Convenient access to 4-O-glycosylated 1,5-anhydro-d-fructoses via disaccharide-derived 2-hydroxyglycal esters

Article abstract of DOI:10.1016/j.tetasy.2011.03.005

Efficient six-step protocols are described for the conversion of common disaccharides, such as maltose, cellobiose, or lactose, into the corresponding 4-O-glycosyl-1,5-anhydro-d-fructoses. Overall yields of 40-45% are favorably compared to the alternative eleven-step procedure from their monosaccharide components (～15%).

Full text of DOI:10.1016/j.tetasy.2011.03.005

Tetrahedron: Asymmetry 22 (2011) 740–744
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
Convenient access to 4-O-glycosylated 1,5-anhydro-D-fructoses via
disaccharide-derived 2-hydroxyglycal esters q
⇑
Toshio Nakagawa, Frieder W. Lichtenthaler
Clemens-Schöpf-Institut für Organische Chemie und Biochemie, Technische Universität Darmstadt, Petersenstraße 22, D-64287 Darmstadt, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Efficient six-step protocols are described for the conversion of common disaccharides, such as maltose,
cellobiose, or lactose, into the corresponding 4-O-glycosyl-1,5-anhydro-D-fructoses. Overall yields of
40–45% are favorably compared to the alternative eleven-step procedure from their monosaccharide
Received 8 March 2011
Accepted 18 March 2011
Available online 19 April 2011
components (ꢀ15%).
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
chemical7 approaches to 1,5-anhydro-
D-fructose, usually charac-
terized as the monohydrate, have been developed, among them
the direct low-temperature deacetylation of the acetylated
hydroxyglucal ester 2 (R = Ac).3b,7b
Thirty years ago, and seven years before being encountered as a
naturally occurring monosaccharide,2 1,5-anhydro-
D
-fructose 1
was synthesized from
D
-glucose in a six-step sequence,3 the first
Either of these simple approaches to 1,5-anhydro-D-fructose,
three steps comprising of the generation of its 2-hydroxyglucales-
ter 2,4 the others being the selective hydroxylaminolysis5 of the
enol ester group 2?3, Zemplén de-O-acylation, and deoximation
(3?4?1) (Scheme 1). Since then, various other enzymatic6 and
should in principle be applicable to any of the common disaccha-
rides, a conjecture which has already been verified8 by conversion
of maltose, cellobiose, and lactose, via their 2-hydroxyglycal esters
5a–5c into the respective perbenzoylated 4-O-glycosyl-1,5-anhy-
drofructoses 6a–6c, with yields in the 70% range for the four steps
involved8 (Scheme 2). The lactose-derived 6c was even deoximated
to the respective, highly crystalline ulose monohydrate 7.8
OH
D-Glucose
O
OH
OBz
HO
Maltose
O
O
RO
BzO
Cellobiose
Lactose
3 steps
1
NaOMe / MeOH
0oC
OBz
MeCHO / H+
~70%
5a-c
4
(R = Ac)
NH2OH
OR
OR
O
OBz
O
a : R = αBz4Glc
b : R = βBz4Glc
c : R = βBz4Gal
O
NH2OH
OR
OR
RO
BzO
RO
RO
N
OH
OR
N
OH
2
6a-c
3a R = Bz
3b R = Ac
NaOMe/
MeOH
OBz
O
BzO
BzO
4 R = H
OH
BzO
O
BzO
O
Scheme 1. Chemical approaches to 1,5-anhydro-
D
-fructose.
BzO
q
OH
Part 45 of the series, Sugar-Derived Building Blocks’; for Part 44, see Ref. 1.
Corresponding author. Tel.: +49 6151 162376.
E-mail address: lichtenthaler@chemie.tu-darmstadt.de (F.W. Lichtenthaler).
