D-Tagatose derivatives from D-fructose by a facile epimerisation procedure

Article abstract of DOI:10.1016/S0008-6215(99)00090-7

1,2-O-Isopropylidene-β-D-fructopyranose was directly converted into 5-O-cyclohexylcarbamoyl-1,2-O-isopropylidene-3,4-O-(2,2,2-trichloroethylidene)-β-D-tagatopyranose by treatment with chloral/N,N'-dicyclohexylcarbodiimide. Subsequent acid-catalysed cleava

Full text of DOI:10.1016/S0008-6215(99)00090-7

www.elsevier.nl/locate/carres
Carbohydrate Research 318 (1999) 167–170
Note
D
-Tagatose derivatives from
D
-fructose by a facile
epimerisation procedureꢀ
Michael Frank, Ralf Miethchen *, Daniela Degenring
Department of Organic Chemistry, Uni6ersity of Rostock, Buchbinderstraße 9, D-18051 Rostock, Germany
Received 9 March 1999; revised 8 April 1999; accepted 9 April 1999
Abstract
1,2-O-Isopropylidene-b-
dene-3,4-O-(2,2,2-trichloroethylidene)-b-
Subsequent acid-catalysed cleavage of the isopropylidene protecting group followed by acetylation afforded,
-tagatopyranose. This
D
-fructopyranose was directly converted into 5-O-cyclohexylcarbamoyl-1,2-O-isopropyli-
D
-tagatopyranose by treatment with chloral/N,N%-dicyclohexylcarbodiimide.
exclusively, 1,2-di-O-acetyl-5-O-cyclohexylcarbamoyl-3,4-O-(2,2,2-trichloroethylidene)-a-
D
product was simultaneously dehydrochlorinated and decarbamoylated to 1,2-di-O-acetyl-3,4-O-ethylidene-a-
topyranose using Bu3SnH/AIBN. © 1999 Elsevier Science Ltd. All rights reserved.
D-taga-
Keywords: Carbohydrates;
D-Fructose; D-Tagatose; Epimerisation; Chloral acetals
tagatose can be easily prepared from
D
-fruc-
The convenient one-pot methodology devel-
oped for the epimerisation of aldopyranoses
using chloral [2] also turned out to be valuable
for epimerisations of cyclitols [1,3]. Generally,
highly carbonyl-active aldehydes like chloral
react in the presence of the co-reagent N,N%-
dicyclohexylcarbodiimide (DCC) with polyols
having a cis–trans sequence of three contigu-
ous hydroxyl groups. Thus, cyclic acetals are
formed with simultaneous inversion of the
configuration at the middle chiral C-atom.
The mechanism of this non-conventional ac-
etalation/epimerisation procedure was de-
scribed in previous reports [1,2]. In this paper,
tose. Moreover, it is demonstrated how some
useful tagatopyranose derivatives are accessi-
ble by variation of the protecting groups.
1,2-O-Isopropylidene-b- -fructopyranose
D
(1) is a suitable starting material for the reac-
tion with chloral/DCC. The compound has
the required cis–trans sequence of three OH
groups and it can be prepared in relatively
large amounts from
steps via the corresponding 1,2:4,5-di-O-iso-
propylidene-b- -fructopyranose [4]; for the
preparation of 1 see also Ref. [5].
In order to convert the -fructose derivative
1 into a -tagatose derivative by inversion of
D-fructose in only two
D
D
D
we show that the rare monosaccharide
D-
the configuration at C-4, compound 1 was
heated for 4–5 h with chloral/DCC in 1,2-
dichloroethane. The 5-O-cyclohexylcarbam-
oyl - 1,2 - O - isopropylidene - 3,4 - O - (2,2,2 - tri-
ꢀ Epimerisation of carbohydrates and cyclitols, Part 16.
For Part 15, see Ref. [1].
* Corresponding author. Tel.: +49-381-498-1764; fax: +
49-381-498-1819.
E-mail address: ralf.miethchen@chemie.uni-rostock.de (R.
Miethchen)
chloroethylidene)-b-
D
-tagatopyranose (2) was
isolated in a yield of 59% after column chro-
0008-6215/99/$ - see front matter © 1999 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(99)00090-7

168
M. Frank et al. / Carbohydrate Research 318 (1999) 167–170
matography (Scheme 1). Because of the new
stereogenic centre of the trichloroethylidene
moiety, a diastereomeric mixture (endo-H/
exo-H form 30:1) was formed; the pure endo-
H
diastereomer was obtained by re-
crystallisation of the diastereomeric mixture
from ethyl acetate or by HPLC fractionation.
