Behavior of free sugar thiosemicarbazones toward heterocyclization reactions

Article abstract of DOI:10.1016/S0008-6215(00)00127-0

The heterocyclization reaction on thiosemicarbazones having the D-galacto, D-gluco and D-manno configuration was studied. We applied two different acetylating conditions, and the reaction products obtained were identified, spectroscopically characterized, and conformationally analyzed. Using experimental data, we discuss a possible mechanistic pathway for heterocyclization and evaluate the influence of several factors, including starting material configuration, pH of reaction medium, and reaction time. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0008-6215(00)00127-0

www.elsevier.nl/locate/carres
Carbohydrate Research 328 (2000) 481–488
Behavior of free sugar thiosemicarbazones toward
heterocyclization reactions
Miriam A. Martins Alho, Norma B. D’Accorso *
CIHIDECAR (CONICET)-Departamento de Qu´ımica Orga´nica, Facultad de Ciencias Exactas y Naturales,
Uni6ersidad de Buenos Aires, 3° Piso, Pabello´n II-Ciudad Uni6ersitaria, CP 1428 Buenos Aires, Argentina
Received 30 November 1999; accepted 26 April 2000
Abstract
The heterocyclization reaction on thiosemicarbazones having the
D
-galacto, D-gluco and D-manno configuration
was studied. We applied two different acetylating conditions, and the reaction products obtained were identified,
spectroscopically characterized, and conformationally analyzed. Using experimental data, we discuss a possible
mechanistic pathway for heterocyclization and evaluate the influence of several factors, including starting material
configuration, pH of reaction medium, and reaction time. © 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Sugar thiadiazolines; Sugar thiosemicarbazones; Acetylation; Heterocyclization; pH influence
1. Introduction
were isolated and characterized in each case.
Using the experimental data, we attempted a
mechanistic explanation for our observations.
The synthesis of (+)- and (−)-5-acet-
amido-3-N-acetyl-2-(1,2,3,4,5-penta-O-acetyl-
D
- galacto - pentitol - 1 - yl) - 1,3,4 - thiadiazoline
[(+)-1 and (−)-1] from penta-O-acetyl-alde-
hydo- -galactose thiosemicarbazone under
2. Results
D
two different acetylating conditions has been
described by Somogyi [1]. In a further publica-
tion, El Ashry et al. [2] reported the synthesis
of the dextrorotatory isomer by direct acetyla-
For our purposes we applied two different
acetylating procedures. Procedure A used an
acetylating mixture of acetic anhydride and
pyridine at room temperature during 24 h,
while Procedure B was carried out by heating
the thiosemicarbazones with boiling acetic
anhydride.
tion of
our attempts to synthesize the analogous com-
pounds from -glucose or -mannose
D
-galactose thiosemicarbazone (2). All
D
D
thiosemicarbazone (3 and 4) using this method
failed, and in addition, the results for com-
pound 2 itself were not reproducible. Accord-
ing to these observations, we decided to study
the reaction products of 2, 3 and 4 under
different acylation conditions. The products
By carrying out Procedure A on compound
2, we could isolate, in several cases, (+)-1 as
a main product, a small proportion of (−)-1,
as well as acetylated pyranose structures,
which were separated and characterized as
4-N-acetyl-1-N-(2,3,4,6-tetra-O-acetyl-b- -
D
galactopyranosyl)thiosemicarbazide (5), and
1,4-N,N-diacetyl-1-N-(2,3,4,6-tetra-O-acetyl-
* Corresponding author. Tel./fax: +54-11-45763346.
E-mail address: norma@qo.fcen.uba.ar (N.B. D’Accorso).
b- -galactopyranosyl)thiosemicarbazide (6).
D
0008-6215/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(00)00127-0

482
M.A. Martins Alho, N.B. D’Accorso / Carbohydrate Research 328 (2000) 481–488
In addition, we obtained spectroscopic evi-
As it was observed for
D
-glucose-4-tolylth-
dence of 1-N-acetyl-1-N-(2,3,4,6-tetra-O-ace-
iosemicarbazone [3], thiosemicarbazones 3
and 4 are b-pyranose structures, even in
aqueous solution, we could not synthesize
them in any other form than these cyclic
structures (determined by 13C NMR, see Sec-
tion 4). As a consequence of that, application
of Procedure A on these compounds allowed
us to isolate and characterize only some acety-
lated pyranose products, but no thiadiazoline
derivatives. These pyranose products were
tyl - b -
D
- galactopyranosyl)thiosemicarbazide
(7). In other cases, when we applied the same
procedure on another batch of 2, no hetero-
cyclic compounds were found, and the main
isolated products were 5 and 6.
