3,4-Anhydro-1,2-O-isopropylidene-β-D-tagatopyranose and 4,5-anhydro- 1,2-O-isopropylidene-β-D-fructopyranose

Article abstract of DOI:10.1016/S0008-6215(99)00272-4

3,4-Anhydro-1,2-O-isopropylidene-β-D-tagatopyranose (8) and 4,5- anhydro-1,2-O-isopropylidene-β-D-fructopyranose (10) have been prepared by treatment of 3,5-di-O-acetyl-1,2-O-isopropylidene-4-O-toluene-p-sulfonyl-β- D-fructopyranose with methanolic sodium methoxide. The structures of 8 and 10 were assigned by ¹H and ¹³C NMR spectroscopy and that of 10 by X-ray crystallography; both exist in half-chair conformations. Compounds 8 and 10 interconvert in aqueous sodium hydroxide, giving a ratio of 1:2 at equilibrium. The monoacetates of 8 and 10 (5-O-acetyl-3,4-anhydro-1,2-O- isopropylidene-β-D-tagatopyranose and 3-O-acetyl-4,5-anhydro-1,2-O- isopropylidene-β-D-fructopyranose) undergo stereospecific epoxide ring opening in 80% acetic acid to give mainly the axial monoacetates 5-O-acetyl- 1,2-O-isopropylidene-β-D-fructopyranose and 4-O-acetyl-1,2-O-isopropylidene- β-D-tagatopyranose, respectively. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0008-6215(99)00272-4

www.elsevier.nl/locate/carres
Carbohydrate Research 323 (2000) 36–43
3,4-Anhydro-1,2-O-isopropylidene-b-
D
-tagatopyranose
and 4,5-anhydro-1,2-O-isopropylidene-b-
D-
fructopyranose
Masoud Ataie a, J. Grant Buchanan a,b,*, Alan R. Edgar a, Richard G. Kinsman b,
Maria Lyssikatou b, Mary F. Mahon b, Pauline M. Welsh b
a Department of Chemistry, Heriot-Watt Uni6ersity, Riccarton, Edinburgh EH14 4AS, UK
b Department of Chemistry, Uni6ersity of Bath, Cla6erton Down, Bath BA2 7AY, UK
Received 4 July 1999; accepted 30 September 1999
Dedicated to Professor Aleksander Zamojski on the occasion of his 70th birthday.
Abstract
3,4-Anhydro-1,2-O-isopropylidene-b-
D
-tagatopyranose (8) and 4,5-anhydro-1,2-O-isopropylidene-b-
D
-fructopyran-
-fruc-
ose (10) have been prepared by treatment of 3,5-di-O-acetyl-1,2-O-isopropylidene-4-O-toluene-p-sulfonyl-b-
D
topyranose with methanolic sodium methoxide. The structures of 8 and 10 were assigned by H and 13C NMR
spectroscopy and that of 10 by X-ray crystallography; both exist in half-chair conformations. Compounds 8 and 10
interconvert in aqueous sodium hydroxide, giving a ratio of 1:2 at equilibrium. The monoacetates of 8 and 10
1
(5-O-acetyl-3,4-anhydro-1,2-O-isopropylidene-b-
dene-b- -fructopyranose) undergo stereospecific epoxide ring opening in 80% acetic acid to give mainly the axial
-taga-
D-tagatopyranose and 3-O-acetyl-4,5-anhydro-1,2-O-isopropyli-
D
monoacetates 5-O-acetyl-1,2-O-isopropylidene-b-
D
-fructopyranose and 4-O-acetyl-1,2-O-isopropylidene-b-
D
topyranose, respectively. © 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Anhydroketose; Epoxide migration; Stereospecific ring opening
preparation and equilibration of the ketose-
derived epoxides 3,4-anhydro-1,2-O-isopro-
1. Introduction
pylidene-b-
D
-tagatopyranose (8) and 4,5-
-fructopyr-
The base-catalysed rearrangement of oxi-
rans (epoxides) bearing a vicinal hydroxyl
group [1,2], known as epoxide migration [3],
has been studied extensively by ourselves [4,5]
and by others [3,6–8]. The rearrangement has
been used for the synthesis of sugars and
polyols by chain extension [9,10]. In continua-
tion of our work on methyl anhydropentopyr-
anosides [11,12], we wished to examine the
anhydro-1,2-O-isopropylidene-b-
D
anose (10). The preparation of the anhydrota-
gatose 8 has previously been described by
Ohle and Schultz [13] and much later by Rao
et al. [14]. Treatment of the sulfonate 2 with
an excess of hot methanolic sodium methoxide
afforded a crystalline epoxide, which was as-
signed the anhydrotagatose structure 8. Such
vigorous conditions are known to cause epox-
ide migration in epoxides containing a vicinal
trans-hydroxyl group. The epoxide isolated by
Ohle might therefore be the anhydrofructose
* Corresponding author. Fax: +44-1225-826-231.
