The fluorination (at C5) of some derivatives of D-glucose

Article abstract of DOI:10.1071/CH03288

The photobromination of various per-esters of β-D-glucopyranose and α- and β-D-glucopyranosyl fluoride has yielded 5-bromo derivatives capable of conversion into the corresponding 5-fluorides. The best reagent for this conversion was found to be silver te

Full text of DOI:10.1071/CH03288

Rapid Communication
CSIRO PUBLISHING
www.publish.csiro.au/journals/ajc
Aust. J. Chem. 2004, 57, 345–353
The Fluorination (at C5) of Some Derivatives of d-Glucose
Brian W. Skelton,A Robert V. Stick,A,B Keith A. Stubbs,A Andrew G. Watts,A and
Allan H. WhiteA
A Chemistry, School of Biomedical and Chemical Sciences M313, University of Western Australia,
Crawley WA 6009, Australia.
B Author to whom correspondence should be addressed (e-mail: rvs@chem.uwa.edu.au).
The photobromination of various per-esters of β-d-glucopyranose and α- and β-d-glucopyranosyl fluoride has
yielded 5-bromo derivatives capable of conversion into the corresponding 5-fluorides. The best reagent for this
conversion was found to be silver tetrafluoroborate in ether/dichloromethane. The single-crystal X-ray structure
determination of penta-O-benzoyl-5-fluoro-α-l-idopyranose is presented.
As well, various approaches to 5-fluoro glycopyranosides have been developed. The photobromination of
phenyl tetra-O-acetyl-β-d-glucopyranoside, followed by fluorination and protecting group removal gave phenyl
5-fluoro-β-d-glucopyranoside. The photobromination of penta-O-acetyl-β-d-glucopyranose, followed by hydro-
lysis and treatment with methanol, gave a hemiacetal that was converted via treatment with diethylaminosulfur
trifluoride followed by protecting group removal into methyl 5-fluoro-β-d-glucopyranoside. Methyl 2,3,4-tri-O-
benzyl-6-deoxy-α- and -β-d-xylo-hex-5-enosides were converted into their corresponding epoxides with either
dimethyldioxirane or 3-chloroperbenzoic acid; subsequent treatment with hydrogen fluoride/pyridine then gen-
erated the 5-fluoro glycopyranosides as the major products. A single-crystal X-ray structure determination of
1,6-anhydro-2,3,4-tri-O-benzyl-5-hydroxy-β-l-idose is presented.
Manuscript received: 31 October 2003.
Final version: 3 January 2004.
The pioneering work of Withers has resulted in the design
and synthesis of fluoro sugars as potent, mechanism-based
inhibitors of various glycosidases and glycanases.[1] For
example, the 2-deoxy-2-fluoro-β-d-glucoside 1, 2-deoxy-
2-fluoro-β-d-glucopyranosyl fluoride 2, and 5-fluoro-β-d-
glucopyranosyl fluoride 3 (Scheme 1) have been used to
inhibit the action of β-glucosidases isolated fromAlcaligenes
faecalis and Agrobacterium sp. and a xylanase/glucanase
from Cellulomonas fimi. In several cases, a fluoro sugar
has been used to label the catalytic nucleophile of various
retaining glycosidases.
questions. For example, the dinitrophenyl β-d-glucoside 1
is a better inhibitor of the β-glucosidase from Aspergillus
faecalis than the β-d-glucosyl fluoride 2 (Ki 0.05 mM for 1,
Ki 0.4 mM for 2).[2,3] Would, then, the phenyl β-d-glucoside
4 be as good an inhibitor of retaining β-glucosidases as the
β-d-glucosyl fluoride 3? Would the methyl β-d-glucoside 5
be capable of acting as an inhibitor of β-glucosidases at all?
We thus set about a synthesis of some of these molecules.
The obvious route to these 5-fluoro derivatives involves a
bromination at C5, followed by halogen (fluorine) exchange.
In the initial stages, we decided to apprentice ourselves
with some simple ‘photobrominations’, that excellent reac-
tion discovered (somewhat serendipitously) by Ferrier and
Furneaux in the mid-1970s.[4,5] Thus, bromination of penta-
O-acetyl-β-d-glucopyranose gave the bromide 6 (Scheme 2),
There seemed a need for better synthetic routes to the 5-
fluoro glycopyranoses. As well, it occurred to us that it may
be advantageous to develop methods for the synthesis
of 5-fluoro glycopyranosides to allow answers to several
OR
OH
OH
O
RO
RO
O
O
HO
HO
HO
HO
OR
X
X
OR
X
HO
F
F
6
7
8
9
R ꢀ Ac
X ꢀ Br
Br
O
NO2
1
2
3
4
5
X ꢀ
X ꢀ F
Bz
O2N
F
4-BrC6H4CO
Br
OPh
OMe
4-O2NC6H4CO
H
Scheme 2.
Scheme 1.
© CSIRO 2004
10.1071/CH03288
0004-9425/04/040345

346
B. W. Skelton et al.
OR
OR
O
RO
RO
O
RO
RO
F
X
OR
OR
X
F
10 R ꢀ Ac
X ꢀ Br
11 R ꢀ Ac
X ꢀ Br
12
11
66
31
32
30
12
19
21
Bz
Ac
Bz
Br
F
13
20
Bz
Ac
Br
F
F(5)
61
10
60
O(6)
F
O(10)
O(1)
O(30)
O(5)
6
Scheme 3.
5
1
4
3
2
F
OAc
OR
F
OBz
O(40)
O(4)
40
BzO
AcO
OAc
OBz
O
O
O(20)
RO
RO
O(2)
20
21
22
O
OR
OR
F
OBz
OBz
OAc
OAc
14
15
17
R ꢀ Ac
41
42
16
18
Bz
4-BrC6H4CO
Scheme 4.
a relatively stable molecule when crystalline but very prone to
decomposition when impure or as an oil.[6] A much more sta-
ble product 7 was obtained when the comparable bromination
was performed on the much less reactive penta-benzoate,[7]
and the trend continued with the penta-4-bromobenzoate
(the bromide 8). Even though we could prepare the crude
penta-4-nitrobenzoate 9, the product had the characteristics
of granite and we were unable to investigate the product(s)
of its photobromination.
Bromination of the β- and α-anomers of tetra-O-acetyl-d-
glucopyranosyl fluoride gave the known bromo fluorides 10
and 11 (Scheme 3), respectively.[8,9] The comparable reac-
tions in the benzoyl series were again much slower: photo-
bromination of tetra-O-benzoyl-β-d-glucopyranosyl fluoride
gave the bromo fluoride 12 in good yield. However, irradi-
ation of tetra-O-benzoyl-α-d-glucopyranosyl fluoride never
gave a clean conversion into the bromide 13. After 48 h at
reflux in carbon tetrachloride in the presence of bromine
and calcium carbonate, large amounts of starting material,
as well as numerous by-products, were observed. The lack of
reactivity of the α-fluoride is most likely a result of the addi-
tive effects of the deactivating benzoyl groups and the axial
anomeric substituent. For the latter, the fluorine atom has an
electron-withdrawing effect on the non-bonding electrons of
O5 (the anomeric effect) that are required for the necessary
stabilization of the intermediate free-radical generated at C5
during the photobromination step.
Fig. 1. Molecular projection of 16.
conformation in the solid state (Fig. 1), also corroborated in
solutionby 1HNMRspectroscopy(smallvaluesfor J1,2, J2,3
and J3,4); substantial differences are found within the pairs of
exocyclic angles at each ring carbon (Table 1). The bromide
8 furnished the fluoride 17 in 80% yield when treated with
silver tetrafluoroborate in ether/dichloromethane.
,
Treatment of the bromide 6 with silver fluoride in aceto-
nitrile resulted in the virtually exclusive formation of the
l-ido fluoride 18. The assignment of the l-ido configuration
to 18 was based mainly on the value of J4,F (11.1 Hz), signif-
icantly smaller than the corresponding value (22.7 Hz) in the
d-gluco fluoride 14. This difference was consistent across all
of the fluorides studied here and greatly aided in the assign-
ment of configuration [d-gluco (J4,F 22.5–23.6 Hz) versus
l-ido (J4,F 6.0–14.2 Hz)]. Somewhat surprisingly, a similar
treatment of the bromide 7 gave the d-gluco fluoride 15 in
79% yield.
Any attempt to rationalize the stereochemical outcome of
the above fluorination reactions, namely retention or inver-
sion of configuration at C5, seems fraught with danger.
