Dimeric structures of 1,5-anhydro-D-fructose

Article abstract of DOI:10.1016/S0008-6215(98)00058-5

Upon treatment with acetic anhydride-pyridine, 1,5-anhydro-D-fructose gave the enolone 1,5-anhydro-4-deoxy-D-glycero-hex-3-eno-pyranose-2-ulose as a product of elimination, which is thus available in an overall yield of 25% from D-glucose. During this reaction, acetylated dimers of 1,5-anhydro-D- fructose were formed in a side reaction, whereas these were the only products when acidic acetylation conditions were applied. The acetylated dimers were isolated and served for an unequivocal structural assignment of dimeric forms of 1,5-anhydro-D-fructose 1 by NOESY experiments. Whereas anhydroketose 1 forms dimers in Me₂SO and pyridine, its C-4 epimer, 1,5-anhydro-D-tagatose, did not form dimeric structures, neither in non aqueous nor in aqueous solvents.

Full text of DOI:10.1016/S0008-6215(98)00058-5

Carbohydrate Research 308 (1998) 195±200
Note
Dimeric structures of 1,5-anhydro-d-fructose1
Stefan Freimund, Sabine Kopper *
Carl v. Ossietzky University of Oldenburg, FB Chemistry, PO Box 2503, D-26121 Oldenburg, Germany
Received 21 July 1997; accepted in revised form 28 February 1998
Abstract
Upon treatment with acetic anhydride-pyridine, 1,5-anhydro-d-fructose gave the enolone
1,5-anhydro-4-deoxy-d-glycero-hex-3-eno-pyranose-2-ulose as a product of elimination, which is
thus available in an overall yield of 25% from d-glucose. During this reaction, acetylated dimers
of 1,5-anhydro-d-fructose were formed in a side reaction, whereas these were the only products
when acidic acetylation conditions were applied. The acetylated dimers were isolated and served
for an unequivocal structural assignment of dimeric forms of 1,5-anhydro-d-fructose 1 by NOESY
experiments. Whereas anhydroketose 1 forms dimers in Me2SO and pyridine, its C-4 epimer,
1,5-anhydro-d-tagatose, did not form dimeric structures, neither in non aqueous nor in aqueous
solvents. # 1998 Elsevier Science Ltd. All rights reserved
Keywords: 1,5-Anhydro-d-fructose; 1,5-Anhydro-d-tagatose; Dimeric structures; NOESY spectroscopy
1,5-Anhydro-d-fructose (1, Scheme 1) is of
interest as a precursor of antibiotics [1±4] and for
various chemical modi®cations. Synthesis of 1 has
been accomplished by enzymatic conversion [1,5±
10] or chemical multi-step reactions [11,12]. We
obtained 1 by enzymatic oxidation of the corre-
sponding 1,5-anhydro-d-hexitol with pyranose-2-
oxidase from the fungus Peniophora gigantea
[13,14], in a procedure that will be published else-
where [15]. We were now able to convert 1 into the
enolone 2 by treatment with acetic anhydride in
pyridine. The enolone 2, a synthetically interesting
product that might be useful in a variety of reac-
tions [16±19], is thus available in a chemoenzy-
matic approach in 25% yield and four steps from
d-glucose.
Besides the enolone 2, dimeric forms of acety-
lated 1,5-anhydro-d-fructose 3 (24%) and 4 (12%)
were obtained when basic reaction conditions were
applied. Upon acidic acetylation conditions
(HClO4), elimination products were not observed,
but solely formation of dimers 3 and 4. The com-
pounds were isolated and served for an unequi-
vocal structural assignment of dimeric forms of
1,5-anhydro-d-fructose (1) by NOE spectroscopy.
In aqueous solution, 1 exclusively exists in a
hydrated pyranoid form 1a [20]. In other solvents
the same compound was reported to form dimers
but the NMR spectra were not assigned due to
their complexity [21]. We found that solutions of
anhydroketose 1 in Me2SO-d6 contained the non
hydrated anhydroketose 1b together with two
dimeric forms 5 and 6 in 15 and 9% proportions,
respectively. In pyridine-d5 dimeric structures pre-
dominated, 5, 6, and 1b were found in 50, 10, and
1
Presented in part at the XVIIth International Carbo-
hydrate Symposium, Milano, Italy, 1996, Abstract BP085.