7
⇑
Scheme 2. Disaccharide-derived 1,5-anhydrofructose derivatives.8
0957-4166/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2011.03.005

T. Nakagawa, F. W. Lichtenthaler / Tetrahedron: Asymmetry 22 (2011) 740–744
741
Apparently unaware of this expedient, large scale adaptable ac-
cess to 4-O-glycosylated 1,5-anhydrofructoses, recently, Agoston
et al. patented7b and published9 another approach comprising of
the glycosylation of a bicyclic 1,5-anhydrofructose derivative with
glycosyl trichloracetimidates, an inconvenient route due to the 11
steps required, three for the donor subtrate;10 five for the accep-
tor;7c and another three for the targets9 with overall yields of
around 15%.
anhydrofructoses 3a and 3b (Table 1), that is, show the same
significant downfield shift caused by the nearly coplanar N–OH
group as seen.1,14
The release of the free glycosyl-(1?4)-1,5-anhydrofructoses 9f–
9h from their oximes proceeded in the expected manner: de-O-
acylation under Zemplén conditions (NaOMe/MeOH, rt, 12–24 h)
smoothly afforded the respective oximes 8f–8h, each isolated
and characterized as E/Z-mixtures, since the basic conditions obvi-
ously induced E?Z equilibrations. The E-isomers thereby consti-
tuted the main components in up to 5:1 preference, yet none
crystallized as was the case with the monosaccharidic 1,5-anhy-
Due to potential medical applications,11 interest in 1,5-anhydro-
D
-fructose has substantially increased recently, an appeal that ex-
tends to its glycosylated analogues, in as much as a (1?3)-1,5-
a
anhydrofructose prepared by enzymatic transglucosylation was
deemed suitable for conjugation with proteins12 and as antioxi-
dant food additives.13 Since their generation via 2-hydroxyglucal
esters from the common disaccharides was considered to be signif-
icantly more simple than glycosylations of 1,5-anhydrofructose
derivatives, we have resumed our earlier work8 and herein describe
the straightforward deprotection of the 4-O-glycosyl-1,5-anhy-
drofructose oximes 6a–6e to the underlying free disaccharides.
dro-D
-fructose E-oximes 4.3 Although separable by chromatogra-
phy due to the sufficiently different Rf values, the E/Z-oxime
mixtures 8f–8h were directly subjected to transoximation by
exposure, in an acetonitrile solution, to excess acetaldehyde in
the presence of 1 M HCl (6 h, rt). The resulting glycosyl-(1?4)-
1,5-anhydrofructoses 9f–9h (Scheme 3) were secured in the form
of glasses or fluffy solids by freeze-drying their Sephadex-purified
aqueous solutions, with yields being in the 75% range only, due to
the somewhat elaborate purification procedure. Analytical values
were uniformly within the acceptable limits for the monohydrate
(C12H20O10ꢁH2O) or 2,2-dihydroxy forms, as indicated in the formu-
lae. Unequivocal structural proof followed from their 1H and 13C
NMR spectra of 9f–9h in D2O (pertinent data shown in Table 2)
which conclusively revealed the signal and coupling patterns ex-
pected for their 4-O-glycosyl and 1,5-anhydrofructose portions.
Moreover, our NMR data for the glycosyl-b(1?4)-1,5-anhydro-
fructose 9g and 9h corresponded well with those obtained for 9g
and 9f prepared via glycosylation of a 4-OH-free 1,5-anhydrofruc-
tose derivative,9 albeit there are minor shift differences. The micro-
analytical data, however, are distinctly different. While our
glycosyl-1,5-anhydrofructoses 9f–9h invariably give satisfactory
C, H values for the monohydrate or 2,2-dihydroxy forms, the
respective Agoston et al.9 products represent dihydrates, as evi-
denced by essentially perfect analytical values for C12H20O10ꢁ2H2O
(or C12H22O11ꢁH2O in their notation), thereby fostering reservations
as to the integrity of these products. Similar reservations also apply
to two 4-O-glycosyl-1,5-anhydro-tagatoses prepared in the same
2. Results and discussion
In view of the most expedient access to 1,5-anhydro-
D-fructose
via de-O-acetylation of 2-acetoxy-
D
-glucal triacetate 2?1,3b,7b it
appeared obvious to generate glycosylated analogues of 1 from
disaccharide-derived 2-hydroxyglucal esters. When applied to
the maltal- and cellobial-derived 5d and 5e, the reaction produced
several products in addition to the desired 4-O-glucosyl-1,5-anhy-
drofructoses 9f and 9g; among them, substantial amounts of D-glu-
cose, obviously originating from base induced b-elimination, a
reaction readily occurring with 2-hydroxyglycal esters under mild
basic conditions.3,5 As chromatographic purification proved quite
laborious, this access was not elaborated.
