Carbohydrate Homologation by the Use of 2-(Trimethylsilyl)thiazole. Preparative Scale Synthesis of Rare Sugars: L-Gulose, L-Idose, and the Disaccharide Subunit of Bleomycin A2

Article abstract of DOI:10.1021/jo970601h

The well established one-carbon homologation method of protected monosaccharides employing 2-(trimethylsilyl)thiazole (2-TST) as a formyl anion equivalent has been used for high yield and multigram scale synthesis of the title rare hexoses from L-xylose. Thus, L-gulose has been obtained by stereoselective anti-addition of 2-TST to aldehydo-L-xylose diacetonide followed by thiazole to formyl conversion of the resulting alcohol. The inversion of configuration at C-1 of this alcohol by an oxidation - reduction sequence prior to the aldehyde releasing from thiazole led to L-idose. The same alcohol was readily elaborated into 1,3,4,6-tetra-O-acetyl-L-gulopyranose whose highly stereoselective glycosidation coupling with 3-O-carbamoyl-2,4,6-tri-O-acetyl-α-D-mannosyl diethyl phosphate afforded the same peracetylated disaccharide subunit employed by Boger and Honda in the total synthesis of the antibiotic bleomycin A₂.

Full text of DOI:10.1021/jo970601h

J . Org. Chem. 1997, 62, 6261-6267
6261
Ca r boh yd r a te Hom ologa tion by th e Use of
2-(Tr im eth ylsilyl)th ia zole. P r ep a r a tive Sca le Syn th esis of Ra r e
Su ga r s: L-Gu lose, L-Id ose, a n d th e Disa cch a r id e Su bu n it of
Bleom ycin A2
Alessandro Dondoni,* Alberto Marra, and Alessandro Massi
Dipartimento di Chimica, Laboratorio di Chimica Organica, Universita`, 44100 Ferrara, Italy
Received April 2, 1997X
The well established one-carbon homologation method of protected monosaccharides employing
2-(trimethylsilyl)thiazole (2-TST) as a formyl anion equivalent has been used for high yield and
multigram scale synthesis of the title rare hexoses from L-xylose. Thus, L-gulose has been obtained
by stereoselective anti-addition of 2-TST to aldehydo-L-xylose diacetonide followed by thiazole to
formyl conversion of the resulting alcohol. The inversion of configuration at C-1 of this alcohol by
an oxidation-reduction sequence prior to the aldehyde releasing from thiazole led to L-idose. The
same alcohol was readily elaborated into 1,3,4,6-tetra-O-acetyl-L-gulopyranose whose highly
stereoselective glycosidation coupling with 3-O-carbamoyl-2,4,6-tri-O-acetyl-R-D-mannosyl diethyl
phosphate afforded the same peracetylated disaccharide subunit employed by Boger and Honda in
the total synthesis of the antibiotic bleomycin A2.
While numerous common sugars are available in kilo-
gram or ton quantities from natural products and there-
fore constitute a convenient source of starting materials
for organic synthesis,1 other special or rare sugars are
the minor yet very important components of biologically
active compounds. A significant example is given by
L-gulose (1) which coupled with a 3-carbamoylmannose
derivative (Figure 1) constitutes the disaccharide subunit
of bleomycin A2,2 the major constituent of a family of
glycopeptide antibiotics capable of mediating the cleavage
of DNA and RNA by a metal-dependent oxidative pro-
cess.3 Quite recently the sugar 1 has been used as
starting material in the synthesis of 1,3-oxathiolane pyri-
midine and purine nucleosides that exhibit very potent
antiviral activity against hepatitis B virus (HBV) and
human immunodeficiency virus (HIV).4 Also the C-2 epi-
mer L-idose (2) is an interesting rare sugar. For instance
quantities of suitable L-idopyranose derivatives were re-
quired in the synthesis of sensitive substrates for R-L-
iduronidase.5 Therefore simple synthetic routes that
make rare sugars readily available at low cost6 and on
meaningful preparative scale are quite useful for the syn-
thesis of carbohydrate-containing natural products and
F igu r e 1.
their analogues. Given the efficient thiazole-based one-
carbon homologation of aldehydo-sugars that exploits 2-
(trimethylsilyl)thiazole (5) as a formyl anion equivalent,7
we have considered this method for rare monosaccharide
synthesis. We report here a preparative scale synthesis
of the hexoses 1 and 2 from the same pentose precursor
L-xylose and the exploitation of a thiazole-bearing key
intermediate for a concise synthesis of the disaccharide
subunit of bleomycin A2 suitably protected for the incor-
poration into the aglycone moiety. The method involves
reactions which are operatively simple and which can be
run for either centigram or gram scale preparations.
Intermediates and final products were obtained in high
yields and required few chromatographic purifications.
X Abstract published in Advance ACS Abstracts, August 1, 1997.
(1) (a) Hanessian, S. Total Synthesis of Natural Products: The
“Chiron” Approach; Pergamon Press: Oxford, 1983. (b) Lichtenthaler,
F. W. In New Aspects of Organic Chemistry I; Yoshida, Z., Shiba, T.,
Ohshiro, Y., Eds.; Kodansha: Tokyo, and VCH: Weinheim, 1989, p
351. (c) Carbohydrates as Organic Raw Materials; Lichtenthaler, F.
W., Ed., VCH: Weinheim, 1991. (d) Carbohydrates as Organic Raw
Materials II; Descotes, G., Ed., VCH: Weinheim, 1993.
(2) (a) Umezawa, H. Pure Appl. Chem. 1971, 28, 665. (b) Takita,
T.; Muraoka, Y.; Nakatani, T.; Fujii, A.; Umezawa, Y.; Naganawa, H.;
Umezawa, H. J . Antibiot. 1978, 31, 801.
(3) (a) Hecht, S. M. Acc. Chem. Res. 1986, 19, 383. (b) Stubbe, J .;
Kozarich, J . W. Chem. Rev. 1987, 87, 1107.
Resu lts a n d Discu ssion
(4) (a) Kim, H. O.; Shanmuganathan, K.; Alves, A. J .; J eong, L. S.;
Beach, J . W.; Schinazi, R. F.; Chang, C.-N.; Cheng, Y.-C.; Chu, C. K.
Tetrahedron Lett. 1992, 33, 6899. (b) Beach, J . W.; J eong, L. S.; Alves,
A. J .; Pohl, D.; Kim, H. O.; Chang, C.-N.; Doong, S.-L.; Schinazi, R. F.;
Cheng, Y.-C.; Chu, C. K. J . Org. Chem. 1992, 57, 2217. (c) J eong, L.
S.; Schinazi, R. F.; Beach, J . W.; Kim, H. O.; Nampalli, S.; Shanmuga-
nathan, K.; Alves, A. J .; McMillan, A.; Chu, C. K.; Mathis, R. J . Med.
Chem. 1993, 36, 181.
Syn th esis of L-Gu lose (1) a n d L-Id ose (2). A clas-
sical method of preparation of L-gulose (1) is that of
(7) For overviews, see: (a) Dondoni, A. Pure Appl. Chem. 1990, 62,
643. (b) Dondoni, A. In Modern Synthetic Methods; Scheffold, R., Ed.;
Verlag Helvetica Chimica Acta: Basel, 1992; p 377. (c) Dondoni, A.
In New Aspects of Organic Chemistry II; Yoshida, Z., Ohshiro, Y., Eds.;
Kodansha: Tokyo, and VCH: Weinheim, 1992; p 105. (d) Dondoni,
A.; Marra, A. In Preparative Carbohydrate Chemistry, Hanessian, S.,
Ed., Marcel Dekker: New York, 1997; p 173.
(5) (a) Stirling, J . L.; Robinson, D.; Fensom, A. H.; Benson, P. F.;
Baker, J . E. Lancet 1978, 147. (b) Hopwood, J . J .; Muller, V.; Smithson,
A.; Baggett, N. Clin. Chim. Acta 1979, 92, 257.