* Corresponding author. Fax: +49(0)441 798 3329;
e-mail: koepper@chemie.uni-oldenburg.de
1
40% ratios, respectively. Although the H and 13C
0008-6215/98/$19.00 # 1998 Elsevier Science Ltd. All rights reserved
PII S0008-6215(98)00058-5

196
S. Freimund, S. Kopper/Carbohydrate Research 308 (1998) 195±200
È
Scheme 1.
NMR spectra of dimers 5 and 6, as well as 3 and 4
were completely assigned, an unequivocal structural
assignment of the dimers was not possible from
these data. We therefore performed NOESY experi-
ments on the isolated acetylated dimers 3 and 4.
Dimeric structures of dicarbonyl carbohydrates
have been assigned in the case of 1,6-anhydro-hex-
osuloses [22±24] and methyl ketosides [25] and the
formation of dimers under acidic or thermal con-
ditions has been reported for ketoaldoses [26] as
well as ketoses [27±30]. The dimers of ꢀ-hydroxy
aldehydes and ketones may theoretically be formed
as symmetrical or unsymmetrical 1,4-dioxane or
dioxolane compounds (Scheme 2).
The NMR spectra of 3 and 4, as well as 5 and 6,
clearly show evidence of the existence of asym-
metric dimers. The C-20 atoms in both acetylated
and non acetylated dimers resonate around
100 ppm, which indicates a 2,6-cyclised pyranoid
anhydroketose and is in agreement with the reso-
nances found for the spiroketal function in dihex-
ulose dianhydrides [28±31]. The C-2 atom of the
other ring shows a relative down®eld shift at
around 108±110 ppm and indicates a quarternary
carbon in a dioxolane type ring. The resonances for
C-3 and C-30 of both rings show di€erences of up
to 12 ppm. This accounts for the fact that O-30 is
involved in the dioxolane ring closure. Similar dif-
ferences in chemical shifts are observed for H-3
and H-30 in the H NMR spectra. The coupling
constants in these spectra show a perfect C1 chair
for the right ring bearing the spiroketal group,
whereas the coupling constants for the primed ring
1
4
0
(J3 ,4 5.8 and 6.7 Hz in 5 and 6, respectively; J3, 4
0
0 0
0
0
6.1 and 4.8 Hz in 3 and 4; as well as J4 ,5 6.3 Hz in
4) indicate a noticeably ¯attened chair for this ring,
a fact that has been previously observed in crystal
structures of pyranoses bearing annelated ®ve-
membered rings [32,33]. From these ®ndings it
could be concluded, that the two distinct dimers
were asymmetric structures with a ®ve-membered
central ring and one spiroketal function and the
four structures A±D had to be considered. The
dioxolane ring may bear a trans-con®guration in
positions 20 and 30 as in A and B, or a cis-relation-
ship as in C and D (Fig. 1).
The data so far accumulated for 4 and 6 did not
allow discrimination between the four possible
structures A±D; we therefore performed NOE
spectroscopy of the isolated acetylated dimers 3
and 4. The NOESY spectra revealed interglycosidic
contacts between H-1e±H-30 (11%) in 3 and
between H-1e±H-40 (31%) in 4. The H-1e±H-40
contact in 4 clearly con®rmed the existence of
structure D for the minor isomer 4 (1.8:1 in Me2SO
and 5:1 in pyridine for 3 and 4, respectively).
Scheme 2.

S. Freimund, S. Kopper/Carbohydrate Research 308 (1998) 195±200
È
197
Fig. 1. Energy minimised (SYBYL) structures of possible asymmetric dimers of 1,5-anhydro-d-fructose (1).
Although the existence of a NOE H-1e±H-30 con-
tact in 3 could be attributed to both isomers A and
C, the relative size of the contact supported struc-
ture C. The relatively small contact observed ®ts
better with the estimated distance H-1e±H-30 of
3.5 A in C than of 2.5 A in A, when compared to
the size of the other contacts in 3 or to the relative
contact H-1e±H-40 of 31% 4 with an estimated
distance H-1e±H-40 of 2.3 A. The relative values of
the numerous intra ring NOE contacts are well in
agreement with these structures and with the ®nd-
the isolated acetylated derivatives, we believe that
the acetylated forms 3 and 4 are derivatives of the
free dimers 5 and 6 in non aqueous solutions. This
is indicated by comparable ratios of free 5 and 6
(1.8:1 in Me2SO; 5:1 in pyridine) and the acetylated
dimers 3 and 4 (2:1) and the very comparable but
striking di€erences in chemical shifts of C-4±C-40
and H-4±H-40 in dimers 3 and 5, but not in 4 and 6.
