Ring-closing olefin metathesis of 3,3′-(D-glycopyranosylidene)bis(1-propene) compounds as a route to anomeric spiro sugars

Article abstract of DOI:10.1002/1099-0690(200108)2001:15<2939::AID-EJOC2939>3.0.CO;2-H

New peracetylated 3,3′-(D-glycopyranosylidene)bis(1-propene), prepared from peracetylated anomeric sugar dihalides and allyltributyltin in excess under UV irradiation conditions (AIBN, ca. 35 °C), lent themselves to ring-closing olefin metathesis in the presence of Grubbs' catalyst (6 mol %) to afford the corresponding peracetylated spiro[1,5-anhydro-D-glycopyranositol-1,4′-cyclopent-1′-enes] (D-gluco: 81%, D-manno: 89%, D-galacto: 72% isolated yield), which could be deacetylated quantitatively. The bis(allylation) reaction (D-gluco: 40%, D-manno: 34%, D-galacto: 24% isolated yield) was in competition with radical-induced rearrangement and elimination reactions. The last process, found to be favored at higher temperature (80 °C), opens an easy route to sugar dienes difficult to prepare otherwise.

Full text of DOI:10.1002/1099-0690(200108)2001:15<2939::AID-EJOC2939>3.0.CO;2-H

FULL PAPER
Ring-Closing Olefin Metathesis of 3,3Ј-(
D
-Glycopyranosylidene)bis(1-propene)
Compounds as a Route to Anomeric Spiro Sugars
Guo-Rong Chen,[a] Zhong Bo Fei,[a] Xiao-Ting Huang,[a] Yu-Yuan Xie,[a] Jin-Lou Xu,[a]
[b]
[b]
[b]
¨
Joelle Gola, Michaela Steng, and Jean-Pierre Praly*
[‡]
´
Dedicated to Gerard Descotes (University of Lyon, France)
Keywords: Allylation / Carbohydrates / Eliminations / Radical reactions
New peracetylated 3,3Ј-(
pene), prepared from peracetylated anomeric sugar dihalides
and allyltributyltin in excess under UV irradiation conditions
D
-glycopyranosylidene)bis(1-pro-
89%,
acetylated quantitatively. The bis(allylation) reaction (
gluco: 40%, -manno: 34%, -galacto: 24% isolated yield)
D-galacto: 72% isolated yield), which could be de-
D-
D
D
(AIBN, ca. 35 °C), lent themselves to ring-closing olefin meta- was in competition with radical-induced rearrangement and
thesis in the presence of Grubbs’ catalyst (6 mol %) to afford elimination reactions. The last process, found to be favored
the corresponding peracetylated spiro[1,5-anhydro- -glyco- at higher temperature (80 °C), opens an easy route to sugar
D
pyranositol-1,4Ј-cyclopent-1Ј-enes] (
D
-gluco: 81%,
D
-manno:
dienes difficult to prepare otherwise.
Introduction
basaccharides,[19,20] azasugars,[21] and spiroacetal derivat-
ives.[22] While this work was in progress, a multi-step
synthesis of a gem-diallyl sugar from a glyconolactone, with
the aid of an allyl Grignard reagent, and its successful
transformation into the corresponding spirocyclopentene
by RCM was reported.[23] We describe here a related ap-
proach, applicable to acetylated sugar derivatives.
In a previous paper, we reported on the stereocontrolled
synthesis of 3-(β--glycopyranosyl)-1-propenes and on ac-
cess to unknown -glycopyranosylidenedienes,[1] based on
free radical reactions applied to anomeric glycosyl dihal-
ides, such as 1Ϫ3. It was of interest to consider treatment
of these with an excess of allyltributyltin, as a possible route
to 3,3Ј-(-glycopyranosylidene)bis(1-propene) derivatives
(‘‘C,C-diallyl glycosides’’). Indeed, such compounds had
been isolated in trace amounts in the course of our previous
investigation.[1] Considering that there are only a few scat-
tered synthetically useful routes, based either on free rad-
ical[2] or on ionic[3,4] approaches to C,C-disubstituted glyco-
pyranosyl compounds, while access to C,C-spirocyclic gly-
cosides[5,6] often involves carbene intermediates,[7Ϫ10] we
anticipated that radical-based allylation of these glycosyl di-
halides would be a useful and straightforward method ap-
plicable to the synthesis of acetylated 3,3Ј-(-glycopyranos-
ylidene)bis(1-propenes). Such compounds appeared ideally
suited for ring-closing olefin metathesis (RCM)[11Ϫ13] to af-
ford new unsaturated spiro sugar derivatives. Indeed, recent
interest in olefin metathesis has inspired varied develop-
ments in carbohydrate chemistry,[14] in particular con-
cerning C-glycosyl compounds,[15] C-linked disacchar-
ides,[16] C-1 glycals,[17] C-glycosylidene compounds,[18] car-
Results and Discussion
Radical-mediated allylation of the -gluco-chlorobromo
sugar 1 was first carried out at ca. 35 °C with an excess of
allyltributyltin (6 equiv.) under UV irradiation conditions
for ca. 3 h with added AIBN. From several attempts with
minor changes in the experimental technique (see Table 1
and Exp. Sect.), it turned out that these conditions pro-
duced a mixture of compounds, separable by careful col-
umn chromatography and eluted in the following order: 7,
9, 12,[1] then 16[1] (Scheme 1). Whereas diene 7 was formed
in small amounts, as observed previously,[1] the desired
compound 9 was obtained in yields not exceeding 40%.
Radical-mediated allylation of 1 was repeated as before, ex-
cept that the temperature was lowered to approximately ca.
0 °C. The diene 7 was formed only in trace amounts at 0
°C, while the (-glucopyranosylidene)bis(1-propene) 9 was
isolated in 24% yield, after repeated chromatographic sep-
arations. The acetylated (β--glucopyranosyl)-1-propene
12, containing an unidentified impurity (ca. 2:1 at 0 °C; ca.
3:1 at 35 °C), was also isolated in 10% yield, while a fourth
compound (8% yield) was identified as 15. The reaction was
also attempted using a 250-W IR sun lamp instead of a
medium-pressure mercury lamp, so that a significantly
higher reaction temperature was reached (80 °C). Unexpec-
tedly, the main compound formed at 80 °C in the presence
[a]
East China University of Sciences and Technology,
P. O. Box 257, 130 Meilong Road, 200237, Shanghai, P. R.
China
[b]
´
´
Universite Claude-Bernard Lyon I, UMR CNRS Ϫ Universite
ˆ
no. 5622, ESCPE Ϫ Lyon, Bat. 308,
43, Boulevard du 11 Novembre 1918, 69622 Villeurbanne,
France
Fax: (internat.) ϩ 33-4/78898914
E-mail: jean-pierre.praly@univ-lyon1.fr
For his pioneering work in synthetic applications of free-radical
reactions and metathesis in carbohydrate chemistry
[‡]
Eur. J. Org. Chem. 2001, 2939Ϫ2946
WILEY-VCH Verlag GmbH, 69451 Weinheim, 2001
1434Ϫ193X/01/0815Ϫ2939 $ 17.50ϩ.50/0
2939

J.-P. Praly et al.
FULL PAPER
Table 1. Allylation of -glucopyranosyl halides 1 and 4: product distribution under various reaction conditions
Entry
Substrate
Conditions[a]
Products
yields (%)[b]
7
T [°C]
Bu3Sn(allyl)[c]
AIBN[d]
9
12
15
16
19
1
2
3
4
5
6
7
8
1
1
1
1
1
4
4
4
ca.
0
2 ϩ 4
2 ϩ 4
4 ϩ 2
6
ϩ
ϩ
ϩ
ϩ
ϩ
ϩ
Ϫ
ϩ
trace
6
15
5
64
80
Ϫ
24
40
24
37
10
Ϫ
10
12
17
11
7
Ϫ
Ϫ
Ϫ
8
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
2
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
63
60
ca. 35
ca. 35
ca. 35
ca. 80
ca. 80
ca. 80
ca. 80
19
14
Ϫ
Ϫ
Ϫ
Ϫ
2 ϩ 3
4
4
1
Ϫ
Ϫ
Ϫ
[a]
The reaction flask was exposed either to unfiltered UV light (reactions performed at ca. 0 °C and ca. 35 °C) or to visible light from a
sunlamp (reactions performed at ca. 80 °C), as described in more detail in the Exp. Sect. Ϫ [b] Isolated yields. Ϫ [c] Allyltri-n-butyltin was
added either all at once or portionwise, as indicated (in equiv. amounts). Ϫ
omitted (Ϫ).
