β-Acarbose. VII approaches towards the synthesis of some N-linked carba-oligosaccharides

Article abstract of DOI:10.1071/CH99030

3,4-Anhydro-1,6-dideoxy-1,6-episelerio-β-D-glucose was treated with cyclohexylamine to afford an amino diol which was subsequently converted into a cyclic carbamate, a compound shown to be a moderately successful gly- cosyl donor. Treatment of the same 3,4-anhydro sugar and the 1,6-epithio analogue with a 1-epivalienamine derivative afforded the corresponding secondary amines which were converted into the analogous cyclic carbamates. The epithio analogue was unsuccessful as a glycosyl donor, failing to glycosylate a carbohydrate alcohol. On the other hand, the episeleno compound appeared to function as a glycosyl donor but decomposition of the product occurred under the conditions necessary for its isolation.

Full text of DOI:10.1071/CH99030

C S I R O P U B L I S H I N G
Australian Journal
of Chemistry
Volume 52, 1999
© CSIRO Australia 1999
A journal for the publication of original research
in all branches of chemistry and chemical technology
w w w. p u b l i s h . c s i ro . a u / j o u rn a l s / a j c
All enquiries and manuscripts should be directed to
The Managing Editor
Australian Journal of Chemistry
CSIRO PUBLISHING
PO Box 1139 (150 Oxford St)
Collingwood
Vic. 3066
Australia
Telephone: 61 3 9662 7630
Facsimile: 61 3 9662 7611
Email: john.zdysiewicz@publish.csiro.au
Published by CSIRO PUBLISHING
for CSIRO Australia and
the Australian Academy of Science

Aust. J. Chem., 1999, 52, 885–894
-Acarbose. VII*
Approaches Towards the Synthesis
of Some N-Linked Carba-Oligosaccharides
Robert V. Stick,A,B D. Matthew G. TilbrookA and Spencer J. WilliamsA
A Department of Chemistry, The University of Western Australia, Nedlands, W.A. 6907.
B Author to whom correspondence should be addressed.
3,4-Anhydro-1,6-dideoxy-1,6-episeleno-␤-D-glucose was treated with cyclohexylamine to afford an amino diol
which was subsequently converted into a cyclic carbamate, a compound shown to be a moderately successful gly-
cosyl donor. Treatment of the same 3,4-anhydro sugar and the 1,6-epithio analogue with a 1-epivalienamine deriva-
tive afforded the corresponding secondary amines which were converted into the analogous cyclic carbamates. The
epithio analogue was unsuccessful as a glycosyl donor, failing to glycosylate a carbohydrate alcohol. On the other
hand, the episeleno compound appeared to function as a glycosyl donor but decomposition of the product occurred
under the conditions necessary for its isolation.
Beginning in the early 1970s, a novel group of secondary
metabolites with inhibitory activity towards -glucosidases
was isolated from culture broths of organisms of the order
Actinomycetales. Many organisms were investigated, includ-
ing representatives of the family Actinoplanacae which
encompasses the genera Actinoplanes, Ampullariella and
Streptosporangium.2
deoxygenation (C 6) (Scheme 1). This paper describes efforts
towards this end which, ultimately, were unsuccessful.
Ring-opening reactions with the amine (6) on the two
epoxides (4) and (5) were seen as key steps in the synthesis
of methyl -acarboside (3). Initially, we chose cyclohexyl-
amine as a more accessible amine for the development of the
methodology necessary for the approach towards the carba-
tetrasaccharide (3). Previously described formative experi-
ments have shown that the epoxide (4) reacts in the desired
fashion with cyclohexylamine or sodium azide, affording the
diaxially-opened amine (7) and azide (8), respectively, in
moderate yields.7,8 These model studies on the epoxide (4)
were to be extended to include the seleno sugar (5). Further-
more, these studies were to investigate the manipulation of an
amine such as (7) into a glycosyl donor which would provide
the desired -selectivity in the glycosylation step.
A prime result of this screening was the identification of a
carba-tetrasaccharide, acarbose (1) (BAY g 5421) which was
shown to exhibit pronounced and potent inhibitory activity
against intestinal -sucraseand also to inhibit starch digestion
in vivo.3,4 Since its discovery, acarbose has been the subject
of prolific investigations which have culminated in its
approval for clinical usage in the management of non-insulin
dependent diabetes mellitus. In addition, acarbose has found
application in the treatment of insulin dependent (type I) dia-
betes, particularly in combination with insulin. Marketed
under a variety of names including Prandase, Precose and
Glucobay by Bayer in a number of countries, acarbose has
been shown to control effectively the metabolism of complex
carbohydrates and thence the production of D-glucose.5
For some time now, we have had an interest in this area as
it seemed that the -linked diastereoisomer of acarbose,
namely -acarbose (2), could be an inhibitor of -glucan
hydrolases.6 More recently, we have improved on our origi-
nal concept and decided to prepare the -anomer of the
methyl glycoside of -acarbose, namely methyl -acarbo-
side (3), it being more substrate-like.
Our initial efforts were directed towards an improvement
in the synthesis of the epoxide (4) and the preparation of the
epoxide (5) and the approach followed the published sequence
starting with D-glucose.7 Treatment of 1,2,3-tri-O-acetyl-D-
glucopyranose with methanesulfonyl chloride in pyridine
gave the dimesylates (9a,b) which were converted, upon pro-
longed treatment with hydrogen bromide in acetic acid, into
the very stable, crystalline bromide (10). Treatment of (10)
with benzyltriethylammonium tetrathiomolybdate provided
the 1,6-epithio mesylate (11).7 Alternatively, treatment of
the bromide (10) with a solution of sodium hydrogen selenide
in ethanol was equally successful, affording the 1,6-epi-
seleno mesylate (12).9
Our approach to methyl -acarboside (3) centred on the
1,6-dideoxy-1,6-epithio- and -1,6-episeleno sugars (4) and
(5), molecules beautifully arranged for the introduction of an
amine (C 4), formation of a glycosidic linkage (C 1) and
The conversion of the two mesylates (11) and (12) into the
epoxides (4) and (5) was next addressed. The procedure
described previously used sodium methoxide to remove the
*Part VI, Aust. J. Chem., 1997, 50, 225.1
Manuscript received 10 March 1999
© CSIRO 1999
0004-9425/99/090885
10.1071/CH99030

R. V. Stick, D. M. G. Tilbrook and S. J. Williams
886

-Acarbose. VII
887
acetyl groups and to form directly the 3,4-anhydro ring; neu-
tralization with acid resin followed by crystallization pro-
vided (4).7 Although effective on a small scale, this
procedure proved unsatisfactory upon scale-up owing to the
formation of copious amounts of a fine precipitate (sodium
mesylate). A better procedure involved the removal of the
bulk of the solvent and extraction of the residue with ethyl
acetate—this gave a good recovery of the epoxides (4) and
(5) which were then easily purified by flash chromatography.
There always existed the possibility that the epoxides (4)
and (5) would undergo a Payne rearrangement upon pro-
longed treatment with base.10 Indeed, when the mesylate
(12) was treated with a large excess of sodium methoxide in
methanol, t.l.c. analysis indicated the gradual conversion of
the initially formed epoxide (5) into another product.
Acetylation of the reaction mixture, after removal of the
The choice of epoxide, either (4) or (5), for the aminoly-
sis reaction now became the next issue. Given the relatively
greater reactivity of the 1,6-episeleno sugar (15) as a glyco-
syl donor, and the easier deoxygenation of the resultant
product to afford a 6-deoxy glycoside when compared to the
epithio sugar (16), it was decided to concentrate efforts in the
model series on the 1,6-episeleno epoxide (5).12 Treatment
of the episeleno sugar (5) with a fivefold excess of cyclo-
hexylamine in butanol afforded the crystalline amino diol
(17), in excellent yield.
The presence of unprotected amino and hydroxy groups in
(17) was the source of concern for the planned glycosylation
step. Triphosgene has been used previously to execute the pro-
tection of the vicinal amino alcohol (18) as the carbamate
(19).13,14 Here, the matter was complicated by the presence of
another hydroxy group at C 2. Upon treatment of the amino
diol (17) with triphosgene, material was formed with variable
chromatographic properties, believed to be the chloroformate
(20) and products derived from its degradation. Upon treat-
ment of this material with sodium methoxide in methanol, the
chromatographic properties improved and resulted in the even-
tual isolation of the carbamate (21), although in a less than sat-
isfactory yield (49%). The identity of the carbamate (21) was
confirmed, in part, by the large magnitude for J3,4, namely 12.5
Hz, indicative of an enforced trans-diaxial arrangement of H3
and H 4, typical of a trans-fused oxazolone ring.13
1
solvent, then provided a compound with an intriguing H
n.m.r. spectrum. Asignal at 3.40 was indicative of a methyl
glycoside, although the relatively small magnitude of the
observed coupling constants for most resonances still indi-
cated the presence of an anhydro ring. The problem was
eventually resolved in favour of the 2,6-episeleno-D-ido sugar
(13); further confirmation was provided by a comparison with
the known 2,6-anhydro-D-ido sugar (14).11 A rationalization
for this transformation is presented in Scheme 2.