An explanation of this behavior was given
by the 13C NMR spectra of different batches
of compound 2, which showed that it can
crystallized as a single pyranose form or as a
mixture of this pyranose and an open-chain
structure, according to its reaction and crys-
tallization conditions, in a variable ratio (see
Section 4). When the CꢀN bond was not
present in the starting material, thiadiazoline
derivatives could not be obtained by this
procedure.
1-N-acetyl-1-N-(2,3,4,6-tetra-O-acetyl-b-
copyranosyl)thiosemicarbazide (8), 4-N-ace-
tyl-1-N-(2,3,4,6-tetra-O-acetyl-b- -glucopyra-
nosyl)thiosemicarbazide (9), 1,4-N,N-diacetyl-
1-N-(2,3,4,6-tetra -O-acetyl-b- -glucopyrano-
syl)thiosemicarbazide (10), and 1-N-acetyl-1-
-mannopyra-
D-glu-
D
D
N-(2,3,4,6-tetra-O-acetyl-b-
D
nosyl)thiosemicarbazide (11). These com-
pounds are summarized in Scheme 1. 1H
NMR data for compounds 5–11 are listed in
13
Table 1, and C NMR data assignments were
made using model compounds (Table 2)1.
When we applied Procedure B on com-
pound 2 (in the pure pyranose form), we
obtained a reproducible result: the main com-
pound was (−)-1. Other isolated products
were penta-O-acetyl-b- -galactopyranose and
D
2-acetamido-5-methyl-1,3,4-thiadiazole. In ad-
dition we had spectroscopic evidence of the
formation of some acetylated derivatives of
(galactopyranosyl)hydrazine by losses of the
thiocarbamoyl moiety. In the literature there
is a similar degradation reported for penta-O-
acetyl-aldehydo-
D
-galactose
semicarbazone,
which in presence of acetic anhydride and
pyridine loses a carbamoyl group, affording
the corresponding N,N-diacetylhydrazine [4].
Application of Procedure B to 3 gave the
same result as that observed in the case of
compound 2: only one thiadiazoline, (−)-5-
acetamido-3-N-acetyl-2-[1,2,3,4,5-penta-O-
acetyl- -gluco-pentitol-1-yl]-1,3,4-thiadiazo-
D
line [(−)-12], peracetylated glucopyranose,
and the above-mentioned thiadiazole with
other degradation products. As expected, Pro-
cedure B applied to compound 4 gave both
possible thiadiazolines, (+)- and (−)-5-
1 For display of the NMR data, the carbohydrate part is
shown with primed numbers.
Scheme 1.

M.A. Martins Alho, N.B. D’Accorso / Carbohydrate Research 328 (2000) 481–488
483
Table 1
1H NMR data for compounds 5, 6 and 8–11
Chemical shifts (l, ppm)
Compounds
5
6
8
9
10
5.95
4.80
5.03
5.33
3.76
4.16
4.13
11
H-1%
5.14
5.15
4.52
5.43
3.89
4.21
4.04
6.41
7.15
5.89
5.07
5.15
5.32
3.99
3.99
3.99
5.84
5.03
5.39
5.04
3.84
4.36
4.08
4.45
4.90
5.22
4.97
3.58
4.19
4.00
6.39
7.27
4.68
H-2%
H-3%
H-4%
H-5%
H-6%a
H-6%b
NH-1
NH-2
NH-4
5.61
5.15
5.25
3.68
4.31
4.10
6.26
7.67
10.17
8.90
8.21
9.21
10.10
12.10
6.64, 6.28
10.07
12.06
Coupling constants (J, Hz)
J1%,2%
J2%,3%
J3%,4%
9.0
10.2
2.2
8.8
10.2
3.1
9.1
9.3
9.5
8.8
9.1
9.3
9.5
9.6
9.6
B1
2.8
9.8
J4%,5%
1.0
7.0
6.3
11.1
B1
10.1
2.2
4.5
9.8
5.4
2.5
12.3
9.5
6.7
5.4
12.1
9.0
6.3
2.5
12.1
J5%,6%a
J5%,6%b
J6%a,6%b
12.6
Table 2
13C NMR data for compounds 5, 6 and 8–11
Chemical shifts (l, ppm)
Compounds
5
6
8
9
10
11
C-1%
C-2%
C-3%
C-4%
C-5%
C-6%
C-3 (CꢀS)
90.3
68.1
71.0
67.1
71.5
61.1
81.1
66.7a
70.5
67.0a
72.4
61.0
80.7
67.6
74.1
68.4
72.4
61.0
89.7
70.7
80.4
88.5
67.4
73.7
69.3
73.7
62.2
182.3
68.0
70.7
66.4
73.4
62.3
184.4
72.6a
68.5
72.8a
61.7
184.3
184.4
183.5
184.9
a These values may be exchanged.