(6) L-Gulose and L-idose cost ca. $200/100 mg.
S0022-3263(97)00601-4 CCC: $14.00 © 1997 American Chemical Society

6262 J . Org. Chem., Vol. 62, No. 18, 1997
Dondoni et al.
Sch em e 1
dichloromethane at 0 °C in about 1 h. Desilylative
workup by treatment with tetrabutylammonium fluo-
ride16 gave a mixture of the expected alcohol anti-6 and
diastereomer syn-6 (not shown) in 95:5 ratio (1H NMR
analysis) and 93% yield. The crystallization of this
mixture from AcOEt-cyclohexane afforded the pure al-
cohol anti-6 in 61% yield from dithioacetal 3. The
protection of the hydroxyl group of anti-6 before proceed-
ing to the formyl group unmasking from the thiazole ring
was recommended by previous work as well.17 Thus
anti-6 was converted into the triethylsilyl ether anti-7
(88%), and this product was subjected to the standard
thiazole-to-formyl deblocking protocol.18 The resulting
crude aldehydo-L-gulose derivative anti-8 was converted
into the free sugar 1 by treatment with aqueous acetic
acid to remove the isopropylidene and silyl protective
groups. The crude L-gulose (1) obtained in this way
appeared to be pure enough as judged by 1H NMR
analysis and the optical rotation value (see Experimental
Section). Purification by chromatography over silica gel
and then Sephadex LH-20 gave analytically pure 1 (0.5
g) in 56% yield (20% from L-xylose).
The inversion of configuration at C-1 of the readily
available alcohol anti-6 by an oxidation-reduction se-
quence provided a tactically simple entry to L-idose (2).
This method was developed in our laboratory in an earlier
thiazole-based synthesis of all possible isomeric tetroses
and pentoses starting from glyceraldehyde.19 Therefore
the crude 95:5 mixture of alcohols anti- and syn-6 (1.6 g)
formed by addition of 5 to aldehydo-L-xylose 4 was
oxidized under Swern-type conditions to give the ketone
9 in nearly quantitative yield (Scheme 2). The reduction
of 9 by sodium borohydride proceeded with good stereo-
selectivity to give a mixture (90%) of the same alcohols
anti- and syn-6 in 9:91 ratio (1H NMR analysis). Because
of the unsuccessful separation of these isomers by
crystallization or chromatography, the above reaction
mixture was silylated and purified by column chroma-
tography on silica gel to afford the pure diastereomer
syn-7 in 82% yield. From this compound the synthesis
followed a parallel line as in Scheme 1, i.e. cleavage of
the thiazole ring to the formyl group to give the protected
aldehydo-L-idose syn-8 (87%) and removal of all protective
groups by acid hydrolysis. The free L-idose, existing as
a mixture of pyranoses (2) and furanoses (2a ), proved to
be a good quality product by 1H NMR analysis and optical
rotation (see Experimental Section). However, the chro-
matographic purification through a column of silica gel
and then Sephadex LH-20 afforded the analytically pure
product in 59% isolated yield (19% overall from L-xylose).
Syn th esis of P er a cetyla ted 2-O-(3-O-Ca r ba m oyl-
r-D-m a n n op yr a n osyl)-L-gu lop yr a n ose (13). The syn-
Sowden and Fischer via nitroaldol condensation of 2,4-
O-benzylidene-L-xylose with nitromethane.8 More re-
cently, special chemical syntheses of 1 and derivatives
in either the furanose or pyranose form have been
reported starting from D-mannose9 and D-glucose.10 Note-
worthy is the gram scale preparation of 1 by stereocon-
trolled ethylhydroborane reduction of D-glucurono-6,3-
lactone.11 Quite interestingly the sodium borohydride
reduction of a derivative of the same sugar lactone
followed by an inversion of configuration at C-5 are the
key steps of a recommended synthesis 12 of the even more
rare sugar L-idose (2). An enzymatic method leading to
1 and 2 based on transketolase-based condensation of
hydroxypyruvic acid with hydroxy aldehydes has been
also reported.13
Our synthesis of both 1 and 2 employs L-xylose as
common starting material and relies on highly stereose-
lective organometallic addition and oxidation-reduction
sequence for the construction of the new hydroxymeth-
ylene group in either R and S configuration. The
aldehydo-L-xylose diacetonide 4 (Scheme 1) was easily
prepared from L-xylose14 in two steps through the corre-
sponding diethyl dithioacetal 3. The Hg(II)-promoted
hydrolysis of 3 afforded the crude aldehyde 4 suitable
for the reaction with 2-(trimethylsilyl)thiazole (5). Guided
by earlier studies carried out in our laboratory,15 the
addition of 5 to 4 (5-6 g) was efficiently carried out in
(15) (a) Dondoni, A.; Fantin, G.; Fogagnolo, M.; Medici, A. Angew.
Chem., Int. Ed. Engl. 1986, 25, 835. (b) Dondoni, A.; Fantin, G.; Foga-
gnolo, M.; Medici, A.; Pedrini, P. J . Org. Chem. 1989, 54, 693 and 702.
(16) Given a misleading statement that recently appeared in the
literature (Parrain, J .-L.; Beaudet, I.; Cintrat, J .-C.; Ducheˆne, A.;
Quintard, J .-P. Bull. Soc. Chim. Fr. 1994, 131, 304), it is worth
recalling here that the addition of the silylthiazole 5 to aldehydes does
not require the activation by fluoride ions nor other catalysts. Treat-
ment of the reaction mixture with n-Bu4NF induces the desilylation
of the O-silyl ether adduct into alcohol. A mechanistic explanation of
this peculiar organosilane addition reaction has been provided, see:
(a) Dondoni, A.; Douglas, A. W.; Shinkai, I. J . Org. Chem. 1993, 58,
3196. (b) Wu, Y.-D.; Lee, J . K.; Houk, K. N.; Dondoni, A. J . Org. Chem.
1996, 61, 1922.
(8) (a) Sowden, J . C.; Fischer, H. O. L. J . Am. Chem. Soc. 1945, 67,
1713. (b) Whistler, R. L.; BeMiller, J . N. Methods Carbohydr. Chem.
1962, 1, 137.
(9) (a) Evans, M. E.; Parrish, F. W. Carbohydr. Res. 1973, 28, 359.
(b) Evans, M. E.; Parrish, F. W. Methods Carbohydr. Chem. 1980, 8,
173. (c) Oshitari, T.; Tomita, M.; Kobayashi, S. Tetrahedron Lett. 1994,
35, 6493.
(10) (a) Santoyo Gonza´lez, F.; Baer, H. H. Carbohydr. Res. 1990,
202, 33. (b) Ding, X.; Kong, F. Carbohydr. Res. 1996, 286, 161.
(11) Dahlhoff, W. V.; Idelmann, P.; Ko¨ster, R. Angew. Chem., Int.
Ed. Engl. 1980, 19, 546.
(12) (a) Blanc-Muesser, M.; Defaye, J . Synthesis 1977, 568. (b)
Blanc-Muesser, M.; Defaye, J .; Horton, D.; Tsai, J .-H. Methods Car-
bohydr. Chem. 1980, 8, 177. For improvements, see: (c) Baggett, N.;
Samra, A. K. Carbohydr. Res. 1984, 127, 149.
(17) We have observed in several instances that the presence of the
free hydroxyl group results in low yield of isolated aldehyde.
(18) Dondoni, A.; Marra, A.; Perrone, D. J . Org. Chem. 1993, 58,
275.
(13) Kobori, Y.; Myles, D. C.; Whitesides, G. M. J . Org. Chem. 1992,
57, 5899.
(14) L-Xylose costs ca. $400/100 g.
(19) Dondoni, A.; Orduna, J .; Merino, P. Synthesis 1992, 201.