However, the reactions with silver tetrafluoroborate must
proceed in solution by abstraction of a bromide ion (to form
silver bromide) and so generate a planar carbenium ion, per-
haps solvated (with ether) on the Re face and associated with a
counter ion (tetrafluoroborate) on the Si face—further reac-
tion of this salt produces the thermodynamically favoured
(axial) fluoride. For the reactions with silver fluoride, the
abstraction of a bromide ion at the surface of the insoluble
reagent again generates a planar carbenium ion but now only
the Re face is available for nucleophilic attack by fluoride ion
in solution, to form the equatorial fluoride; alternatively, the
solvent (acetonitrile) may intercept the carbenium ion from
Having seemingly mastered the art of photobromination,
wenextdecidedtotryourhandatfluorinationofthebromides
so prepared. Thus, treatment of the bromide 6 with silver
tetrafluoroborate in ether gave the fluoride 14 (Scheme 4)
in 61% yield, a result superior to the comparable reaction in
toluene.[8] Treatment ofthelessreactivebromide 7 withsilver
tetrafluoroborate in ether/dichloromethane gave the fluoride
15 in excellent yield (85%). A minor by-product in this reac-
tion was the l-ido epimer 16—a single crystal X-ray structure
1
determination of 16 showed the molecule to exist in a C4

Fluorination of Derivatives of d-Glucose
347
Table 1. Selected non-hydrogen geometries of compounds 16
and 40
O
OAc
CHO
AcO
Atoms
16
40A
Interbond angles [◦]B
114.8(5)
105.8(5)
109.5(5)
114.8(6)
104.0(6)
108.4(9)
112.9(7)
105.6(9)
110.3(8)
113.1(9)
110.4(8)
102.6(7)
111.0(7)
105.8(7)
108.0(9)
120.5(6)
OAc
OAc
C(2)–C(1)–O(5)
C(2)–C(1)–O(1)
O(1)–C(1)–O(5)
C(1)–C(2)–C(3)
C(1)–C(2)–O(2)
C(3)–C(2)–O(2)
C(2)–C(3)–C(4)
C(2)–C(3)–O(3)
C(4)–C(3)–O(3)
C(3)–C(4)–C(5)
C(3)–C(4)–O(4)
C(5)–C(4)–O(4)
C(4)–C(5)–O(5)
C(4)–C(5)–F(5)/O(50)
O(5)–C(5)–F(5)/O(50)
C(5)–O(5)–C(1)
108.7(3)
111.1(3)
106.0(4)
110.9(4)
109.3(4)
108.6(4)
114.1(4)
104.2(4)
112.4(5)
112.2(5)
106.2(4)
111.1(4)
108.2(4)
113.6(5)
109.3(4)
102.0(4)
22
Scheme 5.
OAc
O
AcO
AcO
OPh
X
OAc
23
24
25
X ꢀ H
Br
F
Scheme 6.
Ring torsion angles [◦]
O(5)–C(1)–C(2)–C(3)
C(1)–C(2)–C(3)–C(4)
C(2)–C(3)–C(4)–C(5)
C(3)–C(4)–C(5)–O(5)
C(4)–C(5)–O(5)–C(1)
C(5)–O(5)–C(1)–C(2)
−36(1)
58.0(5)
−34.3(6)
33.1(6)
39(1)
−46(1)
50(1)
OAc
−53.6(6)
O
AcO
AcO
−51(1)
75.7(5)
Br
O
44(1)
−78.5(4)
OAc
A C(1)–O(1)–C(6), O(1)–C(6)–C(5) are 106.9(5), 103.4(4); C(6)–C(5)–
C(4), O(5,50) are 113.8(4), 101.4(5), 109.8(4)◦. C(1)–O(1)–C(6)–C(5)
is −7.6(4); O(1)–C(6)–C(5)–C(4), O(5,50) are −84.2(5), 31.8(4),
147.3(3), and C(6)–O(1)–C(1)–C(2), O(5) are 97.5(4), −20.3(4)◦.
B Ring angles are in bold.
26
Scheme 7.
OR
OR
the Si face (in so doing generating an α-nitrilium ion), forc-
ing the approach of fluoride ion again from the Re face. We
have absolutely no idea for the mechanism of formation of the
axial fluoride 15 from the axial bromide 7 with silver fluoride
in acetonitrile, although participation by an adjacent benzo-
yloxy group cannot be excluded. In fact, such neighbouring
group participation may well offer an alternative explanation
for the formation of some of the above fluorides.
O
O
RO
RO
RO
RO
X
OMe
OR
OR
29 R ꢀ Bz
31 Ac
OH
F
27 R ꢀ Bz
28
30
X ꢀ OH
Bz
Ac
OMe
OMe
Scheme 8.
Finally, treatment of the bromides 10, 11, and 12 with
silver tetrafluoroborate in ether/dichloromethane gave the
fluorides 19, 20, and 21 in good yield (55, 53, and 80%,
respectively), a significant improvement (for 19 and 20) over
the reported method.[8]
ion, is ‘lost’ to the fluorine atom at C5 (n → σ* interaction).
Instead, a slow halogen exchange (F to Br) ensues with the
excess reagent present, to produce the bromide 7.
A somewhat unrelated quirk of these molecules is that
halogenation at C5 prevents selective transformations at
C1: treatment of the fluoride 14 with ammonium carbonate
in dimethylformamide gave (not unexpectedly) the hexosu-
lose 22 (Scheme 5),[8,10] while treatment of the fluoride 15
with hydrogen bromide in glacial acetic acid gave only the
bromide 7!
These last results were typical of many reactions reported
here. A molecule such as 15 essentially has two anomeric
carbon atoms (C1 and C5), with the C5 F bond being par-
ticularly strong owing to the anomeric effect. Any attempt at
generating a carbenium ion at C1, for example by treatment
of 15 with hydrogen bromide, is doomed to failure—a lone
pair of electrons on O5, necessary to stabilize the carbenium
We now turned to the synthesis of the 5-fluoro glyco-
pyranosides. Photobromination of the phenyl β-d-glucoside
23 (Scheme 6) gave the bromide 24 in good yield.[11] Treat-
ment of 24 with silver tetrafluoroborate in ether then gave
the fluoride 25, and deprotection with ammonia in methanol
gave the desired 4 (Scheme 1). Synthesis of the methyl
β-d-glucoside 5, by way of the 5-bromo compound, was
unlikely to succeed since Ferrier and Haines have shown
that standard photobromination conditions convert methyl
tetra-O-acetyl-β-d-glucopyranoside into the bromo lactone
26 (Scheme 7).[12]
Initially, we adopted a rather indirect approach to the
methyl β-d-glucoside 5. Hydrolysis of the 5-bromo derivative
7 (Scheme 2) gave the diol 27 (Scheme 8).[7] The treatment

348
B. W. Skelton et al.
OAc OAc
OAc
Oꢁ
ꢁ
Et2NSF3
CH2Cl2
OMe
AcO
AcO
30
AcO
OMe
O
AcO
OAc
Fꢂ
Fꢂ
31
OAc OAc
OMe
AcO
O
F
OAc
32
Scheme 9.
of 27 with methanol gave the hemiacetal 28,[7] and a subse-
quent treatment with diethylaminosulfur trifluoride (DAST)
gave the 5-fluoro-β-d-glucoside 29. Although such fluori-
nations generally proceed with inversion of configuration,
the anomeric effect probably predominates here and ensures
formation of the more stable (axial) fluoride.
O
OR
RO
RO
O
OMe
OR
RO
RO
OMe
33 R ꢀ Ac
36
Bn
34 R ꢀ Ac
37
Bn
Unfortunately, the attempted debenzoylation of 29 with
sodium methoxide in methanol was not successful. So, in
a parallel sequence, the 5-bromo derivative 6 (Scheme 2)
gave the rather unstable hemiacetal 30. Treatment of 30 with
DAST gave the desired 5-fluoro-β-d-glucoside 31, together
with substantial amounts of what appeared to be two acyclic
fluorides 32 (Scheme 9). The deprotection of 31 proceeded
smoothly to give the methyl β-d-glucoside 5.
We now considered other routes to 5-fluoro glycopyrano-
sides, and the intermediacy of glycos-5-enes and their derived
epoxides seemed an attractive alternative. Thus, the alkenes
33and 34(Scheme10)werepreparedaccordingtoknownand
related methods.[13] Treatment of 33 with iodine in aqueous
acetonitrile[14] indeed gave a desired iodohydrin but of struc-
ture 35 where loss of the aglycon was obvious (Scheme 11).
Initial attempts to epoxidize 33 and 34 using dimethyl-
dioxirane (DMDO) were unsuccessful, so it was decided to
convert the alkenes into their benzyl ether counterparts 36
and 37. Treatment of the alkene 36 with DMDO gave the
l-ido epoxide 38 (Scheme 12), the stereochemistry of which
was assigned from a 1H NMR NOESY spectrum that showed
an interaction between the protons of the methyl group and
a proton at C6. The approach of the reagent, DMDO, to the
Re face of the alkene 36 is to be expected in view of the axial
methoxy group at C1.
Scheme 10.