The two dimeric forms of 1,5-anhydro-d-fructose
formed under acetylating conditions are thus both
cis-con®gured dioxolane-type dimers that differ in
the orientation of the spiroketal function C-2.
4
ing that the right pyranose ring is in a perfect C1
chair conformation in both dimers 3 and 4,
whereas the left ring adopts a slightly distorted
chair conformation. Although the structural
assignment of the dimers was performed solely on
Recently, the NMR data for the C-4 epimer of 1,
1,5-anhydro-d-tagatose (8, Scheme 3) could not be
determined by Barili et al. [34] due to the com-
plexity of the spectra. This led the authors to
Scheme 3.

198
S. Freimund, S. Kopper/Carbohydrate Research 308 (1998) 195±200
È
assume the formation of dimeric or other structures
in solution, which is also supported by our ®nding
that the epimer 1 forms 60% of dimeric structures
in pyridine. We therefore investigated the forma-
tion of dimers of this compound. 1,5-Anhydro-d-
tagatose (8) was obtained by the same enzymatic
oxidation [13,14] and it was found that in aqueous
solutions, both the free ketopyranose 8a and the
hydrated pyranose 8b were present in a 7:3 ratio.
Astonishingly, neither in Me2SO-d6 nor in pyridine-
d5 were dimeric structures detectable. Consequently
1,5-anhydro-d-tagatose (8) exclusively yields the
elimination product 2 under acetylating conditions,
which might be rationalised by the fact that the
trans-diaxial arrangement of H-3 and AcO-4 facil-
itates a ꢁ-elimination through a E2 mechanism.
ꢂ=2.490 and 39.50 ppm for these solvents, respec-
tively. 1D, as well as 2D HMQC and 2D NOESY
spectra were recorded using the standard Bruker
software. For the NOESY spectra mixing times of
300 and 900 ms were used. A baseline correction
was applied in both dimensions and the volume
integration is given relative to the diagonal peak of
H-3.
3,6-Di-O-acetyl-1,5-anhydro-4-deoxy-d-glycero-
hex-3-enopyranose-2-ulose (2), [(2R,3R,4S,5S)-
3,4-diacetoxy-2-acetoxymethyl-tetrahydropyran]-5-
spiro-50-[(10R,20R,30S,70S)-20,70-diacetoxy-10-aceto-
xymethyl-40,60,90-trioxabicyclo[4,3,0]nonane] (3),
and [(2R,3R,4S,5S)-3,4-diacetoxy-2-acetoxymethyl
-tetrahydropyran]-5-spiro-50 -[(10R,20R,30S, 70R)-
20,70 -diacetoxy-10 -acetoxymethyl-40,60,90 -trioxabi-
cyclo[4,3,0]nonane] (4).ÐA soln of 35 mg
(0.21 mmol) of dry 1 in 2 mL of pyridine and
ꢀ
1. Experimental
0.5 mL of Ac2O was kept at 5 C overnight. After
addition of 5 mL of water the solution was extrac-
ted with CH2Cl2, dried over Na2SO4 and evapo-
rated several times with toluene. Column
chromatography using 1:2 AcOEt±petroleum ether
for elution ®rst yielded 21 mg (45%) of syrupy 2:
‰ꢀŠ2D0= 17.7ꢀ (c 0.34 in CH2Cl2); MS: m/z 229 (48)
[M]+, 187 (10) [M±CH2CO]+, 169 (100%) [M±
HOAc]+, 127 (22); 1H NMR and 13C NMR
(CDCl3) signals of the protecting groups: ꢂ 2.128,
2.265 (6 H, CH3CO); 20.25, 20.80 (CH3CO),
168.05, 170.66 (CH3CO); other NMR data see
Tables 1 and 2. Anal. Calcd for C10H12O6 (228.2):
C, 52.63; H, 5.30. Found C, 52.54; H, 5.43.