E-mail address: chsjgb@bath.ac.uk (J.G. Buchanan)
0008-6215/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(99)00272-4

M. Ataie et al. / Carbohydrate Research 323 (2000) 36–43
37
Scheme 1.
10. Rao et al. [14] treated the crystalline epox-
ide with potassium hydrogen fluoride in
refluxing methoxyethanol. The products were
as expected from ring opening of the anhy-
drofructose 10 and the authors suggested that
epoxide migration in 8 took place under the
influence of KHF2. This interpretation is con-
trary to our experience and gave us a further
incentive to re-examine Ohle’s ‘3,4-anhydro-
itself slowly converted into an epoxide of
higher Rf value. Chromatography on silica gel
yielded first a crystalline epoxide, mp 81–
82 °C, in 29% yield, clearly identical to the
epoxide isolated by the earlier workers [13,14],
followed by an isomeric syrupy epoxide (44%).
The latter epoxide, being formed first, was
expected to be 3,4-anhydro-1,2-O-isopropyl-
idene-b- -tagatopyranose (8) and this was evi-
D
1
dent from its H and 13C NMR spectra; the
signals for H-3 and H-4 were shielded relative
to H-5 and H-6 and those for C-3 and C-4
were shielded relative to C-5 and C-6. Corre-
spondingly, the crystalline epoxide that was
formed as the secondary product gave
shielded signals derived from a 4,5-anhydro
ring.
1,2 - O - isopropylidene - b -
D
- tagatopyranose’
and to explore its chemistry in greater detail.
2. Results and discussion
Reaction of 3-O-acetyl-1,2-O-isopropyli-
dene-b- -fructopyranose (1) [13,15–17] with
D
Conclusive proof that the crystalline epox-
ide was 4,5-anhydro-1,2-O-isopropylidene-b-
1.1 equivalents of toluene-p-sulfonyl chloride
in pyridine gave mainly the expected equato-
rial 4-sulfonate (3) as a syrup (Scheme 1). In
the 1H NMR spectrum, the H-4 signal (l 4.93,
J4,3 10.1, J4,5 3.3 Hz) was deshielded relative to
its position in the spectrum of 1. Acetylation
of the crude syrup with acetic anhydride in
pyridine afforded the crystalline 3,5-diacetate
4 (44% overall), whose structure was con-
D
-fructopyranose (10) was obtained by X-ray
crystallography, shown in Fig. 1. The pyran-
ose ring is in a half-chair conformation, H0,
2
in which the anomeric O-2 is axial with C-1
equatorial, known to be a very favourable
arrangement [18]. The puckering parameters
according to Cremer and Pople [19] are Q=
1
firmed by H and 13C NMR. Compound 4
,
0.475 A, [=132.4°, and b=212.66° (the
shape of the isopropylidene ring is described
was treated with methanolic sodium methox-
ide (2 equivalents) at room temperature and
the reaction monitored by thin-layer chro-
matography (TLC). Formation of the dihy-
droxysulfonate was followed by conversion
into an epoxide of low Rf value, which was
,
by the puckering parameters Q2=0.269 A and
b2=212.49°). The hydroxyl proton H-3A (at-
tached to O-3) was located in the penultimate
difference Fourier map and seen to be in-
volved in an intramolecular hydrogen bond to

38
M. Ataie et al. / Carbohydrate Research 323 (2000) 36–43
Fig. 1. ORTEX plot of 10. Thermal ellipsoids are represented at the 30% probability level.
,
series the analogous 3,4-epoxide is disfavoured
and this may be due to a greater freedom for
O-2 and C-1 to become fully axial and equato-
rial, respectively, in 10.
O-2 [O-2···H-3A 1.92(11) A; O-2ꢀH-3AꢀO-3
125(9)°]. The HꢀH dihedral angles are shown
in Table 1, together with the coupling con-
1
stants from the H NMR spectrum. The very
It is well known that the ring opening of an
epoxide can be directed by a neighbouring
trans-acetoxy group [4,5,11,23,24]. It was of
interest to study the reaction of the acetates of
8 and 10 (9 and 11) with 80% acetic acid. The
reaction product from 9 was 5-O-acetyl-1,2-
low coupling between the epoxide protons and
those on C-3 and C-6 is to be expected from
earlier work on methyl 3,4-anhydro-b-ara-
binosides [12,20] and cis-2-tert-butyl-4,5-
epoxytetrahydropyran [21].