[d]
AIBN was either used in catalytic amounts (ϩ) or
of AIBN was the diene 7, isolated in 64% yield, while 9 was
present in small amounts. As seen from Table 1, use of pure
4 instead of 1 produced 7 in an enhanced yield (80%). Inter-
estingly, when 4 was treated at 80 °C under modified condi-
tions (allyltributyltin/4 equiv., no AIBN; allyltributyltin/1
equiv., catalytic AIBN), the known conjugated glucosylid-
enediene 19[1] was formed as the main product (ca. 60%
yield). A similar observation was made in the -galacto
series, in which the -galacto-configured analog[1] 20 (73%
yield) could be obtained from 6 in the absence of AIBN.
These trials showed that competing pathways cannot be
avoided by changes in the reaction temperature. Bis-
(allylation) of 2 and 3 was therefore carried out under
photolytic conditions at ca. 35 °C and, as in the -gluco
series, several compounds were observed: 10, 13, and 17 (-
manno: 34, 1, and 11% isolated yields respectively); 8[1] and
11 (-galacto: 7 and 24% isolated yields, respectively; other
minor components such as 14 and 18[1] were not isolated).
TLC in each case showed significant amounts of unidenti-
fied polar products visible as tails.
While radical-based reduction of the intermediate chlor-
ides 4Ϫ6 and their hydroxylation (due either to hydrolysis
caused by moisture or to reaction of encumbered sugar-
derived radicals with remaining molecular oxygen[24]) can
Scheme 1
explain the occurrence of such products as 3-(β--glycopyr-
anosyl)-1-propene 12Ϫ14 and non-4-ulopyranose derivat- lylation was carried out at room temperature with photo-
ives 16Ϫ18,[1] respectively, the formation and structures of chemical initiation.[27] Since migration of the acetoxy group
7, 8, and 15 deserve some comment. Radical routes to 1,2- from C-5 to C-4 (see Schemes 1 and 3 for numbering of
unsaturated sugars (glycals),[24a] which might explain the compounds 4Ϫ15, and 21Ϫ25) took place under our condi-
formation of 7 and 8, have been found to be efficient for tions, as evidenced by the formation of 15, 1,2-elimination
substrates possessing two vicinal functional groups suscept- might presumably occur for both intermediate radicals 4R
ible to attack by triorganotin radicals, at C-1 and C-2.[25]
A
and 4ЈR (Scheme 2) in the presence of tributyltin radicals,
few examples have demonstrated the formation of unsatur- efficiently producing 7 upon heating (64% from 1, 80% from
ated sugar derivatives by radical-mediated reductive 1,2- 4). It is conceivable that photostimulated electron transfer
elimination involving acetoxy[26] and pivaloyloxy[27] groups. might convert long-lived radical 4R into the corresponding
In particular, attempted radical reduction and allylation of anion, prone to acetoxy group elimination. Of the tri-n-bu-
bromonucleoside derivatives upon treatment either with tri- tyltin halides, formed as common by-products in radical re-
butyltin hydride or with allyltributyltin at 80 °C produced actions, the iodide in particular may participate, as a Lewis
unsaturated by-products (Ͻ 12% yield) as a result of 1,2- acid, in unexpected transformations.[28] Since 4 selectively
elimination of a bromine atom and a pivaloyloxy group. gives 7 (Table 1) upon heating to 80 °C with allyltri-n-bu-
Such a 1,2-elimination was suppressed when the radical al- tyltin and AIBN, and hence in the absence of tri-n-butyltin
2940
Eur. J. Org. Chem. 2001, 2939Ϫ2946

Ring-Closing Olefin Metathesis of 3,3Ј-(D-Glycopyranosylidene)bis(1-propene)
FULL PAPER
bromide, its participation in the 1,2-elimination process giv- radical,[35b] respectively. To the best of our knowledge,[24a]
ing 7 from 1 appears improbable. Under conditions which formation of 15 is the first example of acetoxy group migra-
do not favor formation of tributyltin radicals (no AIBN tion towards a substituted -glucopyranos-1-yl radical.
added or quantity of allyltributyltin reduced to 1 equiv., 80
Ring-closing olefin metathesis of compounds 9Ϫ11 was
°C), isolation of the conjugated sugar-derived dienes 19 (ca. achieved uneventfully, with Grubbs’ catalyst used in cata-
60%) and 20 (73%), as major products formed upon at- lytic amounts (6 mol %) in dichloromethane as the solvent
tempted allylation of 4 and 6 respectively, shows 1,2-elim- and under argon. TLC monitoring showed completion of
ination of hydrochloric acid to be the preferred process. the reactions within ca. 7 h when stirred at room temper-
Elimination of HCl was shown to occur in 4 upon treat- ature, with formation of a single product. The catalyst in
ment with 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) to give the resulting dark solutions was decomposed in the pres-
19,[1] and in a related chlorosugar intermediate produced in ence of air[36] prior to product purification by column chro-
boiling benzene solution.[29]
matography. Spiro sugars 21Ϫ23 were obtained in 81, 89,
and 72% yields, respectively, in good agreement with pub-
lished results.[23] Deacetylation of 21 and 22 was achieved
quantitatively in
a
mixture of MeOH/H2O/NEt3
(Scheme 3).
Scheme 2
On the basis of well-established literature data, the re-
arranged structure of 15 is the result of radical-induced
acetoxy group migration, possibly after homolysis of the
C4ϪCl bond in 4, followed by allylation of the carbon-cent-
ered radical formed at C-5 (4ЈR in Scheme 2). Such re-
arrangements of glycos-1-yl radicals[30,31] have been ex-
ploited as a synthetically useful route[24a] to 2-deoxy
sugars,[32] based on a cis-2,1-migration of acetoxy,[32,33]
chloroacetoxy,[33a] benzoyloxy,[33a] or phosphate groups,[34]
preferably in equatorial orientations at C-2. It was shown
that equatorially oriented groups at C-2 (-gluco, -galacto,
Scheme 3
The 13C NMR resonances of C-3 and C-3Ј in 9 were as-
-xylo series) result in α-configured derivatives gaining sta- signed by comparison with literature values reported for 3-
bilization through the anomeric effect, whereas β-config- (2,3,4,6-tetra-O-acetyl-α--glucopyranosyl)-1-propene and
ured rearranged products are formed in the -manno series the β-anomer 12.[37] The shieldings observed for C-3Ј, C-6,
at much lower rates because of the weaker stabilization of and C-8 in the 3,3Ј-(-glucopyranosylidene)bis(1-propene)
products: No migration of C-2 acetoxy groups has been ob- compound 9, and for the corresponding atoms in 3-(2,3,4,6-
served for -manno configured sugars.[33a,34b] In keeping tetra-O-acetyl-α--glucopyranosyl)-1-propene can be ex-
with these data, no rearranged products were observed plained on the basis of the γ-gauche effect (steric effect),[38]
when the -manno-configured product 10 was prepared defined as ‘‘steric interactions, mostly arising from touching
from 2, which, according to TLC and the isolated product or overlapping of van der Waals radii of closely spaced
distribution (10: 34%; 13: 1%; 17: 11%), reacted at ca. 35 hydrogens, usually cause a shielding of the carbon atoms
°C with a somewhat higher selectivity than 1. Compound attached to these hydrogens’’. On the basis of the signals
15 was assigned a -gluco configuration on the basis of the assigned to C-3 and C-3Ј, the four allylic hydrogen atoms
large vicinal coupling constant between 5-H and 6-H in 9 could be assigned by HSQC correlation: The allylic
(J5,6 ϭ 10.7 Hz), showing their trans-diaxial orientation. hydrogen atoms on the axially oriented branch were found
The C-5 epimer of 15 (-manno configuration) was not to be deshielded in comparison to those on the equatorial
isolated but it may correspond to the unidentified com- allyl group, as was also observed for 12 and its α-anomer.[37]
pound found in admixture with 12 (see above and Exp. Structure assignment for 7, 8Ϫ11, 15, and 21Ϫ25 was based
Sect.). It was probably formed only in trace amounts, in on 2D homonuclear and heteronuclear correlations. For 21,
keeping with the 9:1 and 6:4 diastereoselectivities in favor NOESY correlation allowed unambiguous assignments of
of the -gluco-branched sugars[35] found for allyltributyltin- C-3Јand C-5Ј, the resonances of which can once more be
mediated allylation of 1,3,4,6-tetra-O-acetyl-2-bromo-2-de- interpreted on the basis of the γ-gauche effect. Assignments
oxy-β--glucopyranose (43% total yield),[35a] and methyl proposed for C-3, C-3Ј, and C-5Ј in 10, 11, and 22Ϫ25 fol-
3,4,6-tri-O-acetyl-2-deoxy-α--arabino-hexopyran-2-yloside low those determined for 9 and 21. Homonuclear vicinal
Eur. J. Org. Chem. 2001, 2939Ϫ2946
2941

J.-P. Praly et al.