As a result of the low yield for the production of (21),
attention was directed away from the use of triphosgene to
form the desired oxazolone. McAuliffe and coworkers have
shown that, depending on the conditions, either a cyclic car-
bamate or an O-ethyl carbamate may be formed upon treat-
ment of the amino alcohol (18) with ethyl chloroformate.13
In the event, prolonged treatment of the amino diol (17) with
ethyl chloroformate and triethylamine in dichloromethane
gave a mixture of the carbamate (22) and the amino alcohol
(23). The formation of these two products can be rational-
ized on the basis of two competing pathways for the reaction:
either the C 2 hydroxyl of (17) may react first, providing the
amino alcohol (23) and reducing the reactivity of the amine

R. V. Stick, D. M. G. Tilbrook and S. J. Williams
888
and remaining hydroxy group, or reaction at the C3 hydroxyl
would lead, first, to the carbamate (21) which can react, sub-
sequently, with a further equivalent of ethyl chloroformate to
give (22). The lack of reaction at nitrogen was to be
expected for such a hindered system.13
Under more forcing conditions, with 4-(dimethylamino)-
pyridine as a catalyst, the treatment of the amino diol (17)
with ethyl chloroformate provided an inseparable mixture of
what appeared to be the carbamate (22) and the dicarbonate
(24). The reluctance of the dicarbonate (24) to cyclize to the
carbamate (22) is surprising; upon reflux overnight in
tetrahydrofuran, the composition of the mixture of carbamate
(22) and dicarbonate (24) appeared unchanged.
Although a moderate yield of the carbamate (22) was
achieved with ethyl chloroformate and elaboration of (22)
into the putative donor (25) was an option, it was decided on
the basis of the above result to attempt the direct, selective
functionalization of (17) at O2. Accordingly, the amino diol
(17) was treated with benzoyl chloride at low temperature to
give an excellent yield of the monobenzoate (26), a signifi-
cant result. Subsequent treatment of the amino alcohol (26)
with triphosgene cleanly furnished the putative donor (25) in
an almost quantitative yield.
Two possible alcohols for the glycosylation reaction were
now identified, namely (27) and (28). Of these, the alcohol
(27) was likely to cause fewer problems in the final (depro-
tection) steps of the synthesis; treatment of the diol (29) with
N-acetylimidazole in dry dichloroethane at 80° for 3 days
provided the alcohol (27) in 51% yield.
Now, treatment of a solution of the donor (25) and the
alcohol (27) with a solution of N-iodosuccinimide/trifluo-
romethanesulfonic acid and, after consumption of the donor
(t.l.c.), with tributylstannane, gave the trisaccharide (30) as a
crystalline product, albeit in a poor yield of just 10%, along
with other, unidentified side-products. Nevertheless, given
the presumed low reactivity of the acetylated alcohol (27),
this result was certainly encouraging and the knowledge
gained in the model studies was placed directly into practice
on the actual sequence towards methyl -acarboside (3).
Tetra-O-benzyl-1-epivalienamine (6) was readily avail-
able through the excellent procedure of Nicotra and cowork-
ers15 and the subsequent modifications introduced by
McAuliffe and Stick.16 In order to exploit fully any suc-
cesses which may arise from the work planned, it was
decided to include both epoxides, namely (4) and (5), in the
upcoming investigations. With access to suitable quantities
of the amine (6), the coupling of this compound to the epox-
ides (4) and (5) became the next goal.
In contrast to the earlier work, purification of the product
of aminolysis of the 1,6-epithio epoxide (4) with the 1-epi-
valienamine (6) was not possible by crystallization.17 The
reaction produced a very dark, thick liquid and, although
t.l.c. indicated the formation of a new product, presumably
(31), in addition to the presence of excess epoxide, complete
purification was not possible; ultimately, the partially puri-
fied material was treated directly with benzoyl chloride at
–40° and the reaction mixture allowed to warm. Encour-
agingly, a new compound was formed with a much lower
polarity and this was readily purified, furnishing the amino
alcohol (32). A similar, two-step treatment of the 1,6-epise-
leno epoxide (5) also provided the analogous amino alcohol
(33) via the diol (34).
The conversion of the monobenzoates (32) and (33) into
the cyclic carbamates (35) and (36), respectively, was readily
accomplished upon treatment with triphosgene at low tem-
perature. The 1H n.m.r. spectra of both of these cyclic carba-
mates differed from that of the model carbamate (25) in that
the signal for H1Ј was not readily apparent. For compound
(33), the signal for H 1Ј was eventually located as a complex
multiplet in the range 4.49–4.64, whereas for (32) the signal
for H 1Ј could not be located. Similarly, the signal for C 1Ј in
both compounds showed significant line broadening, result-
ing in a very slow acquisition of the 13C n.m.r. spectra.
Effects such as these have been noted in similar systems17,18
and ascribed to hindered rotation around a carbon–nitrogen
bond, resulting in diastereoisomeric conformations.18
With both glycosyl donors in hand, the more reactive
alcohol (28) was readily prepared. Treatment of a mixture of
the donor (36) and the alcohol (28) with N-iodosuccin-
imide/TfOH resulted in the rapid consumption of the donor
(36) (t.l.c. analysis). Workup of the mixture and the subse-
quent treatment of the residue with tributylstannane resulted
in the eventual isolation of an oil.19 T.l.c. and 1H n.m.r. anal-
ysis at this stage, however, indicated that this oil was com-
posed of a number of compounds. Flash chromatography
resolved the mixture into a major fraction, the 1H n.m.r. spec-
trum of which indicated that it was itself composed of a
number of products, each of which showed characteristic
spectroscopic features suggesting the presence of products
of glycosylation and of 6-deoxy sugar residues. The variety
of spectroscopically similar products resulting from the gly-
cosylation reaction led to the conclusion that some degree of
benzyl ether cleavage had occurred. Indeed, treatment of the
mixture obtained from one such glycosylation attempt with
acetic anhydride and pyridine resulted in the conversion into
a number of unidentified compounds which appeared to have
incorporated acetyl groups (1H n.m.r. spectroscopy).
In one instance, the oil isolated from a glycosylation
attempt was treated with sodium methoxide to remove the
benzoyl group from the presumed product (37), to aid in its
isolation. Following flash chromatography, a poor mass
return of an inseparable mixture of at least four compounds
was evident, with each component exhibiting some spectro-
scopic characteristics expected for the product (38) and also
showing the absence of the benzoyl protecting group.
Alternatively, a portion of the same oil was treated with
sodium hydroxide in a mixture of pyridine and water at 95°
for 4 h, conditions which should effectively remove both the
carbamoyl and benzoyl functions.1,13,17 Again, a poor mass
return was evident and, although only two compounds were
present in the isolated mixture, each possessing spectro-
scopic characteristics expected for the amino diol (39), both
this and the above route appeared unrewarding.
It was believed that the decomposition of material in the
glycosylation reaction performed with the donor (36) may
have arisen from a detrimental degradation of an intermedi-

-Acarbose. VII
889
ate selenenyl triflate or selenenic acid present after the
quenching of the glycosylation reaction. As a result, an alter-
native promoter for the glycosylation reaction was sought
which would not result in the formation of such a potentially
reactive intermediate. It was anticipated that the use of
phenylselenenyl triflate as a promoter would ideally result in
the formation of a mixed diselenide as the product.20,21
Exploratory experiments with the model donor (15) and
alcohol (40) were inconclusive as to the benefits of such a
promoter—whilst glycosylation was apparent, the mass
return was poor and the complete purification of the mixed
diselenide (41) was not achieved.
Under the impression that degradation of the product was
the result of the acid-promoted loss of benzyl ethers, the
replacement of the benzyl ethers of (36) with a less labile pro-
tecting group was attempted. Of particular concern were the
benzyl ethers at the allylic positions of (36)—earlier work had
shown that such benzyl ethers are labile in acetolysis reac-
tions performed with an analogous 1,6-anhydro sugar.17
Recently, a number of Lewis-acid promoted reactions have
been used to remove benzyl ethers, with varying degrees of
efficiency. Olah and coworkers have reported the use of
trimethylsilyl iodide, generated in situ, as a mild but effective
reagent for the removal of benzyl ethers.22 Anhydrous iron(III)
chloride in dichloromethane has been used to good effect for
the debenzylation of some quite complex carbohydrates.23
Similarly, trimethylsilyl triflate in acetic anhydride is estab-
lished as a powerful reagent for the acetolysis of benzyl
ethers.24 Quite recently, iodine in acetic anhydride has been
shown to possess a remarkable ability to cleave benzyl ethers.25
The application of each of these procedures to the modification
of the donor (36) was unsuccessful, in some cases no reaction
was evident and in others incomplete removal of the benzyl
ethers appeared to have occurred. In at least one of these pro-
cedures, there was concern that allylic cleavage may have been
favoured over the desired benzylic cleavage.
The use of base-promoted procedures for the removal of
benzyl ethers has been the subject of fewer studies but has
shown good promise. A procedure modified from that of
Feutrill and Mirrington, using lithium ethanethiolate in hot
dimethylformamide to effect a nucleophilic displacement of
the benzyl ethers of (36), was attempted but resulted in exten-
sive degradation and afforded an intractable, black oil.26
As an aside, the amino alcohol (33) was treated with tri-
butylstannane and , Ј-azobisisobutyronitrile in toluene at
reflux, conditions which were hoped to remove the selenium
atom. Disappointingly, instead of the predicted reduction to
furnish (42), a compound of unidentified composition was
formed; however, it was clear that the 1,6-episeleno ring
remained intact, on the basis of a singlet being present at
5.81 in the 1H n.m.r. spectrum.