The interconversion of 8 and 10 in aqueous
sodium hydroxide by epoxide migration was
studied polarimetrically and by GLC (Scheme
2). At equilibrium the anhydrofructose 10 was
favoured by 2:1. The equilibration was also
conducted on a preparative scale, starting
from 10. Both epoxides were isolated from the
mixture, after chromatography, in a 2:1 ratio.
There is an interesting comparison with the
O-isopropylidene-b- -fructopyranose (5), the
D
axial monoacetate, as would be expected if the
acetoxonium ion 19 is an intermediate [11,25–
27] (Scheme 3). The location of the acetyl
group on O-5 was apparent from the
1
deshielded signal for H-5 in the H NMR
spectrum. O-Deacetylation afforded the
known 1,2-O-isopropylidene-b- -fructopyran-
D
ose (6) [28]. Similarly, 11 gave the axial
monoacetate 12 by participation of the 3-O-
acetyl group; 12 was converted by deacetyla-
analogous methyl 2,3-anhydro-b-
oside (15) and 3,4-anhydro-b- -arabinopyran-
oside (16) [11]1, together with methyl 2,3-
anhydro-6-deoxy-a- -gulopyranoside (17) and
D-lyxopyran-
D
Table 1
L
Selected CꢀH torsion angles (°) and H–H coupling constants
(Hz) for 10
methyl 3,4-anhydro-6-deoxy-a- -galactopyran-
L
oside (18) [22]1. In both of these examples
there is a preference for the 2,3-epoxide at
equilibrium (Scheme 2). In the anhydroketose
Bonds
Angles
3J
H-3ꢀC-3ꢀC-4ꢀH-4
H-4ꢀC-4ꢀC-5ꢀH-5
H-5ꢀC-5ꢀC-6ꢀH-6A
H-5ꢀC-5ꢀC-6ꢀH-6B
105.36
0.14
71.12
46.35
:0
3.9
:0
:0
1 The original papers [11,22] describe the enantiomers of
15–18 but we show these forms for direct comparison with 8
and 10.

M. Ataie et al. / Carbohydrate Research 323 (2000) 36–43
39
Scheme 2.
Scheme 3.
recrystallisation in 39% yield, has now been
tion and removal of the isopropylidene group
into -tagatose, identified by comparison with
an authentic sample.
Ohle and Schultz [13] claimed to have con-
verted the acetate of their crystalline anhy-
droketose, by treatment with acetic anhydride,
acetic acid and a little pyridine, into 3,4,5-tri-
shown to be 3,4,5-tri-O-acetyl-1,2-O-isopro-
D
pylidene-b- -tagatopyranose (14). Although it
D
has a similar melting point and specific rotation
to the fructose derivative 7 [15], it was con-
verted into tagatose by alkaline methanolysis,
1
followed by acidic hydrolysis. The H NMR
spectrum of 14 was as expected of the b- -
D
O-acetyl-1,2-O-isopropylidene-b- -fructopyr-
D
tagato configuration. Using H,H-COSY all of
the signals were assigned; in particular there is
long-range coupling (4J) between H-4 and H-
6eq, which are in a ‘W’ relationship. The spec-
trum differed from that of authentic 7, whose
spectrum agreed with that already published
anose (7) [15]. Since we have now shown that
the crystalline anhydroketose is the 4,5-anhy-
drofructose 10 and not the 3,4-anhydro-
tagatose 8, we have reinvestigated the reac-
tion. The crystalline product, isolated after

40
M. Ataie et al. / Carbohydrate Research 323 (2000) 36–43
[17]. The formation of a product of tagato
configuration is to be expected from some of
our earlier work [29] on the treatment of
acetoxy epoxides with acetic acid and acetic
anhydride.
from O-3. H-3A is involved in intramolecular
hydrogen-bonding to O-4. [H-3A…O-2
1.92(11) A, O-3ꢀH-3AꢀO-2 125(9)°.]
The solution of the structure (SHELX86) [31]
and refinement (SHELX93) [32] converged to a
conventional [i.e., based on 487 with Fo \
4|(Fo)], R1 =0.0911 and wR2 =0.2041. Good-
,
ness-of-fit=0.934.