FULL PAPER
propene) (9), and 4,6,7,9-Tetra-O-acetyl-1,2,3,5-tetradeoxy-5-C-al-
lyl-α-
D-glucopyranos-1-en-4-ulose (15)
couplings showed that compounds 9Ϫ11, 15, and 21Ϫ24
existed in C1-D chair conformations. It is noteworthy that
4
the compounds observed in the course of this study can be
distinguished on TLC plates by characteristic colors visible
during the initial stage of charring (ca. 250 °C) after spray-
Photoinitiated Reaction at ca. 35 °C: 2,3,4,6-Tetra-O-acetyl-1-
bromo-β--glucopyranosyl chloride (1)[39] (222.5 mg, 0.5 mmol),
allyltri-n-butyltin (330 mg, 1 mmol, 2 equiv.) and a catalytic
ing with a 10% H2SO4 solution in 1:1 EtOH/water: 4Ϫ6: amount of AIBN were dissolved in dry, deoxygenated benzene
(5 mL). The mixture was poured into a stoppered tube made of
quartz and argon was introduced. Irradiation with UV light (me-
dium-pressure mercury lamp, Hanovia, 450 W, no filter) was ap-
plied for 20 min whereupon the substrate was converted into the
chloroallyl intermediate 4.[1] After addition of another portion of
allyltri-n-butyltin (660 mg, 2 mmol, 4 equiv.) and AIBN (catalytic),
irradiation was continued for 2 h 20 min. TLC monitoring showed
the complete conversion of 4 (Rf ϭ 0.48, ethyl acetate/petroleum
ether, 1:2). The reaction mixture was stirred overnight in the pres-
brown; 7, 8, 15: black; 9Ϫ11, 21, 22: greenish; 12, 13: violet;
16, 17: beige.
Conclusion
In conclusion, acetylated glycosyl dihalides 1Ϫ3 have
been shown to produce the corresponding 3,3Ј-(-glycopyr-
anosylidene)bis(1-propenes) 9Ϫ11 by free radical mediated ence of potassium fluoride (2.3 g), dissolved in water/acetonitrile
(2:9, 11 mL), after which the white solids were filtered off and
rinsed. After concentration of the organic phase under vacuum, the
residue was applied to a column of silica gel and eluted with ethyl
acetate/petroleum ether, 1:3, to afford fractions corresponding to 7
(10 mg, 6% yield), 9 (84 mg, 40% yield), 12[1,37] (24 mg, ca. 12%
yield), and 16[1] (4 mg, 2% yield). Contamination of 12 by an un-
identified impurity (ratio: ca. 75:25) was detected by 1H NMR.
Several other minor polar compounds following 12 and 16 (Rf ϭ
0.38, and 0.14, respectively, ethyl acetate/petroleum ether, 1:2) were
not isolated. Ϫ Two additional syntheses (A and B) of 9 were at-
allylation in the presence of excess allyltributyltin. The
yields achieved when reactions were carried out at ca. 35 °C
under UV irradiation conditions (-gluco: 40%, -manno:
34%, -galacto: 24%) compared well to that recently re-
ported for a two-step ionic procedure.[23] Lowering the reac-
tion temperature to ca. 0 °C did not enhance the reaction
selectivity, due to competing radical rearrangement, while
performing the reaction at 80 °C on heating with an IR sun
lamp (AIBN added) mainly produced 3,4,6-tri-O-acetyl-1-
allyl-1,5-anhydro-2-deoxy--arabino-hex-1-enitol (7) (64% tempted. They differed in the amount of allyltri-n-butyltin (indic-
ated in equiv.) added first (a) and after (b) conversion of 1 to the
chloroallyl intermediate 4 and in the irradiation time: A: (a) 4
equiv., 20 min; (b) 2 equiv., 6 h; B: (a) 6 equiv., 2 h 50 min. The
quantities/yields recorded for experiments A and B, respectively,
are as follows: 7: 24 mg, 15%; 8 mg, 5%; 9: 50 mg, 24%, 76 mg,
37%; 12: 31 mg, 17%; 21 mg, 11%: 16: 39 mg, 19%; 27 mg, 14%.
isolated yield). This constitutes a synthetically useful ex-
ample of 1,2-elimination of a chlorine atom and an acetoxy
group favored by heating under radical conditions. The
acetylated
3,3Ј-(-glycopyranosylidene)bis(1-propenes)
9Ϫ11 undergo ring-closing metathesis[11] in the presence of
Grubbs’ catalyst to afford spirocyclopentene derivatives in
high yields (-gluco: 81%, -manno: 89%, -galacto: 72%).
The obtained spirobicyclic sugars 21 and 22 were de-
acetylated quantitatively in MeOH/H2O/NEt3 to afford 24
and 25.
Photoinitiated Reaction at ca. 0 °C: Other experiments conducted
with 1 as previously described, but at lower temperature (ca. 0 °C)
mainly produced 9, 12, and the rearranged product 15 (see Table 1).
To maintain the temperature of the reaction medium near 0 °C, we
used different devices, in particular a Dewar flask fitted at the bot-
tom with a double envelope quartz tube in which the quartz tube
containing the reaction mixture could be fitted (with a distance to
the mercury lamp of about 5 cm). The flask was filled with ice/
water, to which crushed ice was added occasionally. In other experi-
ments, the quartz tube containing the reagents was placed near the
mercury lamp, which was immersed in a ice-cooled water bath. The
reaction time was found to be longer when using the Dewar flask
(16 h, 4 h, respectively, depending on the applied conditions), due
to the lower temperature and increased distance between the tube
and the lamp. TLC (petroleum ether/diethyl ether, 1:2) clearly
showed the conversion of 4 (Rf ϭ 0.47) mainly into 9, 12, and 15,
which were somewhat less mobile (Rf ϭ 0.44, 0.41, 0.44, respect-
ively). With this solvent mixture, 9 and 15, having the same mobil-
ity, could not be separated. Efficient chromatographic separation
was achieved with petroleum ether/dichloromethane/diethyl ether,
2:1.5:0.5, in which 9, 12, and 15 have Rf values of 0.37, 0.31, 0.27,
respectively, after two consecutive elutions of the TLC plates.
Experimental Section
General Methods: The NMR spectra were recorded with Bruker
spectrometers DRX 300, DRX 500 for solutions in CDCl3, with
Me4Si as the internal reference. Ϫ Mass spectra were measured
with a FINNIGAN MAT 95 XL spectrometer. Ϫ Optical rotations
were measured with a PerkinϪElmer 241 polarimeter. Ϫ Reactions
were monitored by TLC on silica gel 60 F254 (E. Merck) plates,
developed by exposure to H2SO4 (10% in 1:1 EtOH/H2O) spray
followed by charring (ca. 250 °C). Bromine-containing compounds
were visible on TLC plates as pink spots, after spraying first with
fluorescein (0.1% in ethanol), then with 1:1 H2O2 (30% in water)/
acetic acid, and charring (ca. 250 °C). Ϫ Column chromatography
was performed using Geduran Si 60 silica gel (E. Merck). Allyltri-
butyltin and (benzylidene)bis(tricyclohexylphosphane)ruthenium
(IV) dichloride (Grubbs’ catalyst) were purchased from Aldrich and
Strem Chemicals, respectively. Solvents were distilled before use;
water was distilled twice. Benzene was distilled from CaH2. Dichlo-
romethane for RCM reactions was distilled from CaH2 under ar-
gon.
Reactions Carried out by Heating to 80 °C with a Sunlamp. ؊ a):
Sugar dihalide 1[39] (200 mg, 0.45 mmol), allyltributyltin (300 mg,
0.9 mmol, 2 equiv.) and a catalytic amount of AIBN were dissolved
in dry, deoxygenated benzene (4 mL). The mixture was kept under
argon and was irradiated with a 250-W heat lamp for 2 h, until the
3,4,6-Tri-O-acetyl-1-allyl-1,5-anhydro-2-deoxy-D-arabino-hex-1- substrate had been converted into 4 (Rf ϭ 0.45 petroleum ether/
enitol (7), 3,3Ј-(2,3,4,6-Tetra-O-acetyl-D-glucopyranosylidene)bis(1-
diethyl ether, 1:1). Another portion of allyltributyltin (450 mg,
2942
Eur. J. Org. Chem. 2001, 2939Ϫ2946

Ring-Closing Olefin Metathesis of 3,3Ј-(D-Glycopyranosylidene)bis(1-propene)
FULL PAPER
1.35 mmol, 3 equiv.) and a catalytic amount of AIBN were then 21.1 (acetyl). Ϫ MS (CI, isobutane): m/z ϭ 353 [M Ϫ AcO]ϩ, 293
added to the reaction mixture, which was irradiated for 2 h, until
TLC monitoring showed the complete conversion of 4. Removal of
the organotin was achieved as before, and, after concentration of
the organic phase under vacuum, the syrup was applied to a col-
umn of silica gel and eluted with petroleum ether/diethyl ether, 1:1
to afford 7 (90 mg, 64% yield), 9 (18 mg, 10% yield), and 12 (12 mg,
7% yield). Ϫ b): When pure chloride 4 (0.2 g, 0.45 mmol) was
boiled in benzene (250-W sunlamp, catalytic amount of AIBN, ar-
gon) in the presence of allyltributyltin (4 equiv.), diene 7 was isol-
ated in 80% yield. Ϫ c): Heating 4 for 4:5 h as before [conditions
(b)] but with no added AIBN produced conjugated diene 19[1] (63%
isolated yield). Ϫ d): Diene 19 was obtained in 60% yield from 4
on treatment with allyltributyltin (1 equiv.) and AIBN.
