 1555, 1355, 1249,
1132, 1055, 932, 826, 749, 691; MS m/z 498; [R ]D (c 0.11,
CHCl3) +120.2°. Anal. Calcd for C24H36O6SSi: C, 60.0; H, 7.5;
S, 6.7. Found: C, 59.32; H, 7.59; S, 6.43. For 36 (top-face
diastereomer): 1H NMR (300 MHz, CDCl3) δ 0.00 (s, 6H), 0.04
(s, 3H), 0.96 (m, 2H), 1.4 (d, J ) 6 Hz, 3H), 2.33 (s, 3H), 2.35
(s, 3H), 3.5-4.0 (m, 4H), 4.1 (dd, J ) 4.5, 4.5 Hz, 1H), 4.64
(AB quartet, J ) 10.8 Hz, 1H), 4.7-5.0 (m, 3H), 5.25 (dd, J )
1.2 Hz, 1H), 7.3-7.4 (m, 5H).
R ea ct ion of 3,4,6-Tr i-O-b en zyl-D-glu ca l (10) w it h 3-
(P h th a lim id osu lfen yl)-2,4-p en ta n ed ion e (2). Meth a n ol
a s th e solven t: A sample of glucal 10 (0.41 g, 0.98 mmol),
thiophthalimide (2, 1.2 mmol), and 2,6-lutidine (11 mL, 0.98
mmol) in dry MeOH (6 mL) after 3 days showed by 1H
NMR (CDCl3) complete consumption of tri-O-benzyl-D-glucal
and a mixture of R-gluco cycloadduct 11, â-manno cycload-
duct 23, â-gluco methyl glycoside 39, and R-manno methyl
glycoside 40 (1.5:1:3.5:2.8).The reaction mixture was worked
up as previously described and repeatedly chromatographed
(SiO2; 10% EtOAc/petroleum ether) to afford 11 and 39 in
purified form, whereas 23 and 40 were present as compo-
nents of mixtures: R-gluco cycloadduct 11 (0.11 g, 21%), â-
manno cycloadduct 23 (0.04 g, 6.5%), â-gluco methyl glycoside
39 (0.16 g, 29%), and R-manno methyl glycoside 40 (0.12 g,
21%).
The diastereomeric cycloadducts 25 and 26 were benzylated
under standard conditions27 and afforded derivatives with
properties consistent with tri-O-benzylated adducts 11 and 23.
Rea ction of 3,6-Di-O-ben zyl-D-glu ca l w ith 3-(P h th a l-
im id osu lfen yl)-2,4-p en ta n ed ion e. Glucal 27 (0.07 g, 0.23
mmol), thiophthalimide (2, 0.15 mmol), and 2,6-lutidine (0.03
mL, 0.23 mmol) in CHCl3 (1.2 mL) were stirred at room
temperature for 4 days. Column chromatography on the crude
reaction mixture (SiO2; 20-30% EtOAc in petroleum ether)
gave recovered glycal 27 (0.02 g, 19%), R-gluco cycloadduct
1-O,2-S-(2-acetyl-1-methyl-1,2-ethenediyl)-3,6-tri-O-benzyl-2-
deoxy-2-thio-R-D-glucopyranoside (28, 0.02 g, 34%) as an oil,
and â-manno cycloadduct 1-O,2-S-(2-acetyl-1-methyl-1,2-
ethenediyl)-3,6-tri-O-benzyl-2-deoxy-2-thio-â-D-mannopyrano-
side (29, 0.005 g, 7%) as an oil. For 28: 1H NMR (CDCl3) δ
7.31-7.19 (m, 10H), 5.52 (d, 1H, J ) 3.0 Hz), 4.90 (d, 1H, J )
11.1 Hz), 4.70 (d, 1H, J ) 11.1 Hz), 4.56 (d, 1H, J ) 12.3 Hz),
4.54 (d, 1H, J ) 12.0 Hz), 3.85 (dd, 1H, J ) 9.0, 4.5 Hz), 3.75-
3.63 (m, 3H), 3.40 (dd, 1H, J ) 9.6, 9.3 Hz), 3.11 (dd, 1H, J )
10.8, 3.0 Hz), 2.26 (s, 6H); MS m/z 474 (M + NH4). For 29:
1H NMR (CDCl3) δ 7.40-7.26 (m, 10H), 5.22 (d, 1H, J ) 1.2
Hz), 4.84 (d, 1H, J ) 11.7 Hz), 4.65-4.55 (m, 3H), 4.14-4.11
(m, 1H), 3.82-3.77 (m, 2H), 3.70 (dd, 1H, J ) 4.5, 1.2 Hz),
3.64 (dd, 1H, J ) 9.6, 4.5 Hz), 2.78 (d, 1H, J ) 2.1 Hz), 2.35
(s, 3H), 2.30 (s, 3H); MS m/z 474 (M + NH4). The diastereo-
meric cycloadducts 28 and 29 were benzylated under standard
conditions and afforded derivatives with properties consistent
with tri-O-benzylated adducts 11 and 23.