Carbohydrate Homologation with 2-(Trimethylsilyl)thiazole
J . Org. Chem., Vol. 62, No. 18, 1997 6263
Sch em e 2
F igu r e 2. The Boger and Honda synthesis of the disaccharide
subunit of bleomycin A2 (ref 22).
Sch em e 3
thesis of the carbohydrate moiety of the bleomycin A2
involves the assemblage through 1,2 O-glycosidic R-link-
age of 3-O-carbamoyl-D-mannose (glycosyl donor) and
L-gulose (glycosyl acceptor) and anomeric activation of
the resulting disaccharide for the installation in the
aglycone moiety of the antibiotic. Hence a suitably
protected L-gulopyranose derivative with a free hydroxyl
group at C-2 is required to start the synthesis. Evidently
this problem goes beyond the preparation of the free
L-gulose (1) as the multistep elaboration of this rare
sugar20 does not appear to be a convenient approach.
Although a number of L-gulose derivatives that might
serve as glycosyl acceptors at C-2 were prepared by Hecht
and co-workers more than 10 years ago,21 the synthesis
of the target disaccharide and the completion of the total
synthesis of bleomycin A2 has been reported very recently
by Boger and Honda.22 The peracetylated key disaccha-
ride intermediate 13 was obtained22 by sequential de-
benzylation and acetylation of 12 that was in turn
prepared by stereoselective O-glycosylation of the tetra-
benzyl-L-gulose 10 with the tetraacetyl-D-mannopyrano-
syl diphenyl phosphate 11 (Figure 2). While 10 was
prepared by inversion of the C-5 stereochemistry of
D-mannose and through a series of protection-deprotec-
tion reactions (nine steps, 12% yield), we sought an
alternative synthesis from the L-xylose-thiazole adduct
anti-6. This readily available intermediate in the syn-
thesis of L-gulose (1) appeared suitably tailored for the
preparation of derivatives having the C-2 hydroxyl group
differentially protected from the others. Thus, the alcohol
anti-6 was converted in nearly quantitative yield into the
tert-butyldiphenylsilyl ether 14 (Scheme 3) from which
the aldehydo-L-gulose 15 was obtained by thiazole-to-
formyl cleavage in an exceptionally high yield (96%). As
expected, the tert-butyldiphenylsilyl group could be taken
through the deacetonization of 15 by selective hydrolysis
(AcOH-H2O) to give, however, an inseparable 7:3 mix-
ture of 2-O-silyl-L-gulose derivative23 16 together with
regioisomers in ∼100% overall yield. This detrimental
silyl group migration24 occurred under different condi-
tions25 of acid hydrolysis of 15. Moreover, when crude
16 was subjected to benzylation followed by desilylation,
the major product isolated (51%) was the L-gulofurano-
side derivative26 17 that evidently precluded any further
approach to the target tetrabenzylated L-gulopyranoside
10 by this route.
Therefore we decided to change the protective group
strategy toward the synthesis of a suitable L-gulopyra-
noside derivative that could lead to the peracetylated
disaccharide 13 by a more direct route avoiding the
intermediacy of 12. Thus, the alcohol anti-6 was pro-
(23) The structure of 16 was established by 1H NMR analysis of the
corresponding crude tetra-O-acetyl derivative.
(24) The so-called silicon dance is rather common in polyhydroxy-
lated compounds, see: Arias-Pe´rez, M. S.; Santos, M. J . Tetrahedron
1996, 52, 10785.
(25) Other reaction conditions: 1:1 CF3CO2H-H2O, 20 °C; 4:1
CH3CO2H-H2O, 20 to 80 °C; 4:1 CH3CO2H-CH3OH, 20 °C.
(26) The R-L-furanoside structure of 17 was proven by NOE experi-
ments. Substantial enhancements of the H-1 and H-4 signals were
observed in its 1H NMR spectrum upon irradiation of H-2.
(20) Pozsgay, V.; Ohgi, T.; Hecht, S. M. J . Org. Chem. 1981, 46, 3761.
(21) Katano, K.; Chang, P.-I.; Millar, A.; Pozsgay, V.; Minster, D.
K.; Ohgi, T.; Hecht, S. M. J . Org. Chem. 1985, 50, 5807.
(22) Boger, D. L.; Honda, T. J . Am. Chem. Soc. 1994, 116, 5647.

6264 J . Org. Chem., Vol. 62, No. 18, 1997
Dondoni et al.
Sch em e 4
Sch em e 5
(86% yield) by ammonolysis of the tetraacetyl carbonate
derivative 23.
In conclusion the thiazole-based homologation of L-
xylose set the basis not only for the preparative scale
synthesis of the rare monosaccharides L-gulose (1) and
L-idose (2) but also for an expeditious approach to the
important disaccharide 13. A key intermediate in these
syntheses is the sugar-thiazole adduct anti-6 whose free
hydroxyl group can be manipulated in different ways
while the other protected hydroxyl groups remain unaf-
fected. Hence this stereoselective one-carbon chain-
elongation of sugars offers the advantage over multicar-
bon processes30 to permit a differentiation of the hydroxyl
groups. It is worth noting that anti-6 is conceptually
similar to the allyl alcohol derivative employed as masked
L-gulose glycosyl acceptor in a recent synthetic approach31
to the disaccharide moiety of bleomycin A2 via glycosy-
lation with an activated 3-O-carbamoylmannopyranose
derivative. This suggests the use of anti-6 in this
approach as well.
tected as the O-benzyl ether 18 (97%) by benzylation
(BnBr, NaH, DMF) at 0 °C (Scheme 4). We noticed that
the isolation of this compound, particularly in gram scale
runs, required some simple operations to avoid the
formation of considerable amounts of side products.27
Then compound 18 was converted quite easily into the
tetra-O-acetyl-2-O-benzyl-L-gulopyranoside 20 (70%)
through a one-pot sequence of transformations involving
the cleavage of the thiazole ring into the formyl group
by the standard protocol,18 the removal of the isopropyl-
idene protective groups by acid hydrolysis, and finally
the exhaustive acetylation. Compound 20 proved to be
a mixture of â-L and R-L anomers in ca. 80:20 ratio by
1H NMR analysis, contaminated by ca. 3% of furanosides.
The formation of furanosides could not be suppressed
under different conditions of acetylation nor were these
compounds removable by chromatography. Therefore 20
was subjected to debenzylation by hydrogenolysis over
Pd to give cleanly 1,3,4,5-tetra-O-acetyl-L-gulopyranose
(21) as a ca. 80:20 mixture of â-L and R-L anomers.
Without deliberate purification of this compound,28 the
glycosidation coupling (TMSOTf, 0 °C, CH2Cl2, 10 min)
with the mannosyl diethyl phosphate 25 (Scheme 5)
afforded the peracetylated R-D-linked disaccharide 13 in
a rewarding 90% isolated yield.29 The NMR data for 13
were in agreement with those reported.22 The exclusive
formation of the R-D-linkage in this glycosidation as well
can be attributed to the neighboring acetyl group par-
ticipation in the addition of 21 to the oxycarbenium ion
intermediate derived from 25. The calculated overall
yield of 13 from L-xylose was 23.4% (nine steps). It is
worth noting that this route to 13 features also an
improved synthesis of the 3-O-carbamoylmannopyranosyl
donor 25 since its precursor 24 was prepared in one step
Exp er im en ta l Section
All moisture-sensitive reactions were performed under a
nitrogen atmosphere using oven-dried glassware. All solvents
were dried over standard drying agents32 and freshly distilled
prior to use. Commercially available powdered 4-Å molecular
sieves (50 µm average particle size) were used without further
activation. Flash column chromatography33 was performed on
silica gel 60 (230-400 mesh). Reactions were monitored by
TLC on silica gel 60 F254 with detection by charring with
sulfuric acid. Optical rotations were measured at 20 ( 2 °C
in the stated solvent. 1H (300 MHz), 13C (75 MHz), and 31P
(121 MHz) NMR were recorded at rt for CDCl3 solutions,
unless otherwise specified. Assignments were aided by homo-
and heteronuclear two-dimensional experiments.