General methods.ÐThe reactions were mon-
itored by TLC on silica gel 60 F254 (E. Merck) and
the spots were detected by UV absorption and
staining with sulphuric acid. For column chroma-
tography silica gel 60 (E. Merck) was used. Optical
rotations were determined with a Perkin-Elmer
model 241 MC. Mass spectra were obtained on a
MAT 212 from Finnigan (70 eV, He gas support)
using chemical ionisation. NMR spectra were
recorded on a Bruker AMX 500 spectrometer at
1
500.13 MHz for H and 125.76 MHz for 13C at
300 K. Chemical shifts are referenced to internal
acetone (ꢂ=2.030 and 30.50) for D2O as solvent; in
CDCl3 to ꢂ=7.270 and 77.00 and in Me2SO-d6 to
Further elution with 2:1 EtOAc±petroleum ether
gave 21 mg (36%) of slowly crystallising syrup as a
Table 1
13C NMR chemical shifts ꢂ (ppm)
C-1±C-1'
C-2±C-2'
C-3±C-3'
C-4±C-4'
C-5±C-5'
C-7±C-7'
8b
9b
1bc
5c
72.38
72.61
72.10
69.38
71.17
68.62
71.71
71.43
71.47
69.15
71.92
69.22
208.49
92.40
204.98
107.63
99.87
108.53
100.37
187.65
108.53
105.25
109.41
104.63
76.02
71.86
79.13
73.87
88.57
74.51
87.19
132.75
70.96
82.81
73.32
82.54
72.75
69.41
73.07
69.53
68.21
69.83a
70.02a
143.81
68.56
70.08
68.67
68.67
78.74
80.02
81.25
81.45
80.31
81.51a
80.14a
72.62
76.86a
75.11a
76.46
75.58
61.19
61.45
60.97
61.41
61.79
61.05a
61.17a
64.39
62.72a
63.27a
62.55a
63.25a
6c
2d
3d
4d
Atom numbering according to carbohydrate nomenclature, see also Scheme 1.
a
Signal assignment may be inverted.
In D2O.
In Me2SO-d6.
In CDCl3.
b
c
d

S. Freimund, S. Kopper/Carbohydrate Research 308 (1998) 195±200
È
199
Table 2
1H NMR chemical shifts ꢂ (ppm) and coupling constants (Hz)
H-1e±H-10e J1e,1a H-1a±H-10a H-3±H-30 J3,4 H-4±H-40 J4,5 H-5±H-50 J5,6a J5,6b H-6a±H-60a J6a,6b H-6b±H-60b
8ab,e
8bb
1bc
5c
4.019a
3.573a
4.062a
3.889a
3.949a
3.855a
3.572a
4.462a
3.980
15.0
11.9
14.1
13.1
11.7
11.0
11.6
16.5
12.1
12.0
12.0
12.5
3.904a
3.205a
3.859a
3.428a
3.139a
3.368a
3.388a
4.259a
3.510
4.417
3.520
4.067
3.309
3.810
3.398
3.818
Ð
5.334
4.730
5.243
4.425
3.9
3.8
9.0
9.9
5.8
9.6
6.7
Ð
10.0
6.1
9.7
4.8
4.169
3.702
3.349
3.200
3.863
3.168
3.114
6.608
5.146
5.303
5.079
4.950
0.6
1.1
n.d.
9.0
11.7
9.0
3.896
3.403
3.472
3.033
3.124
ꢁ3.03
3.040
4.803
3.675
3.675
3.675
3.727
6.7
7.9
4.4
n.d.
ꢁ8
n.d.
n.d.
6.1
6.0
AB
2.6
3.8
5.6
4.2
3.571
3.550
3.708
3.643
3.616
3.640
3.617
4.421
4.208
4.150
4.214
4.198
n.d.
11.7
13.1
12.5
12.5
ꢁ12
ꢁ12
11.8
12.3
AB
n.d.
3.503
3.470
3.377
3.343
3.392
3.389
4.251
4.133
4.150
4.162
4.110
AB
n.d.
n.d.
n.d.
n.d.
4.3
3.1
AB
5.5
6c
ꢁ12
2d,f
3d
2.3
10.0
9.1
9.8
6.3
4.550
4.048
4.755
3.896
3.562
3.806
4d
12.5
11.8
6.4
J4,OH-4 OH-6 J6,OH-6 OH-30 J3 ,OH-3 OH-40 J4 ,OH-4
OH-60
0
0
0
0
0
J6 ,OH-6
0
OH-2
J3,OH-3
OH-4
1b
5
6
5.404g
6.484
6.866
5.5
Ð
Ð
5.506
5.065
5.523
5.7
5.6
6.1
4.632
4.498
4.570a
5.3
5.5
5.9
Ð
4.670
n.d.