firmed our previous assignments using model
compounds.
acetamido-3-N-acetyl-2-[1,2,3,4,5-penta-O-
acetyl- -manno-pentitol-1-yl]-1,3,4-thiadiazo-
D
Thiadiazolines derived from
and
(−)-1, (+)-13 and (−)-13] showed a planar
zigzag extended chain as the main conforma-
D
-galactose
-mannose thiosemicarbazone [(+)-1,
line [(+)-13 and (−)-13], in addition to per-
acetylated mannopyranose, 2-acetamido-5-
methyl-1,3,4-thiadiazole and degradation pro-
ducts analogous to those observed for com-
D
1
1
pound 2 and 3. H and 13C NMR data for
tion in solution, as was deduced using the H
compounds (+)-1, (−)-1, (+)-12, (−)-12,
(+)-13 and (−)-13 are listed in Tables 3 and
4. Assignments of 13C NMR signals of com-
pounds (+)-1 and (−)-1 were made by het-
eronuclear COSY experiments, which con-
NMR first-order analysis coupling constant.
Using the same procedure we determined that
(−)-12 showed a deviation from the planar
zigzag extended chain, which corresponded to
1G- using the convention of Angyal et al. [5].

484
M.A. Martins Alho, N.B. D’Accorso / Carbohydrate Research 328 (2000) 481–488
The assigned conformations for these com-
pounds are shown in Fig. 1. Despite the fact
that all spectra were very clear, the absolute
configuration of the heterocyclic C-2 could
not be assigned on the basis of potential inter-
actions between the heterocyclic ring and the
Table 3
1H NMR data for compounds (+)-1, (−)-1, (−)-12, (+)-13 and (−)-13
Chemical shifts (l, ppm)
Compounds
(+)-1
(−)-1
(−)-12
(+)-13
(−)-13
H-2
H-1%
H-2%
H-3%
5.79
4.99
5.39
5.29
5.25
4.19
3.77
8.14
6.06
5.63
5.43
5.29
5.22
4.21
3.81
8.43
5.97
5.52
5.35
5.26
4.96
4.13
3.98
8.91
5.94
5.52
5.37
5.44
5.11
4.20
4.00
8.27
6.09
5.30
5.70
5.41
4.97
4.22
4.03
9.49
H-4%
H-5%a
H-5%b
ꢁNHꢁCOCH3
Coupling constants (J, Hz)
J2,1%
J1%,2%
9.8
1.4
1.6
1.9
2.3
7.3
1.8
9.2
5.6
9.1
J2%,3%
9.9
9.6
3.4
2.0
1.8
J3%,4%
1.9
1.9
7.9
9.4
8.9
J4%,5%a
J4%,5%b
J5%a,5%b
4.8
7.5
11.6
4.9
7.1
11.6
4.3
3.4
12.5
4.9
2.7
12.5
4.9
2.7
12.5
Table 4
13C NMR data for compounds (+)-1, (−)-1, (−)-12, (+)-13 and (−)-13
Chemical shifts (l, ppm)
Compounds
(+)-1
(−)-1
(−)-12
(+)-13
(−)-13
C-2
C-5 (CꢀN)
C-1%
C-2%
C-3%
62.2
148.2
70.6
66.4
67.4
67.4
62.0
67.2
147.3
67.5
67.7
68.0
68.1
62.1
65.1
146.5
67.8a
70.5
66.9
147.9
67.3a
67.6a
67.3a
67.9a
61.9
64.7
147.6
68.8
67.8a
67.7a
68.1a
61.5
68.0a
68.3
C-4%
C-5%
61.2
a These values may be exchanged.
Fig. 1. Conformation of acyclic-sugar derivatives (9)-1, (−)-12 and (9)-13.