2,3:4,5-Di-O-isop r op ylid en e-L-xylose Dieth yl Dith io-
a ceta l (3). To a suspension of L-xylose (5.00 g, 33.3 mmol) in
concentrated HCl (2.0 mL) was added, with vigorous magnetic
stirring, ethanethiol (7.4 mL, 99.9 mmol). Stirring was
continued at room temperature until the two-layer mixture
gave an homogeneous solution (usually after 15 min) which
was diluted with acetone (100 mL). After 5 h, the solution was
neutralized with 28% aqueous NH4OH and coevaporated under
high vacuum several times with toluene in order to remove
water and most of acetone diethyl dithioketal. The residue
was eluted from a column of silica gel with cyclohexane-
CH2Cl2 (from 3:1 to 1:1) to give 3 (7.40 g, 66%) as a syrup:
(27) It is recommended to draw out DMF by repeated washing with
water of Et2O solutions. The evaporation of DMF from the basic
reaction medium under vacuum at 50-60 °C induced deacetonization
with elimination to an allyl alcohol derivative such as:
(28) Crude 21 upon flash chromatography over silica gel gave a
complex mixture of products arising from intramolecular transesteri-
fication reactions.
(29) Our product contained 5% of the R-anomer at the L-gulose
moiety. Since the acceptor 21 was used as a 80:20 mixture of â-L and
R-L anomers, this result indicates that considerable anomerization of
the L-gulose unit took place during the TMSOTf-promoted glycosyla-
tion.
(30) Oshitari, T.; Kobayashi, S. Tetrahedron Lett. 1995, 36, 1089.
(31) For a remarkable synthesis of L-sugars by a two-carbon chain
elongation method see: Ko, S. Y.; Lee, A. W. M.; Masamune, S.; Reed,
L. A.; Sharpless, K. B.; Walker, F. J . Tetrahedron 1990, 46, 245.
(32) Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory
Chemicals, 3rd ed.; Pergamon Press: Oxford, 1988.
(33) Still, W. C.; Kahn, M.; Mitra, A. J . Org. Chem. 1978, 43, 2923.

Carbohydrate Homologation with 2-(Trimethylsilyl)thiazole
J . Org. Chem., Vol. 62, No. 18, 1997 6265
[R]D ) +51.3 (c 1.8, C6H6); lit.34 for the D-isomer: [R]D ) -51.25
(c 2.99, C6H6). 1H NMR (C6D6): δ 4.71 (dd, 1 H, J 1,2 ) 4.5, J 2,3
added. The suspension was stirred at room temperature for
15 min and then concentrated to dryness. To a cooled (0 °C),
stirred suspension of the crude N-methylthiazolium salt in
CH3OH (50 mL) was added NaBH4 (0.42 g, 11.0 mmol). The
mixture was stirred at room temperature for an additional 5
min, diluted with acetone (5 mL), filtered through a pad of
Celite, and concentrated. A solution of the crude mixture of
diastereomeric thiazolidines in CH3CN (45.5 mL) and H2O (4.5
mL) was treated, under vigorous stirring, with CuO (3.18 g,
40.0 mmol) and then CuCl2‚2H2O (0.85 g, 5.0 mmol). The
mixture was stirred at room temperature for 15 min and then
filtered through a pad of Celite and concentrated to remove
acetonitrile and most of the water (bath temperature not
exceeding 40 °C); the brown residue was triturated with Et2O
(4 × 50 mL), and the liquid phase was pipetted and filtered
through a pad (6 x 1.5 cm, d × h) of Florisil (100-200 mesh)
to afford a colorless solution. After a further washing of
Florisil with AcOEt (50 mL), the combined organic phases were
concentrated to yield almost pure (NMR analysis) aldehyde
anti-8 (1.65 g, ∼88%) as a syrup. 1H NMR: δ 9.67 (d, 1 H,
J 1,2 ) 1.6 Hz, H-1), 4.23-3.85 (m, 6 H), 1.42, 1.41, and 1.37 (3
s, 12 H, 4 CH3), 0.98 (t, 9 H, J ) 7.2 Hz, 3 CH3CH2), 0.65 (q,
6 H, 3 CH3CH2).
L-Gu lose (1). A solution of crude anti-8 (1.65 g, ∼4.4 mmol)
in AcOH (24 mL) and H2O (6 mL) was stirred at 100 °C for 1
h and then concentrated by coevaporation with toluene to give
1 slightly contaminated by uncharacterized byproducts: [R]D
after 3 h ) +21.2 (c 1.1, H2O). This mixture was eluted from
a column of silica gel with 85:15 AcOEt-CH3OH and then, in
order to remove contaminating colloidal silica, from a column
of Sephadex LH-20 (2 × 60 cm) with 2:1 CH3OH-CH2Cl2 to
afford pure 1 (0.50 g, 56% from anti-7) as an amorphous
solid: [R]D after 3 h ) +23.4 (c 1.1, H2O); lit.11 [R]D ) +23.3 (c
1.6, H2O).
(2S,3R,4S)-2,3,4,5-Tet r a h yd r oxy-2,3:4,5-d i-O-isop r o-
p ylid en e-1-C-(2-th ia zolyl)-1-p en ta n on e (9). To a cooled
(-78 °C), stirred solution of freshly distilled oxalyl chloride
(0.64 mL, 7.5 mmol) in anhydrous CH2Cl2 (5 mL) was added
dropwise a solution of freshly distilled DMSO (1.06 mL, 15.0
mmol) in anhydrous CH2Cl2 (5 mL). During the addition the
internal temperature was kept below -70 °C and then allowed
to reach -65 °C in 15 min. To this solution was added
dropwise a solution of 6 (1.58 g, 5.0 mmol, 95:5 anti/syn
mixture) in anhydrous CH2Cl2 (15 mL) (internal temperature
not exceeding -50 °C). The mixture was stirred at -50 °C
for 5 min, and then diluted with anhydrous Et3N (3.48 mL,
25.0 mmol), stirred for an additional 5 min, warmed to 0 °C
in 10 min, poured into a 1 M phosphate buffer (pH ) 7), and
extracted with CH2Cl2 (2 × 50 mL). The organic phase was
dried (Na2SO4) and concentrated to give almost pure 9 (1.57
g, ∼100%) suitable for the next step. An analytical sample
was obtained by column chromatography (4:1 cyclohexane-
AcOEt): [R]D ) +40.9 (c 0.9, CHCl3). 1H NMR (C6D6): δ 7.35
and 6.49 (2 d, 2 H, J ) 3.2 Hz, Th), 5.88 (d, 1 H, J 2,3 ) 6.5 Hz,
H-2), 4.42 (ddd, 1 H, J 3,4 ) 3.2, J 4,5a ) J 4,5b ) 7.0 Hz, H-4),
4.32 (d, 1 H, H-3), 3.97 (dd, 1 H, J 5a,5b ) 8.5 Hz, H-5a), 3.78
(dd, 1 H, H-5b), 1.54, 1.53, 1.44, and 1.35 (4 s, 12 H, 4 CH3).
Anal. Calcd for C14H19NO5S: C, 53.66; H, 6.11; N, 4.47.
Found: C, 53.55; H, 6.08; N, 4.46.