I
ꢁI
ꢁ
O
I2
O
AcO
AcO
AcO
AcO
33
MeCN
AcO
AcO
OMe
H2O
OMe
I
I
O
O
AcO
AcO
AcO
H2O
ꢁ CH3OH
OH
AcO
HO
HO
AcO
OAc
OMe
35
Scheme 11.
BnO
BnO
BnO
MeO
OBn
OBn
BnO
O
O
O
O
OMe
38
39
Scheme 12.
OBn
Treatment of the alkene 37 with DMDO gave, as expected,
a mixture of epoxides 39. From a comparison of the 1H NMR
spectrum of this mixture with that of the l-ido epoxide 38,
the d-gluco to l-ido ratio was determined to be 1 : 2. The
epoxides 38 and 39 were rather unstable and were used in
subsequent reactions without purification.
OBn
OH
BnO
AgBF4
ꢂCH3OH
O
38
Et2O/CH2Cl2
(H2O)
MeO
OH
OBn
OBn
OBn
Treatment of the epoxide 38 with silver tetrafluorobo-
rate in ether/dichloromethane gave, after workup, only the
bridged hexosulose 40 (Scheme 13), whose structure was
assigned from a single-crystal X-ray structure determination
(Fig. 2). In potentially a better method of fluorination, the
epoxide 38 was treated with hydrogen fluoride/pyridine. The
OBn
OH
BnO
O
HO
CHO
O
O
OBn
40
Scheme 13.

Fluorination of Derivatives of d-Glucose
349
water and left to stir (15 min), then poured into ice-water, and the resul-
tant gum triturated (EtOAc) to give a fine, colourless powder. Recrys-
tallization of the powder (three times) gave the penta-4-bromobenzoate
as a colourless, microcrystalline powder (5.1 g, 70%), mp 125–128◦C
(EtOAc; lit.[17] 123◦C), [α]D +83.0◦ (lit.[17] +81◦).
32
31
5-Bromo-penta-O-(4-bromobenzoyl)-β-D-glucopyranose 8
30
22
21
A suspension of penta-O-(4-bromobenzoyl)-β-d-glucopyranose (1.0 g,
0.92 mmol) and CaCO3 (1 g) in CCl4 (40 mL) containing Br2 (0.25 mL,
4.8 mmol) was irradiated at reflux untilTLC showed completeconsump-
tion of the starting material (5 h). The mixture was allowed to cool, and
was then filtered through Celite. The filtrate was concentrated to give a
light yellow oil that crystallized. Recrystallization yielded the bromide
8 as a fine, colourless powder (0.89 g, 83%), mp 223–226◦C (EtOAc),
[α]D −13.5◦ (Found: C 41.9, H 2.4. C41H26Br6O11 requires C 41.9,
H 2.2%). δH (300 MHz) 7.89–7.39 (20H, m, Ar), 6.70 (d, J1,2 8.5, H1),
6.24 (t, J2,3 ≈ J3,4 9.8, H3), 5.91 (dd, H2), 5.84 (d, J3,4 9.8, H4), 4.90,
4.67 (ABq, 2H, J 13.5, H6). δC (75.5 MHz) 164.59, 164.47, 164.38,
163.92, 163.38(5C, C=O), 132.09–126.72(Ar), 96.00(C5), 92.37(C1),
71.48, 70.02, 69.56 (C2, C3, C4), 66.61 (C6). m/z (FAB) 1172.6615
O(3)
2
20
40
42
3
6
41
4
O(2)
O(4)
O(50)
O(5)
1
5
O(1)
Fig. 2. Molecular projection of 40.
•
[(M + H)+ requires 1172.6613].
F
F
O
O
BnO
BnO
BnO
BnO
OMe
Tetra-O-benzoyl-5-bromo-β-D-glucopyranosyl Fluoride 12
BnO
BnO
RO
HO
Tetra-O-benzoyl-β-d-glucopyranosyl fluoride[18] (500 mg, 0.74 mmol),
CaCO3 (400 mg), and Br2 (0.20 mL, 3.7 mmol) in CCl4 (40 mL) were
irradiated at reflux until 1H NMR analysis showed complete consump-
tion of the starting material (5 h). The mixture was allowed to cool, then
filtered. Dichloromethane was added and the mixture washed with 1 M
aqueous sodium thiosulfate and saturated aqueous sodium bicarbonate,
and then dried and concentrated to give a colourless foam. Crystalliza-
tion gave the bromide 12 as a microcrystalline solid (420 mg, 84%),
OMe
41 R ꢀ H
42
Ac
43
Scheme 14.
fluoride 41 (Scheme 14) was the sole product, characterized
as the acetate 42. In the 1H NMR spectrum of 42, the value
for J4,F (11.0 Hz) ruled out a d-gluco configuration but was
commensurate with the l-ido configuration.
Treatment of the epoxides 39 with the same reagent gave
just one product, the fluoride 43—in the 1H NMR spectrum,
the value for J4,F (14.2 Hz) again suggested the l-ido config-
uration. Thus, it appears that the treatment of epoxides with
hydrogen fluoride/pyridine is a viable method for the synthe-
sis of 5-fluoro glycosides, but the reasons for the dominant
formation of the l-ido isomers here remain obscure.
During the course of this work, a report appeared on
the generation of epoxides from glycos-5-enes and their
subsequent treatment with hydrogen fluoride/pyridine at low
temperature.[15] Here, the alkenes 36 and 37 were separately
treated with 3-chloroperbenzoic acid, presumably to gen-
erate (mixtures of) epoxides, and these were treated with
hydrogen fluoride/pyridine at –78◦C, to generate the 5-fluoro
l-idosides 41 and 43, respectively, but accompanied in each
case by an amount of the d-gluco isomer.
mp 160–162◦C (Et2O), [α]D +23.5◦ (Found:
C 60.2, H 4.0.
C34H26BrFO9 requires C 60.3, H 3.9%). δH (300 MHz) 8.19–7.28 (20H,
m, Ph), 6.20 (br t, J2,3 ≈J3,4 9.7, H3), 5.99 (dd, J1,2 7.1, J1,F 51.0,
H1), 5.95 (d, H4), 5.78 (ddd, J2,F 5.5, H2), 5.01, 4.67 (ABq, 2H, J
12.3, H6). δC (75.5 MHz) 165.36, 165.17, 164.91, 164.49 (4C, C=O),
133.93–128.00 (Ph), 106.95 (d, J1,F 225, C1), 75.55 (d, J5,F 6.0, C5),
71.19–70.78 (2d, C2, C3), 69.11 (C4), 66.36 (C6). m/z (FAB) 677.0843
•
[(M + H)+ requires 677.0822].
Penta-O-acetyl-5-fluoro-β-D-glucopyranose 14
(a) Silver tetrafluoroborate (47 mg, 0.24 mmol) in dry Et2O (5 mL) was
added dropwise to the bromide 6 (92 mg, 0.20 mmol) in Et2O (5 mL)
with stirring. The solution was left to stir (5 min), and a precipitate
(AgBr) formed immediately. The resultant mixture was filtered through
a plug of silica, the filtrate concentrated, and the residue purified by flash
chromatography (30% EtOAc/petrol) to give a colourless oil that crys-
tallized. Recrystallisation gave the fluoride 14 as a fine powder (49 mg,
61%), mp 86–90◦C (EtOAc/petrol), [α]D −41.6◦ (Found: C 46.9, H
5.0. C16H21FO11 requires C 47.1, H 5.2%). δH (300 MHz) 6.18 (d, J1,2
8.3, H1), 5.48 (dd, J2,3 9.3, J3,4 9.5, H3), 5.30 (dd, J4,F 22.7, H4), 5.22
(dd, H2), 4.38 (dd, J6,6 12.0, J6,F 7.0, H6), 3.99 (dd, J6,F 4.1, H6), 2.13,
2.10, 2.08, 2.05, 2.01 (5s, 15H, Me). δC (125.8 MHz) 169.70, 169.66,
169.33, 169.22 (5C, C=O), 109.50 (d, J5,F 253, C5), 88.59 (d, J1,F 5.7,
C1), 69.76–61.27 (C2, C3, C4), 61.52 (d, J6,F 38.0, C6), 20.69, 20.51,
20.46, 20.39 (5C, Me). δF (282.4 MHz) −131.66 (ddd, J4,F 22.7, J6,F
4.1, 7.0, F5).
Experimental
(b) Silver tetrafluoroborate (55 mg, 2.8 mmol) was added to the bro-
mide 6 (110 mg, 0.23 mmol) in dry toluene (10 mL). The mixture was
left to stir (30 min), and a precipitate formed immediately. The mix-
ture was then processed as in (a) to give the fluoride 14 as a colourless
powder (16 mg, 17%). The spectroscopic data (1H and 13C NMR) were
consistent with those observed above.
General experimental procedures have been given previously.[13] Irradi-
ation of reaction mixtures was effected using two 150-W lamps at 5 cm
from the reaction vessel. 19F NMR spectra were recorded using C6F6
as an external reference.