In the case of disaccharides carrying the glycosyl residue at
positions other than O-4, however, such as the (1?6)-intersaccha-
ridic isomaltose, gentiobiose, or melibiose, the low-temperature
deacytelation procedure may well be used to efficiently generate
1,5-anhydrofructoses with
a-D-glucosyl-, b-D-glucosyl or b-D-
way.9 They analyze for monohydrates, whereas 1,5-anhydro-
D-tag-
galactosyl residues at O-6.
atose exhibits a pronounced tendency to adapt the 2-carbonyl form
in aqueous solution and in the solid state.6d,14 A clarification of
these inconsistencies appears to be de rigueur.
Application of the three-step sequence hydroxylaminolysis, de-
O-acylation and deoximation to any of the disaccharide-derived 2-
hydroxyglucal esters 5a–5e proceeded in a preparatively straight-
forward manner. The established hydroxylaminolysis protocol,5,8
3,5-4 molar hydroxylamine hydrochloride in pyridine at ambient
temperature (24 h) or at 70 °C (4 h), smoothly afforded the glyco-
syl-(1?4)-1,5-anhydrofructose E-oximes 6a–6e isolable in excel-
lent yields of 85–90%.
That these products invariably have their N–OH group oriented
toward the less congested, proanomeric center, that is, are E-
oximes as indicated in the formulae, follows, among other indica-
tions, from the chemical shift for their equatorial H-1, which
appears within the same narrow range as in the acylated 1,5-
3. Conclusion
Efficient, large scale-adaptable protocols are presented for the
conversion of common bulk disaccharides such as maltose, cellobi-
ose, or lactose via their 2-hydroxyglycal esters into the respective
glycosyl-(1?4)-1,5-anhydrofructoses. The yields attainable are in
the 40–45% range for the six steps involved, which compares favor-
ably with the 15% yield over the eleven step procedure,9 when
acquiring these products from glucose and fructose, respectively.
Table 1
Pertinent 1H NMR data (500 MHz) for acylated 1,5-anhydro-D-fructose E-oximes and its 4-O-glycosyl analogs
0
0
Compound
Ra
R0
Solvent
H-1e
H-1a
H-3
H-4
NOH
H-10
J1,1
J3,4
J4,5
8.8
8.1
10.3
8.2
8.1
8.9
7.7
9.5
J1 ,2
Ref.
3b
3a
Bz
Bz
DMSO-d6
CDCl3
CDCl3
CDCl3
CDCl3
DMSO-d6
CDCl3
CDCl3
5.13
5.17
4.92
4.83
4.98
4.88
4.85
4.90
4.28
4.22
4.48
4.23
4.32
4.04
4.20
4.26
6.13
6.02
5.91
6.03
6.18
5.54
5.27
5.73
5.63
5.72
5.45
4.20
4.36
4.94
3.85
4.97
11.35
8.55
8.29
8.54
8.40
11.48
8.23
8.71
—
—
14.8
15.8
16.7
16.5
16.4
15.0
16.5
16.4
8.4
7.2
3.8
4.0
4.5
8.0
4.5
4.4
—
—
3.9
8.0
8.0
—
4.0
8.1
6ab
6bb
6cb
3b
6d
6e
aGlcBz4
Bz
Bz
Bz
Ac
Ac
Ac
5.62
5.11
5.18
—
5.14
4.70
bGlcBz4
bGalBz4
Ac
3b
aGlcAc4
bGlcAc4
a
Abbreviations as in Scheme 3.
Previously,8 only 100 MHz data were given.
b

742
T. Nakagawa, F. W. Lichtenthaler / Tetrahedron: Asymmetry 22 (2011) 740–744
R
R'
OR
OR
a
b
c
d
e
Bz αGlcBz4
Bz βGlcBz4
Bz βGalBz4
Ac αGlcAc4
Ac βGlcAc4
Maltose
Cellobiose
Lactose
O
NH2OH
O
3 steps
75-80%
R'O
RO
R'O
RO
pyridine
85 - 90%
OR
N
OH
5a-e
6a-e
NaOMe
MeOH
OH
OH
O
HO
RO
HO
O
MeCHO
H+
RO
HO
OH
N
OH
OH
8f-h
9f-h
OH
OH
HO
HO
O
O
O
HO
g: R =
;
h: R =
HO
f: R =
;
HO
HO
HO
HO
HO
α-D-Glc
β-D-Gal
β-D-Glc
Scheme 3. Six-step conversion of common disaccharides into 4-O-glycosyl-1,5-anhydrofructoses.