Being mindful of the important reactivity differences
between 1,6-epithio and 1,6-episeleno sugars firmly estab-
lished in preliminary work and having some concern as to the
potential for degradation arising from decomposition of an
intermediate selenenic acid, we now hoped that the glycosyl-
ation reaction using the 1,6-epithio donor (35) would bring
relief to the above unsuccessful glycosylation.27
A solution of the donor (35) and the alcohol (28) in a
mixture of 1,2-dichloroethane and diethyl ether was treated
with N-iodosuccinimide/TfOH as outlined previously. How-
ever, despite the addition of extra triflic acid and warming of
the reaction mixture to room temperature, after 1 h no reac-
tion was evident by t.l.c. With the additional triflic acid
present, decomposition of the donor and alcohol occurred at
an appreciable rate and the reaction was necessarily aban-
doned. The reaction was repeated on another occasion with
no obvious differences. Triethylsilyl triflate has been sug-
gested as an alternative to triflic acid as an efficient reagent
in combination with N-iodosuccinimide for the activation of
n-pentenyl glycosides without concomitant degradation of
acid-sensitive donors or alcohols.28 However, repetition of
the above reaction with triethylsilyl triflate in place of triflic
acid led to no obvious differences in the outcome of the reac-
tion. The lack of reaction was a significant setback, but it
was not entirely unexpected. The relative reactivities of 1,6-
epithio and 1,6-episeleno donors previously demonstrated
have shown the donor (16) to be the poorest of all those
examined and, given a presumed low reactivity of the disac-
charide alcohol (28) and of the donor (35) (resulting from the
presence of the oxazolone ring), even the powerful promoter
system N-iodosuccinimide/TfOH failed to activate the com-
pound towards glycosylation.27 It may also be possible that
the addition of iodonium ions to the electron-rich alkene in
(35) caused a significant reduction in the concentration of such
ions available for promotion of the glycosylation reaction.29
With the availability of methods for the modification of
the protecting groups of the donors (35) and (36) into differ-
ent forms rapidly running short, methods for their reductive
removal were sought. Unsurprisingly, an attempt at hydro-
genolysis of the benzyl ethers of (35) using palladium on
charcoal failed, chalcogens being widely recognized as
potent catalyst poisons.
More reasonably, reduction using lithium in liquid ammonia
appeared an attractive option. Accordingly, the carbamate
(35) was debenzoylated and then treated with lithium metal in
a mixture of liquid ammonia and tetrahydrofuran, the solvent
removed and the residue treated with acetic anhydride and
pyridine. Upon workup, a very low mass return of a product
of intermediate polarity was isolated—spectroscopic analysis
suggested that the product was in fact a mixture of the thioac-
etate (43) and the disulfide (44).
Amuch more polar compound was also isolated. 1H n.m.r.
spectroscopy showed this compound to be homogenous and
it was readily identified as the acetamide (45), further con-
1
firmation being provided by a comparison of the H n.m.r.
spectrum with that of the racemate prepared by Ogawa and
coworkers.30 The degradation resulting in cleavage of the
C 4–N bond of (35) appears to be a result of the 1,6-epithio
bridge, a closely related oxygen-containing analogue being
immune to such a degradation upon treatment with Na/NH3.
The degradation may result from the formation of an
anomeric anion or radical, which then undergoes a fragmen-
tation, ultimately cleaving the C 4–N bond.
The fragmentation of carba-sugars, with a similar result,
has been observed by others. Ogawa and coworkers have

R. V. Stick, D. M. G. Tilbrook and S. J. Williams
890
tallized. Flash chromatography (30–40% EtOAc/petrol) gave the
mesylate (12) as a pale yellow oil (197 mg, 76%). This oil crystallized
to give fine, white needles, m.p. 143–145° (EtOH), [ ]D –68.4° (Found:
C, 34.1; H, 4.5. C11H16O8SSe requires C, 34.1; H, 4.2%). 1H n.m.r.
(300 MHz) 2.06, 2.07, 2s, 6H, Ac; 3.05, s, Ms; 3.10, dd, J5,6 5.9, J6,6
9.9 Hz, H6; 3.31, br d, H 6; 4.54, dd, J3,4 7.0, J4,5 2.3 Hz, H4; 4.86, br
d, J2,3 4.9 Hz, H2; 5.04, br dd, H5; 5.12, dd, H3; 5.72, br s, H1. 13C
n.m.r. (75.5 MHz) 20.81, 2C, COMe; 31.81, C 6; 38.34, Ms; 68.95,
76.65, 77.20, 78.21, 81.91, C1,2,3,4,5; 169.51, 170.11, 2C, CO.
reported that the treatment of per-acetylvalidoxylamine A
(46) with N-bromosuccinimide in aqueous N,N-dimethylfor-
mamide affords, among other products, tetra-O-acetyl-
valienamine (47) and the cyclohexenone (48).31 Whilst the
rationalization for such a degradation undoubtedly differs
from that proposed above, the products demonstrate the
lability of the C 4–N bond under suitable conditions.
The failure of the donors (35) and (36) to glycosylate the
alcohol (28) and provide a novel route to the target com-
pound, methyl -acarboside (3), represented a major blow to
our synthetic plan. One option now was to investigate another
synthesis of epivalienamine in a differently protected form,
perhaps as the di-O-isopropylidene acetal (49). Whilst this
was an option which may have borne fruit, an alternative
synthesis of methyl -acarboside (3) was envisioned by way
of the alkylation of the same tetra-O-benzyl-1-epivalien-
amine (6) with a suitably protected triflate. These investiga-
tions form the basis of the next paper.
3,4-Anhydro-1,6-dideoxy-1,6-epithio- -D-galactose (4)
A suspension of the mesylate (11)7 (12.2 g, 36.0 mmol) in dry
MeOH (50 ml) was treated with sodium metal (998 mg, 43.4 mmol) at
0° for 16 h. The solvent was evaporated from the resultant suspension
and the residue was suspended in EtOAc and filtered off, with repeated
washing of the filter cake with additional EtOAc. The solvent was
evaporated from the filtrate and the residue purified by flash chro-
matography (60% EtOAc/petrol) to give the epoxide (4) as a pale
yellow, crystalline solid (5.11 g, 89%). The 1H n.m.r. spectrum agreed
with that reported in the literature.7
3,4-Anhydro-1,6-dideoxy-1,6-episeleno- -D-galactose (5)
The 1,6-episeleno mesylate (12) (5.32 g) was treated in the same
manner as for the 1,6-epithio mesylate (11) to give the epoxide (5) as a
white, crystalline solid (2.35 g, 83%), m.p. 117–119°, [ ]D –71.0°
Experimental
General experimental procedures have been given previously.16
1
1,2,3-Tri-O-acetyl-4,6-di-O-methylsulfonyl-D-glucose (9a,b)
(Found: C, 35.0; H, 3.8. C6H8O3Se requires C, 34.8; H, 3.9%). H n.m.r.
1,2,3-Tri-O-acetyl-4,6-O-benzylidene-D-glucose7 (20.0 g,50.8 mmol)
and palladium on carbon (10%, 0.4 g) was suspended in a mixture of
EtOAc (300 ml) and MeOH (50 ml) containing two drops of AcCl and
the mixture was treated with hydrogen (6 h). The mixture was filtered
and the solvent evaporated to afford a colourless foam. This foam was
dissolved in pyridine (100 ml) and the solution treated with MsCl
(15.8 ml, 23.3 g, 203 mmol) at 0° for 30 min, then room temperature for
2 h. The mixture was poured into an ice/water slurry (500 ml) which
was stirred vigorously for 10 min and then allowed to stand (10 min).
The supernatant was decanted and the doughy residue was washed with
water and crystallized from EtOH to afford a white powder. The
aqueous supernatant deposited crystals overnight and these were fil-
tered off and washed with EtOH and combined with the first crop to
afford the dimesylate (9a,b) as a white powder (16.5 g, 71%). The 1H
n.m.r. spectrum agreed with that reported in the literature.7
(500 MHz) 2.72–2.77, m, OH; 3.12, dd, J5,6 6.6, J6,6 8.6 Hz, H 6; 3.18,
br d, J3,4 3.5 Hz, H3; 3.31, d, H6; 3.53, dd, J4,5 5.3 Hz, H4; 4.08, br d,
J
2,OH 6.3 Hz, H 2; 5.15, dd, H5; 5.65, br s, H 1. 13C n.m.r. (125.8 MHz)
29.93, C 6; 50.64, 53.67, C 3,4; 67.57, 76.36, 77.02, C 1,2,5.
Methyl 3,4-Di-O-acetyl-2,6-dideoxy-2,6-episeleno- -D-idoside (13)
The mesylate (12) (341 mg, 0.872 mmol) was suspended in dry
MeOH (5 ml) and a solution of NaOMe (1.0 ml of 0.93 M in MeOH,
0.93 mmol) was added at room temperature under nitrogen. The
mixture became clear and turned light orange after 1 min. After 24 h,
more NaOMe (7 ml of 0.93 M in MeOH, 6.5 mmol) was added and the
solution left for another 24 h. T.l.c. at this stage indicated the pres-
ence of one major compound, so the solvent was evaporated and the
residue was treated with pyridine (8 ml) and Ac2O (4 ml) at 0° for 1
h. The solvent was evaporated and the residue subjected to a normal
workup (EtOAc). The residue was purified by flash chromatography
(25% EtOAc/petrol) to give the methyl D-idoside (13) as a clear oil
(64 mg, 23%), [ ]D +44.4° (Found: C, 40.9; H, 5.0. C11H16O6Se
2,3-Di-O-acetyl-4,6-di-O-methylsulfonyl- -D-glucosyl Bromide (10)
A suspension of the triacetate (9a,b) (10.0 g, 21.6 mmol) in dry
AcOH (30 ml) and HBr in AcOH (30%, 10 ml, 50 mmol) was stirred at
room temperature for 2 days. The solution was subjected to the usual
workup (CH2Cl2, excluding the acid wash) and the crystalline residue
was dissolved in EtOAc; crystallization ensued upon the addition of
petrol to give the bromide (10) as chunky crystals (8.06 g, 77%), m.p.