Penta-O-(4-bromobenzoyl)-β-D-glucopyranose
Penta-O-benzoyl-5-fluoro-β-D-glucopyranose 15
d-Glucose (1.2 g, 6.7 mmol) was dissolved in dry pyridine (20 mL)
by heating the mixture on a steam bath. Upon cooling the solution,
4-bromobenzoyl chloride (10.0 g, 45.6 mmol) was added and the mix-
ture held at 60◦C (3 h). The mixture was quenched by the addition of
(a) Silver tetrafluoroborate (100 mg, 0.54 mmol) in dry Et2O (5 mL) was
added dropwise to the bromide 7 (350 mg, 0.45 mmol) in Et2O/CH2Cl2
(1 : 1, 5 mL) with stirring. The mixture was treated as in (a) above to

350
B. W. Skelton et al.
give, after flash chromatography (15% EtOAc/petrol), the fluoride 15
as a colourless oil (275 mg, 85%), [α]D +13◦ (Found: C 68.6, H 4.4.
C41H31FO11 requires C 68.5, H, 4.3%). δH (300 MHz) 8.13–7.22 (25H,
m, Ph), 6.75 (d, J1,2 7.8, H1), 6.27 (t, J2,3 8.8, J3,4 8.9, H3), 6.12 (dd,
J4,F 22.6, H4), 5.97 (dd, H2), 4.63 (dd, J6,6 12.0, J6,F 6.5, H6), 4.54
(dd, J6,F 6.7, H6). δC (75.5 MHz) 165.27, 164.97, 164.03 (5C, C=O),
133.94–128.17 (Ar), 109.87 (d, J5,F 230, C5), 89.73 (d, J1,F 3.0, C1),
70.45, 69.74 (C2, C3), 69.02 (d, J4,F 24.0, C4), 62.60 (d, J6,F 37.0, C6).
12.0, H6), 2.13, 2.10, 2.03, 1.98 (12H, 4s, Me), δF (470.6 MHz) −138.2
(ddd, F1), −123.0 (m, F5).
Tetra-O-benzoyl-5-fluoro-β-D-glucopyranosyl Fluoride 21
Silver tetrafluoroborate (35 mg, 0.18 mmol) was added to the bromide
12 (100 mg, 0.15 mmol) in Et2O/CH2Cl2 (1 : 1, 10 mL) with stirring.
The mixture was treated as in (a) for the preparation of 14 to give, after
flash chromatography (35% EtOAc/petrol), the fluoride 21 as a colour-
less oil (75 mg, 81%), [α]D +33.4◦ (Found: C 66.1, H 4.2. C34H26F2O9
requires C 66.1, H 4.2%). δH (300 MHz) 8.10–7.21 (20H, m, Ar), 6.24–
5.88 (m, H1,3,4), 5.74 (ddd, J1,2 5.3, J2,3 6.6, J2,F 9.8, H2), 4.68 (dd,
J6,6 12.0, J6,F5 6.6, H6), 4.57 (dd, J6,F5 6.4, H6). δC (75.5 MHz) 165.35,
165.27, 164.81 (4C, C=O), 133.79–128.15 (Ar), 109.73 (dd, J5,F5
231, J5,F1 5.5, C5), 104.42 (d, J1,F1 222, C1), 71.23, 69.48, 68.16 (C2,
C3, C4), 62.25 (d, J6,F5 39.2, C6). δF (282.4 MHz) −127.67 (dddd,
J4,F5 23.1, J6,F5 6.4, 6.6, JF1,F5 10.2, F5), −139.78 (ddd, J1,F1 52.0,
•
m/z (FAB) 699.1893 [(M − F)+ requires 699.1866].
Also isolated was the l-ido isomer 16 as colourless plates (25 mg,
8%). δH (500 MHz) 8.12–7.37 (25H, m, Ph), 6.73–6.72 (m, H1), 6.05
(dd, J3,4 4.7, J4,F 6.0, H4), 5.89 (br dd, J2,3 4.5, H3), 5.77 (dd, J1,2 2.4,
H2), 4.78 (dd, J6,6 12.0, J6,F 10.4, H6), 4.67 (dd, J6,F 20.7, H6).
(b) Silver tetrafluoroborate (54 mg, 2.8 mmol) was added to the
bromide 7 (180 mg, 0.23 mmol) in dry toluene (10 mL). The mixture
was treated as in (a) for the preparation of 14 to give the fluoride
15 as a colourless oil (124 mg, 75%). The spectroscopic data (1H and
13C NMR) were consistent with those observed above.
(c) Silver fluoride (76 mg, 0.60 mmol) was added to the bromide 7
(235 mg, 0.30 mmol) in dry MeCN (5 mL). The mixture was treated as
in (a) for the preparation of 14 to give the fluoride 15 as a colourless
oil (170 mg, 79%). The spectroscopic data (1H and 13C NMR) were
consistent with those observed above.
•
J2,F1 9.8, JF1,F5 10.2, F1). m/z (FAB) 617.1676 [(M + H)+ requires
617.1623].
Treatment of 15 with Hydrogen Bromide
The fluoride 15 (540 mg) in HBr/HOAc (30% w/w; 4 mL) was kept at
room temperature (3 days). Usual workup (CH2Cl2) gave a colourless
oil that crystallized on standing. Recrystallization (EtOAc/petrol) gave
the bromide 7 as a fine powder (455 mg, 78%). The spectroscopic data
(1H and 13C NMR) were consistent with those reported.[7]
Penta-O-(4-bromobenzoyl)-5-fluoro-β-D-glucopyranose 17
Silver tetrafluoroborate (40 mg, 0.21 mmol) in dry Et2O (2 mL) was
added dropwise to the bromide 8 (190 mg, 0.16 mmol) in Et2O/CH2Cl2
(1 : 1, 4 mL) with stirring.The mixture was treated as in (a) for the prepa-
ration of 14 to give, after flash chromatography (15% EtOAc/petrol),
a colourless oil that crystallized. Recrystallization gave the fluoride 17
as fine needles (144 mg, 80%), mp 231–233◦C (AcOH), [α]D −15◦
(Found: C 44.2, H 2.5. C41H26Br5FO11 requires C 44.3, H 2.3%).
δH (300 MHz) 7.73–7.30 (20H, m, Ar), 6.49 (d, J1,2 7.8, H1), 5.98
(t, J2,3 8.7, J3,4 9.8, H3), 5.85 (dd, J4,F 22.5, H4), 5.72 (dd, H2),
4.45 (dd, J6,6 12.2, J6,F 7.4, H6), 4.34 (dd, J6,F 6.9, H6). δC
(75.5 MHz) 165.34, 165.29, 165.03, 164.97, 164.04 (5C, C=O),
133.96–128.12 (Ar), 109.87 (d, J5,F 231, C5), 89.73 (d, J1,F 4.3,
C1), 70.43, 69.72 (C2, C3), 69.00 (d, J4,F 24.4, C4), 62.60 (d, J6,F
39.4, C6).
Phenyl Tetra-O-acetyl-5-fluoro-β-D-glucopyranoside 25
Silver tetrafluoroborate (300 mg, 1.50 mmol) in Et2O (15 mL) was
added dropwise to the bromide 24[12] (500 mg, 1.00 mmol) in Et2O
(10 mL) at room temperature. A precipitate formed immediately and
stirring was continued (5 min). The mixture was then filtered through a
plug of silica, subjected to a usual workup (Et2O), and purified by flash
chromatography (40% EtOAc/petrol) to give the fluoride 25 as a colour-
less oil (256 mg, 58%), [α]D +18◦ (Found: C 54.3, H 5.2. C20H23FO10
requires C 54.3, H 5.2%). δH (300 MHz) 7.34–6.99 (20H, m, Ph), 5.60
(d, J1,2 7.7, H1), 5.54 (br dd, J2,3 9.3, J3,4 9.8, J3,F 0.6, H3), 5.41 (dd,
J4,F 22.7, H4), 5.39 (dd, H2), 4.38 (dd, J6,6 11.9, J6,F 8.3, H6), 4.07
(dd, J6,F 4.9, H6), 2.09, 2.08, 2.03, 2.01 (12H, 4s, Me). δC (75.5 MHz)
169.80, 169.76, 169.30, 169.22 (4C, C=O), 156.46, 129.63, 123.50,
116.59 (Ph), 109.64 (d, J5,F 229, C5), 95.77 (d, J1,F 4.5, C1), 70.35,
69.34 (C2, C3), 68.02 (d, J4,F 24.0, C4), 61.78 (d, J6,F 40.0, C6), 20.59,
20.48, 20.41, 20.37 (4C, Me).