Table 2
4.2. 3,6-Di-O-acetyl-4-O-(2,3,4,6-tetra-O-acetyl-a-D-
glucopyranosyl)-1,5-anhydro-D-fructose E-oxime 6d
Significant 1H and 13C NMR data (500 resp. 75,5 MHz in D2O) for the monohydrates of
1,5-anhydro-D-fructose and its 4-O-glycosyl derivatives 9f–9h
Signals
13b
9f
3.69
3.42
3.50
3.48
5.26
12.4
8.1
8.5
9g
9h
To a solution of NH2OHꢁHCl (2.4 g, 35 mmol) in pyridine
(100 mL) were added 6.2 g (10 mmol) of 2-acetoxymaltal hexaace-
tate 5d15 and the mixture was stirred at ambient temperature until
the educt had disappeared (ca. 24 h, Rf = 0.33?0.21 in 2:1 toluene–
EtOAc). Most of the solvent was removed by evaporation in vacuo
(40 °C bath temperature) and the residue was dissolved in a mix-
ture of water and CHCl3 (150 mL each), followed by separation
and extraction of the water layer with CHCl3 (2 ꢂ 100 mL). The
combined CHCl3 extracts were washed with ice-cold 1 M HCl solu-
tion and water, dried (MgSO4), and taken to dryness in vacuo. The
sirupy residue crystallized on trituration with EtOH: 4.75 g (81%)
of 6d as colorless needles (2 crops); mp 168–170 °C;
H-1e
H-1a
H-3
H-4
H-10
J1,1
3.76
3.70
3.41
3.59
3.30
3.46
3.56
3.44
—
12.1
8.9
3.60
ꢀ3.50
a
a
4.40
4.35
12.1
12.0
a
a
J3,4
J4,5
a
a
10.1
—
0
0
J1 ,2
3.8
7.9
7.8
C-1
C-2
C-3
C-4
C-5
C-6
C-10
73.3
93.9
78.4
70.4
82.2
62.0
—
73.0
93.5
78.9
78.7
81.5
63.0
99.7
71.8
93.1
79.3
79.7
75.7
61.1
103.0
72.0
93.2
79.0
79.4
75.9
61.5
103.7
½
a 2D0
ꢃ
¼ ꢄ61:4 (c 1.0, CHCl3); 1H NMR (500 MHz, CDCl3): d 1.95–
2.03 (six 3H-s, 6 OAc), 3.58 (ddd, 1H, H-5), 3.85 (dd, 1H, H-4),
3.96 (ddd, 1H, H-50), 4.01 and 4.17 (two 1H-dd, 6-H2), 4.11 and
a
Not unequivocably assignable due to complex overlap of multiplets.
4.31 (two dd, 60-H2), 4.20 (1H-d, H-1 ), 4.81 (1H-dd, H-20), 4.85
a
(1H-dd, H-1b), 4.98 (1H-dd, H-40), 5.27 (1H-d, H-3), 5.31 (1H-dd,
H-30), 5.41 (1H-d, H-10), 8.23 (1H-s, NOH); J1,1 = 16.5, J3,4 = 4.5,
0
0
0
0
J4,5 = 7.7, J5,6 = 2.3 and 4.5, J6,6 = 12.4; J1 ,2 = 4.0, J2 ,3 = 10.4,
4. Experimental
4.1. General
J3 ,4 = J4 ,5 = 9.6, J5 ,6 = 2.7 and 6.3, J6 ,6 = 12.1 Hz. 13C NMR
(75.5 MHz, CDCl3): d 20.9–21.2 (6 AcCH3), 62.1 and 64.1 (C-6, C-
60), 63.1 (C-1), 68.6–70.4 (C-5, C-20–C-40), 72.1 (C-3), 75.4 and
77.1 (C-4, C40), 95.8 (C-10), 151.9 (C-2). Anal. Calcd for
0
0
0
0
0
0
0
0
Melting points were determined with a Bock hot-stage micro-
scope and are uncorrected. Optical rotations were measured at
20 °C with a Perkin-Elmer 241 polarimeter using a cell of 1 dm
path length. 1H and 13C NMR spectra were recorded on Bruker
ARX 300 and Avance 500 instruments. Mass spectra were acquired
on Varian MAT 311 spectrometer, microanalyses on a Perkin-Elmer
240 elemental analyzer. Analytical thin layer chromatography
(TLC) was performed on Kieselgel 60 F254 plastic sheets (Merck,
Darmstadt) with detection by UV light or by spraying with 50% sul-
furic acid and charring at 140 °C for 5 min. Column chromatogra-
phy was performed on Silica Gel 60 (Merck, 63–200 ppm) using
the specified eluents.