143–145° (EtOAc/petrol), [ ]D +172.6° (Found: C, 30.0; H, 4.2. Calc.
for C12H19BrO11S2: C, 29.8; H, 4.0%). 1H n.m.r. (500 MHz) 2.09,
2.10, 2s, 6H, Ac; 3.09, 3.09, 2s, 6H, Ms; 4.36, ddd, J4,5 10.1, J5,6 2.1, 3.5
Hz, H 5; 4.42, dd, J6,6 11.6 Hz, H 6; 4.48, dd, H 6; 4.81, dd, J1,2 4.1, J2,3
1
requires C, 40.9; H, 5.0%). H n.m.r. (500 MHz) 2.05, 2.07, 2s, OAc;
2.88, ddd, J4,6 1.0, J5,6 4.0, J6,6 10.8 Hz, H 6; 2.91, dd, J5,6 2.0 Hz, H6;
3.09, dd, J1,2 2.1, J2,3 1.9 Hz, H 2; 3.40, s, OMe; 4.46, ddd, J4,5 4.2 Hz,
H 5; 5.12, dd, J1,3 0.92 Hz, H 1; 5.33, ddd, J3,4 5.1 Hz, H 4; 5.42, ddd,
H 3. 13C n.m.r. (125.8 MHz) 15.53, C 6; 20.87, 21.08, 2C, COMe;
29.31, C 2; 55.26, OMe; 66.26, 72.15, 75.14, C 3,4,5; 100.54, C 1;
169.71, 170.54, 2C, CO.
4-Cyclohexylamino-1,6-episeleno-1,4,6-trideoxy- -D-glucose (17)
10.0 Hz, H2; 4.85, dd, J3,4 9.3 Hz, H 4; 5.63, dd, H 3; 6.59, d, H1. 13
C
A solution of the epoxide (5) (694 mg, 3.56 mmol) and cyclohexyl-
amine (1.66 g, 1.9 ml, 16.8 mmol) in BuOH (5 ml) was heated (100°)
under nitrogen for 16 h. The solvent was evaporated with the aid of
toluene. The crystalline residue so obtained was purified by flash chro-
matography (60% EtOAc/petrol, then 10% EtOH/1% Et2NH/EtOAc) to
give the amine (17) as a pale brown, crystalline solid (904 mg, 87%) which
crystallized as pale yellow needles, m.p. 163–164° (CH2Cl2/Pri2O), [ ]D
–39.0° (Found: C, 47.1; H, 7.0. C12H21NO3Se requires C, 47.1; H,
6.9%). 1H n.m.r. (500 MHz) 0.90–1.30, 1.56–2.03, 2m, 10H, cyclo-
hexyl CH2; 2.54, tt, J 3.6, 10.5 Hz, cyclohexyl CH; 2.88, d, J3,4 1.8 Hz,
H4; 3.26, dd, J5,6 7.6, J6,6 9.6 Hz, H6; 3.37, br d, H 6; 3.68, br s, H2;
3.76, d, H3; 4.80, br d, J5,6 7.5 Hz, H 5; 5.90, br s, H1. 13C n.m.r.
(125.8 MHz) 25.02, 25.10, 25.79, 30.45, 33.52, 34.11, 6C, C 6, cyclo-
hexyl CH2; 53.82, cyclohexyl CH; 57.74, C 4; 70.28, 72.16, 80.41,
81.07, C 1,2,3,5.
n.m.r. (125.8 MHz) 20.52, 20.74, 2C, COMe; 37.73, 38.67, 2C, Ms;
65.41, C 6; 69.01, 70.29, 71.41, 71.62, C 2,3,4,5; 85.53, C 1; 169.58,
169.85, 2C, CO.
2,3-Di-O-acetyl-1,6-dideoxy-1,6-episeleno-
4-O-methylsulfonyl- -D-glucose (12)
A mixture of sodium borohydride (47 mg, 1.2 mmol) and selenium
(79 mg, 1.0 mmol) in EtOH (5 ml) was stirred under nitrogen until all
of the selenium had dissolved and a clear solution had formed. The
bromide (10) (320 mg, 0.66 mmol) was added, along with dry
dimethylformamide (5 ml). The mixture was left to stir overnight and
then the mixture was poured into water and extracted with EtOAc. The
combined extracts were washed successively with water and brine, then
dried and the solvent evaporated to give a pale yellow oil which crys-

-Acarbose. VII
891
1
3,4-O,N-Carbonyl-4-cyclohexylamino-1,6-episeleno-
above. Partial H n.m.r. (500 MHz) for (24): 2.48–2.58, m, cyclo-
hexyl CH; 2.69–2.71, m, H 4; 5.82, br s, H1. Partial 13C n.m.r.
(125.8 MHz) for (24): 14.14, CH3; 24.39, 24.61, 30.79, 33.39, 33.47,
6C, C 6, cyclohexyl CH2; 53.27, cyclohexyl CH; 55.01, C 4; 64.30,
64.42, CH2CH3; 72.67, 76.25, 76.89, 82.40, C 1,2,3,4,5; 153.94,
154.26, 2C, CO.
1,4,6-trideoxy- -D-glucose (21)
Triphosgene (149 mg, 540 µmol) was added to a solution of the
amine (17) (153 mg, 500 µmol) and pyridine (480 µl) in CH2Cl2 (10 ml)
at –78°. The mixture was allowed to warm to room temperature, then
water (0.5 ml) was added and the mixture stirred for a further 10 min.
The reaction mixture was subjected to the usual workup (CH2Cl2) and
the residue dissolved in MeOH (5 ml). The solution was treated with a
small piece of sodium. After 30 min, t.l.c. indicated a homogenous
product, so the solution was neutralized with resin (Dowex-50, H+) and
the solvent evaporated. Flash chromatography of the residue (30–40%
EtOAc/petrol) gave the carbamate (21) as a clear oil (82 mg, 49%)
which crystallized as needles, m.p. 194–195° (EtOH), [ ]D –35° (Found:
C, 47.0; H, 5.6. C13H19NO4Se requires C, 47.0; H, 5.8%). max (NaCl)
1736 cm–1. 1H n.m.r. (500 MHz) 1.05–1.43, 1.67–2.02, 2m, 10H,
cyclohexyl CH2; 3.16–3.22, m, 2H, H 6; 3.61–3.67, m, cyclohexyl CH;
3.66, dd, J3,4 12.5, J4,5 4.7 Hz, H 4; 3.98, dd, J2,3 8.6 Hz, H3; 4.18, d,
H2; 5.04, dd, J5,6 1.6, 3.4 Hz, H5; 5.75, s, H 1. 13C n.m.r. (125.8 MHz)
25.47, 25.93, 29.72, 32.67, 35.86, 6C, C 6, cyclohexyl CH2; 55.36,
cyclohexyl CH; 58.73, C 4; 76.47, 77.07, 79.68, 81.42, C 1,2,3,5;
159.28, CO.
Attempted Formation of the Carbamate (22)
A solution of a mixture of the oxazolone (22) and the dicarbonate
(24) (50 mg, 3: 2) in dry tetrahydrofuran was heated at reflux under N2
overnight. The solvent was evaporated to afford an oil (49 mg). The
composition of the material was shown by 1H n.m.r. (300 MHz) to be
essentially unchanged from that of the starting mixture.
2-O-Benzoyl-4-cyclohexylamino-1,6-episeleno-
1,4,6-trideoxy- -D-glucose (26)
Benzoyl chloride (320 µl, 2.8 mmol) was added slowly to a stirred
suspension of the amine (17) (610 mg, 2.0 mmol), 4-(dimethyl-
amino)pyridine (24 mg, 0.20 mmol) and Et3N (840 µl, 6.0 mmol) in dry
CH2Cl2 (20 ml) at –50° under N2. The solution was allowed to warm to
–20° over the space of 1 h, at which stage t.l.c. indicated almost com-
plete consumption of the starting material. MeOH (1 ml) was added
and the solution was left to warm to room temperature. The solution
was then diluted with CH2Cl2 and washed with water, then dried and the
solvent evaporated. The residue was purified by flash chromatography
(30–50% EtOAc/petrol plus 1% Et3N) to give the benzoate (26) as an
off-white solid (690 mg, 84%) that crystallized as white needles, m.p.
136–137° (CH2Cl2/Pri2O), [ ]D +10.7° (Found: C, 55.9; H, 6.2.
C19H25NO4Se requires C, 55.7; H, 5.9%). 1H n.m.r. (500 MHz)
1.07–1.33, 1.55–1.90, 2m, 10H, cyclohexyl CH2; 2.63, tt, J 3.7, 10.1
Hz, cyclohexyl CH; 2.83, dd, J3,4 2.4, J4,5 5.2 Hz, H4; 3.25, dd, J5,6 6.1,
3,4-O,N-Carbonyl-4-cyclohexylamino-1,6-episeleno-2-O-
ethoxycarbonyl-1,4,6-trideoxy- -D-glucose (22) and
4-Cyclohexylamino-2-O-ethoxycarbonyl-1,6-episeleno-
1,4,6-trideoxy- -D-glucose (23)
Ethyl chloroformate (143 µl, 1.50 mmol) was added to a solution of
the amino diol (17) (153 mg, 500 µmol) and Et3N (420 µl, 3.0 mmol) in
dry CH2Cl2 (10 ml) at –30° under N2. The solution was allowed to
warm to room temperature and was left for 8 h. More ethyl chlorofor-
mate (50 µl, 0.17 mmol) was added and the solution was left overnight.