Penta-O-acetyl-5-fluoro-α-L-idopyranose 18
Silver fluoride (53 mg, 0.42 mmol) was added to the bromide 6 (130 mg,
0.28 mmol) in dry MeCN (5 mL). The mixture was treated as in (a)
for the preparation of 14 to give, after flash chromatography (35%
EtOAc/petrol), the fluoride 18 as a colourless oil (87 mg, 76%). δH
(300 MHz) 6.11 (dd, J1,2 3.1, J1,F 1.4, H1), 5.47 (dd, J3,4 6.0, J4,F 11.1,
H4), 5.32–5.22 (m, H2,3), 4.31 (dd, J6,6 11.9, J6,F 15.2, H6), 4.24 (dd,
J6,F 23.0, H6), 2.12, 2.10, 2.08, 2.07, 2.03 (15H, 5s, Me).
Phenyl 5-Fluoro-β-D-glucopyranoside 4
Ammonia was passed over a stirred solution of the fluoride 25 (200 mg)
in dry MeOH (5 mL) at 0◦C (5 min) and the solution kept at this temper-
ature (1 h). The solution was then concentrated and the residue purified
by flash chromatography (EtOH/EtOAc/H2O, 1 : 6 : 0.5) to give the
β-D-glucoside 4 as a colourless oil (98 mg, 79%) (Found: C 52.4, H
5.5. C12H13FO6 requires C 52.6, H 5.5%). δH (300 MHz) 7.31–6.95
(m, Ph), 5.31 (d, J1,2 8.0, H1), 4.58–3.58 (5H, m, H2, H3, H4, H6). δC
(75.5 MHz) 158.57, 130.28, 123.21, 117.24 (Ph), 113.76 (d, J5,F 226,
C5), 99.07 (d, J1,F 4.6, C1), 74.10, 73.47 (C2, C3), 70.84 (d, J4,F 28.0,
C4), 62.85 (d, J6,F 34.0, C6).
Tetra-O-acetyl-5-fluoro-β-D-glucopyranosyl Fluoride 19
Silver tetrafluoroborate (130 mg, 0.66 mmol) in Et2O (15 mL) was
added dropwise to the bromide 10 (235 mg, 0.55 mmol) in Et2O (5 mL)
with stirring. The mixture was treated as in (a) for the preparation of 14
to give, after flash chromatography (35% EtOAc/petrol), the fluoride 19
as a colourless oil (111 mg, 55%). δH (300 MHz)[19] 5.65 (dd, J1,F 49.0,
H1), 5.45 (dd, J4,F 22.1, H4), 5.40 (dd, J3,4 9.2, H3), 5.20 (ddd, J1,2 6.0,
J2,3 8.0, J2,F 9.0, H2), 4.34 (dd, JF,6 7.1, H6), 4.08 (dd, JF,6 4.6, J6,6
12.0, H6), 2.15, 2.13, 2.09, 2.01 (12H, 4s, Me). δF (470.6 MHz) −143.1
(ddd, F1), −128.3 (m, F5).
Methyl Tetra-O-benzoyl-5-fluoro-β-D-glucopyranoside 29
Diethylaminosulfur trifluoride (0.5 mL, 4 mmol) was added to the
methyl β-d-glucoside 28[11] (780 mg, 1.25 mmol) in CH2Cl2 (10 mL) at
−10◦C and the solution kept at this temperature (20 min). Usual workup
(CH2Cl2) followed by flash chromatography (25% EtOAc/petrol) gave
the fluoride 29 as fine needles (320 mg, 41%), mp 138–139◦C (MeOH),
[α]D +25◦ (Found:C66.7, H4.6. C35H29FO10 requiresC66.9, H4.7%).
δH (300 MHz) 8.00–7.19 (20H, m, Ph), 6.15 (t, J2,3 9.5, J3,4 9.9, H3),
5.97 (dd, J4,F 22.5, H4), 5.66 (dd, J1,2 7.8, H2), 5.24 (d, H1), 4.63
(dd, J6,6 12.0, J6,F 7.6, H6), 4.53 (dd, J6,F 7.9, H6), 3.59 (s, OMe).
δC (75.5 MHz) 165.52, 165.45, 165.05, 165.01 (4C, C=O), 133.63–
128.28 (Ph), 109.35 (d, J5,F 229, C5), 99.15 (d, J1,F 2.0, C1), 71.16,
Tetra-O-acetyl-5-fluoro-α-D-glucopyranosyl Fluoride 20
Silver tetrafluoroborate (195 mg, 1.00 mmol) in Et2O (15 mL) was
added dropwise to the bromide 11 (350 mg, 0.82 mmol) in Et2O (10 mL)
with stirring. The mixture was treated as in (a) for the preparation of 14
to give, after flash chromatography (35% EtOAc/petrol), the fluoride 20
as a colourless oil (160 mg, 53%). δH (300 MHz)[19] 5.81 (dd, J1,F 53.5,
H1), 5.70 (dd, J2,3 ≈ J3,4 9.8, H3), 5.30 (dd, J4,F 22.0, H4), 5.00 (ddd,
J1,2 3.0, J2,F 24.0, H2), 4.25 (dd, JF,6 6.5, H6), 4.01 (dd, JF,6 3.0, J6,6

Fluorination of Derivatives of d-Glucose
351
69.57 (C2, C3), 69.48 (d, J4,F 25.0, C4), 62.95 (d, J6,F 36.0, C6), 57.54
J1,2 8.0, H2), 4.88 (t, J4,5 9.4, H4), 4.44 (d, H1), 3.57 (s, OMe), 3.56–
3.49 (m, H5), 3.29 (dd, J5,6 2.7, J6,6 10.9, H6), 3.15 (dd, J5,6 8.5, H6),
2.08, 2.06, 1.95 (9H, Me). δC (75.5 MHz) 169.99, 169.38 (3C, C=O),
101.34 (C1), 73.38, 72.41, 72.25, 71.46 (C2, C3, C4, C5), 57.14 (OMe),
20.69, 20.57 (3C, Me), 2.93 (C6).
•
(OMe). m/z (FAB) 629.1832 [(M + H)+ requires 629.1823].
Methyl Tetra-O-acetyl-5-hydroxy-β-D-glucopyranoside 30
The hydrate of 2,3,4,6-tetra-O-acetyl-hexos-5-ulose[20] (1.60 g) was
stirred at 40◦C in dry MeOH (30 mL; 2 h).The solution was then concen-
trated to give the methyl β-d-glucoside 30 as a pale yellow oil (1.74 g),
which was used without purification. δH (300 MHz) 5.45 (t, J2,3 9.9, J3,4
7.1, H3), 5.21 (d, J1,2 7.1, H1), 5.13 (d, H4), 4.91 (dd, H2), 4.08, 3.94
(2H, ABq, J 11.8, H6), 3.40 (s, OMe), 2.09, 2.00, 1.98, 1.91 (12H, 4s,
Me). δC (75.5 MHz) 170.92, 170.35, 170.05, 169.68 (4C, C=O), 95.10
(C5), 90.50 (C1), 73.10, 69.78, 69.64 (C2, C3, C4), 64.89 (C6), 56.54
Methyl Tri-O-acetyl-6-deoxy-β-D-xylo-hex-5-enopyranoside 34
1,8-Diazabicyclo[5.4.0]undec-7-ene (13.0 mL, 87.2 mmol) was added
to the above iodide (7.50 g, 17.4 mmol) inTHF (60 mL), and the solution
heated at reflux (2 h). The mixture was then concentrated, subjected to
a usual workup (EtOAc), and purified by flash chromatography (40%
EtOAc/petrol) to give the alkene 34 as a colourless oil (3.63 g, 69%),
[α]D −34.6◦ (lit.[22] −34.8◦). δH (300 MHz) 5.68 (dt, J3,4 8.9, J4,6 1.8,
1.8, H4), 5.08 (dd, J2,3 5.6, H3), 5.00 (dd, J1,2 4.4, H2), 4.74 (t, J6,6
1.8, H6), 4.65 (d, H1), 4.47 (t, H6), 3.51 (OMe), 2.16, 2.09, 2.02 (9H,
Me). δC (75.5) 170.14, 169.69, 169.64 (3C, C=O), 150.93 (C5), 101.26
(C1), 94.70 (C6), 72.48, 72.22, 68.46 (C2, C3, C4), 56.73 (OMe), 20.97
(3C, Me).
•
(OMe), 20.55, 20.41 (4C, Me). m/z (FAB) 361.1156 [(M − OH)+
requires 361.1135].