) 7.5 Hz, H-2), 4.31 (ddd, 1 H, J 3,4 ) 2.5, J 4,5a ) J 4,5b
)
7.5 Hz, H-4), 4.21 (dd, 1 H, H-3), 4.02 (dd, 1 H, J 5a,5b ) 7.5
Hz, H-5a), 3.94 (d, 1 H, H-1), 3.77 (dd, 1 H, H-5b), 2.78-2.48
(m, 4 H, 2 CH3CH2), 1.49, 1.45, 1.43, and 1.33 (4 s, 12 H, 4
CH3), 1.10 and 1.08 (2 t, 6 H, J ) 7.5 Hz, 2 CH3CH2). 13C
NMR selected data: δ 110.0 and 109.5 (2 OCO), 27.3, 27.1,
26.1, 25.6, 25.3, and 24.9 (4 CH3, 2 CH3CH2), 14.3 and 14.2 (2
CH3CH2).
2,3:4,5-Di-O-isop r op ylid en e-a ld eh yd o-L-xylose (4).
A
stirred solution of 3 (6.00 g, 17.8 mmol) in acetone (60 mL)
was diluted with H2O (6.0 mL) and treated with yellow
mercury(II) oxide (8.88 g, 41.0 mmol) and mercury(II) chloride
(8.71 g, 32.1 mmol). The mixture was stirred at 55 °C for 2 h
and then cooled to room temperature, filtered through a pad
of Celite, and concentrated. The residue was suspended in
CH2Cl2 (3 × 100 mL) and filtered through a pad of Celite. The
solution was washed with 1 M aqueous KI (100 mL), dried
(Na2SO4), and concentrated to give almost pure aldehyde 4
(3.53 g, ∼86%): [R]D ) +26.1 (c 1.9, EtOH); lit.35 [R]D ) +25.6
(c 3.2, EtOH). 1H NMR: δ 9.79 (s, 1 H, H-1), 4.25 (ddd, 1 H,
J 3,4 ) 5.0, J 4,5a ) J 4,5b ) 6.8 Hz, H-4), 4.24 (d, 1 H, J 2,3 ) 7.1
Hz, H-2), 4.14 (dd, 1 H, H-3), 4.08 (dd, 1 H, J 5a,5b ) 8.6 Hz,
H-5a), 3.88 (dd, 1 H, H-5b), 1.53, 1.45, 1.42, and 1.39 (4 s, 12
H, 4 CH3).
(2S,3R,4R,5S)-1,2:3,4-Di-O-isop r op ylid en e-5-C-(2-t h i-
a zolyl)p en ta n e-1,2,3,4,5-p en tol (a n ti-6). To a cooled (-20
°C), stirred solution of crude 4 (3.53 g, ∼15.3 mmol) in
anhydrous CH2Cl2 (60 mL) was added 2-(trimethylsilyl)thi-
azole (5, 3.2 mL, 19.9 mmol) during 15 min. The solution was
stirred at 0 °C for an additional 1 h and then concentrated. A
solution of the residue in anhydrous THF (60 mL) was treated
with n-Bu4NF‚3H2O (4.84 g, 15.3 mmol) at room temperature
for 30 min and then concentrated. The residue was dissolved
in CH2Cl2 (300 mL), washed with H2O (3 × 50 mL), dried
(Na2SO4), and concentrated to give 6 (4.50 g, 80% from 3, anti/
syn ) 95:5) as a white solid. Repeated crystallization of this
mixture with AcOEt-cyclohexane afforded pure anti-6 (3.42
g, 61% from 3): mp 146-148 °C; [R]D ) +18.5 (c 1.1, CHCl3).
1H NMR: δ 7.79 and 7.38 (2 d, 2 H, J ) 3.2 Hz, Th), 5.13 (dd,
1 H, J 1,OH ) 4.0, J 1,2 ) 6.0 Hz, H-1), 4.33 (dd, 1 H, J 2,3 ) 7.5
Hz, H-2), 4.20 (dd, 1 H, J 3,4 ) 3.8 Hz, H-3), 3.90 (dd, 1 H, J 4,5a
) 6.5, J 5a,5b ) 7.5 Hz, H-5a), 3.85 (ddd, 1 H, J 4,5b ) 6.5 Hz,
H-4), 3.73 (dd, 1 H, H-5b), 3.55 (d, 1 H, OH), 1.48, 1.46, 1.42,
and 1.34 (4 s, 12 H, 4 CH3). Anal. Calcd for C14H21NO5S: C,
53.32; H, 6.71; N, 4.44. Found: C, 53.43; H, 6.73; N, 4.43.
(2S,3R,4S,5S)-1,2:3,4-Di-O-isop r op ylid en e-5-C-(2-t h i-
a zolyl)-5-O-(tr ieth ylsilyl)p en ta n e-1,2,3,4,5-p en tol (a n ti-
7). To a stirred solution of anti-6 (2.00 g, 6.34 mmol) and
4-(N,N-dimethylamino)pyridine (0.20 g, 1.64 mmol) in pyridine
(15 mL) was added triethylsilyl chloride (3.2 mL, 19.0 mmol).
The mixture was stirred at room temperature for an additional
30 min, diluted with CH3OH (2 mL), and, after 30 min,
concentrated. The residue was suspended in CH2Cl2 (200 mL),
washed with H2O (30 mL), dried (Na2SO4), and concentrated.
The residue was filtered through a short column (5 × 8 cm, d
× h) of silica gel with 4:1 cyclohexane-Et2O (containing 0.3%
of Et3N) to give anti-7 (2.40 g, 88%) as a syrup: [R]D ) -9.9 (c
1.1, CHCl3). 1H NMR: δ 7.78 and 7.34 (2 d, 2 H, J ) 3.2 Hz,
Th), 5.22 (d, 1 H, J 1,2 ) 4.0 Hz, H-1), 4.42 (dd, 1 H, J 2,3 ) 7.0
Hz, H-2), 4.17-4.13 (m, 1 H, H-3), 3.83-3.79 (m, 1 H, H-5a),
3.69-3.61 (m, 2 H, H-4, H-5b), 1.42, 1.39, 1.35, and 1.30 (4 s,
12 H, 4 CH3), 0.94 (t, 9 H, J ) 7.2 Hz, 3 CH3CH2), 0.64 (q, 6
H, 3 CH3CH2). Anal. Calcd for C20H35NO5SSi: C, 55.91; H,
8.21; N, 3.26. Found: C, 56.07; H, 8.23; N, 3.25.
The oxidation of a ca. 1:1 anti/syn-6 mixture led to similar
results.
(2S,3R,4R,5R)-1,2:3,4-Di-O-isop r op ylid en e-5-C-(2-t h i-
a zolyl)p en ta n e-1,2,3,4,5-p en tol (syn -6). To a cooled (-78
°C), stirred solution of crude 9 (1.57 g, ∼5.0 mmol) in CH3OH
(25 mL) was added NaBH4 (208 mg, 5.5 mmol). The mixture
was stirred at -78 °C for an additional 40 min and then
diluted with acetone (2 mL), warmed to room temperature,
and concentrated. The residue was suspended in CH2Cl2 (150
mL), washed with H2O (2 × 20 mL), dried (Na2SO4), and
concentrated to give a 91:9 mixture of syn-6 and anti-6 (1.42
g, 90%). 1H NMR: δ 7.78 and 7.36 (2 d, 2 H, J ) 3.2 Hz, Th),
5.03 (dd, 1 H, J 1,OH ) 7.2, J 1,2 ) 3.9 Hz, H-1), 4.47 (dd, 1 H,
J 2,3 ) 7.8 Hz, H-2), 4.21 (dd, 1 H, J 3,4 ) 4.0 Hz, H-3), 4.07
3,4:5,6-Di-O-isopr opyliden e-2-O-(tr ieth ylsilyl)-a ldeh ydo-
L-gu lose (a n ti-8). A mixture of anti-7 (2.15 g, 5.0 mmol),
activated 4-Å powdered molecular sieves (10.0 g), and anhy-
drous CH3CN (50 mL) was stirred at room temperature for
10 min, and then methyl triflate (0.74 mL, 6.5 mmol) was
(34) Kochetkov, N. K.; Dmitriev, A. Tetrahedron 1965, 21, 803.