Rea ction of 3,4-Di-O-ben zyl-D-glu ca l w ith 3-(P h th a l-
im id osu lfen yl)-2,4-p en ta n ed ion e. A mixture of glucal 30
(0.09 g, 0.28 mmol), thiophthalimide (2, 0.28 mmol), and 2,6-
lutidine (0.033 mL, 0.28 mmol) in CHCl3 (1.6 mL) was stirred
for 3 days at room temperature. Column chromatography on
the crude product mixture (20-40% EtOAc in petroleum ether)
gave starting glycal 30 (0.02 g) and then R-gluco cycloadduct
1-O,2-S-(2-acetyl-1-methyl-1,2-ethenediyl)-3,4,-tri-O-benzyl-2-
deoxy-2-thio-R-D-glucopyranoside (31, 0.054 g, 42%) as a clear
oil: 1H NMR (CDCl3) δ 7.31-7.18 (m, 10H), 5.52 (d, 1H, J )
3.0 Hz), 4.86-4.76 (m, 3H), 4.63 (d, 1H, J ) 10.8 Hz), 3.82-
3.75 (m, 3H), 3.61 (ddd, 2H, J ) 19.2, 9.6, 9.0 Hz), 3.11 (dd,
1H, J ) 9.9, 3.0 Hz), 2.24 (s, 6H); MS m/z 474 (M + NH4). The
â-manno cycloadduct, 1-O,2-S-(2-acetyl-1-methyl-1,2-ethene-
diyl)-3,4-tri-O-benzyl-2-deoxy-2-thio-â-D-mannopyranoside (32,
0.01 g, 6%) was also obtained as an inseparable mixture with
the major cycloadduct 31: 1H NMR (CDCl3) δ 5.23 (d, 1H, J 1,2
) 1.2 Hz) (C1-H). The mixture of diastereomeric cycloadducts
31 and 32 was benzylated under standard conditions and
afforded derivatives with properties consistent with tri-O-
benzylated adducts 11 and 23.
Met h yl 3,4,6-Tr i-O-b en zyl-2-d eoxy-2-t h io-3-(p en t a n e-
2,4-d ion e)-â-D-glu cop yr a n ose (39): IR (thin film) 1574 cm-1
:
1H NMR (CDCl3) δ 7.41-7.16 (m, 15H), 4.91-4.78 (m, 3H),
4.64-4.52 (m, 3H), 4.34 (d, 1H, 8.3 Hz), 3.73-3.63 (m, 4H),
3.49-3.37 (m, 1H), 3.43 (s, 3H), 2.72 (dd, 1H, J ) 8.4, 13.5
Hz), 2.40 (s, 6H); MS (DEP/CI) m/z 596 (M + NH4).
Rea ction of th e Tr i-O-ben zylglu cose r-Glu co Cycloa d -
d u ct (11) w it h Th iop h t h a lim id e (2) a n d 2,6-Lu t id in e.
Meth a n ol a s solven t. Cycloadduct 11 (0.05 g, 0.08 mmol),
thiophthalimide (2, 0.08 mmol), and 2,6-lutidine (0.01 mL, 0.08
mmol) in MeOH (0.5 mL) gave only recovered 11 after being
allowed to react for 3 days.
Rea ction of 3-((ter t-Bu tyld im eth ylsilyl)oxy)-4-m eth -
oxym eth yl-D-r iba l w ith 3-(P h th a lim id osu lfen yl)-2,4-p en -
tan edion e. Major adduct (top-face) 1-O,2-S-(2-acetyl-1-methyl-
1,2-ethenediyl)-3-((tert-butyldimethylsilyl)oxy)-4-methoxymethyl-
2-deoxy-2-thio-â-D-arabinofuranoside (22â): 1H NMR CDCl3
δ 7.68-7.34 (m, 10H, Ph), 5.55 (d, J ) 4.89 Hz, 1H, C1′), 4.33
(d, J ) 2.32, 2H, -OCH2O-), 4.22-4.18 (m, 1H, C4′), 4.08 (dd,
J ) 7.99, 2.05 Hz, 1H, C3′), 3.69 (dd, J ) 8.02, 3.16 Hz, 1H,
C2′), 3.29-3.03 (abq, J ) 11.08, 6.27, 2.91 Hz, 2H, C5′,C5′′), 3.18
(s, 3H, -OCH3), 2.15 (s, 3H, -CH3), 2.11 (s, 3H, -CH3), 1.07
(s, 9H, t-Bu); 13C NMR (CDCl3) δ 195.02, 161.39, 135.95,
135.76, 132.91, 132.52, 130.07, 127.76, 127.68, 103.44, 100.62,
96.37, 85.45, 73.69, 67.46, 55.06, 48.20, 29.49, 26.84, 22.02,
19.19. Anal. Calcd: C, 63.60; H, 6.86; S, 6.06. Found: C,
63.38; H, 6.87; S, 6.17.