[M Ϫ AcO Ϫ AcOH]ϩ, 233 [M Ϫ 179]ϩ.
3,3Ј-(2,3,4,6-Tetra-O-acetyl- -mannopyranosylidene)bis(1-propene)
D
(10): In a quartz tube (13 mm external diameter), a mixture con-
taining sugar dihalide 2[39] (158.6 mg, 0.35 mmol), allyltributyltin
(0.44 mL, 1.42 mmol, 4 equiv.), and a catalytic amount of AIBN,
dissolved in dry, deoxygenated benzene (3.6 mL), was stirred over-
night at room temperature under argon. After the mixture had been
cooled with ice/water, the tube was mounted near a medium-pres-
sure mercury lamp (ca. 1 cm distance, no filter), and irradiation was
continued for 2.5 h under argon. TLC monitoring (ethyl acetate/
petroleum ether, 1:2) showed no traces of 2 after 1 h, and almost
complete transformation of the chloroallyl intermediate 5 (Rf ϭ
0.24) to give 10 (Rf ϭ 0.26), and trace amounts of 13 (Rf ϭ 0.20).
The organotin compounds were removed by filtration of the solids
produced on stirring for 1 h with KF (1.07 g) dissolved in acetonitr-
ile/water (5:1, 3.75 mL). The liquid phase was concentrated under
Compound 7: Colorless oil, Rf ϭ 0.52 (ethyl acetate/petroleum
ether, 1:2). Ϫ [α]2D5 ϭ ϩ3 (c ϭ 0.8, CHCl3). Ϫ IR (NaCl, film): ν˜ ϭ
1750 cmϪ1 (CϭO), 1645, 1675 cmϪ1 (CϭC). Ϫ 1H NMR
(200.1 MHz, CDCl3): δ ϭ 5.80 (ddt, 1 H, J2,1(Z) ϭ 17.0 Hz, vacuum and the residue was dissolved in acetonitrile (5 mL). Wash-
J2,1(E) ϭ 10.2 Hz, J2,3a ϭ J2,3b ϭ 6.7 Hz, 2-H), 5.30 (t, 1 H, J6,7 ing with hexanes (3 ϫ 5 mL) and concentration of the acetonitrile
ϭ
3.5 Hz, 6-H), ca. 5.15 (t, 1 H, J7,8 ϭ 5.3 Hz, 7-H), 5.0Ϫ5.2 [m, 2 phase afforded a residue (171 mg), which was applied to a column
H, 1-H(E), 1-H(Z)], 4.68 (d, 1 H, J5,6 ϭ 3.4 Hz, 5-H), 4.42 (dd, 1 of silica gel and eluted with ethyl acetate/petroleum ether, 1:3.
H, J9a,9b ϭ 11.6 Hz, 9-Ha), 4.27 (dt, 1 H, J8,9a ϭ 5.6 Hz, J8,9b
ϭ
Compound 10 (50 mg, 0.12 mmol, 34% yield) was obtained first as
3.1 Hz, 8-H), 4.16 (dd, 1 H, 9-Hb), 2.83 (d, 2 H, 3-Ha, 3-Hb), 2.09, a clear syrup, followed by 13[1] (1.3 mg, 0.035 mmol, 1%), and the
2.07, 2.04 (3 s, 9 H, acetyl). Ϫ 13C NMR (75.5 MHz, CDCl3): δ ϭ ulose derivative 17[1] (15 mg, 0.04 mmol, 11%).
156.2 (C-4), 133.2 (C-2), 118.1 (C-1), 94.9 (C-5), 74.5 (C-8), 68.5
Compound 10: [α]2D2 ϭ ϩ18.5 (c ϭ 1.1, acetone). Ϫ 1H NMR
(C-6), 67.6 (C-7), 61.8 (C-9), 38.1 (C-3), 171.0, 170.9, 170.1 (Cϭ
(300.1 MHz, CDCl3): δ ϭ 5.75 (dddd, 1 H, J2,1(Z) ϭ 17.6 Hz,
O), 21.5, 21.2, 21.1 (acetyl). Ϫ C15H20O7 (312.32): calcd. C 57.69,
J2,1(E) ϭ 9.6 Hz, J2,3a ϭ 5.9 Hz, J2,3b ϭ 7.9 Hz, 2-H), 5.61 (ddt, 1
H 6.45, O 35.86; found C 58.08, H 6.80.
H, J2Ј,1Ј(Z) ϭ 17 Hz, J2Ј,1Ј(E) ϭ 10.0 Hz, J2Ј,3Јa ϭ 7.4 Hz, J2Ј,3Јb
7.9 Hz, 2Ј-H), 5.35 (dd, 1 H, J5,6 ϭ 3.3, J6,7 ϭ 10.0 Hz, 6-H),
ether, 1:2). Ϫ [α]2D5 ϭ ϩ62 (c ϭ 1.1, CHCl3). Ϫ IR (NaCl, film): 5.25Ϫ5.20 [dm, 1 H, J1(E),3a ഠ 1.3 Hz, 1-H(E)], 5.25Ϫ5.20 [dm, 1
ν˜ H, J1(Z),3a ഠ 1.3 Hz, 1-H(Z)], 5.22 (t, 1 H, J7,8 ϭ 10.0 Hz, 7-H),
1750 cmϪ1 (CϭO), 1645 cmϪ1 (CϭC). 1H NMR
ϭ
Compound 9: Colorless oil, Rf ϭ 0.44 (ethyl acetate/petroleum
ϭ
Ϫ
(200.1 MHz, CDCl3): δ ϭ 5.76 (m, 2 H, 2-H, 2Ј-H), 5.34 (t, 1 H, 5.21 (d, 1 H, 5-H), 5.11 [dm, 1 H, J1Ј(E),1Ј(Z) ഠ 1.3 Hz, J1Ј(E),3Јa
J5,6 ϭ 9.8, J6,7 ϭ 9.6 Hz, 6-H), 5.12 (d, 1 H, 5-H), 5.08Ϫ5.14 [m, 1.3 Hz, 1Ј-H(E)], 5.06 [dm, 1 H, J1Ј(Z),1Ј(E) ഠ 1.3 Hz, J1Ј(Z),3Јa
ഠ
ഠ
4 H, 1-H(E), 1-H(Z), 1Ј-H(E), 1Ј-H(Z)], 4.99 (t, 1 H, J7,8
10.1 Hz, 7-H), 4.12 (d, 2 H, 9-Ha, 9-Hb), 3.84 (dt, 1 H, J8,9a
J8,9b ϭ 3.8, J8,7 ϭ 10.1 Hz, 8-H), 2.79 (ddt, 1 H, J3Јa,1Ј(E)
ϭ
ϭ
ϭ
1.3 Hz, 1Ј-H(Z)], 4.19 (dd, 1 H, J9a,9b ϭ 12.0 Hz, J8,9a ϭ 5.7 Hz,
9-Ha), 4.13 (dd, 1 H, J8,9b ϭ 2.8 Hz, 9-Hb), 3.89 (ddd, 1 H, 8-H),
2.66 (ddm, 1 H, J3a,3b ϭ 15 Hz, 3-Ha), 2.40 (ddm, 1 H, 3-Hb), 2.40
J3Јa,1Ј(Z) ഠ 1 Hz, J3Јa,2Ј ϭ 6.3 Hz, J3Јa,3Јb ϭ 15.7 Hz, 3Ј-Ha), 2.44 (ddm, 1 H, J3Јa,3Јb ϭ 14 Hz, 3Ј-Ha), 2.30 (ddm, 1 H, 3Ј-Hb), 2.17,
(ddt, 1 H, J3a,1(E) ϭ J3a,1(Z) ഠ 1 Hz, J3a,2 ϭ 6.3 Hz, J3a,3b
14.6 Hz, 3-Ha), 2.37 (ddt, 1 H, J3Јb,1Ј(E) ϭ J3Јb,1Ј(Z) ഠ 1 Hz, J3Јb,2Ј
7.5 Hz, 3Ј-Hb), 2.11 (ddd, 1 H, J3b,1 ഠ 1 Hz, J3b,2 ഠ 7.5 Hz, 3-Hb),
2.08, 2.03, 2.01, 1.98 (4 s, 12 H, acetyl). Ϫ 13C NMR (50.3 MHz, NMR (75.5 MHz, CDCl3): δ ϭ 131.7, 130.4 (C-2, C-2Ј), 119.6,
CDCl3): δ ϭ 131.7, 131.0 (C-2, C-2Ј), 119.3, 119.2 (C-1, C-1Ј), 72.4 119.4 (C-1, C-1Ј), 78.4 (C-4), 70.7, 70.2, 70.0, 66.5 (C-5, C-6, C-7,
ϭ
2.10, 2.06, 1.97 (4 s, 12 H, acetyl). Proton resonances for the two
ϭ
allyl groups have been assigned tentatively, based on the chemical
shift values found for 9, 13,[1] and the α-anomer of 13.[40] 13C
Ϫ
(C-6), 72.1 (C-5), 69.3, 69.1 (C-7, C-8), 62.7 (C-9), 40.9 (C-3), 34.0 C-8), 63.4 (C-9), 38.5, 34.1 (C-3, C-3Ј), 170.7, 170.3, 170.2, 169.8
(C-3Ј), 170.7, 170.5, 169.5, 169.1 (CϭO), 20.8, 20.8, 20.7, 20.6 (CϭO), 20.8, 20.8, 20.7, 20.6 (acetyl). Ϫ MS (CI, NH3): m/z ϭ 430
(acetyl). The C-4 signal, which was hidden by the CDCl3 reson-
[M ϩ 18]ϩ, 413 [M ϩ 1]ϩ. Ϫ C20H28O9 (412.44): calcd. C 58.24,
ances at 50 MHz, was visible at δ ϭ 77.8 at 75.47 MHz. ϪMS (CI, H 6.84, O 34.91; found C 58.44, H 6.92, O 34.62.