The
maximum
and
3. Experimental
minimum residual densities were 0.352 and
−0.525 e A−3, respectively. The asymmetric
,
General methods.—1H NMR spectra were
recorded at 400 MHz (Jeol EX400) and at 100
MHz (Jeol JNM); 13C NMR spectra at 100
MHz (Jeol EX400). Optical rotations were
measured with a Thorn–Bendix TBL-NPL
polarimeter type 143D and a cell of 1 cm path
length. Adsorption chromatography was con-
ducted on silica gel (E. Merck) Kieselgel 60–
70 230 mesh ASTM and TLC with Kieselgel
60F 254 as adsorbent. On TLC, sugar deriva-
tives were visualised by ethanolic H2SO4, fol-
lowed by heating; toluenesulfonates by
alcoholic diphenylamine, followed by expo-
sure to UV light [30]; epoxides by sodium
iodide and methyl red, followed by heating
[23]. GLC was carried out using a Perkin–
Elmer F11 gas chromatograph with a 3% SE
30 column (1 m), employing an inlet pressure
of 20 psi and a column temperature of 100–
110 °C.
unit (shown in Fig. 1), along with the labelling
scheme used, was produced using ORTEX [33].
Final fractional atomic coordinates and
isotropic thermal parameters, bond distances
and angles and tables of anisotropic tempera-
ture factors are available as supplementary
data (see Section 4).
3,5-Di-O-Acetyl-1,2-O-isopropylidene-4-O-
toluene-p-sulfonyl-i- -fructopyranose (4).—
D
The acetate (1) [15] (8.61 g) in pyridine (25
mL) was stirred at room temperature (rt) and
toluene-p-sulfonyl chloride (6.9 g, 1.1 equiv)
was added. The mixture was left overnight at
rt and the product isolated using CH2Cl2 to
yield a crude syrup (11.34 g). TLC showed a
major spot corresponding to a monosulfonate
(3) (Rf 0.64), together with a small amount of
1
disulfonate (Rf 0.75). H NMR (270 MHz,
CDCl3): l 7.80 (d, 2 H, J 7.7 Hz, Ar), 7.35 (d,
2 H, Ar) 5.40 (d, 1 H, J3,4 10.1 Hz, H-3), 4.93
(dd, 1 H, J4,5 3.3 Hz, H-4), 4.16 (m, 1 H, H-5),
3.33 (1 H, br s, OH, D2O-exchangeable), 2.43
(s, 3 H, Ts), 1.80 (s, 3 H, Ac), 1.45, 1.37 (2 s,
ea 3 H, Me2C).
X-ray crystallography.—A crystal of 10
(approximate dimensions 0.20×0.22×0.30
mm) was used for data collection. Crystallo-
graphic measurements were made at 293(2) K
on a CAD4 automatic four-circle diffractome-
ter in the range 2.06BqB21.91°. Data (728
reflections) were corrected for Lorentz and
polarization but not for absorption. Despite
several recrystallisation attempts, it was not
possible to grow larger crystals. This was un-
fortunate given the weak diffracting ability of
the sample selected.
The above syrup was acetylated with an
excess of Ac2O in pyridine. Isolation using
CH2Cl2 and crystallisation from EtOH gave
the diacetate 4 (6.67 g, 37% from 1), mp
113–114 °C. Recrystallised from EtOH 4 had
mp 114–116 °C, [h]D −121° (c 1.00, CHCl3);
1H NMR (400 MHz, CDCl3) l 7.75 (d, 2 H, J
8.3 Hz, Ar), 7.34 (d, 2 H, Ar), 5.34 (d, 1 H,
J3,4 10.3 Hz, H-3), 5.23 (m, 1 H, H-5), 5.01
(dd, 1 H, J4,5 3.9 Hz, H-4), 4.00 (d, 1 H, J6a,6b
13.2 Hz, H-6a), 3.94 and 3.90 (ABq, 2H, J1a,1b
9.3 Hz, H-1a, H-1b), 3.77 (dd, 1 H, J6b,5 2.0
Hz, H-6b) 2.43 (s, 3 H, Ts), 2.04 and 1.93 (2 s,
ea 3 H, 2×Ac), 1.45 and 1.39 (2 s, ea 3 H,
Crystal data: C9H14O5, M=202.20, mono-
clinic, a=9.002(3), b=5.553(1), c=10.590(4)
3
,
,
A, i=111.28(3)°, U=493.3(3) A , space
group P21, Z=2, Dc=1.361 g cm−3, v (Mo
Ka)=0.111 mm−1, F(000)=216.