Water (0.5 ml) was added to the solution and the mixture was stirred for
10 min, before being subjected to the usual workup (CH2Cl2). The
residue was purified by flash chromatography (20–50% EtOAc/petrol)
to give firstly the oxazolone (22) as a clear oil (117 mg, 59%), [ ]D
–31.1° (Found: C, 47.4; H, 5.8. C16H23NO6Se requires C, 47.5; H,
J
6,6 9.3 Hz, H6; 3.29, dd, J5,6 1.1 Hz, H6; 3.82, m, H3; 4.94, br d, H 5;
5.11, d, J2,3 3.5 Hz, H2; 5.96, br s, H1; 7.43–7.47, 7.57–7.61,
8.03–8.06, 3m, Ph. 13C n.m.r. (125.8 MHz) 24.68, 24.75, 25.99, 32.64,
33.45, 34.05, 6C, C 6, cyclohexyl CH2; 53.47, cyclohexyl CH; 57.58,
C 4; 68.89, 77.64, 77.81, 83.37, C 1,2,3,5; 128.45–129.77, 133.44, Ph;
165.80, CO.
1
5.7%). H n.m.r. (500 MHz) 1.05–1.42, 1.67–2.02, 2m, 10H, cyclo-
hexyl CH2; 1.32, t, J 7.1 Hz, CH3; 3.18, m, 2H, H 6; 3.62–3.68, m,
cyclohexyl CH; 3.73, dd, J3,4 12.6, J4,5 4.7 Hz, H 4; 4.11, dd, J2,3 9.0 Hz,
H3; 4.18–4.26, m, CH2CH3; 4.96, d, H 2; 5.02, ddd, J5,6 2.5, 2.5 Hz,
H5; 5.77, s, H1. 13C n.m.r. (125.8 MHz) 14.10, CH3; 25.42, 25.91,
29.65, 32.61, 35.84, 6C, C 6, cyclohexyl CH2; 55.40, cyclohexyl CH;
58.55, C 4; 64.86, CH2CH3; 72.38, 77.99, 79.98, 81.00, C 1,2,3,5;
154.18, OC=O; 158.36, NC=O.
2-O-Benzoyl-3,4-O,N-carbonyl-4-cyclohexylamino-
1,6-episeleno-1,4,6-trideoxy- -D-glucose (25)
Triphosgene (148 mg, 500 µmol) was added to a solution of the ben-
zoate (26) (409 mg, 1.00 mmol) and pyridine (242 µl, 3.00 mmol) in dry
CH2Cl2 (10 ml) at –78° under N2. The solution was allowed to warm to
room temperature over 1 h, then water (0.5 ml) was added and the solu-
tion stirred (1 min). The mixture was subjected to the usual workup
(CH2Cl2) and the residue was purified by flash chromatography
(10–30% EtOAc/petrol) to give the carbamate (25) as a white, crys-
talline solid (414 mg, 95%) that crystallized as white, lustrous, long
plates, m.p. 164–165° (EtOAc/petrol), [ ]D +25.4° (Found: C, 55.1; H,
5.3. C20H23NO5Se requires C, 55.1; H, 5.3%). max (KBr) 1751, 1711
cm–1. 1H n.m.r. (500 MHz) 1.06–1.16, 1.30–1.47, 1.67–2.06, 3m,
10H, cyclohexyl CH2; 3.22–3.27, m, 2H, H 6; 3.65–3.73, m, cyclohexyl
CH; 3.82, dd, J3,4 12.5, J4,5 4.7 Hz, H4; 4.32, dd, J2,3 9.2 Hz, H3;
5.08–5.12, m, H5; 5.25, d, H 2; 5.83, s, H 1; 7.42–7.48, 7.56–7.65,
8.04–8.08, 3m, Ph. 13C n.m.r. 25.43, 25.90, 29.68, 32.65, 35.76, 6C,
C 6, cyclohexyl CH2; 55.42, cyclohexyl CH; 58.84, C 4; 72.42, 78.41,
78.91, 79.88, C 1,2,3,5; 128.37–129.85, 133.68, Ph; 158.52, NCO;
165.91, COPh.
Next to elute was the amino alcohol (23) as needles (45 mg, 24%),
m.p. 76–80°, [ ]D –37.6° (Found: C, 48.0; H, 7.0. C15H25NO5Se
requires C, 47.6; H, 6.7%). 1H n.m.r. (500 MHz)
1.02–1.28,
1.57–1.88, 2m, 10H, cyclohexyl CH2; 1.31, t, J 7.1 Hz, CH3; 2.55, tt, J
3.7, 10.3 Hz, cyclohexyl CH; 2.72, dd, J3,4 6.4, J4,5 2.6 Hz, H 4;
3.15–3.20, m, 2H, H 6; 3.63, br t, H 3; 4.21, q, CH2CH3; 4.72, br d, J2,3
4.6 Hz, H2; 4.83, dt, J5,6 2.5, 4.4 Hz, H5; 5.85, s, H 1. 13C n.m.r.
(125.8 MHz) 14.15, CH3; 24.75, 24.84, 25.91, 33.11, 33.23, 34.30,
6C, C 6, cyclohexyl CH2; 53.70, cyclohexyl CH; 57.60, C 4; 64.46,
CH2CH3; 68.60, 77.53, 81.68, 83.71, C 1,2,3,5; 154.52, CO.
3,4-O,N-Carbonyl-4-cyclohexylamino-1,6-episeleno-2-O-ethoxy-
carbonyl-1,4,6-trideoxy- -D-glucose (22) and 4-Cyclohexylamino-2,3-
di-O-ethoxycarbonyl-1,6-episeleno-1,4,6-trideoxy- -D-glucose (24)
Ethyl chloroformate (278 µl, 2.92 mmol) was added to an ice-cold
solution of the amine (17) (149 mg, 487 µmol), Et3N (400 µl, 2.9 mmol)
and 4-(dimethylamino)pyridine (60 mg, 0.5 mmol) in dry CH2Cl2
(5 ml) under N2. The solution was allowed to warm to room tempera-
ture and was left for 3 h. The solution was quenched by the addition of
water (1 ml) and the mixture was stirred for 5 min, before being sub-
jected to the usual workup (CH2Cl2). The residue was purified by flash
chromatography (15–50% EtOAc/ petrol) to give a clear oil (163 mg)
that appeared to be a mixture of two compounds, the oxazolone (22) and
the dicarbonate (24), in a ratio of 3: 2, respectively, as determined by 1H
n.m.r. spectroscopy and by a comparison with the spectra recorded
Methyl 2,3,6-Tri-O-acetyl-4-O-(2,3,6-tri-O-
acetyl- -D-glucopyranosyl)- -D-glucoside (27)
AcCl (110 µl, 1.5 mmol) was added slowly to an ice-cold solution
of imidazole (200 mg, 3.0 mmol) in dry 1,2-dichloroethane (10 ml)
under N2. The mixture was filtered after 1 min and a portion of the fil-
trate (9.0 ml of 0.15 M, 13 mmol) was added to the diol (29)32 (762 mg,
1.34 mmol) and the solution was heated in a sealed ampoule at 80° for
3 days. The mixture was cooled to room temperature and subjected to
the usual workup (CH2Cl2) to give a clear oil which was purified by
flash chromatography (50–60% EtOAc/petrol) to give the alcohol (27)

R. V. Stick, D. M. G. Tilbrook and S. J. Williams
892
as a clear syrup (412 mg, 51%), [ ]D –44.7° (Found: C, 49.4; H, 6.0.
C25H36O17 requires C, 49.3; H, 6.0%). 1H n.m.r. (500 MHz) 2.02, 2.02,
2.03, 2.04, 2.11, 2.12, 6s, 18H, Ac; 3.22, d, J4Ј,OH 4.5 Hz, OH;
3.43–3.46, m, H 5Ј; 3.46, s, OMe; 3.50, ddd, J3Ј,4Ј 9.8, J4Ј,5Ј 9.8 Hz, H4Ј;
3.57, ddd, J4,5 9.9, J5,6 2.1, 5.0 Hz, H 5; 3.77, dd, J3,4 9.4 Hz, H4; 4.09,
dd, J6,6 12.0 Hz, H6; 4.20, dd, J5Ј,6Ј 2.0, J6Ј,6Ј 12.5 Hz, H 6Ј; 4.36, d, J1,2
7.9 Hz, H 1; 4.47, d, J1Ј,2Ј 7.9 Hz, H 1Ј; 4.51, dd, H 6; 4.57, dd, H6Ј;
4.83, dd, J2Ј,3Ј 9.4 Hz, H2Ј; 4.87, dd, J2,3 9.6 Hz, H 2; 4.98, dd, H 3Ј;
5.16, dd, H 3. 13C n.m.r. (125.8 MHz) 20.53, 20.55, 20.70, 20.74,
20.75, 20.82, 6C, COMe; 56.98, OMe; 61.80, C 6; 62.55, C 6Ј; 68.11,
C 4Ј; 71.48, C 2,2Ј; 72.56, 72.64, C3,5; 74.23, C5Ј; 75.06, C 3Ј; 76.44,
C 4; 100.86, C 1Ј; 101.36, C 1; 169.26, 169.69, 169.88, 170.34, 171.07,
171.85, 6C, CO.
material. EtOH (1 ml) was added and the solution was stirred for 5
min. The solution was diluted with CH2Cl2 and washed with water,
then dried and the solvent evaporated. The dark brown residue was
purified by flash chromatography (50–75% CHCl3/petrol plus 1%
Et3N) to afford a solid residue which was crystallized to yield the
amino alcohol (32) as pale brown needles (891 mg, 30% over the two
steps), m.p. 139–140° (CH2Cl2/Pri2O), [ ]D –56.6° (Found: C, 72.0; H,
6.3. C48H49NO8S requires C, 72.1; H, 6.2%). 1H n.m.r. (500 MHz)
2.82, dd, J3,4 4.6, J4,5 1.7 Hz, H4; 2.99, d, J6,6 9.8 Hz, H6; 3.07, dd, J5,6
6.3 Hz, H 6; 3.40, d, J3,OH 6.4 Hz, OH; 3.46, J1Ј,6Ј 9.2 Hz, H 1Ј; 3.54, J5Ј,6Ј
9.9 Hz, H6Ј; 3.72–3.76, m, H3; 3.77, br d, J7Ј,7Ј 11.9 Hz, H7Ј; 3.89, dd,
J4Ј,5Ј 7.7 Hz, H5Ј; 4.19, br d, H7Ј; 4.35, br d, H4Ј; 4.42–5.03, m, 8H,
CH2Ph; 4.80–4.83, m, H5; 4.96–4.98, m, H2; 5.59, br s, H2Ј; 5.61, br s,
H1; 7.25–7.38, 7.50–7.55, 8.01–8.04, 3m, Ph. 13C n.m.r. (125.8 MHz)
36.08, C6; 57.11, C1Ј; 58.99, C4; 70.09, C7Ј; 70.16, C3; 72.44, 74.60,
75.24, 75.41, 4C, CH2Ph; 77.25, C 2; 80.07, C4Ј; 80.96, C5; 82.05, C 1;
83.37, C 6Ј; 85.00, C5Ј; 127.64, C2Ј; 127.66–138.36, C 3Ј, Ph; 165.75, CO.