(35) Bourne, E. J .; McSweeney, G. P.; Wiggins, L. F. J . Chem. Soc.
1952, 3113.
(ddd, 1 H, J 4,5a ) J 4,5b ) 6.5 Hz, H-4), 4.01 (dd, 1 H, J 5a,5b
)
8.1 Hz, H-5a), 3.81 (dd, 1 H, H-5b), 3.49 (d, 1 H, OH), 1.48,
1.46, 1.42, and 1.35 (4 s, 12 H, 4 CH3).
(36) Wiggins, L. F. Methods Carbohydr. Chem. 1962, 1, 140.

6266 J . Org. Chem., Vol. 62, No. 18, 1997
Dondoni et al.
(2S ,3R ,4S ,5R )-1,2:3,4-D i-O -is o p r o p y lid e n e -5-C-(2-
th ia zolyl)-5-O-(tr ieth ylsilyl)p en ta n e-1,2,3,4,5-p en tol (syn -
7). A 91:9 syn/anti-6 mixture (1.26 g, 4.0 mmol) was silylated
as described for the preparation of anti-7. Column chroma-
tography of the residue (85:15 cyclohexane-Et2O) gave first
a 2:1 anti/syn-7 mixture (0.30 g). Eluted second was pure syn-7
(1.29 g, 82%) as a syrup: [R]D ) +50.1 (c 0.9, CHCl3). 1H
NMR: δ 7.78 and 7.34 (2 d, 2 H, J ) 3.2 Hz, Th), 5.18-5.12
(m, 1 H, H-1), 4.24-4.18 (m, 2 H, H-2, H-3), 3.93 (dd, 1 H,
J 4,5a ) 6.6, J 5a,5b ) 7.8 Hz, H-5a), 3.83-3.76 (m, 1 H, H-4),
3.70 (dd, 1 H, J 4,5b ) 7.8 Hz, H-5b), 1.42, 1.39, 1.35, and
1.30 (4 s, 12 H, 4 CH3), 0.94 (t, 9 H, J ) 7.5 Hz, 3 CH3CH2),
0.65 and 0.64 (2 q, 6 H, 3 CH3CH2). Anal. Calcd for
C20H35NO5SSi: C, 55.91. H, 8.21; N, 3.26. Found: C, 55.91;
H, 8.20; N, 3.25.
3,4:5,6-Di-O-isopr opyliden e-2-O-(tr ieth ylsilyl)-a ldeh ydo-
L-id ose (syn -8). Treatment of syn-7 (1.29 g, 3.0 mmol) as
described for the preparation of anti-8 afforded almost pure
aldehyde syn-8 (0.98 g, ∼87%) as a syrup. 1H NMR: δ 9.72
(s, 1 H, CHO), 4.27 (dd, 1 H, J 2,3 ) 3.2, J 3,4 ) 7.9 Hz, H-3),
4.23 (ddd, 1 H, J 4,5 ) 4.3, J 5,6a ) 6.7, J 5,6b ) 7.0 Hz, H-5), 4.14
(dd, 1 H, H-4), 4.13 (d, 1 H, H-2), 4.04 (dd, 1 H, J 6a,6b ) 8.2
Hz, H-6a), 3.87 (dd, 1 H, H-6b), 1.43, 1.42, 1.41, and 1.36 (4 s,
12 H, 4 CH3), 0.97 (t, 9 H, J ) 7.2 Hz, 3 CH3CH2), 0.64 (q, 6
H, 3 CH3CH2).
H2O (30 mL), and extracted with Et2O (2 × 100 mL). The
combined organic phases were dried (Na2SO4) and concen-
trated. A solution of the crude tetrabenzyl derivative in
anhydrous THF (20 mL) was treated with n-Bu4NF‚3H2O (517
mg, 1.64 mmol) at room temperature for 1.5 h and then
concentrated. The residue was dissolved in CH2Cl2 (80 mL),
washed with H2O (2 × 10 mL), dried (Na2SO4), and concen-
trated to give crude 17. The residue was eluted from a column
of silica gel with 9:1 cyclohexane-AcOEt to afford 17 (443 mg,
51% from 14) as a syrup: [R]D ) -0.8 (c 0.4, CHCl3). 1H NMR
(C6D6): δ 7.50-7.05 (m, 20 H, 4 Ph), 5.13 and 4.37 (2 d, 2 H,
J ) 10.6 Hz, PhCH2), 5.09 (d, 1 H, J 1,2 ) 4.6 Hz, H-1), 4.96
and 4.51 (2 d, 2 H, J ) 12.1 Hz, PhCH2), 4.89 and 4.49 (2 d, 2
H, J ) 11.7 Hz, PhCH2), 4.55 and 4.28 (2 d, 2 H, J ) 11.6 Hz,
PhCH2), 4.18 (dd, 1 H, J 3,4 ) 4.2, J 4,5 ) 8.5 Hz, H-4), 4.07 (ddd,
1 H, J 5,6a ) 3.3, J 5,6b ) 4.1 Hz, H-5), 3.80 (dd, 1 H, J 6a,6b
)
11.6 Hz, H-6a), 3.79 (dd, 1 H, J 2,3 ) 4.5 Hz, H-3), 3.67 (dd, 1
H, H-6b), 3.56 (dd, 1 H, H-2), 2.00 (bs, 1 H, OH). 1H NMR
(C6D6 + Cl3CC(O)NCO) selected data: δ 8.00 (s, 1 H, NH),
4.99 (d, 1 H, J 1,2 ) 4.6 Hz, H-1), 4.50 (dd, 1 H, J 5,6a ) 2.1,
J 6a,6b ) 11.6 Hz, H-6a), 4.41 (dd, 1 H, J 5,6b ) 4.5 Hz, H-6b),
4.15 (dd, 1 H, J 3,4 ) 4.4, J 4,5 ) 8.5 Hz, H-4), 4.10 (ddd, 1 H,
H-5), 3.84 (dd, 1 H, J 2,3 ) 4.6 Hz, H-3), 3.49 (dd, 1 H, H-2).
Anal. Calcd for C34H36O6: C, 75.53; H, 6.71. Found: C, 75.76;
H, 6.79.
L-Id op yr a n ose a n d L-Id ofu r a n ose (2 a n d 2a ). Aldehyde
syn-8 (0.98 g, ∼2.6 mmol) was treated with 80% AcOH as
described for the synthesis of 1 to afford 2 and 2a contami-
nated by trace amounts of uncharacterized byproducts: [R]D
after 3 h ) -13.9 (c 0.9, H2O). Similar chromatographic
purification gave pure 2 and 2a (0.32 g, 59% from syn-7) as
an amorphous solid: [R]D after 3 h ) -15.8 (c 1.1, H2O); lit.12a
[R]D ) -13.0 (c 1.4, H2O); lit.12c [R]D ) -21.0 (c 1.0, H2O); lit.36
for the D-isomer [R]D ) +16 ( 1 (c 2.3, H2O).