The structure of 2 (endo-H form) is sup-
ported by characteristic H and 13C NMR
1
signals for the carbamoyl and chloral acetal
function. The chemical shifts of H-3 (l 4.40
ppm) and H-4 (l 4.55 ppm) show the presence
of an acetal at these positions. Furthermore,
the coupling constants between H-3/H-4 and
H-4/H-5 are in agreement with the configura-
tion shown. The singlets of the trichloroethyli-
dene acetal protons were used to determine
the ratio of the endo-H/exo-H diastereomers.
As already reported in previous papers (e.g.,
Refs. [1–3]) the acetal proton of the endo-H
form (l 5.54 ppm) is always downfield shifted
compared with that of the exo-H form (l 5.31
ppm).
Scheme 1. (i) Chloral/DCC (ClCH2CH2Cl); (ii) TFA/H2O;
(iii) Ac2O/pyridine; (iv) Bu3SnH/AIBN (toluene).
i.e., a conformational interconversion from
5
5
2C5 (2) to C2 (4) has occurred. The C2 con-
formation should be thermodynamically fa-
voured, since the anomeric effect works
effectively together with the favourable equa-
torial orientation of the C-1 acetoxymethyl
group. In compound 2 (2C5) this argument is
likewise fulfilled.
The -tagatose derivative 2 was deprotected
D
stepwise: at first the carbamoyl group by heat-
ing in methanolic sodium methoxide and then
the acid-stable trichloroethylidene function by
dehydrochlorination to an ethylidene acetal
followed by acid-catalysed deacetalation. In
order to avoid any generation of furanose
derivatives, the sequence described in Scheme
1 was realised. An anomeric mixture of 3 was
generated by treatment of 2 with aqueous
trifluoroacetic acid (TFA). Further conven-
tional acetylation of 3 afforded 1,2-di-O-
acetyl-5-O-cyclohexylcarbamoyl-3,4-O-(2,2,2-
In order to confirm the a configuration of
compound 4, a NOE measurement was car-
ried out, which showed the correlation of the
axially arranged H-6 to both CH3 groups of
the acetyl functions (Scheme 2). This result fits
only with the a anomer. As an additional
indication, no interaction was found between
H-4 and the two exocyclic H-1 and H-1%.
Furthermore, only a very weak correlation
was observed between H-3 and H-1/H-1%. Fi-
nally, the correlation between the acetal-H
trichloroethylidene)-a- -tagatopyranose (4) in
D
a yield of 93%. No traces of the b anomer
were detected after a reaction time of 48 h
under standard conditions [6]. It is noticeable
that two intermediates, probably monoacetyl
derivatives, could be detected by TLC before
completion. Comparison of the 1H NMR
spectra of 2 and 4 shows that several coupling
constants are significantly different. Thus, the
coupling constants of compound 2 (J4,5 5.5,
J5,6a 1.5, and J5,6b 2.1 Hz) indicate a trans-
diequatorial arrangement of H-4 and H-5,
whereas the corresponding data of compound
4 (J4,5 7.3, J5,6ax 10.4, J5,6eq 5.8 Hz) indicate a
trans-diaxial arrangement of these protons,
Scheme 2. NOE experiment: correlation illustrated in a
molecule fragment of 1,2-di-O-acetyl-5-O-cyclohexylcarb-
amoyl-3,4-O-(2,2,2-trichloroethylidene)-a-
(4).
D-tagatopyranose

M. Frank et al. / Carbohydrate Research 318 (1999) 167–170
169
N,N%-dicyclohexyl urea was removed by filtra-
tion, the organic phase was separated, and the
aqueous phase was washed with CH2Cl2 (2×
15 mL). The combined extracts were washed
with water (2×20 mL), dried (Na2SO4) and
concentrated under reduced pressure. The
residue was purified by column chromatogra-
phy (Rf 0.35, 5:1 heptane–EtOAc) giving 1.27
g (59.1%) of the diastereomeric 30:1 endo-H/
exo-H mixture of 2. The pure endo-H form
was obtained by recrystallisation from EtOAc
or by HPLC: 1.10 g (51.2%) of 2 as colourless
crystals; mp 198.5–201 °C; [h]2D5 −32.4° (c
and H-5 confirms the endo-H arrangement of
the cyclic acetal function in 4 (Scheme 2).