NH3): m/z ϭ 413 [M ϩ 1]ϩ, 430 [M ϩ 18]ϩ. Ϫ C20H28O9 (412.44):
calcd. C 58.24, H 6.84, O 34.91; found C 57.85, H 7.12.
3,4,6-Tri-O-acetyl-1-allyl-1,5-anhydro-2-deoxy-
D
-lyxo-hex-1-enitol
-galactopyranosylidene)bis(1-
propene) (11): 2,3,4,6-Tetra-O-acetyl-1-bromo-β--galactopyrano-
(8) and 3,3Ј-(2,3,4,6-Tetra-O-acetyl-
D
Compound 15: Colorless oil. Ϫ [α]2D5 ϭ ϩ65 (c ϭ 0.5, CHCl3). Ϫ
1H NMR (500.1 MHz, CDCl3): δ ϭ 5.73 (m, 2 H, 2-H, 2Ј-H), 5.30 syl chloride (3)[39] (183.8 mg, 0.41 mmol), allyltri-n-butyltin
(dd, 1 H, J6,7 ϭ 9.3, J5,6 ϭ 10.7 Hz, 6-H), 5.18 [m, 2 H, 1-H(E), 1-
(0.25 mL, 0.84 mmol, 2 equiv.) and a catalytic amount of AIBN
H(Z)], 4.99 (dd, 1 H, J7,8 ϭ 10.3 Hz, 7-H), 4.95 [m, 2 H, 1Ј-H(E), were dissolved in dry, deoxygenated benzene (4 mL). The mixture
1Ј-H(Z)], 4.27 (dd, 1 H, J8,9a ϭ 4.6 Hz, J9a,9b ϭ 12.3 Hz, 9-Ha), was poured into a stoppered tube made of quartz (13 mm diameter)
4.12 (dd, 1 H, J8,9b ϭ 2.3 Hz, 9-Hb), 3.83 (ddd, 1 H, 8-H), 3.58
(ddt, 1 H, J3a,3b ϭ 14.4 Hz, J3a,2 ϭ 4.7 Hz, J3a,1(E) ϭ J3a,1(Z)
and argon was introduced. Irradiation with UV light (medium-
pressure mercury lamp, Hanovia, 450 W, no filter) was applied for
ϭ
2.0 Hz, 3-Ha), 2.61 (dd, 1 H, J3b,2 ϭ 9.6 Hz, 3-Hb), 2.42 (dddd, 1 25 min, after which the substrate had been converted into the chlo-
H, J3Јa,3Јb ϭ 13 Hz, J3Јa,2Ј ഠ 6 Hz, J3Јa,1Ј(E) ഠ 2 Hz, J3Јa,5 ഠ 3.5 Hz, roallyl intermediate 6. After addition of another portion of allyltri-
3Ј-Ha), 2.18 (m, 1 H, 3Ј-Hb), 2.15 (m, 1 H, 5-H), 2.12, 2.10, 2.04, n-butyltin (0.5 mL, 1.68 mmol, 4 equiv.) and AIBN (catalytic), irra-
1.98 (4 s, 12 H, acetyl). Ϫ 13C NMR (125.8 MHz, CDCl3): δ ϭ diation was continued for 2 h 50 min. TLC monitoring showed the
136.7 (C-2Ј), 132.4 (C-2), 119.8 (C-1), 115.6 (C-1Ј), 107.1 (C-4), complete conversion of 6, to give a mixture containing at least 7
73.3 (C-6), 70.2 (C-8), 69.6 (C-7), 62.4 (C-9), 45.3 (C-5), 39.7 (C- products. The reaction mixture was stirred overnight in the pres-
3), 32.6 (C-3Ј), 171.2, 170.9, 170.4, 168.7 (CϭO), 22.6, 21.3, 21.2, ence of potassium fluoride (1.84 g), dissolved in 2:9 water/aceto-
Eur. J. Org. Chem. 2001, 2939Ϫ2946
2943

J.-P. Praly et al.
FULL PAPER
nitrile (8 mL), after which the white solids were filtered off and mol %) was then added under argon. After the reaction mixture
rinsed. After concentration of the organic phase under vacuum, the
had been stirred at room temperature for 7 h, TLC monitoring
residue was dissolved in acetonitrile (15 mL), which was then showed the complete transformation of the starting material (Rf ϭ
washed with hexane (3 ϫ 15 mL). The acetonitrile phase was con-
centrated under vacuum, and the residue was applied to a column
of silica gel and eluted with a 1:1 diethyl ether/petroleum ether
mixture. Two fractions were collected. The first one (Rf ϭ 0.56,
diethyl ether/petroleum ether, 1:1, two successive elutions) was
identified as 3,4,6-tri-O-acetyl-1-allyl-1,5-anhydro-2-deoxy--lyxo-
hex-1-enitol (8) (8.7 mg, 7% yield), obtained as a colorless oil. The
0.37, ethyl acetate/petroleum ether, 3:7) into a new product (Rf ϭ
0.28, ethyl acetate/petroleum ether, 3:7), which appeared green on
the TLC plate after spraying with diluted H2SO4 and charring. The
resulting black solution was concentrated under reduced pressure.
The residual dark oil was taken up in diethyl ether, stirred overnight
under air to decompose the catalyst,[36] and filtered. After removal
of the solvent in vacuum followed by silica gel chromatography
second fraction (Rf ϭ 0.44, diethyl ether/petroleum ether, 1:1, two (100 mL of ethyl acetate/petroleum ether, 3:7, then 200 mL of ethyl
successive elutions) was a mixture containing two compounds acetate/petroleum ether, 1:2), 21 was obtained as a white solid
(50 mg, ratio ca. 8:1). They were separated by chromatography with (22.8 mg, 81% yield). Ϫ Colorless needles (diethyl ether/petroleum
a column of silica gel and eluted with petroleum ether/dichlorome-
thane/diethyl ether, 2:1.5:0.5. The main compound was identified
as the ‘‘C,C-diallyl -galactoside’’ 11 (40.6 mg, 24%) with Rf ϭ
ether), m.p. 106Ϫ107 °C. Ϫ [α]2D2 ϭ ϩ42.3 (c ϭ 1, acetone). Ϫ IR
Ϫ1
(KBr): ν ϭ 1750 cm (CϭO), 1725 cmϪ1 (CϭC). Ϫ 1H NMR
˜
(500.1 MHz, CDCl3): δ ϭ 5.63 (m, 2 H, 1Ј-H, 2Ј-H), 5.19 (m, 2 H,
0.45, after two successive elutions of the same plate with petroleum 2-H, 3-H), 5.09 (m, 1 H, 4-H), 4.24 (dd, 1 H, J5,6a ϭ 4.4 Hz,
ether/dichloromethane/diethyl ether, 2:1.5:0.5. The minor impure
fraction (6 mg) which was more polar and the other compounds
were not identified. Treatment of 6 as for 4 under conditions (c)
(see above, 2 h heating to 80 °C) produced the -galacto analog[1]
20 in 73% isolated yield.