Methyl O-(2-O-Benzoyl-3,4-O,N-carbonyl-4-cyclohexylamino-4,6-
dideoxy- -D-glucosyl)-(1 4)-O-(2,3,6-tri-O-acetyl- -D-glucosyl)-
(1 4)-2,3,6-tri-O-acetyl- -D-glucoside (30)
A solution of N-iodosuccinimide (145 mg, 645 µmol) and TfOH
(15 µl) in dry 1,2-dichloroethane/Et2O (1 : 1, 4 ml) at 0° under N2 was
added to a solution of the donor (25) (241 mg, 553 µmol) and alcohol
(27) (281 mg, 461 µmol) in dry 1,2-dichloroethane (3 ml) at 0° under
N2. The mixture was left to stand for 1 h, then was quenched by the
addition of saturated aqueous sodium bicarbonate solution (1 ml) and
aqueous sodium thiosulfate solution (0.5 M, 1 ml). The organic phase
was separated, dried and evaporated to give a residue which was dis-
solved in toluene (3 ml) and treated with Bu3SnH (0.4 ml) and ␣,␣Ј-
azobisisobutyronitrile (1 mg) at reflux under N2 for 30 min. The
solvent was evaporated and the residue was dissolved in MeCN. The
MeCN solution was washed repeatedly with petrol (5 times) and the
solvent was evaporated. The residue was subjected to flash chro-
matography (60% EtOAc/petrol) to give a semicrystalline residue
(120 mg) which was purified by crystallization to give the trisaccharide
(30) as plates (45 mg, 10%), m.p. 151–153° (CH2Cl2/EtOH), [ ]D
–28.7° (Found: C, 55.8; H, 6.3. C45H59NO22 requires C, 55.9; H, 6.2%).
2-O-Benzoyl-3,4-O,N-carbonyl-1,6-epithio-4-[(1R,4R,5S,6S)-4,5,6-
tribenzyloxy-3-(benzyloxymethyl)cyclohex-2-enyl]amino-1,4,6-
trideoxy- -D-glucose (35)
Triphosgene (213 mg, 0.717 mmol) was added to a solution of the
amino alcohol (32) (1.15 mg, 1.44 mmol) and pyridine (350 µl,
4.3 mmol) in dry CH2Cl2 (20 ml) at –78° under N2. The mixture was
allowed to warm to 0°, water (0.5 ml) was added and the mixture was
stirred for 10 min. The mixture was diluted with CH2Cl2 and then
washed with water. The organic layer was dried and the solvent evap-
orated to leave an oil which was purified by flash chromatography
(10–30% EtOAc/petrol) to afford the carbamate (35) as a colourless
foam (895 mg, 75%), [ ]D –70.2° (Found: C, 71.1; H, 5.7. C49H47NO9S
requires C, 71.2; H, 5.7%).
(NaCl) 1769, 1721 cm–1. 1H n.m.r.
max
(500 MHz) 2.74, d, J6,6 9.4 Hz, H6; 2.92, dd, J5,6 4.1 Hz, H6;
3.25–3.35, m, H 4; 3.90–3.96, m, H 5Ј,6Ј; 3.95, br d, J7Ј,7Ј 12.5 Hz, H 7Ј;
4.13, br t, J2,3 Ϸ J3,4 9.3 Hz, H 3; 4.19, br d, H7Ј; 4.30–4.35, m, H4Ј;
4.46–5.01, m, 8H, CH2Ph; 4.90, br d, H2; 5.02, t, J4,5 Ϸ J5,6 4.1 Hz, H5;
5.40, s, H 1; 5.67, br s, H2Ј; 7.25–7.53, 7.61–7.66, 8.09–8.13, 3m, Ph.
The signal for H1Ј could not be located. 13C n.m.r. (125.8 MHz)
39.35, C 6; 57.06, C 1Ј; 58.47, C 4; 69.36, C 7Ј; 72.29, 74.58, 74.81,
75.26, 4C, CH2Ph; 73.10, C 3; 77.58, C 5; 77.83, C 2; 79.03, C 4Ј;
84.01, C 1; 77.25, 84.22, C 5Ј,6Ј; 123.04, C 2Ј; 127–138.97, C 3Ј, Ph;
157.67, NCO; 165.74, COPh. High-resolution mass spectrum (f.a.b.)
m/z 826.3059 [C49H48NO9S (M+H)+• requires 826.3050].
(KBr) 1755 cm–1. 1H n.m.r. (500 MHz) 1.10–1.29, 1.61–1.88,
max
2m, 10H, cyclohexyl CH2; 1.50, d, J5,6 6.2 Hz, 3H, H 6A; 1.93, 1.98,
2.00, 2.01, 2.02, 2.10, 6s, Ac; 3.19–3.27, m, cyclohexyl CH; 3.42, dd,
J
3,4 11.7, J4,5 8.9 Hz, H 4A; 3.45, s, OMe; 3.51, ddd, J4,5 9.8, J5,6 2.3, 4.6
Hz, H 5C; 3.56, ddd, J4,5 9.8, J5,6 2.0, 4.8 Hz, H 5B; 3.71, dd, J3,4 9.6 Hz,
H4B; 3.82, dq, H 5A; 3.84, dd, J3,4 9.5 Hz, H 4C; 4.06, dd, J6,6 12.0 Hz,
H6B; 4.09, dd, J2,3 10.2 Hz, H 3A; 4.24, dd, J6,6 12.3 Hz, H 6C; 4.30, dd,
H6C; 4.35, d, J1,2 7.9 Hz, H 1B; 4.46, d, J1,2 7.9 Hz, H 1C; 4.52, dd, H 6B;
4.73, d, J1,2 4.9 Hz, H1A; 4.79, dd, J2,3 9.4 Hz, H 2C; 4.84, dd, J2,3 9.6
Hz, H2B; 5.11, dd, H3C; 5.13, dd, H 3B; 5.20, dd, H 2A; 7.44–7.50,
7.56–7.61, 8.02–8.05, 3m, Ph. 13C n.m.r. (125.8 MHz) 20.50, 20.58,
20.69, 20.85, 20.92, 6C, COMe; 21.44, C 6A; 25.10, 25.74, 26.17,
29.72, 5C, cyclohexyl CH2; 55.78, cyclohexyl CH; 56.96, OMe; 61.70,
C 4A; 61.74, C 6B; 62.23, C 6C; 71.49, C 2B; 71.99, C 2C; 72.38, C 3B;
72.57, 72.62, C 5B,5C; 73.74, 73.78, C 2A,3C; 73.96, C 5A; 75.94, C 4C;
76.39, C4B; 76.47, C3A; 100.32, C 1C; 101.35, C 1B; 102.65, C1A;
128.55, 128.60, 129.95, 133.71, Ph; 157.45, NCO; 164.88, COPh;
169.45, 169.60, 169.78, 170.32, 170.39, 6C, COMe.
2-O-Benzoyl-1,6-episeleno-4-[(1R,4 R,5 S,6 S)-4,5,6-tribenzyloxy-3-
(benzyloxymethyl)cyclohex-2-enyl]amino-1,4,6-trideoxy- - -glucose (33)
D
(A) A solution of tetra-O-benzyl-1-epivalienamine (6)15 (6.17 g,
11.5 mmol) and the epoxide (5) (4.77 g, 23.0 mmol) in BuOH (25 ml)
was heated at 100° for 2.5 days. The solvent was evaporated and the
residue purified by flash chromatography (20–10% petrol/CHCl3 plus
1% Et3N) to give the impure amino diol (34) as a black oil (5.00 g).
(B) A solution of some of the above amino diol (34) (1.98 g), pyri-
dine (1.11 ml) and 4-(dimethylamino)pyridine (32 mg) in CH2Cl2
(40 ml) was treated with benzoyl chloride (430 µl, –40 to –5°). At this
stage t.l.c. indicated the presence of unreacted starting material. The
solution was treated with more benzoyl chloride (100 µl, –30 to –5°),
whereupon t.l.c indicated the absence of the starting material. The reac-
tion mixture was quenched and subjected to a workup as for (31) above.
The dark brown residue after workup was purified by flash chromato-
graphy (25–40% CHCl3/petrol plus 1% Et3N) to afford the amino
alcohol (33) as a light brown oil (1.58 g, 41% over the two steps).