C
24H33NO16 (591.51): C, 48.73; H, 5.62; N, 2.37. Found: C, 48.61;
H, 5.57; N, 2.29.
4.3. 3,6-Di-O-acetyl-4-O-(2,3,4,6-tetra-O-acetyl-b-D-
glucopyranosyl)-1,5-anhydro-D-fructose E-oxime 6e
Hydroxylaminolysis of 2-acetoxy-cellobial hexaacetate 5a16
(10.2 g, 16.4 mmol) was effected by stirring in pyridine (100 mL)
containing NH2ꢁHCl (10.0 g, 0.15 mol) for about 24 h at room tem-
perature (Rf = 0.27?0.16 in 2:1 toluene–EtOAc). Work-up as de-
scribed above for 6d gave a sirupy residue crystallizing from

T. Nakagawa, F. W. Lichtenthaler / Tetrahedron: Asymmetry 22 (2011) 740–744
743
EtOH: 8.24 g (85%) of 6e in two crops; colorless needles of mp 159–
4.6. 4-O-(b-
monohydrate 9g
D-Glucopyranosyl)-1,5-anhydro-D-fructose
161 °C; ½a 2D0
ꢃ
¼ ꢄ25:9 (c 1.0, CHCl3); 1H NMR (500 MHz, CDCl3): d
2.00–2.10 (six 3H-s, 6 Ac-CH3), 3.51 (1H-ddd, H-5), 3.78 (1H-ddd,
H-50), 3.83 (1H-dd, H-4), 4.09 and 4.11 (two 1H-dd, H-6a and H-
To a stirred solution of cellobiose-derived oxime hexaacetate 6e
(4.15 g, 7.0 mmol) in dry methanol (200 mL) was added 2.0 mL of a
commercial 30% (5.56 M) methanolic sodium methoxide solution
and the mixture was stirred at ambient temperature. After about
3 h, (TLC monitoring in 5:3:1 i-PrOH/EtOAc/water revealed prod-
ucts at Rf = 0.62 and 0.71) the solution was deionized with a small
amount of Amberlite IR 120 (H+ form) and, subsequently, subjected
to charcoal treatment. Removal of the solvent in vacuo gave the sir-
upy 8g (2.1 g, 88%) as an approximate 3:1 mixture (1H NMR) of E/Z-
isomers. (major) 3.8 Hz-d for H-3 at 5.16, 15.5 Hz-d for H-1b at
4.95 ppm; minor: H-3 as 3.4 Hz-d at 5.02, H-1b as 15.9 Hz-d at
4.93 ppm).
6b), 4.26 (1H-d, H-1a), 4.30 and 4.31 (two 1H-dd, H-6b and H-
6b0), 4.70 (1H-d, H-10), 4.90 (1H-d, H-1b), 4.97 (1H-dd, H-20), 5.08
(1H-t, H-30), 5.19 (1H-t, H-4), 5.73 (1H-d, H-3), 8.71 (1H-s, NOH);
J1,1 = 16.4, J3,4 = 4.4, J4,5 = 8.1, J5,6 = 2.6 and 4.9, J6,6 = 11.9,
0
0
0
0
0
0
0
0
0
0
0
0
J1 ,2 = 8.1, J2 ,3 = J3 ,4 = J4 ,5 = 9.5, J5 ,6 = 4.9 and 6.1, J6 ,6 = 12.0 Hz.
13C NMR (75.5 MHz, CDCl3): d 20.5–20.7 (AcCH3), 61.8, 63.1 and
63.4 (C-1, C-6, C-60), 71.0, 71.4, 72.1, 72.8 (C-3, C-20, C-30, C-50),
76.5 (C-5), 79.0 (C-4), 101.4 (C-10), 151.5 (C-2). Anal. Calcd for
C24H33NO16 (591.51): C, 48.73; H, 5.62; N, 2.37. Found: C, 48.69;
H, 5.51; N, 2.30.