Methyl Tetra-O-acetyl-5-fluoro-β-D-glucopyranoside 31
Diethylaminosulfur trifluoride (0.3 mL, 2 mmol) was added to the
β-d-glucoside 30 (0.4 g, 1 mmol) in dry CH2Cl2 (15 mL) at 0◦C. The
solution was kept at this temperature (5 min), then subjected to a usual
workup (CH2Cl2), and the residue purified by flash chromatography
(30% EtOAc/petrol). First to elute was the β-d-glucoside 31 as a colour-
less oil (165 mg, 43%) (Found: C 47.4, H 5.6. C15H21FO10 requires
C 47.4, H 5.6%). δH (300 MHz) 5.43 (t, J2,3 9.9, J3,4 9.9, H3), 5.31 (dd,
J4,F 22.8, H4), 5.10 (dd, J1,2 8.0, H2), 4.90 (d, H1), 4.33 (dd, J6,6 11.9,
J6,F 7.2, H6), 4.03 (dd, J6,F 4.9, H6), 3.52 (s, OMe), 2.09, 2.07, 2.06,
1.99 (12H, 4s, Me). δC (125.8 MHz) 169.84, 169.34, 169.32 (4C, C=O),
109.21 (d, J5,F 227, C5), 98.59 (d, J1,F 5.0, C1), 70.55, 69.40 (C2, C3),
68.17 (d, J4,F 25.0, C4), 61.89 (d, J6,F 38.0, C6), 57.44 (OMe), 20.59,
20.54, 20.48, 20.40 (4C, Me). δF (470.5 MHz) −130.85 (ddd, J4,F 22.8,
J6,F 4.9, 7.2, F5). m/z (FAB) 381.1193 [(M + H)+• requires 381.1197].
Next to elute were the hex-5-uloses 32 as an inseparable mixture
(1 : 1.3) (165 mg, 40%). δH (300 MHz) 5.78 (dd, J2,3 6.7, H3), 5.72 (dd,
J2,3 6.7, H3), 5.50 (d, J3,4 3.1, H4), 5.44 (d, J3,4 3.1, H4), 5.27–5.18
(m, H2), 5.22 (dd, J1,2 2.5, J1,F 66.0, H1), 5.16 (dd, J1,2 4.0, J1,F 66.0,
H1), 4.85, 4.83, 4.73, 4.71 (2H, 2ABq, J 17.3, H6), 3.57 (d, JOMe,F 1.2,
OMe), 3.51 (d, JOMe,F 1.2, OMe), 2.19, 2.15, 2.08 (12C, 3s, Me). δC
(75.5 MHz) 197.36, 197.28 (C5), 169.67, 169.56, 169.38 (4C, C=O),
110.19, 109.81 (2d, J1,F 223, C1), 70.23, 69.90 (2d, J2,F 25.0, 29.0, C2),
74.55, 67.62, 67.56 (C3, C4), 66.63, 66.59 (C6), 57.79, 57.49 (OMe),
20.44, 20.39, 20.27 (4C, Me). δF (470.5 MHz) −135.9 (br dd, J1,F 66.0,
J2,F 8.0, F1), −143.17 (br dd, J1,F 66.0, J2,F 16.0, F1). m/z (FAB)
Tri-O-acetyl-6-deoxy-5-hydroxy-6-iodo-β-D-glucopyranose 35
Iodine (508 mg, 2.00 mmol) was added to the alkene 33[13] (300 mg,
1.0 mmol) in MeCN/H2O (30 mL; 2 : 1) at −15◦C and the solution left
to stir at this temperature (30 min). The solution was then concentrated,
subjected to a usual workup (CH2Cl2), and the residue purified by flash
chromatography (50% EtOAc/petrol) to give the diol 35 as colourless
cubes (360 mg, 84%), mp 132–138◦C (EtOH), [α]D +23◦ (Found: C
33.4, H 4.1. C12H17IO9 requires C 33.4, H 4.0%). δH (500 MHz) 7.28
(d, OH), 7.26 (s, OH), 5.55 (dd, J2,3 9.9, J3,4 9.8, H3), 5.38 (d, H4), 5.31
(dd, J1,OH 7.8, J1,2 8.0, H1), 5.03 (dd, H2), 3.59 (s, OMe), 3.56, 3.47
(2H,ABq, J 10.9, H6), 2.26, 2.24, 2.18 (9H, Me). δC (75.5 MHz) 167.60,
167.38, 167.19 (3C, C=O), 91.97 (C5), 88.48 (C1), 71.29, 69.00, 68.94
(C2, C3, C4), 18.82, 18.63 (3C, Me), 8.17 (C6). m/z (FAB) 414.9914
•
[(M − OH)+ requires 414.9890].
Methyl Tri-O-benzyl-6-deoxy-β-D-xylo-hex-5-enopyranoside 37
The alkene 34 (7.20 g, 23.8 mmol) in MeOH (50 mL) was treated with
Na (3 mg) and the solution kept at room temperature (30 min). Concen-
tration of the mixture then gave a pale yellow oil that was dissolved
in DMF (100 mL) and treated with NaH (3.8 g, 95 mmol; 60% suspen-
sion in oil) with vigorous stirring (30 min). Benzyl bromide (8.76 mL,
73.8 mmol) was added dropwise and stirring continued (2 h). Water
(10 mL) was then added dropwise to the mixture and stirring continued
(2 h). Concentration of the mixture, followed by usual workup (EtOAc)
and flash chromatography (10% EtOAc/petrol), gave the alkene 37 as
a colourless oil (9.46 g, 89%), [α]D −10.8◦. δH (300 MHz) 7.40–7.29
(15H, m, Ph), 4.84–4.67 (8H, m, H6,CH2Ph), 4.65 (d, J1,2 6.1, H1),
4.10 (dt, J4,6 ≈ J4,6 1.4, J3,4 7.5, H4), 3.67 (dd, J2,3 6.9, H3), 3.60
(s, OMe), 3.59 (dd, H2). δC (75.5 MHz) 153.84 (C5), 138.23–127.67
(Ph), 103.99 (C6), 94.61 (C1), 82.87, 81.45, 78.28 (C2, C3, C4),
74.27, 73.86, 73.04 (3C, CH2Ph), 56.90 (OMe). m/z (FAB) 447.2154
•
361.1156 [(M − F)+ requires 361.1135].
Methyl 5-Fluoro-β-D-glucopyranoside 5
Ammonia was bubbled through a solution of 31 (62 mg) in dry
MeOH (5 mL) at 0◦C (5 min) and the solution held at this temperature
(30 min). The solution was then concentrated and the residue purified
by flash chromatography (EtOH/EtOAc/H2O, 1 : 6 : 0.5) to give the
β-d-glucoside 5 as a colourless oil (26 mg, 74%) (Found: C 39.5,
H 6.3. C7H13FO6 requires C 39.6, H 6.2%). δH (300 MHz) 4.98 (d,
J1,2 8.2, H1), 4.01–3.55 (5H, m, H2, H3, H4, H6), 3.45 (s, OMe).
δC (75.5 MHz) 111.45 (d, J5,F 229, C5), 99.28 (d, J1,F 4.6, C1), 74.58,
74.46 (C2, C3), 71.87 (d, J4,F 29.0, C4), 62.65 (d, J6,F 35.0, C6),
57.82 (OMe).
•
[(M + H)+ requires 447.2171].
Methyl 5,6-Epoxy-tri-O-benzyl-β-L-idopyranoside 38
DMDO in acetone (100 mL of 0.1 M) was added to the alkene 36[13]
(780 mg, 1.75 mmol) in CH2Cl2 (20 mL) at room temperature and the
mixture kept (10 min). Concentration of the mixture gave the epoxide
38 as a colourless oil (795 mg). δH (500 MHz) 7.45–7.38 (15H, m, Ph),
4.93–4.64 (7H, m, H1, CH2Ph), 4.08 (t, J2,3 9.6, J3,4 9.7, H3), 3.87
(d, H4), 3.69 (dd, J1,2 3.7, H2), 3.46 (s, OMe). δC (75.5 MHz) 138.48–
127.55 (Ph), 99.26 (C1), 81.43 (C5), 79.85, 78.99, 77.30 (C2, C3, C4),
75.83, 74.89, 73.63 (3C, CH2Ph), 56.80 (OMe), 50.09 (C6).The epoxide
38 was used without purification.
Methyl Tri-O-acetyl-6-deoxy-6-iodo-β-D-glucopyranoside
Triphenylphosphine(24.3 g, 92.8 mmol), imidazole(6.32 g, 92.8 mmol),
and I2 (19.6 g, 77.3 mmol) were added to a suspension of methyl
β-d-glucopyranoside (15.0 g, 77.3 mmol) in toluene (200 mL) with vig-
orous stirring. The mixture was held at 70◦C (2 h), then quenched
with water, and stirring continued (30 min). The mixture was concen-
trated, and then taken up in pyridine (100 mL) and Ac2O (100 mL).
The solution was left to stand at room temperature (4 h), and then
concentrated. Usual workup (EtOAc) followed by flash chromatogra-
phy (20% EtOAc/petrol) gave the title iodide as fine needles (23.6 g,
71%), mp 112–113◦C (Pr2i O/petrol; lit.[21] 114–115◦C), [α]D +2.0◦
(lit.[21] +1.0◦). δH (300 MHz) 5.19 (t, J2,3 9.4, J3,4 9.4, H3), 4.97 (dd,
Methyl 5,6-Epoxy-tri-O-benzyl-α-L-idopyranoside
and Methyl 5,6-Epoxy-tri-O-benzyl-β-D-glucopyranoside
DMDO in acetone (100 mL of 0.1 M) was added to the alkene 37
(500 mg, 1.12 mmol) in CH2Cl2 (20 mL) at room temperature and

352
B. W. Skelton et al.
the mixture kept (10 min). Concentration of the mixture gave the
epoxides 39 as an inseparable mixture (d-gluco/l-ido 1 : 2; 480 mg).