Crystallization of a portion of this oil yielded a microcrystalline
powder, m.p. 123–129° (CH2Cl2/Pri2O), [ ]D –69.3° (Found: C, 68.0;
H, 5.8. C48H49NO8Se requires C, 68.1; H, 5.8%). 1H n.m.r. (500 MHz)
2.79, dd, J3,4 2.6, J4,5 6.5 Hz, H 4; 3.07, d, J6,6 9.2 Hz, H6; 3.13, dd,
2-O-Benzoyl-1,6-epithio-4-[(1R,4R,5S,6S)-4,5,6-tribenzyloxy-3-
(benzyloxymethyl)cyclohex-2-enyl]amino-1,4,6-trideoxy- -D-glucose
(32)
(A) A solution of tetra-O-benzyl-1-epivalienamine (6)15 (2.74 g,
5.11 mmol) and the epoxide (4) (1.64 g, 10.2 mmol) in BuOH (8 ml)
was heated at 100° for 2.5 days. The solvent was evaporated with the
aid of toluene and the residue purified by flash chromatography
(50–10% petrol/CHCl3 plus 1% Et3N) to give the impure amino diol
(31) as a black oil (2.69 g).
(B) Benzoyl chloride (530 µl) was added to a solution of some of
the above amino diol (31) (1.96 g), pyridine (690 µl) and 4-(dimethyl-
amino)pyridine (35 mg) in CH2Cl2 (50 ml) at –30°. The solution was
allowed to warm to 0°, at which stage t.l.c. indicated the presence of
unreacted starting material. The solution was cooled again (–20°) and
more benzoyl chloride (100 µl) was added. The solution was allowed
to warm to 0°, whereupon t.l.c indicated the absence of the starting
J
5,6 5.9 Hz, H6; 3.45, br d, J1Ј,6Ј 8.5 Hz, H 1Ј; 3.52, dd, J5Ј,6Ј 9.9 Hz, H6Ј;
3.65, br m, H3; 3.81, br d, J7Ј,7Ј 11.8 Hz, H7Ј; 3.90, dd, J4Ј,5Ј 7.7 Hz,
H5Ј; 4.20, br d, H 7Ј; 4.35, br d, H4Ј; 4.43–5.02, m, 8H, CH2Ph;
4.86–4.88, m, H5; 5.08, dd, J1,2 0.8, J2,3 4.6 Hz, H2; 5.62, br s, H2Ј;
5.92, s, H 1; 7.26–7.43, 7.53–7.57, 8.03–8.06, 3m, Ph. 13C n.m.r.

-Acarbose. VII
893
(125.8 MHz) 33.44, C 6; 57.03, C 1Ј; 59.14, C 4; 69.91, C 3; 70.14,
C 7Ј; 72.53, 74.62, 75.26, 75.37, 4C, CH2Ph; 77.95, C 1; 79.33, C 2;
80.04, C 4Ј; 82.83, C 6Ј; 82.96, C 5; 84.98, C 5Ј; 126.17, C 2Ј;
127.68–129.77, 133.39, 138.20, 138.36, C 3Ј, Ph; 166.06, CO.
tial degradation had occurred to give poorly defined material (t.l.c.), in
addition to some residual starting materials. The reaction mixture was
quenched with saturated aqueous NaHCO3 (1 ml) and aqueous Na2S2O3
(0.5 M, 1 ml), the organic layer was separated, dried and the solvent
evaporated. The residue was subjected to flash chromatography, result-
ing in the recovery of some starting material (1H n.m.r.).
2-O-Benzoyl-3,4-O,N-carbonyl-1,6-episeleno-
4-[(1 R,4 R,5 S,6 S)-4,5,6-tribenzyloxy-3-(benzyloxymethyl)-
cyclohex-2-enyl]amino-1,4,6-trideoxy- -D-glucose (36)
Attempts at Protecting Group Manipulation
of the 1,6-Episeleno Carbamate (36)
The amino alcohol (33) (847 mg) was treated in the same manner as
for the amino alcohol (32) to give, after flash chromatography (5–30%
EtOAc/petrol), the carbamate (36) as a clear glass (726 mg, 83%), [ ]D
–65.4° (Found: C, 67.2; H, 5.5. C49H47NO9Se requires C, 67.4; H,
5.4%). max (NaCl) 1770, 1719 cm–1. 1H n.m.r. (500 MHz) 2.97, d,
(A) Trimethylsilyl chloride (17 µl, 137 µmol) was added to a solution
of sodium iodide (21 mg, 137 µmol) and the carbamate (36) (20 mg,
23 µmol) in dry MeCN (1.5 ml) and the resultant solution was stirred for
2 h at room temperature under N2, then heated under reflux for 5 h. T.l.c.
indicated that a complex mixture containing a considerable amount of
starting material was present and the reaction was abandoned.
(B) A mixture of anhydrous FeCl3 (59 mg, 370 µmol) and the carba-
mate (36) (20 mg, 23 µmol) in dry CH2Cl2 was stirred at 0° for 1 h and
then at room temperature for 2 h whilst under N2. Water (0.1 ml) was
added and the mixture was stirred for 5 min. Pyridine (3 ml), acetic
anhydride (2 ml) and 4-(dimethylamino)pyridine (2 mg) were added
and the mixture was stirred for 10 min. The reaction mixture was sub-
jected to the usual workup (EtOAc) to afford a light orange residue,
shown to consist of a complex mixture of products (t.l.c.).
J
6,6 9.1 Hz, H 6; 3.05, dd, J5,6 4.3 Hz, H 6; 3.90–4.00, m, H 5Ј,6Ј; 3.98, br
d, J7Ј,7Ј 12.6 Hz, H7Ј; 4.11, br t, J2,3 Ϸ J3,4 10.3 Hz, H 3; 4.18, br d, H 7Ј;
4.31–4.35, m, H 4Ј; 4.33, m, H4; 4.46–4.98, m, 8H, CH2Ph; 4.49–4.64,
m, H 1Ј; 5.04, t, J4,5 4.3 Hz, H 5; 5.04–5.08, m, H 2; 5.66, br s, H2Ј;
5.76, s, H 1; 7.26–7.65, 8.08–8.11, 2m, Ph. 13C n.m.r. (125.8 MHz)
35.89, C 6; 57.21, 59.17, C 4,1Ј; 69.37, C 7Ј; 72.28, 74.87, 75.30, 4C,
CH2Ph; 72.88, C 3; 77.56, 84.28, C 5Ј,6Ј; 122.81, C 2Ј; 127.60–129.88,
133.72, 137.74–139.04, Ph; 157.65, NCO; 165.87, COPh.
Attempted Glycosylation of the Alcohol (28) with the Carbamate (36)
(C) Trimethylsilyl triflate (80 µl) was added to a solution of the car-
bamate (36) (21 mg) in acetic anhydride (1 ml) and the reaction mixture
allowed to stand at 15° for 5 h. Aqueous sodium carbonate (2 M, 1 ml)
was added to the reaction mixture, followed by the usual workup
(EtOAc). Flash chromatography of the residue after evaporation of the
solvent afforded a major fraction (10 mg) which was shown to be a
complex mixture (1H n.m.r.).
A solution of N-iodosuccinimide (105 mg, 468 µmol) and TfOH
(20 µl) in dry Et2O/1,2-dichloroethane (1: 1, 4 ml) at 0° was added to a
solution of the carbamate (36) (350 mg, 400 µmol) and the alcohol (28)
(300 mg, 334 µmol) in dry 1,2-dichloroethane (3 ml) at 0°. The solu-
tion immediately went dark brown and was left at room temperature for
20 min. The solution was then quenched with saturated aqueous
NaHCO3 (1 ml) and aqueous Na2S2O3 (0.5 M, 2 ml), the organic layer
was separated, dried and the solvent evaporated. The crude residue was
dissolved in dry toluene (5 ml) and treated with Bu3SnH (800 µl,
3.0 mmol) and ␣,␣Ј-azobisisobutyronitrile (1 mg) at reflux for 2 h. The
solvent was evaporated and the residue partitioned between MeCN and
petrol. The MeCN layer was separated and evaporated and the residue
subjected to flash chromatography (7–9% EtOAc/petrol). The residue
after flash chromatography was taken up in pyridine/acetic anhydride
(1:1, 2 ml) and left to stand for 2 h. The solvent was evaporated to give
an oil which was purified by flash chromatography (7–9% EtOAc/petrol)
(D) The carbamate (36) (40 mg) was dissolved in acetic anhydride
(1 ml) and iodine (20 mg) was added. The reaction mixture was
allowed to stand overnight under N2. MeOH (1 ml) was added and the
mixture was allowed to stand at room temperature for 10 min. Workup
(EtOAc) afforded a light brown oil which was shown to be a complex
mixture (1H n.m.r.).
(E) Butyllithium (860 µl of 1.4 M in pentane, 1.2 mmol) was added
slowly to a solution of ethanethiol (88 µl, 1.2 mmol) in tetrahydrofuran
(1 ml) at room temperature. A white precipitate formed and the solvent
was evaporated under a stream of N2. The off-white solid was dissolved
in dry dimethylformamide (1 ml) and the carbamate (36) (44 mg,
50 µmol) was added. The solution was heated at 100° for 2 h, then the
solution was cooled and subjected to a workup (EtOAc) to provide a
black oil. T.l.c. showed this oil to contain u.v.-active material.
1
to give a clear oil (96 mg). Partial H n.m.r. (500 MHz): 5.41, 5.48,
5.52, 5.54, 4 br s. Partial 13C n.m.r. (125.8 MHz): 18.77, 19.03, 19.23,
20.48, 20.55, 21.41, 22.08, Me; 100.11, 100.63, 102.14, 102.25, 104.55,
104.58, 100.92, 101.38, CH.