M.A. Martins Alho, N.B. D’Accorso / Carbohydrate Research 328 (2000) 481–488
485
carbohydrate chain. All efforts to obtain crys-
tals for X-ray diffraction were unsuccessful.
Thus this configuration could not be assigned.
However, we carried out molecular calcula-
tions which led us to formulate an hypothesis
about their absolute configuration (see Section
3).
can be concluded that the inversion of the
configuration on C-2 has a strong influence on
the products obtained.
When we studied the consequences of heat-
ing time on compounds 3 and 4, we found
that for compound 3, the results of a longer
reaction time was an increase in the yield of
heterocyclic products. Meanwhile for com-
pound 4, this increase in yield was parallel to
a loss of (−)-13. This fact could indicate that
an excess of energy could effect a conversion
of one isomer into the other. Thus (−)-13
could be a kinetic product that can be con-
verted into the thermodynamic (+)-13.
All acetylated pyranose structures showed
4
the C1 as the main conformation in solution
1
that was determined by H NMR first-order
analysis of the coupling constants, selective
1
irradiation, as well as by H NMR spectral
simulations [6].
Taking in account that (+)-1 was synthe-
sized by the lowest energy procedure, we as-
sume that (+)-1 could be the kinetic product,
and (−)-1 the thermodynamic one. In order
to estimate the conversion time between two
isomers, we applied Procedure B on pure (+
3. Discussion
Comparing the behavior of 2, 3 and 4 to-
ward Procedure A, we can conclude that hete-
rocyclization occurs only if the starting
material has a CꢀN bond. When we studied
the same thiosemicarbazones with Procedure
B, we found that even in the case where the
substrate has a single pyranosidic form, thia-
diazolines could be obtained.
1
)-1 and found using H NMR that at 30 min
both thiadiazolines were present. So we inves-
tigated the reaction product of thiosemicarba-
zone 2 by Procedure B at 30 min, and we
found, that even when the reaction was very
incomplete, the two corresponding thiadiazoli-
nes were present in a 1:1 ratio.
Summing up these experimental data we
can conclude the following:
1. The cyclization of free sugar thiosemicar-
bazones can be made in either acidic or
basic media.
2. If the reaction is carried out in basic me-
dia, the thiadiazolines can be obtained
only if the CꢀN is present, but not with the
cyclic form of thiosemicarbazones.
3. If the reaction is carried out in acidic
media, these conditions promote pyranose
ring opening, and the cyclation yields both
thiadiazolines.
4. One of these thiadiazolines might be a
kinetic product and the other the thermo-
dynamic one.
5. Isolation of both thiadiazolines depends on
the carbohydrate under reaction and the
interconversion speed between the kinetic
and thermodynamic products.
The main differences between the two meth-
ods are the pH and the temperature. Taking in
account that a pyranosidic form is a hemiac-
etal, acidic media in Procedure B can promote
pyranose ring opening, which allows the hete-
rocyclization reaction to proceed. Obtaining
2-acetamido-5-methyl-1,3,4-thiadiazole under
these conditions is a direct consequence of the
loss of the nitrogen-containing part. A similar
loss was described in the transglycosylation of
glycosylamines [7], and it occurs probably by
formation of
D-hexopyranosylium ion. Ac-
cording to this theory, the possibility of ob-
taining the a- and the b-per-O-acetyl
hexopyranose from that intermediate is the
same, but we could isolate only the b anomer.
When we applied Procedure B to compound
6, using GC with corresponding standards, we
found that both anomers were formed. Isola-
tion of only one of them may be a direct
consequence of separative procedures, which
includes further crystallization that may favor
the isolation of a single annomer.
As we previously noted, the absolute
configuration of the heterocyclic C-2 in thiadi-
azoline derivatives could not be determined.
Attempting to build some theoretical supposi-
As mentioned above, Procedure B carried
out on 2 and 3 gave only one of the two
possible thiadiazolines, but 4 gave both, and it

486
M.A. Martins Alho, N.B. D’Accorso / Carbohydrate Research 328 (2000) 481–488
R configuration on C-2 might be attributed to
compound (+)-1.
Similar calculations performed with the S
isomer show that rotamers A and B have an
energetic difference of only 0.5 kcal/mol, while
the difference with rotamer C is near 3 kcal/
mol, and the configuration on C-2 of com-
pound (−)-1 which has a J2,1% of 1.6 Hz, could
be assigned to the S isomer.