δH (300 MHz) d-gluco: 7.39–7.28 (15H, m, Ph), 4.95–4.69 (6H, m,
CH2Ph), 4.68 (d, J1,2 8.3, H1), 3.99 (d, J3,4 9.3, H4), 3.92 (dd, J2,3 8.5,
H3), 3.59 (dd, H2), 3.52 (s, OMe), 2.97, 2.89 (2H, ABq, J 5.5, H6);
l-ido: 7.39–7.28 (15H, m, Ph), 4.97–4.63 (6H, m, CH2Ph), 4.56 (d, J1,2
6.2, H1), 3.99 (d, J3,4 9.0, H4), 3.73 (dd, J2,3 8.3, H3), 3.64 (dd, H2),
3.54 (s, OMe), 3.13, 2.87 (2H, ABq, J 5.2, H6).
Methyl 2,3,4-Tri-O-benzyl-5-fluoro-β-L-idopyranoside 41
and Methyl 2,3,4-Tri-O-benzyl-5-fluoro-α-D-glucopyranoside
(a)3-Chloroperbenzoicacid(170 mg, 1.0 mmol)wasaddedtothealkene
36 (460 mg, 1.0 mmol) in a mixture of CH2Cl2 (10 mL) and 0.5 M
NaHCO3 solution (10 mL), and the mixture stirred at room tempera-
ture (5 min). Separation of the organic phase followed by evaporation
of the solvent gave the epoxide 38 as an inseparable mixture (2 : 1,
1H NMR) with methyl 5,6-anhydro-tri-O-benzyl-β-d-glucopyranoside
(480 mg).
1,6-Anhydro-tri-O-benzyl-5-hydroxy-β-L-idopyranose 40
(b) Hydrogen fluoride/pyridine (0.5 mL) was added to the above
mixture of epoxides (480 mg) in CH2Cl2 (5 mL) at −78◦C, and the solu-
tion stirred (10 min). Water was added and the usual workup (CH2Cl2)
followed by flash chromatography (EtOAc/petrol; 1 : 4) furnished the
β-l-idoside 41 as a colourless oil (123 mg, 24%), [α]D +13.5◦. m/z
The epoxide 38 (460 mg, 1.00 mmol) in CH2Cl2 (20 mL) was added
dropwise to a stirred solution of AgBF4 (195 mg, 1.00 mmol) in Et2O
(30 mL) and stirring continued (10 min). The mixture was then filtered
through a plug of silica and the filtrate treated with cation-exchange
resin (Amberlite IR-120, H+). The mixture was filtered and the filtrate
concentrated to give a pale yellow solid. Recrystallization then gave the
hexosulose 40 as fine needles (383 mg, 85%), mp 137–141◦C (Et2O),
[α]D +43◦ (Found: C 72.4, H 6.1. C27H28O6 requires C 72.3, H 6.3%).
δH (500 MHz) 7.36–7.27 (15H, m, Ph), 5.28 (d, J1,2 1.9, H1), 4.86,
4.83 (ABq, J 11.0, CH2Ph), 4.88, 4.79 (ABq, J 11.5, CH2Ph), 4.72,
4.67 (ABq, J 11.8, CH2Ph), 4.21 (d, J6,6 8.1, H6), 3.80 (t, J2,3 7.8,
J3,4 8.1, H3), 3.69 (dd, H4), 3.56 (dd, H2), 3.34 (dd, J4,6 1.9, H6).
δC (125.8 MHz) 138.36–127.66 (Ph), 103.36 (C5), 98.44 (C1), 82.57
(C2), 82.05 (C3), 81.92 (C4), 75.53, 7•4.69, 73.00 (3C, CH2Ph), 68.61
(C6). m/z (FAB) 447.1790 [(M − H)+ requires 447.1807].
•
481.2037 [(M − H)+ requires 481.2026].
Next to elute was methyl 2,3,4-tri-O-benzyl-5-fluoro-α-D-
glucopyranoside as a colourless oil (250 mg, 50%), [α]D +8.8◦ (Found:
C 69.9, H 6.6. C28H31FO6 requires C 69.7, H 6.5%). δH (600 MHz)
7.38–7.29 (15H, m, Ph), 5.02–4.68 (6H, m, CH2Ph), 4.72 (d, J1,2 3.7,
H1), 4.34 (dd, J2,3 9.7, J3,4 9.7, H3), 3.70 (dd, J4,F 23.0, H4), 3.58–3.50
(3H, m, H2,6), 3.44 (s, OMe), 2.44 (br s, OH). δC (150.8 MHz) 138.47–
127.71 (Ph), 112.15 (d, J5,F 232, C5), 99.77 (C1), 79.07, 77.72 (C2, C3),
76.53 (d, J4,F 25.0, C4), 75.95, 75.35, 74.46 (3C, CH2Ph), 62.90 (d, J6,F
38.0, C6), 56.09 (OMe). m/z 481.2034 [(M − H)+• requires 481.2026].
Methyl 2,3,4-Tri-O-benzyl-5-fluoro-α-L-idopyranoside 43
and Methyl 2,3,4-Tri-O-benzyl-5-fluoro-β-D-glucopyranoside
Methyl 2,3,4-Tri-O-benzyl-5-fluoro-β-L-idopyranoside 41
(a) 3-Chloroperbenzoic acid (190 mg, 1.1 mmol) was added to the
alkene 37 (500 mg, 1.1 mmol) in a mixture of CH2Cl2 (10 mL) and
0.5 M NaHCO3 solution (10 mL), and the mixture stirred at room
temperature (5 min). Separation of the organic phase followed by evap-
oration of the solvent gave the epoxides 39 as an inseparable mixture
(ratio of d-gluco to l-ido 2 : 1; 517 mg).
Hydrogen fluoride/pyridine (0.7 mL) was added to a stirred solution
of the epoxide 38 (240 mg) in CH2Cl2 (50 mL) at 0◦C. The solution
was kept at this temperature (5 min), then poured into cold water, and
subjected to a usual workup (CH2Cl2). Flash chromatography (20%
EtOAc/petrol) then gave the fluoride 41 as a colourless oil (210 mg,
88%). δH (300 MHz) 4.85 (d, J1,2 3.1, H1), 4.81–4.68 (6H, m, CH2Ph),
4.05–3.80 (5H, m, H2,3,4,6), 3.52 (s, OMe), 2.56 (br m, OH). δC
(75.5 MHz) 138.00–127.67 (Ph), 113.31 (d, J5,F 218, C5), 98.25 (d, J1,F
4.4, C1), 80.05 (d, J4,F 32.0, C4;), 77.76 (d, J3,F 6.0, C3), 77.63 (C2),
74.07, 73.82, 73.54 (3C, CH2Ph), 64.00 (d, J6,F 30.6, C6), 56.86 (OMe).
(b) Hydrogen fluoride/pyridine (0.5 mL) was added to the epox-
ides 39 (500 mg) in CH2Cl2 (5 mL) at −78◦C and the solution stirred
(10 min). Water was added and the usual workup (CH2Cl2) fol-
lowed by flash chromatography (EtOAc/toluene, 1 : 9) furnished the
α-l-idoside 43 as colourless plates (140 mg, 27%), mp 132–135◦C
(Pri2O), [α]D +16.6◦ (Found: C 69.8, H 6.5. C28H31FO6 requires C
•
69.7, H 6.5%). m/z 481.2038 [(M − H)+ requires 481.2026].