In the final least-squares cycles all atoms
were allowed to vibrate anisotropically. Hy-
drogen atoms were included at calculated po-
sitions where relevant, except for H-3A, which
13
Me2C). C NMR (100 MHz, CDCl3) l 169.9
and 169.4 (2×MeCO), 144.8 and 133.5 (Ar
,
was located and refined at a distance of 1.08 A
quaternary), 129.3 and 127.6 (Ar methine),

M. Ataie et al. / Carbohydrate Research 323 (2000) 36–43
41
tion using CHCl3 gave the acetate (11), (75
mg, 62%), mp 80–81 °C, [h]D −87.8° (c 3.4,
112.1 (Me2C),104.2 (C-2), 75.3 (C-4), 71.5 (C-
1), 69.4 (C-5), 66.3 (C-3), 61.8 (C-6), 25.7 and
25.6 (Me2C), 21.3 MeAr). Anal. Calcd for
C20H26O10S: C, 52.39; H,5.72; S, 6.99. Found:
C 52.47; H,5.69; S, 6.87.
1
CHCl3); (lit.: mp 80–81 °C, [h]D −28.6°); H
NMR (100 MHz, CDCl3): l 4.84 (s, 1 H,
H-3), 3.70–4.20 (4 H, m, H-4, H-5, CH2), 3.16
(s, 2 H, CH2), 2.10 (s, 3 H, Ac), 1.46 and 1.38
(2 s, ea 3 H, Me2C). Anal. Calcd. for
C11H16O6: C, 54.09; H, 6.60. Found: C, 54.56;
H, 6.90.
4,5-Anhydro-1,2-O-isopropylidene-i-
topyranose (10) and 3,4-anhydro-1,2-O-iso-
propylidene-i- -tagatopyranose (8).—The
D
-fruc-
D
sulfonate 4 (24 g), in MeOH (150 mL), was
treated with a solution of NaOMe [from Na
(2.46 g)] in MeOH (210 mL) at rt for 28 h.
TLC throughout the period indicated the
rapid conversion into a monotosyl diol, fol-
lowed by the formation of an epoxide of low
Rf value, which was slowly converted into
another epoxide of higher Rf value. Solid CO2
was then added and the mixture evaporated to
dryness. The residue was extracted with hot
benzene and the filtered extract evaporated to
yield a syrup, which was chromatographed on
silica gel. Toluene–EtOAc (9:1) eluted first the
anhydrofructose (10) (3.08 g, 29%), mp 78–
79 °C, [h]D −86° (c 1.0, CHCl3) (lit. [13] mp
Acetylation and hydrolysis of 3,4-anhydro-
1,2 - O - isopropylidene - i -
D
- tagatopyranose
(8).—The anhydro compound 8 (370 mg) was
acetylated using Ac2O in pyridine and the
product isolated using CHCl3. The crude
1
syrupy 5-acetate (9) (410 mg, 92%) showed H
NMR (100 MHz, CDCl3); l 2.10 (s, 3 H, Ac),
1.46 (s, 6 H, Me2C). The syrup (370 mg) was
treated with 80% HOAc at rt for 3 days. After
evaporation of the solution, the crude prod-
uct was chromatographed on silica gel
(15 g). Toluene–EtOAc (1:1) eluted 5-O-
acetyl-1,2-O-isopropylidene-b- -fructopyran-
D
ose (5), which crystallised from EtOAc–light
1
81–82 °C, [h]D −80.7°, CHCl3); H NMR
petroleum (340 mg, 86%), mp 103–104 °C,
1
(400 MHz, CDCl3): l 4.09 (d, 1 H, J6a,6b 13.3
Hz, H-6a), 4.03 and 4.01 (ABq, 2 H, J1a,1b 9.0
Hz, H-1a, H-1b), 4.03 (d, 1 H, H-6b), 3.70 (d,
1 H, J3,OH 11.7 Hz, H-3), 3.28 (d, 1 H, J4,5 3.9
Hz, H-4), 3.20 (d, 1 H, H-5), 2.30 (d, 1 H,
JOH,3 11.7 Hz, OH-3), 1.54 and 1.44 (2 s, ea. 3
H, Me2C); 13C NMR (100 MHz, CDCl3): l
112.5 (Me2C), 101.4 (C-2), 72.6 (C-1), 63.3
(C-3), 58.6 (C-6), 53.9 (C-5), 49.5 (C-4), 26.6
and 26.0 (Me2C); Anal. Calcd for C9H14O5: C,
53.46; H, 6.98. Found: C, 53.48; H,7.01.