The acid-stable trichloroethylidene group of
4 was converted into an acid-labile ethylidene
acetal by a radical dehydrohalogenation using
Bu3SnH/AIBN [7]. On heating compound 4
with this reagent in toluene, the carbamoyl
group was removed as well (Scheme 1). How-
ever, for a complete decarbamoylation, an
excess of the hydride is essential. The 1,2-di-
O-acetyl-3,4-O-ethylidene-b- -tagatopyranose
D
(5) was isolated in a yield of 72%. The cou-
pling constants J3,4 5.5 Hz and J4,5 4.0 Hz of
compound 5 indicate an equilibrium between
1
0.98, CHCl3); H NMR (250 MHz, CDCl3): l
2
5
5.54 (s, 1 H, acetal-H), 5.09 (ddd, 1 H, J5,6a 1.5
Hz, J5,6b 2.1 Hz, H-5), 4.70 (d, 1 H, JNH,CH 7.9
the two conformations C5 and C2 (Scheme
1).
4
Hz, N–H), 4.55 (ddd, 1 H, J4,5 5.5 Hz, J4,6
1.8 Hz, H-4), 4.40 (d, 1 H, J3,4 5.2 Hz, H-3),
2
1. Experimental
4.20 (dd, 1 H, J6,6 13.4 Hz, H-6a), 4.10 (d, 1
2
H, J1a,1b 10.4 Hz, H-1a), 4.05 (d, 1 H, H-1b),
General.—Column chromatography: E.
Merck Silica Gel 60 (63–200 mm); thin-layer
chromatography (TLC): E. Merck Silica Gel
60 F254 foils; HPLC: Knauer equipment, Ver-
tex column B31-Y520 (Eurosher 100-15, 15
mm) 5:1 heptane–EtOAc, detection by refrac-
tive index. NMR: AC 250 and ARX 300;
internal standard TMS. Melting points were
measured using a Leitz polarising microscope
(Laborlux 12 Pol) equipped with a hot stage
(Mettler FP 90). 1,2-O-Isopropylidene-b-
fructopyranose (1) [4] was prepared from the
corresponding 1,2:4,5-di-O-isopropylidene
derivative [5] by selective acid hydrolysis fol-
lowing the procedure reported by Licht-
enthaler et al. [8] using a mixture of H2SO4 (1
M, 10 mL), MeOH (120 mL) and H2O (100
mL) at room temperature (rt).