Deoximation of the E/Z-mixture 8g (2.0 g, 5.9 mmol) was ef-
fected by stirring in acetonitrile (40 mL) with acetaldehyde
(1 mL, 17.5 mmol) and M HCl (8 mL) at rt for 5 h, followed by dilu-
tion with water (15 mL) neutralization by briefly stirring with a ba-
sic resin, and in vacuo evaporation to give a sirup. Purification by
elution from a silica gel column (2 ꢂ 20 cm) with 7:3 n-propa-
nol–water and in vacuo evaporation of the product-carrying elu-
4.4. 4-O-(a-D-Glucopyranosyl)-1,5-anhydro-D-fructose E/Z-
oxime 8f
Commercial 30% methanolic sodium methoxide (1.3 mL) was
added with stirring to a solution of maltose-derived oxime hexa-
acetate 6d (3.75 g, 6.3 mmol) in methanol (150 mL) TLC monitor-
ing after about 2 h with 8:5:1 iPrOH/EtOAc/water showed
complete conversion of the educt into an approximate 2.5:1 mix-
ture of E/Z-isomers (1H NMR) of Rf = 0.54 (tentatively the E-isomer)
and 0.27, respectively. The solution was then deionized by briefly
stirring with Amberlite 120 (H+ form) filtration and washing of
the resin with methanol. The filtrate and washings were subjected
to a brief charcoal treatment followed by an in vacuo removal of
the solvent, finally at 0.1 Torr: 1.78 g (83%) of E/Z-8f as a fluffy so-
ates gave 1.27 g (63 %) of 9g as a fluffy solid. ½a D20
¼ ꢄ21:2 (c 1.0,
ꢃ
water); 1H NMR (500 MHz, D2O, 3 h after dissolution): d 4.40
(1H-d, J1 ,2 = 7.9 Hz, H-10), 3.90 and 3.75 (two 2H-m, 6-H2, 60-H2),
0
0
3.70 and 3.42 (two 1H-d, J1,1 = 12.1 Hz, H-1b and H-1a), 3.60
(2H-m, H-3, H-4), 3.40 (br 4H-m, H-5, H-30–H-50), 3.26 (1H-m, H-
13
20). C NMR (D2O): d 103.0 (C-10), 93.1 (C-2), 79.7 and 79.4 (C-3,
C-4), 76.2 (C-3, C-50), 75.7 (C-5), 74.0 (C-20), 71.8 (C-1), 70.3 (C-
40), 60.9 (C-6, C-60). Anal. Calcd for C12H20O10ꢁH2O (342.30): C,
42.10; H, 6.48. Found: C, 42.02; H, 6.50.
lid. 1H NMR (500 MHz, D2O): 5.20 (1H-d, J1 ,2 = 3.8 Hz, H-10), 5.12
(1H-d, J3,4 = 3.8 Hz, H-3), 4.90 and 4.20 (two 1H-d, J1,1 = 16.1 Hz,
0
0
Whilst these 1H and 13C NMR data correspond fairly well with
those reported by Agoston et al.9 (there were some minor chemical
shift differences) their microanalytical data (found:9 C, 40.09; H,
6.65) do not, as these are distinctly too low; however, they match
those required for C12H22O10.H2O (sum formula given in Ref. 9),
which factually is a dihydrate (calcd for C12H18O9ꢁ2H2O: C, 40.00;
H, 4.71). These peculiar incongruities remain to be clarified.
De-O-benzoylation of oxime hexabenzoate 6b, directly followed
by deoximation in a procedure as detailed for the galactosyl ana-
logue (see Section 4.7) afforded 9f in 61% yield.
1-Hb, and 1-Ha) for the major, conceivably the E-isomer; other sig-
nals not interpreted due to E/Z-mixture and extensive overlap.
Anal. Calcd for C12H21NO10 (339.29): C, 42.48; H, 6.24; N, 4.13.
Found: C, 42.29; H, 6.17; N, 4.06.
The E/Z-isomeric oximes, due to sufficiently different mobilities
(Rf = 0.54 and 0.27, resp.), may be separated by chromatography on
a silica column, yet the pure isomers tend to equilibrate upon re-
moval of the elution solvents. Thus, the E/Z-mixture was subjected
directly to deoximation (?9f, see below).