[h]D −153.7° (c 1.06, CHCl3); H NMR (100
MHz, CDCl3); l 5.15 (m, 1 H, J4,5:J5,6a
:
J5,6b:2 Hz, H-5), 2.13 (s, 3 H, Ac), 1.49 and
1.45 (2 s, ea. 3 H, Me2C). The acetate 5 (190
mg) was treated with methanolic NaOMe (6
mL) [from Na (20 mg)] for 1 h at rt. After
deionisation using Zeo–Karb 225 (NH4+)
resin the evaporated residue was recrystallised
from EtOAc to give 1,2-O-isopropylidene-b-
D-fructopyranose (6) (130 mg, 81%), mp 120–
122 °C, indistinguishable from an authentic
sample prepared by deacetylation of 1 (lit. [28]
mp 121–122 °C).
Further elution yielded, after evaporation,
the syrupy anhydrotagatose (8) (4.70 g, 44%),
1
Hydrolysis of 3-O-acetyl-4,5-anhydro-1,2-
[h]D −22° (c 1, CHCl3); H NMR (400 MHz,
O-isopropylidene-i- -fructopyranose (11).—
D
CDCl3): l 4.08 (ABq, 2 H, H-1a, H-1b), ꢀ4.02
(bd, 1 H, H-5), 4.00 (d, 1 H, J6a,6b 12.1 Hz, H-
6a), 3.56 (d, 1 H, H-6b), 3.37 (bs, 1 H, H-4),
3.29 (d, 1 H, J3,4 3.9 Hz, H-3), 2.24 (d, 1 H, J
The acetate 11 (500 mg) was treated with 80%
HOAc (10 mL) at rt for 3 days. TLC indicated
that the product was largely a single monoac-
etate. After evaporation, the crude product
was chromatographed on silica gel (10 g),
eluting with toluene–EtOAc (1:1). 4-O-
13
8.6 Hz, OH-5), 1.49 (s, 6 H, Me2C); C NMR
(100 MHz, CDCl3): l 111.0 (Me2C), 100.1 (C-
2), 73.3 (C-1), 62.9 (C-5), 62.1 (C-6), 53.9 (C-3),
51.8 (C-4), 26.2 and 25.5 (Me2C); Anal. Found:
C, 53.60; H, 7.06.
Acetyl-1,2-O-isopropylidene-b- -tagatopyra-
D
nose (12) was isolated as a syrup (290 mg,
1
54%); H NMR (100 MHz, CDCl3) l 5.04
3-O-Acetyl-4,5-anhydro-1,2-O-isopropyli-
(dd, J3,4 3.0, J4,5 2.6 Hz, H-4), 2.06 (s, 3 H,
Ac), 1.44 and 1.36 (2 s, ea. 3 H, Me2C); Anal.
Calcd. for C11H18O7: C, 50.38; H, 6.92.
dene-i- -fructopyranose (11).—The anhy-
D
drofructose (10) (100 mg) was treated for 20 h
with Ac2O (0.5 mL) in pyridine (1 mL). Isola-

42
M. Ataie et al. / Carbohydrate Research 323 (2000) 36–43
Found: C, 50.37; H, 6.73. Deacetylation of the
acetate 12 (270 mg) with methanolic NaOMe
and deionisation and evaporation as for 6
Ac2O (2.0 mL), glacial HOAc (4.0 mL) and
three drops of dry pyridine and the mixture
was heated at 100 °C for 7 h and left for 40 h
at rt. The mixture was evaporated to a syrup
and extracted from water with CHCl3. TLC of
the evaporated CHCl3 solution showed that
the major product corresponded to a triac-
etate, together with more polar material. The
syrup was therefore treated with Ac2O in
pyridine and the product isolated using
CHCl3. Crystallised from EtOH, 3,4,5-tri-O-
above gave syrupy 1,2-O-isopropylidene-b- -
D
tagatopyranose (13) (135 mg, 60%). The acetal
(135 mg) was treated with 0.05 M H2SO4 (2
mL) at 40 °C for 3 h, neutralised with BaCO3
and filtered through Celite. After evaporation
the syrup was crystallised from EtOH to give
D-tagatose, mp 113 °C, indistinguishable from
an authentic sample. GLC analysis of the
pertrimethylsilyl ether of the product showed
a single peak (retention time 7.1 min), identi-
cal to that of the perTMS ether of authentic
tagatose (1 m column of 3% SE 30, at 110 °C,
carrier gas pressure 20 psi).