J6a,6b ϭ 12.2 Hz, 6-Ha), 4.08 (dd, 1 H, J5,6b ϭ 2.4 Hz, 6-Hb), 3.83
(ddd, 1 H, 5-H), 2.85 (m, 4 H, 3Ј-Ha, 3Ј-Hb, 5Ј-Ha, 5Ј-Hb), 2.02,
2.01, 1.98, 1.96 (4 s, 12 H, acetyl). Ϫ 1H NMR (500.1 MHz, [D6]a-
cetone): δ ϭ 5.65 (m, 1 H, 1Ј-H), 5.64 (m, 1 H, 2Ј-H), 5.19 (t, 1 H,
J
J
3,4 ϭ 9.8 Hz, 3-H), 5.11 (d, 1 H, J2,3 ϭ 9.8 Hz, 2-H), 5.04 (t, 1 H,
4,5 ϭ 9.9 Hz, 4-H), 4.21 (dd, 1 H, J5,6a ϭ 4.6 Hz, J6a,6b ϭ 12.2 Hz,
Compound 8: Colorless oil. Ϫ Rf ϭ 0.56 (diethyl ether/petroleum
ether, 1:1, after two successive elutions). Ϫ [α]2D2 ϭ ϩ2.3 (c ϭ 0.5,
acetone). Ϫ 1H NMR (200.1 MHz, CDCl3): δ ϭ 5.84 (ddt, 1 H,
J1(E),2 ϭ 10.1 Hz, J1(Z),2 ϭ 16.8 Hz, J2,3a ϭ 6.7 Hz, J2,3b ϭ 6.7 Hz,
2-H), 5.54 (m, 1 H, 6-H), 5.41 (dt, 1 H, J6,7 ϭ 4.5, J7,8 ϭ 1.6 Hz,
J ഠ 1.2 Hz, 7-H), 5.20Ϫ5.09 [m, 2 H, 1-H(E), 1-H(Z)], 4.56 (m, 1
6-Ha), 4.05 (dd, 1 H, J5,6b ϭ 2.5 Hz, 6-Hb), 3.98 (ddd, 1 H, 5-H),
2.73 (s, 2 H, 5Ј-Ha, 5Ј-Hb), 2.52 (q, 2 H, 3Ј-Ha, 3Ј-Hb), 2.02, 2.01,
1.98, 1.96 (4 s, 12 H, acetyl). Ϫ 13C NMR (125.8 MHz, CDCl3):
δ ϭ 171.2, 170.8, 170.2, 169.9 (CϭO), 128.3 (C-1Ј), 127.9 (C-2Ј),
85.4 (C-1), 73.1 (C-2), 72.8 (C-3), 70.6 (C-5), 69.2 (C-4), 62.9 (C-6),
45.6 (C-3Ј), 36.5 (C-5Ј), 21.2, 21.1, 21.0, 21.0 (acetyl). Ϫ C18H24O9
(384.286): calcd. C 56.25, H 6.29, O 37.46; found C 56.39, H 6.37,
O 37.77.
H, J
ϭ 3 Hz, J5,7 ϭ 1 Hz, 5-H), 4.31 (m, 2 H, 8-H, 9-Ha), 4.21
5,6
(dd, 1 H, J8,9b ϭ 9 Hz, J9a,9b ϭ 15 Hz, 9-Hb), 2.85 (d, 2 H, 3-Ha,
3-Hb), 2.14, 2.10, 2.05, (3 s, 9 H, acetyl). Ϫ 13C NMR (125.8 MHz,
CDCl3): δ ϭ 155.9 (C-4), 133.3 (C-2), 118.1 (C-1), 94.8 (C-5), 73.3
(C-8), 65.1 (C-6), 64.2 (C-7), 62.1 (C-9), 38.1 (C-3), 171.1, 170.8,
170.6 (CϭO), 21.3, 21.2, 21.2 (acetyl). Ϫ CIMS (isobutane): m/z
(%) ϭ 313 (2) [M ϩ H]ϩ, 253 (50) [M ϩ H Ϫ AcOH]ϩ, 193 (100)
[M ϩ H Ϫ 2 AcOH]ϩ. Ϫ HRMS: C15H21O7 [M ϩ H]: calcd.
313.12873; found 313.12861.
2,3,4,6-Tetra-O-acetylspiro[1,5-anhydro-D-mannitol-1,4Ј-cyclopent-
1Ј-ene] (22): This compound was prepared from 10 (89% isolated
yield) as described for 21. Colorless oil. Ϫ [α]2D2 ϭ ϩ23.1 (c ϭ 1.4,
1
acetone). Ϫ H NMR (500.1 MHz, [D6] acetone): δ ϭ 5.68 (m, 2
H, 1Ј-H, 2Ј-H), 5.28 (d, 1 H, J2,3 ϭ 3.1 Hz, 2-H), 5.27 (t, 1 H,
J3,4 ϭ 10.1, J4,5 ϭ 10.1 Hz, 4-H), 5.19 (dd, 1 H, 3-H), 4.18 (dd, 1
Compound 11: Colorless oil. Ϫ Rf ϭ 0.45 (petroleum ether/dichloro-
methane/diethyl ether, 2:1.5:0.5, after two successive elutions). Ϫ
[α]2D2 ϭ ϩ87.1 (c ϭ 0.88, acetone). Ϫ 1H NMR (500.1 MHz,
CDCl3): δ ϭ 5.84 (m, 2 H, 2-H, 2Ј-H), 5.39 (dd, 1 H, J6,7 ϭ 3.4,
J7,8 ϭ 0.8 Hz, 7-H), 5.36 (d, 1 H, J5,6 ϭ 10.5 Hz, 5-H), 5.19 [m, 2
H, 1-H(E), 1-H(Z)]*, 5.18 (dd, 1 H, 6-H), 5.12 [br. d, 1 H, J1(E),2 ϭ
10.3 Hz, 1-H(E)]*, 5.08 [br. d, 1 H, J1(Z),2 ϭ 17.5 Hz, 1-H(Z)]*,
4.10 (dd, 1 H, J8,9a ϭ 8.6 Hz, J9a,9b ϭ 13.0 Hz, 9-Ha), 4.04 (m, 2
H, 8-H, 9-Hb), 2.72 (dd, 1 H, J2Ј,3Јa ϭ 6.2 Hz, Jgem ϭ 15.6 Hz, 3Ј-
Ha), 2.45 (dd, 1 H, J2,3a ϭ 6.7 Hz, Jgem ϭ 14.6 Hz, 3-Ha), 2.33
(dd, 1 H, J2Ј,3Јb ϭ 7.7 Hz, 3Ј-Hb), 2.14 (dd, 1 H, J2,3b ϭ 5.8 Hz, 3-
Hb), 2.16, 2.05, 2.03, 1.97 (4 s, 12 H, acetyl). Ϫ 13C NMR
(125.8 MHz, CDCl3): δ ϭ 132.3 (C-2), 131.8 (C-2Ј), 119.4 (C-1Ј),
119.2 (C-1), 78.5 (C-4), 70.2 (C-6), 69.8 (C-5), 68.4 (C-7), 68.2 (C-
8), 62.5 (C-9), 41.6 (C-3), 34.0 (C-3Ј), 170.9, 170.8, 170.8, 169.7
(CϭO), 21.3, 21.2, 21.1, 21.0 (acetyl). Ϫ *The assignments for 1Ј-
H(E), 1Ј-H(Z), 1-H(E), and 1-H(Z) may be reversed. The assign-
ments of 3-Ha, 3-Hb, 3Ј-Ha, and 3Ј-Hb were made on the basis of
the assignments found for 9. Ϫ C20H28O9 (412.44): calcd. C 58.24,
H 6.84, O 34.91; found C 58.28, H 6.84, O 33.93.
H, J5,6a ϭ 5.2 Hz, J6a,6b ϭ 12.1 Hz, 6-Ha), 4.07 (dd, 1 H, J5,6b
ϭ
2.6 Hz, 6-Hb), 3.97 (ddd, 1 H, 5-H), 2.86 (br. d, J5Јa,5Јb ഠ 17 Hz,
5Ј-Ha), 2.55 (br. d, 1 H, 5Ј-Hb), 2.49 (br. d, 1 H, J3Јa,3Јb ഠ 18 Hz,
3Ј-Ha), 2.43 (br. d, 1 H, 3Ј-Hb), 2.16, 2.04, 2.02, 1.93 (4 s, 12 H,
acetyl). Ϫ 13C NMR (75.5 MHz, [D6]acetone): δ ϭ 128.4 (C-1Ј),
127.6 (C-2Ј), 85.7 (C-1), 73.2 (C-2), 71.4 (C-5), 71.0 (C-3), 66.3 (C-
4), 63.4 (C-6), 43.9 (C-3Ј), 39.0 (C-5Ј), 170.5, 170.3, 170.0, 169.7
(CϭO), 20.3, 20.2, 20.15, 20.1 (acetyl). Ϫ C18H24O9 (384.286):
calcd. C 56.25, H 6.29, O 37.46; found C 56.19, H 6.55, O 37.16.