(1S,2R,3R,4S)-1-O-Ben zyl-2,3:4,5-d i-O-isop r op ylid en e-
1-C-(2-th ia zolyl)p en ta n e-1,2,3,4,5-p en tol (18). To a cooled
(0 °C), stirred solution of anti-6 (1.58 g, 5.0 mmol) in DMF (15
mL) were added portionwise NaH (0.40 g, 10.0 mmol, of a 60%
dispersion in oil) and, after 30 min, benzyl bromide (0.89 mL,
7.5 mmol). The mixture was stirred at room temperature for
30 min and then treated with CH3OH (1 mL), stirred for an
additional 10 min, diluted with H2O (30 mL), and extracted
with Et2O (2 × 100 mL). The combined organic phases were
dried (Na2SO4) and concentrated. The residue was eluted from
a column of silica gel with 9:1 toluene-AcOEt to give 18 (1.97
g, 97%) as a white solid: mp 80-81 °C (from i-Pr2O-
cyclohexane); [R]D ) -32.3 (c 1.1, CHCl3). 1H NMR: δ 7.83
and 7.42 (2 d, 2 H, J ) 3.2 Hz, Th), 7.38-7.29 (m, 5 H, Ph),
4.88 (d, 1 H, J 1,2 ) 4.8 Hz, H-1), 4.71 and 4.52 (2 d, 2 H, J )
12.1 Hz, PhCH2), 4.42 (dd, 1 H, J 2,3 ) 7.0 Hz, H-2), 4.04 (dd,
1 H, J 3,4 ) 4.5 Hz, H-3), 3.98 (ddd, 1 H, J 4,5a ) 6.2, J 4,5b ) 7.3
Hz, H-4), 3.87 (dd, 1 H, J 5a,5b ) 8.2 Hz, H-5a), 3.75 (dd, 1 H,
H-5b), 1.42, 1.38, 1.35, and 1.28 (4 s, 12 H, 4 CH3). Anal. Calcd
for C21H27NO5S: C, 62.20; H, 6.71; N, 3.45. Found: C, 62.37;
H, 6.72; N, 3.44.
2-O-Ben zyl-3,4:5,6-d i-O-isop r op ylid en e-a ld eh yd o-L-gu -
lose (19). Treatment of 18 (1.63 g, 4.0 mmol) as described
for the preparation of anti-8 gave almost pure aldehyde 19
(1.26 g, ∼90%) as a syrup. 1H NMR: δ 9.71 (d, 1 H, J 1,2 ) 2.0
Hz, H-1), 7.44-7.26 (m, 5 H, Ph), 4.76 and 4.59 (2 d, 2 H, J )
11.9 Hz, PhCH2), 4.26 (dd, 1 H, J 2,3 ) 5.2, J 3,4 ) 7.5 Hz, H-3),
4.12 (ddd, 1 H, J 4,5 ) 4.1, J 5,6a ) J 5,6b ) 6.5 Hz, H-5), 4.03 (dd,
1 H, H-4), 3.96 (dd, 1 H, J 6a,6b ) 7.9 Hz, H-6a), 3.85 (dd, 1 H,
H-2), 3.83 (dd, 1 H, H-6b), 1.43, 1.41, 1.38, and 1.37 (4 s, 12
H, 4 CH3).
1,3,4,6-Tetr a -O-a cetyl-2-O-ben zyl-L-gu lop yr a n ose (20).
A solution of crude 19 (1.26 g, ∼3.6 mmol) in AcOH (20 mL)
and H2O (5 mL) was stirred at 100 °C for 40 min and then
concentrated by coevaporation with toluene to give crude 2-O-
benzyl-L-gulose as a 79:17:4 mixture of â-pyranose, R-pyranose,
and furanose forms. 1H NMR (D2O) selected data: δ 7.33-
7.06 (m, Ph), 5.21 (d, J 1,2 ) 4.0 Hz, H-1f), 5.06 (d, J 1,2 ) 3.8
Hz, H-1Rp), 4.78 (d, J 1,2 ) 8.2 Hz, H-1âp), 4.56 and 4.52 (2 d,
J ) 11.0 Hz, PhCH2âp), 3.99 (dd, J 2,3 ) 3.2, J 3,4 ) 3.4 Hz,
H-3âp), 3.82 (ddd, J 5,4 ) 1.0, J 5,6a ) J 5,6b ) 6.3 Hz, H-5âp),
3.61 (dd, H-4âp), 3.55 (d, 2 H-6âp), 3.36 (dd, H-2âp). A solution
of the residue and 4-(N,N-dimethylamino)pyridine (0.44 g, 3.6
mmol) in pyridine (9 mL) and acetic anhydride (9 mL) was
kept at room temperature for 6 h and then concentrated. The
residue was eluted from a column of silica gel with 85:15
toluene-AcOEt to give 20 (1.22 g, 70% from 18) as a 77:20:3
mixture of â-pyranose, R-pyranose, and furanose forms. 1H
NMR selected data: δ 7.36-7.25 (m, Ph), 6.26 (d, J 1,2 ) 3.8
Hz, H-1Rp), 6.23 (d, J 1,2 ) 2.4 Hz, H-1f), 5.93 (d, J 1,2 ) 8.5 Hz,
H-1âp), 5.49 (dd, J 2,3 ) 3.5, J 3,4 ) 3.8 Hz, H-3âp), 4.99 (dd,
(2S,3R,4S,5S)-1,2:3,4-Di-O-isop r op ylid en e-5-O-(ter t-bu -
tyld ip h en ylsilyl)-5-C-(2-th ia zolyl)p en ta n e-1,2,3,4,5-p en -
tol (14). To a warmed (80 °C), stirred solution of anti-6 (630
mg, 2.00 mmol) and imidazole (817 mg, 12.00 mmol) in
anhydrous DMF (5 mL) was added tert-butyldiphenylsilyl
chloride (15.4 mL, 6.00 mmol). The mixture was stirred at 80
°C for 18 h and then cooled to room temperature, treated with
CH3OH (1 mL), stirred for an additional 30 min, diluted with
H2O (20 mL), and extracted with Et2O (2 × 125 mL). The
combined organic phases were dried (Na2SO4) and concen-
trated. The residue was eluted from a column of silica gel with
cyclohexane-Et2O (4:1 and then 1.5:1) to give 14 (1.095 g, 99%)
as a syrup: [R]D ) +1.7 (c 1.7, CHCl3). 1H NMR: δ 7.80-
7.74, 7.68-7.65, 7.60-7.54, and 7.52-7.26 (4 m, 12 H, 2 Ph
and Th), 5.11 (d, 1 H, J 1,2 ) 3.0 Hz, H-1), 4.24 (dd, 1 H, J 2,3
)
8.0 Hz, H-2), 4.02 (dd, 1 H, J 3,4 ) 4.2 Hz, H-3), 3.68-3.46 (m,
3 H, H-4, H-5a, H-5b), 1.36, 1.33, 1.29, and 1.05 (4 s, 12 H, 4
CH3), 1.11 (s, 9 H, t-Bu). Anal. Calcd for C30H39NO5SSi: C,
65.07; H, 7.10; N, 2.53. Found: C, 65.24; H, 7.11; N, 2.51.
2-O-(ter t-Bu t yld ip h en ylsilyl)-3,4:5,6-d i-O-isop r op yl-
id en e-a ld eh yd o-L-gu lose (15). Treatment of 14 (1.00 g, 1.81
mmol) as described for the preparation of anti-8 gave almost
pure aldehyde 15 (0.85 g, ∼96%) as a syrup. 1H NMR: δ 9.60
(s, 1 H, H-1), 7.75-7.68, 7.67-7.60, and 7.51-7.35 (3 m, 10
H, 2 Ph), 4.23 (dd, 1 H, J 2,3 ) 3.2, J 3,4 ) 8.0 Hz, H-3), 4.11 (d,
1 H, H-2), 3.77 (dd, 1 H, J 4,5 ) 3.6 Hz, H-4), 3.83 (ddd, 1 H,
J 5,6a ) 6.4, J 5,6b ) 7.8 Hz, H-5), 3.71 and 3.66 (2 dd, 2 H, J 6a,6b
) 8.2 Hz, H-6a, H-6b), 1.39, 1.37, and 1.34 (3 s, 12 H, 4 CH3),
1.14 (s, 9 H, t-Bu).
2-O-(ter t-Bu tyld ip h en ylsilyl)-L-gu lop yr a n ose (16). Al-
dehyde 15 (800 mg, ∼1.64 mmol) was treated with 80% AcOH
as described for the synthesis of 1 to give a ∼7:3 mixture of
16 and regioisomers (686 mg, ∼100%). 1H NMR (acetone-d6
+ D2O) selected data: δ 5.05 (d, 0.5 H, J 1,2 ) 8.2 Hz, H-1âp),
4.82 (d, 0.5 H, J 1,2 ) 3.6 Hz, H-1Rp).