4. Experimental
General methods.—1H and 13C NMR spec-
tra were recorded at 200 or 300 MHz and 50
or 75 MHz, respectively, in CDCl3 with Me4Si
as internal standard. Mass spectra were per-
formed with a Shimadzu QP-5000 by electron-
impact ionization and fast-atom bombard-
ment mass spectra (FABMS) in a ZAB-
SEQ4F. Optical rotations were recorded at
20 °C, and the melting points are uncorrected.
Elemental analyses were performed at the
UMYMFOR, Facultad de Ciencias Exactas y
Naturales, University of Buenos Aires, Buenos
Aires, Argentina. Chromatographic purifica-
tion was performed on Silica Gel G using
mixtures of benzene and EtOAc as the eluent.
Procedure A.—Aldose thiosemicarbazone (1
g, 3.95 mmol) was suspended in 125 mL of
pyridine, and 25 mL of Ac2O were added. The
mixture was stirred for 24 h at rt, at the end of
which time the reaction was stopped by addi-
tion of EtOH. The mixture was concd under
diminished pressure to a thick syrup, with
occasional addition of water and toluene to
eliminate all AcOH residues.
By application of this procedure to com-
pounds 2, 3 and 4 we could isolate, by chro-
matographic purification, compounds (9)-1,
5, 6, 8, 9, 10, 11, and obtained spectroscopic
evidence for 7.
Procedure B.—Aldose thiosemicarbazone (1
g, 3.95 mmol) was suspended in 44 mL of
Ac2O, and the mixture was stirred at reflux
using a sand bath during a variable time
period. The reaction was stopped by addition
of EtOH. The mixture was concd under dimin-
ished pressure to a thick syrup, with occasional
addition of water and toluene to eliminate
all AcOH residues. The syrup was dissolved
in EtOH and filtered to eliminate the
Fig. 2. Rotamers for (2R)-1 and (2S)-1.
tions and applying an approximation used
before [8], we calculated the lowest energy
conformation for both isomers of compound 1
using molecular mechanics and semiempirical
methods [9]. As the result of both calculations,
we came to the same conclusion: the lowest
energy conformation for the 2R and 2S iso-
mers of compound 1 show a gauche relation-
ship between heterocyclic H-2 and H-1%.
According to this calculation, they might have
a small coupling constant, but (+)-1 has a J2,1%
value of 9.8 Hz that indicates an anti relation-
ship. As the observed J2,1% is an average be-
tween coupling constants of closer rotamers
(in regard to the energetic level), we must to
suppose that there is a greater contribution to
the rotameric equilibrium of an anti form,
which could not be avoided.
In regard to that, we used PCMODEL™ [9],
which can calculate a theoretical coupling con-
stant for each pair of protons, and we per-
formed the minimization of three rotamers for
2R and 2S isomer of 1. We arrived at the
following conclusions:
For the R isomer, rotamer B is the lowest
energy structure (see Fig. 2), but the difference
of energy with rotamer C is about 1 kcal/mol,
while the difference with rotamer A is about 4
kcal/mol. Thus rotamers B and C must make
the major contribution to the equilibrium mix-
ture. Taking into account that rotamers B and
C have a J2,1% bigger than (−)-1 (1.6 Hz), the

M.A. Martins Alho, N.B. D’Accorso / Carbohydrate Research 328 (2000) 481–488
487
2-acetamido-5-methyl-1,3,4-thiadiazole. It was
then concd, and the residue was purified by
column chromatography.
compound 2, but they could only be obtained
in the pyranose form.
D-Glucose
thiosemicarbazone
(3).—13C
By application of this procedure to com-
pounds 2–4 we could isolate compounds (−)-
1, (−)-12, (+)-13, (−)-13, 2-acetamido-5-
methyl-1,3,4-thiadiazole, and the correspond-
NMR (Me2SO-d6): l 181.5 (CꢀS), 90.9 (C-1),
77.7 (C-3 or C-5), 77.2 (C-3 or C-5), 71.5
(C-2), 70.8 (C-4), 62.4 (C-6).
D
-Mannose thiosemicarbazone (4).—13C
ing penta-O-acetyl-
(+) - 5 - Acetamido - 3 - N - acetyl - 2 - [1,2,3,-
4,5-penta-O-acetyl- -galacto-pentitol-1-yl]-
D
-hexopyranose.