2,3,4,6-Tetra-O-acetylspiro[1,5-anhydro-D-galactitol-1,4Ј-cyclopent-
1Ј-ene] (23): This compound was prepared from 11 (72% isolated
yield) as described for 21. White solid (diethyl ether/petroleum
ether), m.p. 108Ϫ109 °C. Ϫ [α]2D2 ϭ ϩ50.7 (c ϭ 0.4, CH2Cl2). Ϫ
1H NMR (500.1 MHz, CDCl3): δ ϭ 5.56 (m, 2 H, 1Ј-H, 2Ј-H), 5.37
(d, 1 H, J2,3 ϭ 10.8 Hz, 2-H), 5.34 (dd, 1 H, J3,4 ϭ 3.3, J4,5
ϭ
0.9 Hz, 4-H), 4.97 (dd, 1 H, 3-H), ca. 4.0 (m, 3 H, 5-H, 6-Ha, 6-
Hb), ca. 2.5 (m, 4 H, 3Ј-Ha, 3Ј-Hb, 5Ј-Ha, 5Ј-Hb), 2.10, 1.96, 1.94,
1.91 (4 s, 12 H, acetyl). Ϫ 13C NMR (100 MHz, CDCl3): δ ϭ 170.4,
170.4, 170.0, 170.0 (CϭO), 127.9, 127.5 (C-1Ј, C-2Ј), 85.6 (C-1),
2,3,4,6-Tetra-O-acetylspiro[1,5-anhydro-
D-glucitol-1,4Ј-cyclopent- 70.1 (C-3), 70.0 (C-2), 68.7 (C-5), 67.9 (C-4), 61.8 (C-6), 45.4 (C-
1Ј-ene] (21): 3,3Ј-(-Glucopyranosylidene)bis(1-propene)
9
3Ј), 35.9 (C-5Ј), 20.7, 20.7, 20.6, 20.6 (acetyl). Ϫ MS CI (isobutane)
C18H24O9 (384): m/z (%) ϭ 385 (96) [M ϩ H]ϩ, 325 (100) [M ϩ H
Ϫ AcOH]ϩ, 265 (18) [M ϩ H Ϫ 2 AcOH]ϩ. Ϫ HRMS calcd. for
(30.2 mg, 0.073 mmol) was placed in a 10-mL flask. After the flask
had been flushed with argon, oxygen-free CH2Cl2 (3 mL) was ad-
ded by syringe. Grubbs’ catalyst (3.83 mg, 4.6 ϫ 10Ϫ3 mmol, 6 C18H25O9 [M ϩ H]: 385.149857, found 385.14975.
2944
Eur. J. Org. Chem. 2001, 2939Ϫ2946

Ring-Closing Olefin Metathesis of 3,3Ј-(
Spiro[1,5-anhydro-
21 (44 mg, 0.11 mmol) was dissolved in an 8:1:1 MeOH/NEt3/H2O
mixture (6 mL). After stirring at room temperature for ca. 20 h,
deacetylation was complete, as shown by TLC (Rf ϭ 0.3, AcOEt/
MeOH/H2O, 12:5:1). Concentration under reduced pressure pro-
duced compound 24 as a colorless oil (24 mg, quantitative yield).
Ϫ [α]2D2 ϭ ϩ39.8 (c ϭ 1.27, MeOH). Ϫ 1H NMR (300.1 MHz,
D
-Glycopyranosylidene)bis(1-propene)
FULL PAPER
[10]
´
L. Somsak, J.-P. Praly, G. Descotes, Synlett 1992, 119Ϫ120.
R. H. Grubbs, S. Chang, Tetrahedron 1998, 54, 4413Ϫ4450.
H. E. Blackwell, D. J. O’Leary, A. K. Chatterjee, R. A. Wash-
enfelder, D. A. Bussmann, R. H. Grubbs, J. Am. Chem. Soc.
2000, 122, 58Ϫ71.
D. M. Lynn, B. Mohr, R. H. Grubbs, L. M. Henling, M. W.
Day, J. Am. Chem. Soc. 2000, 122, 6601Ϫ6609.
R. Roy, S. K. Das, Chem. Commun. 2000, 519Ϫ529.
R. Dominique, B. Liu, S. K. Das, R. Roy, Synthesis 2000,
862Ϫ868.
D
-glucitol-1,4Ј-cyclopent-1Ј-ene] (24): Compound
[11]
[12]
[13]
[14]
[15]
CD3OD): δ ϭ 5.52 (m, 2 H, 1Ј-H, 2Ј-H), 3.66 (dd, 1 H, J5,6a
ϭ
2.4 Hz, J6a,6b ϭ 11.7 Hz, 6-Ha), 3.53 (dd, 1 H, J5,6b ϭ 5.2 Hz, 6-
Hb), 3.32 (ddd, 1 H, J4,5 ϭ 9.3 Hz, 5-H), 3.25Ϫ3.18 (m, 3 H, 2-H,
3-H, 4-H), 2.66, 2.49, 2.33 (dt, dt, dd, 1 H, 1 H, 2 H, Jgem ϭ 17 Hz,
Jgem ϭ 15 Hz, Jgem ϭ 17 Hz, 3Ј-Ha, 3Ј-Hb, 5Ј-Ha, 5Ј-Hb)*. Ϫ 13C
NMR (75.5 MHz, CD3OD): δ ϭ 129.6, 129.5 (C-1Ј, C-2Ј)*, 87.5
(C-1), 77.2, 76.4, 76.2, 72.5 (C-2, C-3, C-4, C-5)*, 63.6 (C-6), 47.0
(C-3Ј), 36.1 (C-5Ј). Ϫ *These assignments are not precisely estab-
lished. Ϫ MS CI (isobutane): m/z (%) ϭ 217 (47) [M ϩ H]ϩ, 199
(44) [M ϩ H Ϫ H2O]ϩ, 181 (100) [M ϩ H Ϫ 2 H2O]ϩ. Ϫ HRMS:
C10H17O5 [M ϩ H]: calcd. 217.107599; found 217.10764.
[16]
[17]
[18]
[19]
[20]
[21]
[22]
M. H. D. Postema, D. Calimente, Tetrahedron Lett. 1999, 40,
4755Ϫ4759.
D. Calimente, M. H. D. Postema, J. Org. Chem. 1999, 64,
1770Ϫ1771.
O. Dirat, T. Vidal, Y. Langlois, Tetrahedron Lett. 1999, 40,
4801Ϫ4802.
P. Kapferer, F. Sarabia, A. Vasella, Helv. Chim. Acta 1999,
82, 645Ϫ656.
L. Ackermann, D. El Tom, A. Fürstner, Tetrahedron 2000,
56, 2195Ϫ2202.
U. K. Pandit, H. S. Overkleeft, B. C. Borer, H. Bieräugel, Eur.
J. Org. Chem. 1999, 959Ϫ968.
M. A. Leeuwenburgh, C. C. M. Appeldoorn, P. A. V. van
Hooft, H. S. Overkleeft, G. A. van der Marel, J. H. van Boom,
Eur. J. Org. Chem. 2000, 873Ϫ877.
A.-M. Periers, P. Laurin, Y. Benedetti, S. Lachaud, D. Ferroud,
A. Iltis, J.-L. Haesslein, M. Klich, G. L’Hermite, B. Musicki,
Tetrahedron Lett. 2000, 41, 867Ϫ871.