acetyl-1,2-O-isopropylidene-b- -tagatopyran-
D
ose (14) (277 mg, 39%), had mp 95–96 °C,
1
[h]D −124° (c 0.89, CHCl3); H NMR (400
MHz, CDCl3) l 5.28 (td, 1 H, J4,5 3.8 Hz,
H-4), 5.16 (d, 1 H, J3,4 3.35 Hz, H-3), 4.86 (m,
1 H, H-5), 4.29 (dd, 1 H, J6ax,6eq 13.1, J6ax,5 1.8
Hz, H-6ax), 3.97 (d, 1 H, J1a,1b:9 Hz, H-1a),
Equilibration studies of 3,4-anhydro-1,2-O-
isopropylidene-i- -tagatopyranose (8) and 4,5-
D
3.90 (d, 1 H, H-1b), 3.71 (dt, 1 H, J6eq,5
:
anhydro-1,2-O-isopropylidene-i-D-fructopyran-
J6eq,4:1.5 Hz, H-6eq), 2.14, 2.11, and 2.05 (3
s, ea. 3 H, 3×Ac), 1.50 and 1.40 (2 s, ea. 3 H,
Me2C); Anal. Calcd for C15H22O9: C, 52.02;
H, 6.40. Found: C, 51.95; H, 6.38.
ose (10)
Polarimetric studies. Each anhydrocom-
pound was dissolved in 0.05 M NaOH (c 1.0)
at 25 °C and the changes in optical rotation
with time were measured. The 3,4-anhydro-
tagatose 8 {[h]D −20° (H2O)} showed [h]D
−28° after 5 min and −51° (constant) after
12 h. The 4,5-anhydrofructose 10 {[h]D −66°
(H2O)} showed −59° after 5 min and −51°
(constant). after 12 h. The ratio at equilibrium
is therefore 2:1 in favour of 10.
3,4,5-Tri-O-acetyl-1,2-O-isopropylidene-i-
D
-fructopyranose (7) [15].—1H NMR (400
MHz, CDCl3) l 5.37 (d, 1 H, J3,4 10.5 Hz,
H-3), 5.34 (m, 1 H, H-5), 5.29 (dd, 1 H, J4,3
3.1 Hz, H-4), 4.10 (bd, J6a,6b 13.3 Hz, H-6a),
3.98 (d, 1 H, J1a,1b 9.4 Hz, H-1a), 3.93 (d, 1 H,
H-1b), 3.76 (dd, 1 H, J6b,5 1.2 Hz, H-6b), 2.14,
2.09, and 1.99 (3 s, ea. 3 H, 3×Ac), 1.47 and
1.41 (2 s, ea. 3 H, Me2C).
GLC studies. The mixture of epoxides was
isolated by deionisation and evaporation and
subjected to GLC using a 1 m column of SE
30 with inlet pressure of 20 psi and a column
temperature of 100°. The retention times were
1.5 and 2.5 min for 8 and 10, respectively, and
the ratio of 8:10 was 1:2.
Equilibration and isolation of epoxides. The
anhydrofructose 10 (300 mg) was treated with
0.05 M NaOH (2 mL) and the reaction fol-
lowed by TLC. After 12 h there was no fur-
ther change, the mixture was neutralised with
diluted HCl, evaporated and the syrupy
residue subjected to chromatography on silica
gel. Toluene–EtOAc (9:1) eluted the anhy-
drofructose 10 (180 mg); further elution with
toluene–EtOAc (1:1) eluted the syrupy anhy-
drotagatose 8 (100 mg).
O-Deacetylation of 3,4,5-tri-O-acetyl-1,2-
O-isopropylidene-i- -tagatopyranose, followed
D
by acid hydrolysis.—The triacetate 14 (500
mg) was treated with methanolic NaOMe [15
mL, Na (50 mg)] until deacetylation was com-
plete (TLC). The solution was deionised using
Zeo–Karb 225 (NH4+) resin and evaporated
to give 13 as a syrup. The acetal 13 (200 mg)
was treated with 0.05 M H2SO4 (5 mL) at
40 °C for 3 h, neutralised (BaCO3), filtered
(Celite) and evaporated to give -tagatose (50
D
mg), mp 113 °C, indistinguishable from an
authentic sample.
4. Supplementary material
3,4,5-Tri-O-acetyl-1,2-O-isopropylidene-i-
D
-tagatopyranose (14).—The acetoxyepoxide
Full crystallographic details, excluding
structure features, have been deposited with
11 (500 mg) was dissolved in a mixture of

M. Ataie et al. / Carbohydrate Research 323 (2000) 36–43
43
[14] G.V. Rao, L. Que Jr., L.D. Hall, T.P. Fondy, Carbohydr.
Res., 40 (1975) 311–321.