Rea ction of 3-((((Tr im eth ylsilyl)eth yl)oxy)m eth yl)-4-
O-ben zyl-D-r h a m n a l w ith 3-(P h th a lim id osu lfen yl)-2,4-
p en t a n ed ion e. 3-(Phthalimidosulfenyl)-2,4-pentanedione
(0.3928 g, 0.88 mmol, 2.0 equiv) in 2.2 mL of dry CHCL3 under
N2 atmosphere with powdered 4A molecular sieves (activated,
0.144 g) in a dry 25-mL flask plus rhamnal (34, 0.1516 g, 0.43
mmol, 1 equiv) and dry 2,6-lutidine (0.05 mL, 0.43 mmol, 1
equiv) gave, after 24 h, 0.1531 g (75%) of a mixture of 109.2
mg (56%) of 1-O,2-S-(2-acetyl-1-methyl-1,2-ethenediyl)-3-
Minor adduct (bottom-face) 1-O,2-S-(2-acetyl-1-methyl-1,2-
ethenediyl)-3-((tert-butyldimethylsilyl)oxy)-4-methoxymethyl-
2-deoxy-2-thio-R-D-ribofuranoside (22r): 1H NMR CDCl3
δ
7.82-7.35 (m, 10H, Ph), 5.42 (d, J ) 3.79 Hz, 1H, C1′), 4.36-
4.28 (m, 3H, -OCH2O-, C4′), 4.21 (bs, 1H, C3′), 3.51 (dd, J )
5.78, 1.91 Hz, 1H, C2′), 3.17-2.45 (AB quartet, J ) 12.0, 5.68,
3.17 Hz, 2H, C5′, C5′′), 3.13 (s, 3H, -OCH3), 2.37 (s, 3H, CH3),
2.35 (s, 3H, -CH3), 1.03 (s, 9H, t-Bu); 13C NMR (CDCl3) δ
195.14, 159.61, 136.57, 136.37, 135.74, 135.43, 135.38, 134.01,
132.99, 130.59, 130.46, 130.03, 128.20, 127.94, 105.47, 98.58,
(27) Czernecki, S.; Georgoulis, C.; Provelenghiou, C. Tetrahedron
Lett. 1976, 3535.

Novel Heterocycloaddition Reaction of Glycals
J . Org. Chem., Vol. 63, No. 19, 1998 6679
96.92, 88.81, 73.83, 67.03, 55.57, 44.28, 30.26, 27.24, 23.12,
19.55. Anal. Calcd: C, 63.60; H, 6.86; S, 6.06. Found: C,
63.44; H, 6.95; S, 6.32.
2.32 (s, 3H, CH3); 13C NMR (CDCl3) δ 195.65, 161.33, 103.86,
100.51, 96.86, 83.86, 72.98, 68.09, 55.41, 47.69, 29.87, 22.23.
To further establish their identity, both isomers of 20 were
silylated under standard conditions to yield the corresponding
isomers of 21.
Rea ction of 4-Meth oxym eth yl-D-r iba l w ith 3-(P h th a l-
im id osu lfen yl)-2,4-p en ta n ed ion e. Major adduct (bottom-
face) 1-O,2-S-(2-acetyl-1-methyl-1,2-ethenediyl)-4-methoxy-
methyl-2-deoxy-2-thio-â-D-ribofuranoside (20r): 1H NMR CDCl3
δ 5.48 (d, J ) 3.78 Hz, 1H, C1′), 4.63 (s, 2H, -OCH2O), 4.48
(q, 2.23 Hz, 1H, C4′), 4.28-4.23 (bdd, 1H, C3′), 3.76 (d, J )
3.56 Hz, 2H, C5′,C5′′), 3.73 (dd, J ) 3.82, 0.92 Hz, 1H, C2′), 3.36
(s, 3H, -OCH3), 2.41 (s, 3H, CH3), 2.23 (s, 3H, CH3); 13C NMR
(CDCl3) δ 194.63, 160.11, 106.01, 98.42, 96.79, 86.61, 71.78,
67.35, 55.41, 45.38, 30.16, 22.46.
Minor adduct (top-face) 1-O,2-S-(2-acetyl-1-methyl-1,2-
ethenediyl)-4-methoxymethyl-2-deoxy-2-thio-â-D-arabinofura-
noside (20â): 1H NMR CDCl3 δ 5.54 (d, J ) 4.85 Hz, 1H, C1′),
4.66 (s, 2H, -OCH2O), 3.69-3.68 (m, 2H, C3′, C4′), 3.61 (d, J
) 4.39 Hz, 2H, C5′, C5′′), 3.60 (dd, J ) 8.81, 3.96 Hz, 1H, C2′),
3.38 (s, 3H, -OCH3), 2.98 (bd, 1H, C3′-OH), 2.35 (s, 3H, CH3),
Ack n ow led gm en t. The work has been supported at
Hunter by the NIH (GM 51216 and RR03037). Also NY
State grants from the GRI and HEAT initiatives funded
NMR equipment and computer hardware used in this
project. We are grateful to all of these agencies. We are
indebted to Professor G. Capozzi and his group at the
University of Florence (Italy) for generously sharing
their results on cycloadditions of diacylthiones and for
allowing us to develop their pioneering discovery of the
cycloaddition which produced adduct 11.
J O981047U