3.68 (ddd, 1 H, H-6b), 3.53–3.36 (m, 1 H,
cyclohexyl-CH), 1.97–1.83 (m, 2 H, cyclo-
hexyl-CH2), 1.74–1.57 (m, 3 H, cyclohexyl-
CH2), 1.38–1.05 (m, 5 H, cyclohexyl-CH2),
1.51, 1.43 (2 s, 3 H, isopropyl-CH3). 13C NMR
(63 MHz, CDCl3): l 154.0 (NH–CꢀO), 113.4
(C(CH3)2), 107.8 (acetal-C), 102.5 (C-2), 99.2
(CCl3), 74.9, 72.4, 67.3 (C-3, C-4, C-5), 73.2
(C-1), 59.3 (C-6), 50.1 (cyclohexyl-CH), 33.3,
25.4, 25.4, 24.7, 24.7 (5 cyclohexyl-CH2), 27.3,
25.2 (2 C(CH3)2). Anal. Calcd for
C18H26Cl3NO7 (474.76): C, 45.54; H, 5.52; N,
2.95. Found: C, 45.52; H, 5.46; N, 2.96.
D-
(R)-5-O-Cyclohexylcarbamoyl-3,4-O-(2,2,-
2 -trichloroethylidene)-h,i- -tagatopyranose-
D
(3).—A solution of 2 (250 mg, 0.53 mmol) in
aq TFA (60% v/v, 10 mL) was stirred for 12 h
at rt. After evaporation of the solvents under
reduced pressure, the residue was twice co-dis-
tilled with toluene (2×10 mL) under reduced
pressure and subsequently purified by column
chromatography (Rf 0.35, 1:1 toluene–EtOAc)
yielding 215 mg (93.9%) of the syrupy
anomeric mixture 3 (a/b=45.3:54.7); 1H
NMR (250 MHz, CD3OD): l 5.60 (s, 1 H, a
acetal-H), 5.50 (s, 1.21 H, b acetal-H). 13C
NMR (63 MHz, CD3OD): l 156.9, 156.5 (a/b
NH–CꢀO), 109.1, 107.6 (a/b acetal-C), 100.9,
100.4 (a/b CCl3), 96.7, 96.4 (C-2), 79.5, 78.2,
76.2, 73.9, 69.3, 69.1 (a/b C-3, C-4, C-5), 66.4,
65.9 (a/b C-1), 58.8 58.7 (a/b C-6), 51.3, 51.2
(R)-5-O-Cyclohexylcarbamoyl-1,2-O-iso-
propylidene-3,4-O-(2,2,2-trichloroethylidene)-
i-
D
-tagatopyranose (2).—To a solution of
1,2-O-isopropylidene-b-
D
-fructopyranose (1)
[5] (1.0 g, 4.54 mmol) in dry 1,2-
dichloroethane (15 mL), chloral (2.34 g, 15.89
mmol) and DCC (2.35 g, 11.35 mmol) were
sequentially added. Subsequently, the mixture
was refluxed under stirring for 4–5 h (TLC
control). After cooling to rt and addition of
CH2Cl2 (20 mL) and 10% aq AcOH (30 mL),
the reaction mixture was shaken for about 30
min to destroy excess DCC. The precipitated

170
M. Frank et al. / Carbohydrate Research 318 (1999) 167–170
phase was separated, washed twice with water
(15 mL), dried (Na2SO4), and concentrated
under reduced pressure. The residue was
purified by column chromatography (Rf 0.17,
2:1 toluene–EtOAc) giving 1.05 g (72.4%) of 5
as a colourless syrup, [h]2D2 +38.2° (c 1.34,
(a/b cyclohexyl-CH), 33.9, 33.8, 26.4, 25.9, 25.7,
25.2 (a/b cyclohexyl-CH2). Anal. Calcd for
C15H22Cl3NO7 (434.70): C, 41.45; H, 5.10; N,
3.22. Found: C, 41.11; H, 5.33; N, 3.06.
(R)-1,2-Di-O-acetyl-5-O-cyclohexylcarba-
moyl-3,4-O-(2,2,2-trichloroethylidene)-h- -ta-
D
1
CHCl3); H NMR (250 MHz, CDCl3): l 5.08
gatopyranose (4).—The a/b anomeric mixture
of 3 (100 mg, 0.23 mmol) was treated with a
mixture of Ac2O (5 mL) and pyridine (5 mL)
under stirring at rt. First two monoacetylation
products were detectable by TLC (Rf 0.08 and
0.185, 5:1 toluene–EtOAc), the diacetylation
reaction being completed after 48 h (Rf of 4:
0.30). After evaporation of the solvents under
reduced pressure, the residue was co-distilled
withtoluene(2×10mL)andpurifiedbycolumn
chromatography with the above solvent mixture
to give 102 mg (93.2%) of4 as colourless crystals,
mp 196–198 °C (EtOAc); [h]2D5 −20.80° (c 0.95,
CHCl32); 1H NMR (300 MHz, C6D6): l 5.24 (d,
1 H, J1a,1b 11.7 Hz, H-1a), 5.16 (s, 1 H,
acetal-H), 5.10 (m, 1 H, J5,6ax 10.4 Hz, J5,6eq 5.8
Hz, H-5), 4.90 (d, 1 H, J3,4 5.4 Hz, H-3), 4.85
(d, 1 H, H-1b), 4.74 (dd, 1 H, J4,5 7.4 Hz, H-4),
4.12 (d, 1 H, JNH,CH 7.9 Hz, N–H), 4.09 (dd,
1 H, 2J6a,6b 11.3 Hz, H-6e), 3.48 (dd, 1 H, H-6a),
3.46 (m, 1 H, cyclohexyl-CH), 1.80–1.65 (m, 2
H, cyclohexyl-CH2), 1.70, 1.63 (2 s, 6 H,
C(O)CH3), 1.46–1.28 (m, 3 H, cyclohexyl-CH2),
1.09–0.65 (m, 5 H, cyclohexyl-CH2). 13C NMR
(63 MHz, CDCl3): l 169.6, 168.1 (2 C(O)CH3),
154.0 (NH–CꢀO), (106.4 (C-CCl3), 101.1 (C-2),
98.8 (CCl3), 77.6 (C-4), 75.2 (C-3), 67.2 (C-5),
62.5 (C-1), 60.2 (C-6), 50.1 (cyclohexyl-CH),
33.3, 25.4, 25.4, 24.7, 24.7 (5 cyclohexyl-CH2),
21.7, 20.5 (2 C(O)CH3). Anal. Calcd for
C19H26Cl3NO9 (518.77): C, 43.99; H, 5.05; N,
2.70. Found: C, 44.09; H, 5.08; N, 2.77.
(q, 1 H, Jacetal-H,acetal-CH 4.8 Hz, acetal-H), 4.81
3
(dd, 1 H, J4,5 4.0 Hz, H-4), 4.41 (d, 1 H, J3,4 5.5
Hz, H-3), 4.41 (dd, 1 H, 2J6a,6b 11.1 Hz, H-6a),
2
4.35 (d, 1 H, J1a,1b 11.9 Hz, H-1a), 4.34 (ddd,
1 H, J5,6a 4.1 Hz, J5,6b 6.9 Hz, H-5), 4.23 (d, 1
H, H-1b), 4.16 (dd, 1 H, H-6b), 3.29 (br, 1 H,
OH), 2.11, 2.07 (2 s, 6 H, C(O)CH3), 1.31 (d,
3 H, acetal-CH3). 13C NMR (63 MHz, CDCl3):
l 171.0, 170.7 (2 C(O)CH3), 103.7 (acetal-C),
103.6 (C-2), 84.2, 80.0, 78.1 (C-3, C-4, C-5), 64.7,
62.0 (C-1/C-6), 20.8, 20.8 (2 C(O)CH3), 19.6
(acetal-CH3). Anal. Calcd for C12H18O8
(290.27): C, 49.65; H, 6.25. Found: C, 49.51; H,
6.39.
Acknowledgements
We are grateful to the Deutsche Forschungs-
gemeinschaft and the Fonds der Chemischen
Industrie for financial support.
References
[1] M. Frank, R. Miethchen, H. Reinke, Eur. J. Org. Chem.,
(1999) 1259–1263.
[2] (a) R. Miethchen, D. Rentsch, Liebigs Ann. Chem., (1994)
1191–1197. (b) R. Miethchen, D. Rentsch, M. Frank, J.
Carbohydr. Chem., 15 (1996) 15–31. (c) C. Zur, A.O.
Miller, R. Miethchen, J. Fluorine Chem., 90 (1998) 67–76.
[3] M. Frank, R. Miethchen, Carbohydr. Res., 313 (1998)
49–53.
[4] T.C. Irvine, C.S. Garrett, J. Chem. Soc., 97 (1910) 1277–
1284.
(R)-1,2-Di-O-acetyl-3,4-O-ethylidene-h- -
D
[5] (a) R.F. Brady Jr., Carbohydr. Res., 15 (1970) 35–40. (b)
T. Bieg, S. Wieslaw, Synthesis, 1 (1985) 76–77. (c) C.-K.
Lee, Carbohydr. Res., 170 (1987) 255–262. (d) J. Kang,
G.J. Lim, S.K. Yoon, M.Y. Kim, J. Org. Chem., 60 (1995)
564–577. (e) J.M. Garc´ıa Ferna´ndez, C. Ortiz Mellet, A.
Moreno Mar´ın, J. Fuentes, Carbohydr. Res., 274 (1995)
263–268.
[6] (a) T.W. Greene, P.G.M. Wuts, Protecti6e Groups in
Organic Synthesis, vol. 2, Wiley, New York, 1990. (b) P.J.
Kocienski, Protecting Groups, Thieme, Stuttgart, 1994.
[7] (a) S. Jacobsen, F. Sløk, Acta Chem. Scand., 47 (1993)
1012–1018. (b) R.K. Freidlina, R.G. Gasanov, N.
Kuzmina, E.T. Chukovskaya, Russ. Chem. Re6., 54 (1985)
662–675. (c) W.P. Neumann, Synthesis, (1987) 665–683.
[8] F.W. Lichtenthaler, S. Hahn, F.-J. Flath, Liebigs Ann.
Chem., (1995) 2081–2088.
tagatopyranose (5).—A solution of 4 (2.6 g, 5.01
mmol), Bu3SnH (5.25 g, 18.04 mmol) and AIBN
(100 mg) in dry toluene (35 mL) was heated at
80 °C under stirring (Ar atmosphere). After 6
h, Bu3SnH (1.46 g, 5.01 mmol) and AIBN (30
mg) was added and heating was continued for
further 6 h. An intermediate (Rf 0.48, 2:1
toluene–EtOAc) detected during this period
had now disappeared. After cooling down, the
solution was shaken with a saturated aq KF (30
mL) for 30 min and the precipitated Bu3SnF was
removed by filtration. Subsequently, the organic