Ben zyl 2,3,5-Tr i-O-ben zyl-r-L-gu lofu r a n osid e (17). To
a cooled (0 °C), stirred solution of crude 16 and regioisomers
(686 mg, ∼1.64 mmol) in DMF (10 mL) were added portionwise
NaH (393 mg, 9.84 mmol, of a 60% dispersion in oil) and, after
30 min, benzyl bromide (935 µL, 7.87 mmol). The mixture was
stirred at room temperature for 30 min and then treated with
CH3OH (1 mL), stirred for an additional 10 min, diluted with

Carbohydrate Homologation with 2-(Trimethylsilyl)thiazole
J . Org. Chem., Vol. 62, No. 18, 1997 6267
J 4,5 ) 1.4 Hz, H-4âp), 4.67 and 4.53 (2 d, J ) 11.9 Hz,
PhCH2âp), 4.32 (ddd, J 5,6a ) 5.8, J 5,6b ) 7.0 Hz, H-5âp), 4.15
(dd, J 6a,6b ) 11.0 Hz, H-6aâp), 4.09 (dd, H-6bâp), 3.88 (dd,
H-2âp), 2.15, 2.11, 2.09, and 2.05 (4 s, 4 CH3âp). 13C NMR
selected data: δ 91.4 (C-1âp), 91.0 and 89.4 (C-1Rp, C-1f). Anal.
Calcd for C21H26O10: C, 57.53; H, 5.98. Found: C, 57.68; H,
5.98.
When the acetylation of crude 2-O-benzyl-L-gulose was
carried out without 4-(N,N-dimethylamino)pyridine the fura-
nose tetraacetates were formed at a larger extent. Also other
acetylation procedures (THF, Ac2O, Et3N, rt; Ac2O, AcONa,
130 °C) gave appreciable amounts of acetylated L-gulofura-
noses.
1,3,4,6-Tetr a -O-a cetyl-L-gu lop yr a n ose (21). A vigorously
stirred mixture of 20 (1.32 g, 3.0 mmol), 10% palladium on
activated carbon (0.66 g), and AcOEt (20 mL) was degassed
under vacuum and saturated with hydrogen (by a H2-filled
balloon) three times. The suspension was stirred at room
temperature for 2 h under a slightly positive pressure of H2
2,4,6-Tr i-O-a cet yl-3-O-ca r b a m oyl-r-D-m a n n op yr a n o-
syl Dieth yl P h osp h a te (25). To a cooled (-78 °C), stirred
solution of hemiacetal 24 (254 mg, 0.73 mmol) in anhydrous
THF (20 mL) was added n-BuLi (0.6 mL, 0.87 mmol of a 1.6
M solution in hexanes). The solution was stirred at -78 °C
for 10 min before the addition of freshly distilled diethyl
chlorophosphate (126 µL, 0.87 mmol). The reaction mixture
was stirred at -78 °C for an additional 10 min and then poured
into a mixture of AcOEt (8 mL) and saturated aqueous
NaHCO3 (5 mL) with vigorous stirring. The organic layer was
separated, washed with brine (3 mL), dried (Na2SO4), and
concentrated. The residue was eluted from a column of silica
gel with 6:4 AcOEt-cyclohexane (containing 5% of Et3N) to
give 25 (261 mg, 74%) as a white solid: mp 121-122 °C (from
AcOEt-cyclohexane); [R]D ) +0.3 (c 1.2, CHCl3). 1H NMR: δ
5.64 (dd, 1 H, J 1,2 ) 1.8, J 1,P ) 6.8 Hz, H-1), 5.38-5.24 (m, 3
H), 4.73 (bs, 2 H, NH2), 4.31 (dd, 1 H, J ) 4.8, J ) 11.9 Hz,
H-6a), 4.25-4.07 (m, 6 H), 2.17, 2.09, and 2.07 (3 s, 9 H, 3
CH3), 1.36 (t, 6 H, J ) 7.0 Hz, 2 CH3CH2). 31P NMR: δ -3.14.
Anal. Calcd for C17H28NO13P: C, 42.07; H, 5.81; N, 2.89.
Found: C, 42.06; H, 5.81; N, 2.88.
(balloon) and then filtered through
a pad of Celite and
concentrated to afford 21 (0.99 g, 95%) as 77:20:3 mixture of
â-pyranose, R-pyranose, and furanose forms. 1H NMR selected
data: δ 6.21 (d, J 1,2 ) 3.9 Hz, H-1Rp), 6.16 (bs, H-1f), 5.86 (d,
J 1,2 ) 8.5 Hz, H-1âp), 5.35 (dd, J 2,3 ) 4.9, J 3,4 ) 4.6 Hz, H-3âp),
1,3,4,6-Tet r a -O-a cet yl-2-O-(2,4,6-t r i-O-a cet yl-3-O-ca r -
ba m oyl-r-D-m a n n op yr a n osyl)-â-L-gu lop yr a n ose (13). To
a cooled (0 °C), stirred solution of 21 (50 mg, 0.14 mmol) and
25 (82 mg, 0.17 mmol) in anhydrous CH2Cl2 (1 mL) was added
TMSOTf (47 µL, 0.26 mmol). The mixture was stirred at 0 °C
for 10 min and then poured into a mixture of AcOEt (5 mL)
and saturated aqueous NaHCO3 (5 mL) with vigorous stirring.
The organic layer was separated, washed with brine (3 mL),
dried (Na2SO4), and concentrated. The residue was eluted
from a column of silica gel with 7:3 AcOEt-cyclohexane to give
a 95:5 mixture of 13 and its R-L-anomer (86 mg, 90%) as a
colorless foam. The NMR data for 13 were in agreement with
those reported.22 1H NMR selected data for the R-L-anomer:
δ 6.29 (d, 1 H, J 1,2 ) 3.5 Hz, H-1).
5.03 (dd, J 4,5 ) 1.4 Hz, H-4âp), 4.30 (ddd, J 5,6a ) 5.8, J 5,6b
)
7.0 Hz, H-5âp), 4.18 (dd, J 6a,6b ) 11.0 Hz, H-6aâp), 4.11 (dd,
H-6bâp), 3.98 (dd, H-2âp), 2.34 (bs, OH), 2.18, 2.13, 2.12, and
2.04 (4 s, 4 CH3âp). 13C NMR selected data: δ 92.3 (C-1âp),
90.6 (C-1Rp). Anal. Calcd for C14H20O10: C, 48.28; H, 5.79.
Found: C, 48.15; H, 5.78.
2,4,6-T r i-O -a c e t y l-3-O -c a r b a m o y l-r-D -m a n n o p y r a -
n ose (24). Ammonia was bubbled at room temperature
through a solution of 23 (315 mg, 0.62 mmol) in anhydrous
THF (12 mL) until the TLC analysis (9:1 CH2Cl2-acetone)
revealed that the starting material and the 1,2,4,6-tetra-O-
acetyl-3-O-carbamoyl derivative intermediate had disappeared
(usually 40 min). The solution was concentrated, and the
residue was eluted by a column of silica gel with 6:4 cyclo-
hexane-AcOEt to give 24 (186 mg, 86%) as a colorless syrup.
The 1H NMR data for 24 were in agreement with those
reported.22
Ack n ow led gm en t. This work was supported by the
Progetto Strategico Tecnologie Chimiche Innovative
(CNR, Rome). We thank Mr. P. Formaglio (University
of Ferrara, Italy) for NMR measurements.
Prolonged treatment with ammonia (∼1.5 h) led to lower
yields of 24 due to further cleavage of acetyl groups.
J O970601H