Spiro[1,5-anhydro-D-mannitol-1,4Ј-cyclopent-1Ј-ene] (25): Com-
pound 22 (48.3 mg, 0.13 mmol) was dissolved in an 8:1:1 MeOH/
NEt3/H2O mixture (6 mL). After the mixture had been stirred for
ca. 6 h at room temperature, TLC showed no traces of the starting
material (Rf ഠ 0.3, ethyl acetate/petroleum ether, 1:2), which had
been converted into a polar compound (Rf ϭ 0, ethyl acetate/petro-
leum ether, 1:2, and Rf ϭ 0.66, AcOEt/EtOH/H2O, 7:5:1). Concen-
tration under reduced pressure provided compound 25 as an oil
(29 mg, quantitative yield). Ϫ [α]2D2 ϭ ϩ95.4 (c ϭ 1.5, MeOH). Ϫ
1H NMR (300.1 MHz, CD3OD): δ ϭ 5.69 (m, 2 H, 1Ј-H, 2Ј-H),
3.81Ϫ3.63 (m, 5 H, 2-H, 3-H, 4-H, 6-Ha, 6-Hb), 3.41 (m, 1 H, 5-
H), 2.84, 2.67, 2.48, 2.27 (4 br. d, 4 H, Jgem ഠ 16 Hz, 3Ј-Ha, 3Ј-
Hb, 5Ј-Ha, 5Ј-Hb)*. Ϫ 13C NMR (75.5 MHz, CD3OD): δ ϭ 128.6,
127.4 (C-1Ј, C-2Ј)*, 86.8 (C-1), 75.5, 75.2, 72.9, 67.3 (C-2, C-3, C-
4, C-5)*, 62.2 (C-6), 43.7 (C-3Ј), 38.5 (C-5Ј). Ϫ *These assignments
are not precisely established. Ϫ MS EI (70 eV): m/z (%) ϭ 199 (3)
[M Ϫ OH]ϩ, 181 (3) [M Ϫ H2O Ϫ OH]ϩ.
[23]
[24] [24a]
J.-P. Praly, Adv. Carbohydr. Chem. Biochem. 2000, 56,
65Ϫ151, and references therein, Ϫ [24b] J. Desire, J. Prandi, Eur.
´
´
[24c]
J. Org. Chem. 2000, 3075Ϫ3084, and references therein, Ϫ
´
´
´
´
L. Somsak, L. Kovacs, M. Toth, E. Ösz, L. Szilagyi, Z. György-
´
´
deak, Z. Dinya, T. Docsa, B. Toth, P. Gergely, J. Med. Chim.,
accepted.
[25] [25a]
D. H. R. Barton, D. O. Jang, J. Cs. Jaszberenyi, Tetrahed-
[25b]
ron 1993, 49, 7193Ϫ7214. Ϫ
A. Fernandez-Mayoralas, A.
`
Marra, M. Trumtel, A. Veyrieres, P. Sinay¨, Carbohydr. Res.
1989, 188, 81Ϫ95.[25c] J. J. Patroni, R. V. Stick, D. M. G. Til-
brook, B. W. Skelton, A. H. White, Aust. J. Chem. 1989, 42,
2127Ϫ2141. Ϫ [25d] T.-S. Lin, J.-H. Yang, M.-C. Liu, J.-L. Zhu,
Tetrahedron Lett. 1990, 31, 3829Ϫ3832. Ϫ [25e] F. Santoyo-Gon-
´
´
´
zalez, F. G. Calvo-Flores, F. Hernandez-Mateo, P. Garcıa-Men-
Acknowledgments
Stipends in the framework of a collaboration (TEMPRA program)
´
´
doza, J. Isac-Garcıa, M. D. Perez-Alvarez, Synlett 1994,
454Ϫ456.
´
ˆ
between ‘‘la Region Rhone-Alpes’’ and Shanghai City (P. R.
China) (to G.-R. C. and Z.-B. F.) and of the Erasmus exchange
program between the University of Stuttgart and ESCPE Ϫ Lyon
(to M. S.), as well as support from the National Natural Science
Foundation of China, are gratefully acknowledged.
[26]
[27]
[28]
[29]
[30]
A. Ghosez, T. Göbel, B. Giese, Chem. Ber. 1988, 121,
1807Ϫ1811.
Y. Itoh, K. Haraguchi, H. Tanaka, K. Matsumoto, K. T. Naka-
mura, T. Miyasaka, Tetrahedron Lett. 1995, 36, 3867Ϫ3870.
S. W. T. Choe, M. E. Jung, Carbohydr. Res. 2000, 329,
731Ϫ744.
J.-P. Praly, Z. El Kharraf, G. Descotes, Carbohydr. Res. 1992,
232, 117Ϫ123.
[1]
J.-P. Praly, G.-R. Chen, J. Gola, G. Hetzer, Eur. J. Org. Chem.
Houben Weyl, Methoden der Organischen Chemie, C-Radikale
(Eds.: M. Regitz, B. Giese), vol. E19a, part 1 and 2, Georg
Thieme Verlag, Stuttgart, New York, 1989.
A. L. J. Beckwith, D. Crich, P. J. Duggan, Q. Yao, Chem. Rev.
1997, 97, 3273Ϫ3312.
2000, 2831Ϫ2838.
J. Dupuis, B. Giese, J. Hartung, M. Leising, J. Am. Chem. Soc.
[2]
1985, 107, 4332Ϫ4333.
[31]
[3]
F. Nicotra, L. Panza, G. Russo, J. Org. Chem. 1987, 52,
5627Ϫ5630.
[32] [32a]
[4]
B. Giese, K. S. Gröninger, T. Witzel, H.-G. Korth, R.
K. C. Nicolaou, C.-K. Hwang, M. E. Duggan, J. Chem. Soc.,
Sustmann, Angew. Chem. Int. Ed. Engl. 1987, 26, 233Ϫ234. Ϫ
Chem. Commun. 1986, 925Ϫ926.
L. A. Paquette, U. Dullweber, L. D. Cowgill, Tetrahedron Lett.
[32b] B. Giese, S. Gilges, K. S. Gröninger, C. Lamberth, T. Wit-
[5]
[32c]
zel, Liebigs Ann. Chem. 1988, 615Ϫ617. Ϫ
Gröninger, Org. Synth. 1990, 69, 66Ϫ71. Ϫ
B. Giese, K. S.
B. Giese, B.
1993, 34, 8019Ϫ8022.
L. A. Paquette, M. J. Kinney, U. Dullweber, J. Org. Chem.
[32d]
[6]
Kopping, C. Chatgilialoglu, Tetrahedron Lett. 1989, 30,
681Ϫ684. Ϫ [32e] B. Quiclet-Sire, S. Z. Zard, J. Am. Chem. Soc.
1996, 118, 9190Ϫ9191.
1997, 62, 1713Ϫ1722.
[7]
C. Waldraff, B. Bernet, A. Vasella, Helv. Chim. Acta 1997, 80,
1882Ϫ1900, and references therein.
[8]
[33] [33a]
`
J.-P. Praly, C. Di Stefano, G. Descotes, R. Faure, Tetrahedron
H.-G. Korth, R. Sustmann, K. S. Gröninger, M. Leisung,
[33b]
Lett. 1994, 35, 89Ϫ92, and references therein.
S. E. Mangholz, A. Vasella, Helv. Chim. Acta 1995, 78,
1020Ϫ1035, and references therein.
B. Giese, J. Org. Chem. 1988, 53, 4364Ϫ4369. Ϫ
S. Yam-
[9]
ago, H. Miyazoe, J.-i. Yoshida, Tetrahedron Lett. 1999, 40,
2339Ϫ2342.
Eur. J. Org. Chem. 2001, 2939Ϫ2946
2945

J.-P. Praly et al.
FULL PAPER
[34] [34a]
[38]
A. Koch, C. Lamberth, F. Wetterich, B. Giese, J. Org.
E. Breitmaier, W. Voelter, Carbon-13 NMR Spectroscopy, High-
Resolution Methods and Applications in Organic Chemistry and
Biochemistry, 3rd ed., VCH, Weinheim, 1987; for γ-gauche ef-
fect, see p. 115.
[34b]
Chem. 1993, 58, 1083Ϫ1089. Ϫ
A. Koch, B. Giese, Helv.
[34c]
Chim. Acta 1993, 76, 1687Ϫ1701. Ϫ
D. Crich, Q. Yao, J.
[34d]
Am. Chem. Soc. 1993, 115, 1165Ϫ1166. Ϫ
D. Crich, Q.
[39]
[40]
Yao, G. F. Filzen, J. Am. Chem. Soc. 1995, 117, 11455Ϫ11470.
J.-P. Praly, L. Brard, G. Descotes, L. Toupet, Tetrahedron 1989,
[35] [35a]
H.-G. Korth, R. Sustmann, B. Giese, B. Rückert, K. S.
45, 4141Ϫ4152.
[35b]
Gröninger, Chem. Ber. 1990, 123, 1891Ϫ1898. Ϫ
B. Giese,
K. C. Nicolaou, C.-K. Hwang, M. E. Duggan, J. Am. Chem.
Soc. 1989, 111, 6682Ϫ6690.
Angew. Chem. Int. Ed. Engl. 1989, 28, 969Ϫ980.
Y.-J. Hu, R. Roy, Tetrahedron Lett. 1999, 40, 3305Ϫ3308.
D. Horton, T. Miyake, Carbohydr. Res. 1988, 184, 221Ϫ229.
[36]
[37]
Received December 1, 2000
[O00619]
2946
Eur. J. Org. Chem. 2001, 2939Ϫ2946