Ð
5.4
n.d.
Ð
Ð
Ð
4.486
4.512a
Ð
6.3
5.8
4.903 5.5
5.186 5.6
Atom numbering according to carbohydrate nomenclature, see also Scheme 1; n.d., not determined.
a
Signal assignment may be inverted.
In D2O.
b
c In Me2SO-d6.
d
In CDCl3.
e J1a,5=1.8 Hz.
J1e,3=1.0 Hz.
OH-3.
f
g
mixture of 3 and 4 in the ratio 2:1 (NMR): mp 60±
ꢀ
(28), H-1e±H-40 (31), H-10a±H-10e (107), H-10a±H-30
(7), H-10a±H-50 (29), H-30±H-40 (23), H-30±H-50
(18), H-40±H-50 (65), H-40±H-60a (17), H-40±H-60b
(28).
2 from acetylation of 1,5-anhydro-d-tagatose
(8).ÐA total of 19 mg (0.12 mmol) of 8 were
treated with Ac2O-pyridine as described above.
Separation by column chromatography (1:2
EtOAc±petroleum ether) yielded 17 mg (64%) of a
colourless syrup that was identical to compound 2.
62 C; MS: 517 (38) [M±HOAc]+, 457 (14) [M±
+
+
.
.
.
2 HOAc] , 397 (3) [M±3 HOAc] , 337 (4) [M±
+
+
.
4 HOAc] , 277 (2) [M±5 HOAc] , 229 (100) [M/
2+1±HOAc]. Anal. Calcd for C24H32O16 (576.5):
C, 50.00; H, 5.59. Found C, 50.13; H, 5.41.
For structural assignment using NOE spectros-
copy an analytical sample was separated by HPLC
on LiChrosorb Si 60 (E. Merck, 1:1 EtOAc±petro-
leum ether). 1H NMR and 13C NMR signals of the
protecting groups: 3: ꢂ 2.025, 2.046, 2.059, 2.094,
2.116, 2.191 (18 H, CH3CO); 20.52, 20.62, 20.73,
20.75, 20.78, 21.67 (CH3CO); 169.04, 169.16,
169.24, 169.99, 170.53, 170.64 (CH3CO). 4: ꢂ 2.024,
2.076, 2.095, 2.098, 2.100, 2.120 (18 H, CH3CO,
20.61, 20.73 (2Â), 20.79, 20.89, 21.66 (CH3CO),
168.87, 169.28, 169.37, 169.71, 170.57, 170.73
(CH3CO). For other NMR data see Tables 1 and
2. NOE contacts (% relative to the e€ect H-1a±H-
1e): 3: H-1a±H-3 (29), H-1a±H-5 (21), H-4±H-5
(26), H-4±H-6a (16), H-4±H-6b (28), H-1e±H-30
(11), H-10a±H-10e (119), H-10a±H-30 (5), H-10a±H-
50 (56), H-30±H-40 (55), H-30±H-50 (30), H-40±H-60a
(63). 4: H-1a±H-3 (23), H-1a±H-5 (29), H-3±H-4
(18), H-4±H-5 (66), H-4±H-6a (17), H-4±H-6b
Acknowledgements
This work was supported by the `Bundesminister
fur Forschung und Technologie' (Grant No.
0319516A).
References
[1] G. Deeux, R. Baute, M.-A. Baute, M. Atfani,
and A. Carpy, Phytochemistry, 26 (1987) 1391±1393.
[2] M.-A. Baute, G. Deeux, R. Baute, A. Badoc, J.
Vercauteren, J.-M. Leger, and A. Neveu, Phy-
tochemistry, 30 (1991) 1419±1423.

200
S. Freimund, S. Kopper/Carbohydrate Research 308 (1998) 195±200
È
[3] M.-A. Baute, G. Deeux, J. Vercauteren, R.
Baute, and A. Badoc, Phytochemistry, 33 (1993)
[18] P. Jarglis and F.W. Lichtenthaler, Angew. Chem.,
94 (1982) 140±141.
41±45.
[4] A. Broberg, L. Kenne, and M. Pedersen, Phy-
tochemistry, 41 (1996) 151±154.
[5] T. Taguchi, M. Haruna, and J. Okuda, Biotechnol.