[15] E. Fischer, H. Noth, Bericht, (1918) 321–352.
[16] P.A.J. Gorin, L. Hough, J.K.N. Jones, J. Chem. Soc.,
(1955) 2699–2705.
[17] T. Maeda, K. Tori, S. Satoh, K. Tokuyama, Bull. Chem.
Soc. Jpn., 42 (1969) 2635–2647.
[18] E.L. Eliel, N.L. Allinger, S.J. Angyal, G.A. Morrison,
Conformational Analysis, Interscience, New York, 1965,
pp. 408–409.
the Cambridge Crystallographic Data Centre.
These data may be obtained, on request, from
the CCDC, 12 Union Road, Cambridge, CB2
1EZ, UK (Tel.: +44-1223-336408; fax: +44-
1223-336033; e-mail: deposit@ccdc.cam.ac.uk)
quoting CCDC 135058.
[19] D. Cremer, J.A. Pople, J. Am. Chem. Soc., 97 (1975)
1354–1358.
[20] M. Chmielewski, A. Zamojski, Rocz. Chem., 46 (1972)
2039–2050.
[21] G. Catelani, L. Monti, P. Tognetti, Carbohydr. Res., 97
(1981) 189–197.
[22] J. Jary, K. Capek, Collect. Czech. Chem. Commun., 31
(1966) 315–320.
References
[1] E.P. Kohler, N.K. Richtmyer, W.F. Hester, J. Am.
Chem. Soc., 53 (1931) 205–221.
[2] W.H.G. Lake, S. Peat, J. Chem. Soc., (1938) 1069–1074.
[3] S.J. Angyal, P.T. Gilham, J. Chem. Soc., (1957) 3691–
3699.
[4] (a) J.G. Buchanan, Chem. Ind. (London), (1954) 1484–
1485. (b) J.G. Buchanan, J. Chem. Soc., (1958) 995–
1000, 2511–2516, and subsequent papers.
[5] J.G. Buchanan, H.Z. Sable, in B.S. Thyagajaran (Ed.),
Selecti6e Organic Transformations, Vol. 2, Wiley–Inter-
science, New York, 1972, pp. 1–95.
[23] J.G. Buchanan, J.C.P. Schwarz, J. Chem. Soc., (1962)
4770–4777.
[24] W.R. Roush, R.J. Brown, M. DiMare, J. Org. Chem., 48
(1983) 5083–5093.
[25] J.G. Buchanan, R. Fletcher, J. Chem. Soc., (1965) 6316–
6323.
[6] (a) F.H. Newth, J. Chem. Soc., (1956) 441–447. (b) F.H.
Newth, Q. Re6. Chem. Soc., 13 (1959) 30–47.
[7] G.B. Payne, J. Org. Chem., 27 (1962) 3819–3822.
[8] N.R. Williams, Ad6. Carbohydr. Chem. Biochem., 25
(1970) 109–179.
[9] T. Katsuki, A.W.M. Lee, P. Ma, V.S. Martin, S. Masa-
mune, K.B. Sharpless, D. Tuddenham, F.J. Walker, J.
Org. Chem., 47 (1982) 1373–1378.
[10] J.S. Brimacombe, R. Hanna, A.K.M.S. Kabir, J. Chem.
Soc., Perkin Trans. 1, (1987) 2421–2426.
[11] J.G. Buchanan, R. Fletcher, J. Chem. Soc. C, (1966)
1926–1931.
[26] J.F. King, A.D. Allbutt, Can. J. Chem., 48 (1970) 1754–
1769.
[27] P. Deslongchamps, Stereoelectronic Effects in Organic
Chemistry, Pergamon, 1983, pp. 82–85.
[28] D.J. Bell, J. Chem. Soc., (1947) 1461–1464.
[29] J.G. Buchanan, J. Conn, A.R. Edgar, R. Fletcher, J.
Chem. Soc., C, (1971) 1515–1521.
[30] M. Jackson, L.D. Hayward, J. Chromatogr., 5 (1961)
166–169.
[31] G.M. Sheldrick, Acta Crystallogr., Sect. A, 46 (1990)
467–73.
[32] G.M. Sheldrick, SHELXL-93, a computer program for
crystal structure refinement, University of Go¨ttingen,
1993.
[12] J.G. Buchanan, R. Fletcher, K. Parry, W.A. Thomas, J.
Chem. Soc. B, (1969) 377–385.
[13] H. Ohle, C.A. Schultz, Bericht, 71 (1938) 2302–2315.
[33] P. McArdle, J. Appl. Crystallogr., 27 (1994) 438.
.