4.5. 4-O-(
monohydrate 9f
a-D-Glucopyranosyl)-1,5-anhydro-D-fructose
4.7. 4-O-(b-
monohydrate 9h
D-Galactopyranosyl)-1,5-anhydro-D-fructose
Acetaldehyde (0.57 mL, 10 mmol) and 1 M HCl (5 mL) was
added to a suspension of the E/Z-oxime mixture 8f (1.20 g,
3.5 mmol) in acetonitrile (25 mL), followed by stirring at ambient
temperature for 5 h. Dilution with water (10 mL), neutralization
by briefly stirring with a basic resin, in vacuo evaporation to a
small volume (ꢀ2 mL), elution from an LH 20 Sephadex column
(25 ꢂ 2 cm) with water, and lyophilization of the product-carrying
eluates gave the disaccharide monohydrate 9f as a colorless foam
(630 mg, 74%). 1H NMR (500 MHz, D2O, 5 h after solution for equil-
Lactose-derived oxime hexabenzoate 6c (2.40 g, 2.5 mmol) was
stirred in 30 mL of a 0.1 M methanolic NaOMe solution (prepared
by adding 5.4 mL of commercial NaOMe/MeOH to 25 mL of MeOH)
resulting in a clear solution after about 30 min. Stirring was contin-
ued for another 3 h (TLC monitoring), and the solution was diluted
with 100 mL of MeOH, and neutralized by stirring with Dowex
50 WX8, H+ form, for 10 min. Filtration, washing of the resin with
MeOH, and in vacuo evaporation of the filtrate and washings gave
a sirup, comprising an approximate 3:2 E/Z-oximes 8h (740 mg,
87%); 1H NMR in D2O: H-3as 3.8 Hz-d at 5.10 ppm for the major
and 3.3 Hz-d at 4.94 for the minor isomer.
ibration): d 5.29 (1H-d, J1 ,2 = 3.8 Hz, H-10), 3.80 and 3.68 (two 2H-
m, 6-H2, 60-H2, 3.71 and 3.42 (two 1H-d, J1,1 = 12.4 Hz, H-1b, and
0
0
H1
a
), ꢀ3.60 (br m, H-4, H-5, H-20, H-30), 3.50 (14-d, J3,4 = 8.1 Hz,
The product was directly subjected to deoximation with acetal-
dehyde (0.5 mL) in acetonitrile (20 mL) and 4 mL of M HCl (5 h, rt),
followed by dilution with water (10 mL), neutralization with a ba-
sic resin and in vacuo evaporation to a sirup. Purification by elution
from a silica gel column (2 ꢂ 20 cm) with 7:3 n-PrOH/water and in
vacuo evaporation of the product-carrying fractions left 570 mg of
13
H-3), 3.35 (1H-m, H-40). C NMR (75.5 MHz, D2O): d 99.7 (C-10),
80.1 (C-5), 78.9 and 78.7 (C-3, C-4), 73.6 (C-30), 73.0 (C-1), 72.4
and 72.2 (C-2, C-50), 70.1 (C-40), 61.8 and 61.3 (C-6, C-60). Anal.
Calcd for C12H20O10ꢁH2O (342.30): C, 42.10; H, 6.48. Found: C,
41.82; H, 6.39.
9h (67%, based on 6c) as a glass; ½a D25
ꢃ
¼ ꢄ22:3 (c 0.8, water); 1H
De-O-benzoylation of maltose-derived oxime hexabenzoate 6a
(?8a) and subsequent deoximation with acetaldehyde in a proce-
dure identical to the galactosyl analogue 6c (see Section 4.7) affor-
ded 9f in 66% yield.
NMR (500 MHz, D2O): d 4.35 (d, 1H, J1 ,2 = 7.9 Hz, H-10), 3.78–
0
0
3.62 (complex m, 6-H2, 60-H2, H-5, H-50), 3.59 and 3.30 (two 1H-
d, J1,1 = 12.0 Hz, H-1b and H-1
a
), ꢀ3.50 (3H-m, H-3, H-30, H-40),

744
T. Nakagawa, F. W. Lichtenthaler / Tetrahedron: Asymmetry 22 (2011) 740–744
ꢀ3.40 (2H-m, H-4, H-20). 13C NMR (75.5 MHz, D2O): d 103.7 (C-10),
93.2 (C-2), 79.4 (C-4), 79.0 (C-3), 75.9 (C-5, C-50), 73.1 (C-30), 72.0
(C-1), 71.4 (C-20), 69.2 (C-40), 61.5 and 61.2 (C-6, C-60). Anal. Calcd
for C12H20O10ꢁH2O (342.30): C, 42.10; H, 6.48. Found: C, 41.98; H,
6.40.