NMR (Me2SO-d6): l 181.4 (CꢀS), 87.7 (C-1),
78.3 (C-5), 74.4 (C-3), 70.2 (C-2), 68.2 (C-4),
62.8 (C-6).
D
1,3,4-thiadiazoline [(+)-1].—The yield of this
compound was variable in accord with the
proportion of the open-chain form. The better
yield was 1.23 g (57.2%); mp 210–212 °C. (lit.
4-N-Acetyl-1-N-(2,3,4,6-tetra-O-acetyl-i-
D
-
galactopyranosyl)thiosemicarbazide (5).—
Yield: 0.30 g (16.5%); mp 122–123 °C (sealed
capillary), [h]2D0 −73.6° (c 1, CHCl3), umax
=
1
1
217 °C [1], 214 °C [2]); H NMR (CDCl3): l
272 nm (EtOH); H NMR (CDCl3): l 2.45–
2.16–1.96 (7 s, 21 H, ꢁCH3); 13C NMR
(CDCl3): l 170.2–168.9 (4 s, CꢀO), 23.1–20.5
(4 s, CH3ꢁ); EIMS: m/z 547 (M+), 362, 331,
186, 144, 102 (100%).
1.93 (5 s, 15 H, ꢁCH3); C NMR (CDCl3): l
13
178.9–170.0 (5 s, CꢀO), 25.7–20.6 (4 s,
ꢁCH3); FABMS (glycerol): m/z 464 [M+H]+;
EIMS: m/z 463 [M+], 404, 388, 331, 43
(100%); Anal. Calcd for C17H25N3O10S: C,
44.06; H, 5.40; N, 9.07; S, 6.91. Found: C,
43.77; H, 5.53; N, 8.75; S, 6.86.
(−) - 5 - Acetamido - 3 - N - acetyl - 2 - [1,2,3,-
4,5-penta-O-acetyl- -galacto-pentitol-1-yl]-
D
1,3,4-thiadiazoline [(−)-1].—By application
of Procedure B over 4 h on -galactose
D
1,4 - N,N - Diacetyl - 1 - N - (2,3,4,6 - tetra - O-
thiosemicarbazone in the pyranose form, 0.56
g (26%) of the title compound could be ob-
acetyl-i- -galactopyranosyl)thiosemicarbazide
D
(6).—Yield: 0.5 g (25.0%); mp 103–105 °C;
1
tained; mp 234 °C (lit. 234 °C [2]); H NMR
[h]2D0 −14.8° (c 1, CHCl3); umax =276 nm
(CDCl3): l 2.19–1.93 (7 s, 21 H, ꢁCH3); 13C
NMR (CDCl3): l 170.4–168.5 (7 s, CꢀO),
23.1–20.5 (6 s, ꢁCH3); EIMS: m/z 547 [M+],
331, 186, 144, 102, 43 (100%).
(EtOH); H NMR (CDCl3): l 2.19–1.91 (6 s,
1
18 H, ꢁCH3); 13C NMR (CDCl3): l 172.5–
169.8 (5 s, CꢀO), 24.0–20.6 (3 s, ꢁCH3);
EIMS: m/z 505 [M+], 331, 43 (100%); Anal.
Calcd for C19H27N3O11S: C, 45.15; H, 5.35; N,
8.32; S, 6.34. Found: C, 45.34; H, 5.73; N,
8.63; S, 6.25.
D
-Galactose thiosemicarbazone (2).—Syn-
thesis was carried out in two different ways:
(a) as described in the literature [10], and (b)
by reflux a mixture of thiosemicarbazide and
1-N-Acetyl-1-N-(2,3,4,6-tetra-O-acetyl-i- -
D
D-galactose in EtOH. In the first case, we
glucopyranosyl)thiosemicarbazide (8).—Yield:
could obtain the thiosemicarbazone 2 as a
mixture of pyranose and open-chain form.
When the second method was used, and the
mixture was allowed to crystallize at rt, we
obtained 2 only as a pyranose form.
Pyranose form: 13C NMR (Me2SO-d6): l
181.4 (CꢀS), 91.8 (C-1), 76.7 (C-5), 74.0 (C-3),
68.9 (C-2 or C-4), 68.8 (C-2 or C-4), 61.7
(C-6).
0.17 g (9.1%); mp 217 °C; [h]2D0 67.7° (c 1,
1
CHCl3); umax =216 and 250 nm (EtOH); H
NMR (CDCl3): l 2.12–2.04 (4 s, 15 H, ꢁCH3);
13C NMR (CDCl3): l 173.4–169.1 (5 s, CꢀO),
20.8–20.4 (3 s, ꢁCH3); EIMS: m/z 463 [M+],
331,
43
(100%);
Anal.