Appl. Biochem., 18 (1993) 275±283.
[6] M.-A. Baute, R. Baute, and G. Deeux, Phy-
tochemistry, 27 (1988) 3401±3403.
[7] T. Nakamura, A. Naito, Y. Takahashi, and H.
Akanuma, J. Biochem., 99 (1986) 607±613.
[8] S. Yu, L. Kenne, and M. Pedersen, Biochim. Bio-
phys. Acta, 1244 (1995) 1±9.
[9] S. Yu, K. Bojsen, K.M. Kragh, M.Bojko, J. Niel-
sen, J. Marcussen, and T.M.I.E. Christensen, PCT
Int. Appl. WO 95 10,616 (1995); Chem. Abstr., 123
(1995) 226024f.
[10] S. Yu, K. Bojsen, and J. Marcussen, PCT Int.
Appl. WO 96 12,026 (1996); Chem. Abstr., 125
(1996) 56375t.
[11] F.W. Lichtenthaler, E.S.H. El Ashry, and V.H.
Gockel, Tetrahedron Lett., 21 (1980) 1429±1432.
[12] T. Nakamura, A. Takahashi, and K. Kato, Jpn.
Kokai Tokkyo Koho JP 63 72,696 [88 72,696]
(1988); Chem. Abstr., 109 (1988) 110835c.
[13] A. Huwig, A. Brans, H.J. Danneel, S. Kopper, and
F. Gi€horn, Meded. Fac. Landbouwwet., Univ.
Gent, 57 (1992) 1749±1753.
[19] F. W. Lichtenthaler, P. Jarglis, and W. Hempe,
Liebigs Ann. Chem., (1983) 1959±1972.
[20] T. Taguchi, M. Haruna, and J. Okuda, Biotechnol.
Appl. Biochem., 18 (1993) 275±283.
[21] F.W. Lichtenthaler, E.S.H. El Ashry, and V.H.
Gockel, Tetrahedron Lett., 21 (1980) 1429±1432.
[22] K. Heyns and P. Koll, Chem. Ber., 104 (1971)
2553±2566; 3835±3841.
[23] K. Heyns and P. Koll, Chem. Ber., 106 (1973) 611±
632.
[24] K. Heyns, W.-D. Soldat, and P. Koll, Chem. Ber.,
108 (1975) 3611±3618.
[25] H.-M. Liu, Y. Sato, and Y. Tsuda, Chem. Pharm.
Bull., 41 (1993) 491±501.
[26] P.A. Garcia Ruiz, I. Cartagena Travesedo, and A.
Soler Martinez, Can. J. Chem., 62 (1984) 561±564.
[27] M. Manley-Harris and G.N. Richards, Adv. Car-
bohydr. Chem. Biochem., 52 (1997) 207±266.
[28] J. Defaye, A. Gadelle, and C. Pedersen, Carbohydr.
Res., 136 (1985) 53±65.
[29] J.M. Garcia Fernandez, A. Gadelle, and J. Defaye,
Carbohydr. Res., 265 (1994) 249±269.
[30] H.-Y. Li, H. Hagiwara, W. Zhu, C. Yokoyama,
and N. Harada, Carbohydr. Res., 299 (1997) 301±
305.
[31] J. Defaye and J.M. Garcia Fernandez, Carbohydr.
Res., 237 (1992) 223±247.
[14] A. Huwig, H.J. Danneel, and F. Gi€horn, J. Bio-
technol., 32 (1994) 309±315.
[15] S. Freimund, A. Huwig, F. Gi€horn, and S. Kop-
per, Chem. Eur. J., in press.
[16] W.G. Dauben, B.A. Kowalczyk, and F.W. Lich-
tenthaler, J. Org. Chem., 55 (1990) 2391±2398.
[17] M. Brehm, W.G. Dauben, P. Kohler, and F.W.
Lichtenthaler, Angew. Chem., 99 (1978) 1318±1319.
[32] J. Kovacs, I. Pinter, D. Abeln, J. Kopf, and P.
Koll, Carbohydr. Res., 257 (1994) 97±106.
[33] P. Koll, W. Saak, S. Pohl, B. Steiner, and M.
Koos, Carbohydr. Res., 254 (1994) 183±194.
[34] P.L. Barili, G. Berti, G. Catelani, F. D'Andrea,
and L. Miarelli, Carbohydr. Res., 274 (1995) 197±
208.