6. (a) Baute, A.; Baute, R.; Deffieux, G. Phytochemistry 1988, 27, 3401; (b) Taguchi,
T.; Haruna, M.; Okuda, J. J. Biotechnol. Appl. Biochem. 1993, 18, 275; (c) Yu, S.;
Kenne, M.; Pedersén, M. Biochim. Biophys. Acta 1995, 1244, 1; (d) Freimund, S.;
Huwig, A.; Giffhorn, F.; Köpper, S. Chem. Eur. J. 1998, 4, 2442; (e) Fujisue, M.;
Yoshinaga, K.; Muroya, K.; Abe, J.; Hizukuri, S. J. Appl. Glycosci. 1999, 46, 439; (f)
Yu, S. Zuckerindustrie (Berlin) 2004, 129, 26.
7. (a) Nakamura, T.; Takahashi, A.; Kato, K. (Nippon Kagaku Co.) Jap. Pat. 63-
072696, 1988, Appl. 1986; Chem. Abstr. 1988, 510835.; (b) Lundt, I.; Stütz, A.;
Dekany, G.; Thiem, J.; Agoston, K.; Andreassen, M. PCT Int. Appl. WO 2005
21114; Chem. Abstr. 2005, 144, 51836.; (c) Dekany, G.; Lundt, I.; Niedermair, F.;
Bichler, S.; Spreitz, J.; Sprenger, F. K.; Stütz, A. E. Carbohydr. Res. 2007, 342,
1249.
8. Lichtenthaler, F. W.; Kaji, E.; Weprek, S. J. Org. Chem. 1985, 50, 3505.
9. Agoston, K.; Dékány, G.; Lundt, I. Carbohydr. Res. 2009, 344, 1014.
10. Schmidt, R. R.; Stumpp, M. Liebigs Ann. Chem. 1983, 1249.
11. Fiskesund, R.; Abeyama, K.; Yoshinaga, K.; Abe, J.; Yuan, Y.; Yu, S. Planta Med.
2010, 76, 1635.
The NMR data reported by Agoston et al.9 for 9h corresponded
reasonably well to those described above, yet their microanalytical
data (found:9 C, 39.94; H, 6.72) are too low by an intolerable 5%.
Curiously, they correspond perfectly to C12H22O10ꢁH2O,9 which de
facto is a dihydrate C12H20O9ꢁ2H2O (calcd C, 40.00; H, 6.71).
References
1. Lergenmüller, M.; Kläres, U.; Lichtenthaler, F. W. Carbohydr. Res. 2009, 344,
2127.
2. Deffieux, G.; Baute, R.; Baute, M.-A.; Atfani, M.; Carpy, A. Phytochemistry 1987,
26, 1391.
12. Yoshinaga, K.; Abe, J.; Tanimoto, T.; Koizumi, K.; Hizukuri, S. Carbohydr. Res.
2003, 338, 2221.
13. Richard, G.; Yu, S.; Monsan, P.; Remaud-Simeon, M.; Morel, S. Carbohydr. Res.
2005, 340, 395.
3. (a) Lichtenthaler, F. W.; El Ashry, E. S. H.; Göckel, V. H. Tetrahedron Lett. 1980,
21, 1429; (b) Brehm, M.; Göckel, V. H.; Jarglis, P.; Lichtenthaler, F. W.
Tetrahedron: Asymmetry 2008, 19, 358.
4. Blair, M. G. Methods Carbohydr. Chem. 1963, 2, 412.
5. (a) Lichtenthaler, F. W.; Jarglis, P. Tetrahedron Lett. 1980, 21, 1425; (b) Jarglis, P.;
Lichtenthaler, F. W. Angew. Chem., Int. Ed. Engl. 1982, 21, 141; Angew. Chem.
Suppl. 1982, 175.
14. Jarglis, P.; Göckel, V.; Lichtenthaler, F. W. Tetrahedron: Asymmetry 2009, 20,
952.
15. Lainé, V.; Coste-Sarguet, A.; Gadelle, A.; Defaye, J.; Perly, B.; Djedaini-Pilard, F. J.
Chem. Soc., Perkin Trans. 2 1995, 1479.
16. (a) Xu, X.; Tan, Q.; Hayashi, M. Synthesis 2008, 770; (b) Rao, D. R.; Lerner, L. M.
Carbohydr. Res. 1972, 22, 345.