Calcd
for
C17H25N3O10S: C, 44.06; H, 5.40; N, 9.07.
Found: C, 43.93; H, 5.72; N, 9.35.
4-N-Acetyl-1-N-(2,3,4,6-tetra-O-acetyl-i- -
D
Open-chain form: 13C NMR (Me2SO-d6): l
178.2 (CꢀS), 148.6 (CꢀN, C-1), 71.7 (C-4),
70.3 (C-3 or C-5), 70.1 (C-3 or C-5), 69.4
(C-2), 63.4 (C-6).
glucopyranosyl)thiosemicarbazide (9).—Yield:
0.60 g (30.2%); mp 69–71 °C; [h]2D0 −71.9°
(c 1, CHCl3); umax=226 and 268 nm (EtOH);
1H NMR (CDCl3): l 2.41–1.94 (5 s, 15 H,
ꢁCH3); 13C NMR (CDCl3): l 178.5–169.5
(5 s, CꢀO), 25.3–20.5 (3 s, ꢁCH3); EIMS:
m/z 463 [M+], 331, 43 (100%); Anal.
D
-Glucose and -mannose thiosemicarbazone
D
(3) and (4).—The title compounds were syn-
thesized by both procedures cited earlier for

488
M.A. Martins Alho, N.B. D’Accorso / Carbohydrate Research 328 (2000) 481–488
Calcd for C17H25N3O10S·C4H8O2: C, 45.74; H,
5.99; N, 7.62. Found: C, 45.62; H, 5.72; N, 7.35.
1,4 - N,N - Diacetyl - 1 - N - (2,3,4,6 - tetra - O-
EIMS m/z 547 [M+], 186, 144 (100%), 102, 43;
Anal. Calcd for C21H29N3O12S·CH4O: C, 45.60;
H, 5.70; N, 7.25. Found: C, 45.81; H, 5.44; N,
7.21.
acetyl - i -
D
- glucopyranosyl)thiosemicarbazide
(−)-5-Acetamido-3-N-acetyl-2-[1,2,3,4,5-
(10).—Yield: 0.85 g (42.8%); mp: 99–101 °C;
[h]D 14.6° (c 1, CHCl3); umax =224 and 274 nm
penta-O-acetyl- -manno-pentitol-1-yl]-1,3,4-
D
1
thiadiazoline [(−)-13].—Yield: 0.17 g (7.8%);
(EtOH); H NMR (CDCl3): l 2.16–1.34 (4 s,
1
mp: 98–100 °C; [h]2D0 –121.9° (c 1, CHCl3); H
18 H, ꢁCH3); 13C NMR (CDCl3): l 170.9–
169.3 (6 s, CꢀO), 26.8–20.6 (5 s, ꢁCH3); EIMS:
m/z 505 [M+], 331, 43 (100%); Anal. Calcd for
C19H27N3O11S·C4H8O2: C, 46.54; H, 5.90; N,
7.08. Found: C, 46.50; H, 5.77; N, 6.77.
NMR (CDCl3): l 2.22–2.04 (7 s, 21 H, ꢁCH3);
13C NMR (CDCl3): l 170.6–168.9 (6 s, CꢀO),
23.0–20.5 (4 s, ꢁCH3); EIMS: m/z 186, 144
(100%), 102, 43; Anal. Calcd for C21H29N3O12-
S·C4H8O2: C, 47.24; H, 5.83; N, 6.61; S, 5.04.
Found: C, 47.04; H, 5.66; N, 6.83; S, 4.81.
4-N-Acetyl-1-N-(2,3,4,6-tetra-O-acetyl-i- -
D
mannopyranosyl)thiosemicarbazide
(11).—
Yield: 0.40 g (20.2%) (syrup); 1H NMR
(CDCl3): l 2.47–2.02 (5 s, 15 H, ꢁCH3); 13C
NMR (CDCl3): l 178.1–169.3 (5 s, CꢀO),
25.2–20.7 (5 s, ꢁCH3); EIMS: m/z 463 [M+],
331, 43 (100%); Anal. Calcd for C17H25N3O10S:
C, 44.06; H, 5.40; N, 9.07. Found: C, 44.32; H,
5.25; N, 8.93.