Attempted Glycosylation of (40) with (15) Using Phenylselenenyl
Triflate
(A) A mixture of powdered 4 Å molecular sieves (0.5 g) and silver
triflate (77 mg, 0.30 mmol) in dry 1,2-dichloroethane (3 ml) was stirred
at 0° under N2 for 10 min. Phenylselenenyl chloride (58 mg, 0.30 mmol)
was added and the mixture was stirred for a further 5 min.
Treatment of the 1,6-Episeleno Monobenzoate (33) with
Tributylstannane
A solution of the monobenzoate (33) (64 mg, 74 µmol), tributyl-
stannane (120 µl, 450 µmol) and ␣,␣Ј-azobisisobutyronitrile in toluene
(3 ml) was heated at reflux for 30 min under N2. More tributylstannane
(50 µl, 160 µmol) was added and the solution was heated under reflux
for 30 min and then additional tributylstannane (50 µl, 160 µmol) was
added and reflux continued (30 min). The solvent was evaporated and
the residue was partitioned between MeCN and petrol. The MeCN
layer was evaporated to give a residue which was purified by flash chro-
matography (10–30% EtOAc/petrol plus 1% Et3N) to afford a clear oil
(11 mg). Partial 1H n.m.r. (200 MHz): 5.64, br s, 1H; 5.81, s, 1H.
(B) A solution of the tribenzoate (15) (129 mg, 0.24 mmol) and the
alcohol (40) (52 mg, 0.20 mmol) in dry 1,2-dichloroethane (3 ml) at 0°
was added to the solution generated in (A) under N2. Stirring was contin-
ued for 5 min, at which stage t.l.c. indicated that no tribenzoate remained.
Saturated NaHCO3 solution (1 ml) was added and the mixture was stirred
for 30 min. The mixture was filtered through Celite (EtOAc) and the
organic phase was washed with water, then dried and the solvent evapo-
rated. The residue was purified by flash chromatography to afford a
yellow oil (84 mg), shown to contain two spectroscopically similar prod-
ucts. Partial 1H n.m.r. (200 MHz): 1.20, 1.28, 1.43, 3s, 12H, Me; 4.88,
d, J1,2 8.2 Hz, H 1; 4.94, d, J1,2 8.2 Hz, H1.
Treatment of the Carbamate (35) with Sodium Metal in Liquid
Ammonia: N-[(1 R,4R,5S,6S)-4,5,6-Triacetoxy-3-
(acetoxymethyl)cyclohex-2-enyl]acetamide (45)
A small piece of sodium metal was added to a solution of the carba-
mate (35) (100 mg, 121 µmol) in dry tetrahydrofuran/MeOH (4 ml,
1: 1) and the mixture allowed to stand for 1 h. The solvent was evapo-
rated and the residue was dried under high vacuum for 2 h. Lithium
metal (50 mg, 8 mmol) was dissolved in NH3 (25 ml) at –78° and a solu-
tion of the above residue in dry tetrahydrofuran (4 ml) was added drop-
wise to the ammoniacal solution. The blue mixture was stirred at –78°
for 2 h, then NH4Cl (100 mg) was added and the ammonia allowed to
Attempted Glycosylation of the Alcohol (28) with the Carbamate (35)
A solution of N-iodosuccinimide (63 mg, 280 µmol) and TfOH
(10 µl) in dry Et2O/1,2-dichloroethane (1: 1, 4 ml) at 0° was added to a
solution of the carbamate (35) (198 mg, 240 µmol) and the alcohol (28)
(179 mg, 200 µmol) in dry 1,2-dichloroethane (3 ml) at 0°. T.l.c. indi-
cated no change after 5 min, so additional TfOH (10 µl) was added and
the solution left at room temperature for 1 h. After this time, substan-

R. V. Stick, D. M. G. Tilbrook and S. J. Williams
894
11
12
evaporate. The residue was treated with pyridine (4 ml), Ac2O (4 ml)
and 4-(dimethylamino)pyridine (50 mg) overnight at room tempera-
ture. The solvent was evaporated and the residue was partitioned
between water and EtOAc. The organic layer was separated and
washed with water and brine and then dried. The solvent was evapo-
rated and the residue purified by flash chromatography (40–100%
EtOAc/petrol) to afford, first, a mixture of the thioacetate (43) and the
disulfide (44) as an oil (13 mg). Partial 1H n.m.r. (200 MHz): 2.35, s,
SAc; 2.80–2.96, m, CH2S; 3.31, dd, J1,1 12.3, J1,2 8.6 Hz, H1; 4.24, dd,
Köll, P., Tayman, F. S., and Heyns, K., Chem. Ber., 1979, 112, 2305.
Stick, R. V., Tilbrook, D. M. G., and Williams, S. J., Tetrahedron
Lett., 1997, 38, 2741.
13
McAuliffe, J. C., Skelton, B. W., Stick, R. V., and White,A. H., Aust.
J. Chem., 1997, 50, 209.
14
15
Bracher, F., and Litz, T., J. Prakt. Chem., 1995, 337, 516.
Nicotra, F., Panza, L., Ronchetti, F., and Russo, G., Gazz. Chim. Ital.,
1989, 119, 577.
16
17
18
McAuliffe, J. C., and Stick, R. V., Aust. J. Chem., 1997, 50, 193.
McAuliffe, J. C., and Stick, R. V., Aust. J. Chem., 1997, 50, 219.
Knapp, S., Naughton, A. B. J., and Dhar, T. G. M., Tetrahedron Lett.,
1992, 33, 1025.
J1,2 5.8 Hz, H 1; 5.93, br s, H 2Ј.
Next to elute was the amide (45) as a pale yellow oil (24 mg, 51%)
which crystallized as shards, m.p. 137–140° (EtOH/Pri2O), [ ]D
–108.9°. 1H n.m.r. (500 MHz) 1.96, s, NHAc; 2.02, 2.05, 2.05, 2.05,
4s, 12H, OAc; 4.37, br d, J7,7 13.0 Hz, H7; 4.66, br d, H7; 4.85, br t,
19
Stick, R. V., Tilbrook, D. M. G., and Williams, S. J., Aust. J. Chem.,
1999, 52, 895.
20
21
J
1,6 Ϸ J1,NH 8.1 Hz, H 1; 5.10, dd, J5,6 10.8 Hz, H 6; 5.36, dd, J4,5 8.1 Hz,
Ito, Y., and Ogawa, T., Tetrahedron Lett., 1988, 29, 1061.
Ito, Y., Ogawa, T., Numata, M., and Sugimoto, M., Carbohydr. Res.,
1990, 202, 165.
H5; 5.71–5.78, m, H 2,4, NH. 13C n.m.r. (125.8 MHz) 20.60, 20.62,
20.69, 20.72, 4C, OCOMe; 23.21, NHCOMe; 50.40, C 1; 62.66, C 7;
70.25, 71.23, 72.14, C 4,5,6; 128.60, C 2; 132.20, C 3; 169.79, 169.96,
170.16, 170.37, 4C, OC=O; 171.23, NHC=O.
22
23
Olah, G. A., Narang, S. C., Gupta, B. G. B., and Malhotra, R., J. Org.
Chem., 1979, 44, 1247.
Debenham, J. S., Rodebaugh, R., and Fraser-Reid, B., J. Org. Chem.,
1997, 62, 4591.
References
24
25
26
27
1
2
Alzeer, J., and Vasella, A., Helv. Chim. Acta, 1995, 78, 177.
Kartha, K. P. R., and Field, R. A., Tetrahedron, 1997, 53, 11753.
Feutrill, G. I., and Mirrington, R. N., Aust. J. Chem., 1972, 25, 1719.
Stick, R. V., Tilbrook, D. M. G., and Williams, S. J., Aust. J. Chem.,
1997, 50, 237.
McAuliffe, J. C., and Stick, R. V., Aust. J. Chem., 1997, 50, 225.
Truscheit, E., Frommer, W., Junge, B., Müller, L., Schmidt, D. D.,
and Wingender, W., Angew. Chem., Int. Ed. Engl., 1981, 20, 744.
Schmidt, D. D., Frommer, W., Junge, B., Müller, L., Wingender, W.,
Truscheit, E., and Schäfer, D., Naturwissenschaften, 1977, 64, 535.
Puls, W., Keup, U., Krause, H. P., Thomas, G., and Hoffmeister, F.,
Naturwissenschaften, 1977, 64, 536.
3
4
28
29
30
31
32
Fraser-Reid, B., Udodong, U. E., Wu, Z., Ottosson, H., Merritt, J. R.,
Rao, C. S., Roberts, C., and Madsen, R., Synlett, 1992, 927.
Rodebaugh, R., and Fraser-Reid, B., J. Am. Chem. Soc., 1994, 116,
3155.
5
6
7
McAuliffe, J., and Hindsgaul, O., Chem. Ind., 1997, 170.
Stick, R. V., Top. Curr. Chem., 1997, 187, 187.
Toyokuni, T., Ogawa, S., and Suami, T., Bull. Chem. Soc. Jpn, 1983,
56, 1161.
Driguez, H., McAuliffe, J. C., Stick, R. V., Tilbrook, D. M. G., and
Williams, S. J., Aust. J. Chem., 1996, 49, 343.
8
9
Ogawa, S., Miyamoto, Y., and Nakajima, A., Chem. Lett., 1989,
725.
Fürst, A., and Plattner, Pl. A., Helv. Chim. Acta, 1949, 32, 275.
Klayman, D. L., and Griffin, T. S., J. Am. Chem. Soc., 1973, 95, 197.
Payne, G. B., J. Org. Chem., 1962, 27, 3819.
10
Takeo, K., Fukatsu, T., andYasato, T., Carbohydr. Res., 1982, 107, 71.