Methyl 6-O-Acetyl-tri-O-benzyl-5-fluoro-β-L-idopyranoside 42
Next to elute was methyl 2,3,4-tri-O-benzyl-5-fluoro-β-D-gluco-
pyranoside asacolourlessoil(200mg, 38%), [α]D +3.9◦. δH (600 MHz)
7.38–7.28 (15H, m, Ph), 4.95–4.75 (7H, m, H1, CH2Ph), 4.00 (dd, J2,3
9.5, J3,4 9.8, H3), 3.78 (dd, J4,F 23.6, H4), 3.74–3.68 (2H, m, H6),
3.59 (s, OMe), 3.51 (dd, J1,2 8.2, H2), 2.38 (br s, OH). δC (150.8 MHz)
138.26–127.68 (Ph), 111.35 (d, J5,F 227, C5), 101.72 (d, J1,F 4.0, C1),
81.62, 80.61 (C2, C3), 76.97 (d, J4,F 24.0, C4), 75.91, 75.39, 74.84
(3C, CH2Ph), 62.65 (d, J6,F 35.0, C6), 57.68 (OMe). m/z 481.2029
Acetic anhydride (1 mL) was added to the fluoride 41 (130 mg) in
CH2Cl2 (10 mL) and pyridine (2 mL) at room temperature and the mix-
ture kept (20 min). Concentration of the mixture, followed by flash
chromatography (25% EtOAc/petrol) of the residue, gave the acetate
42 as a colourless oil (112 mg, 79%). δH (300 MHz) 7.43–7.25 (15H,
m, Ph), 4.93, 4.53 (ABq, J 12.1, CH2Ph), 4.83 (d, J1,2 3.0, H1), 4.81,
4.69 (ABq, J 11.3, CH2Ph), 4.62 (s, CH2Ph), 4.56 (dd, J6,6 12.2, J6,F
16.6, H6), 4.30 (dd, J6,F 21.0, H6), 3.95 (dd, J2,3 7.8, J3,4 5.1, H3),
3.82 (dd, J1,2 3.0, H2), 3.76 (dd, J4,F 11.0, H4), 3.48 (s, OMe), 2.08
(s, Me). δC (75.5 MHz) 170.17 (C=O), 138.02–127.75 (Ph), 112.07
(d, J5,F 218, C5), 98.32 (d, J1,F 4.4, C1), 77.43 (d, J4,F 32.0, C4), 77.32
(C2), 77.11 (d, J3,F 4.7, C3), 74.15, 73.60, 73.42 (3C, CH2Ph), 64.20
(d, J6,F 25.9, C6), 56.70 (OMe), 20.69 (Me).
•
[(M − H)+ requires 481.2026].
Structure Determination of Compounds 16 and 40
Both determinations were of less than optimal standard, by virtue of
crystal quality and bulk, presenting difficulties involving compromises
in procedure as noted; in neither case did the determination lead to
assignment of absolute configuration (in both cases, assumed from the
chemistry) and Friedel data were combined in the merging process. In
both cases monochromatic MoKα radiation sources (λ 0.71073 Å) were
employed.
For 16, a full sphere of data was measured at approximately 295 K
using a single counter instrument (2θ/θ scan mode), Nt(otal) reflections
merging to N unique (Rint cited), No with I > 3σ(I) being considered
‘observed’ and used in the full matrix least-squares refinement with-
out absorption correction, refining anisotropic displacement parameter
Methyl 2,3,4-Tri-O-benzyl-5-fluoro-α-L-idopyranoside 43
Hydrogen fluoride/pyridine (0.8 mL) was added to the epoxides 39 (ratio
of d-gluco to l-ido 1 : 2; 310 mg) in CH2Cl2 (50 mL) at 0◦C. The solu-
tion was kept (5 min), then poured into cold water, and subjected to
a usual workup (CH2Cl2). Flash chromatography (20% EtOAc/petrol)
then gave the fluoride 43 as a colourless oil (278 mg, 84%). δH
(300 MHz) 7.39–7.28 (15H, m, Ph), 4.88 (br dd, J1,2 2.1, J1,F 2.0, H1),
4.80–4.56 (6H, m, CH2Ph), 4.27 (dd, J3,4 8.9, J4,F 14.2, H4), 3.95 (br dd,
J6,6 12.1, J6,F 9.0, H6), 3.88 (dd, J2,3 5.0, H3), 3.82–3.73 (m, H2,6),
3.52 (s, OMe), 2.34 (br s, OH). δC (75.5 MHz) 138.12–127.71 (Ph),
112.56 (d, J5,F 231, C5), 102.32 (C1), 82.32, 79.73 (C2, C3), 79.40 (d,
J4,F 26.0, C4), 74.78, 73.76, 73.02 (3C, CH2Ph), 63.05 (d, J6,F 38.0,
C6), 56.21 (OMe).
forms for C, F, O, (x, y, z, Uiso
) being constrained at estimates. For
H
40, data were measured at approximately 153 K using a Bruker AXS
CCD area-detector instrument and processed similarly (observed crite-
ria: F > 4σ(F); (x, y, z, Uiso
) refined). Conventional residuals R, Rw
H
are quoted at convergence [weights: (σ2(F) + 0.0004F2)−1]; neutral

Fluorination of Derivatives of d-Glucose
353
atom scattering factors were employed within the Xtal 3.7 program
system.[16] Pertinent results are given below and in the Figures (which
show 50% for 16, and 20% for 40, probability amplitude displacement
ellipsoids for C, O, (F), hydrogen atoms having arbitrary radii of 0.1 Å).
CIFs are deposited with the Cambridge Crystallographic Data Centre.
[4] R. J. Ferrier, R. H. Furneaux, J. Chem. Soc., Perkin Trans. 1
1977, 1996.
[5] L. Somsák, R. J. Ferrier, Adv. Carbohydr. Chem. Biochem.
1991, 49, 37.
[6] R. Blattner, R. J. Ferrier, J. Chem. Soc., Perkin Trans. 1 1980,
1523.
Crystal/Refinement Data
[7] R. J. Ferrier, P. C. Tyler, J. Chem. Soc., Perkin Trans. 1
1980, 1528.
[8] J. D. McCarter, S. G. Withers, J. Am. Chem. Soc. 1996, 118,
241. doi:10.1021/JA952732A
[9] J. P. Praly, G. Descotes, Tetrahedron Lett. 1987, 28, 1405.
doi:10.1016/S0040-4039(00)95938-0
[10] R. J. Ferrier, S. R. Haines, J. Chem. Soc., Perkin Trans. 1 1984,
1675.
[11] R. J. Ferrier, P. C. Tyler, J. Chem. Soc., Perkin Trans. 1 1980,
2762.
[12] R. J. Ferrier, S. R. Haines, J. Chem. Soc., Perkin Trans. 1 1984,
1689.
[13] J. C. McAuliffe, R. V. Stick, Aust. J. Chem. 1997, 50, 193.
doi:10.1071/CH96151
[14] A. M. Sanseverino, M. C. S. de Mattos, Synthesis 1998, 1584.
doi:10.1055/S-1998-2187
[15] M. C. T. Hartman, J. K. Coward, J. Am. Chem. Soc. 2002, 124,
10036. doi:10.1021/JA0127234
[16] The Xtal 3.7 System (Eds S. R. Hall, D. J. du Boulay, R. Olthof-
Hazekamp) 2001 (University of Western Australia: Perth).
[17] A. J. Eagle, T. M. Herrington, N. S. Isaacs, J. Chem. Res. (M)
1993, 2663.
[18] G. H. Posner, S. R. Haines, Tetrahedron Lett. 1985, 26, 5.
doi:10.1016/S0040-4039(00)98451-X
[19] J. D. McCarter, Ph.D. Thesis 1995 (University of British
Columbia: Vancouver, BC).
16, C41H31FO11, M 718.7. Triclinic, space group P1 (C11, no. 1),
a 12.658(6), b 10.520(3), c 7.469(5) Å, α 105.20(4), β 90.70(5),
γ 110.85(3)◦, V 890.7(9) Å3. Dc (Z 1) 1.340 g cm−3. µMo 1.0 cm−1
;
specimen: 0.70 × 0.42 × 0.10 mm3. 2θmax 50◦; Nt 4058, N 2573 (Rint
0.031), No 1786; R 0.055, Rw 0.061. |&ρmax| 0.32(3) e Å−3. CCDC no.
220141.
40, C27H28O6, M 448.5. Monoclinic, space group C2 (C23, no. 5),
a 20.665(7), b 5.911(2), c 19.155(7) Å, β 102.511(6)◦, V 2284 Å3. Dc
(Z 4)1.304 g cm−3. µMo 0.92 cm−1;specimen:0.40 × 0.05 × 0.05 mm3.
2θmax 58◦; Nt 12547, N 3051 (Rint 0.056), Nt 2200; R 0.060, Rw 0.059.
|&ρmax| 0.4(1) e Å−3. CCDC no. 220142.
Acknowledgments
We thank the Australian Research Council for financial
support, and Professor Steve Withers for suggestions and
discussion. K.A.S. thanks the University ofWesternAustralia
for the assistance of a Hackett Postgraduate Scholarship.
References
[1] S. J. Williams, S. G. Withers, Carbohydr. Res. 2000, 327, 27.
doi:10.1016/S0008-6215(00)00041-0
[2] S. G. Withers, I. P. Street, P. Bird, D. H. Dolphin, J. Am. Chem.
Soc. 1987, 109, 7530.
[3] S. G. Withers, K. Rupitz, I. P. Street, J. Biol. Chem. 1988, 263,
7929.
[20] E. W. Baxter, A. B. Reitz, J. Org. Chem. 1994, 59, 3175.
[21] F. Cramer, G. Mackensen, K. Sensse, Chem. Ber. 1969, 102, 494.
[22] B. Helferich, E. Himmen, Chem. Ber. 1928, 61, 1825.