Acknowledgements
The authors thank the Consejo Nacional de
Investigaciones Cient´ıficas y Te´cnicas, Agencia
de Promocio´n Cient´ıfica y Tecnolo´gica and the
Universidad de Buenos Aires for financial sup-
port. Dr N.B. D’Accorso is a research member
of CONICET. The author are indebted to
Professor Galbis Pe´rez (Sevilla, Espan˜a) for the
COSY spectrum of compound (−)-1 and to
Lic. Tettamanzzi of Universidad de Buenos
Aires for the COSY spectrum of compound
(+)-1. The authors are also very grateful to
UMYMFOR (CONICET-FCEyN UBA) for
(−)-5-Acetamido-3-N-acetyl-2-[1,2,3,4,5-
penta-O-acetyl- -gluco-pentitol-1-yl]-1,3,4-
D
thiadiazoline [(−)-12].—By application of
Procedure B during 24 h on -glucose thiosem-
D
icarbazone in the pyranose form, 0.69 g (32%)
of the title compound was obtained; mp: 98–
100 °C; [h]2D0 −152.1° (c 1, CHCl3); umax =218
1
and 286 nm (EtOH); H NMR (CDCl3): l
2.14–1.93 (6 s, 21 H, ꢁCH3); 13C NMR
(CDCl3): l 170.7–168.7 (7 s, CꢀO), 23.2–20.5
(5 s, ꢁCH3); EIMS: m/z 186 (100%), 144, 102;
Anal. Calcd for C21H29N3O12S: C, 46.07; H,
5.30; N, 7.68. Found: C, 46.34; H, 5.07; N,
7.82.
1
13
the recording of the H and C NMR, as well
as for the microanalyses and mass spectra.
References
(+)- and (−)-5-Acetamido-3-N-acetyl-2-
[1] L. Somogyi, Carbohydr. Res., 75 (1979) 325–330.
[2] E.S.H. El Ashry, M.A. Nassr, Y. El Kilany, A. Mousaad,
Bull. Chem. Soc. Jpn., 60 (1987) 3405–3509.
[3] H.J. Tweeddale, J.W. Redmond, J. Carbohydr. Chem., 17
(1998) 27–38.
[4] L. Somogyi, Carbohydr. Res., 64 (1978) 289–292.
[5] S.J. Angyal, R. Le Fur, D. Gagnaire, Carbohydr. Res., 23
(1972) 121–134.
[1,2,3,4,5-penta-O-acetyl- -manno-pentitol-1-
D
yl]-1,3,4-thiadiazoline [(+)-13 and (−)-
13].—By using Procedure B for 2 h on -man-
D
nose thiosemicarbazone, 0.50 g (23%) of the
title compounds were obtained in a 2:1 ratio.
At a reaction time of 4 h, the global yield rises
to 0.65 g (30.1%) with a 4:1 ratio. After a reflux
of 15 h, only (+)-13 could be isolated in a 35%
yield (0.76 g).
[6] M. Saunders, K.E. Gilbert, J.J. Gajewski, PMR-Spectrum
Simulator, Serena Software.
´
[7] R. Bogna´r, P. Na´na´si, E. Nemes-Na´na´si, J. Chem. Soc.,
(1955) 193–195.
[8] (a) J.M.J. Tronchet, Br. Baehler, A. Jorretand, F. Perret,
Hel6. Chim. Acta., 54 (1971) 1660–1676; (b) D. Horton,
M. Nakadate, J.M.J. Tronchet, Carbohydr. Res., 7 (1968)
56–65.
[9] HyperChem™ Molecular Modeling System, Hypercube
Inc. and Autodesk Inc., 1993. PCMODEL for Windows 5.1,
Serena Software, 1994.
(+) - 5 - Acetamido - 3 - N - acetyl - 2 - [1,2,3,-
4,5-penta-O-acetyl- -manno-pentitol-1-yl]-
D
1,3,4-thiadiazoline [(+)-13].—Yield: 0.76 g
(35%); mp: 115–117 °C; [h]2D0 373.2° (c 1,
1
CHCl3); H NMR (CDCl3): l 2.21–1.72 (7 s,
21 H, ꢁCH3); 13C NMR (CDCl3): l 170.5–
168.7 (7 s, CꢀO), 23.4–20.7 (3 s, 21 H, ꢁCH3);
[10] C. Neuberg, W. Niemann, Ber. Dtsch. Chem. Ges., 35
(1902) 2049–2056.