Sulfuric acid-catalyzed acetolysis of anomeric methyl 2,3,4,6-tetra-O-acetyl-D-mannopyranosides: Kinetics and mechanism

Article abstract of DOI:10.1016/S0008-6215(99)00308-0

The kinetics of the acetolysis and accompanying anomerization of methyl 2,3,4,6-tetra-O-acetyl-α- and -β-D-mannopyranosides at different concentrations of sulfuric acid in acetic anhydride-acetic acid mixtures were studied. The progress of the reactions was followed by gas chromatography, and the rate constants of the partial reactions were calculated on the basis of the time-dependent product distribution obtained. The mechanisms of the reactions involved are discussed. The involvement of unstable ionic intermediates is taken into account in the evaluation of the kinetic results, and simplified and extended models are used in the mathematical treatment of the results. A fourth-order Runge-Kutta algorithm is used to calculate rate constants. Acetolysis was found to be faster for mannosides than for glucosides relative to their anomerization. The β-mannopyranoside prefers endocyclic CO-bond rupture, while in the α anomer the endocyclic and exocyclic cleavages are comparatively rapid. Copyright (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0008-6215(99)00308-0

www.elsevier.nl/locate/carres
Carbohydrate Research 325 (2000) 16–29
Sulfuric acid-catalyzed acetolysis of anomeric methyl
2,3,4,6-tetra-O-acetyl-
D
-mannopyranosides: kinetics and
mechanism
Jo´zef Kaczmarek a, Zbigniew Kaczyn´ski a, Zygmunt Trumpakaj a, Janusz Szafranek a,*,
Magda Bogalecka b, Harri Lo¨nnberg c
a Department of Chemistry, Uni6ersity of Gdan´sk, PL-80952 Gdan´sk, Poland
b Department of Chemistry, Gdynia Maritime Academy, PL-81225 Gdynia, Poland
c Department of Chemistry, Uni6ersity of Turku, FIN-20014 Turku, Finland
Received 10 May 1999; revised 25 October 1999; accepted 22 November 1999
Abstract
The kinetics of the acetolysis and accompanying anomerization of methyl 2,3,4,6-tetra-O-acetyl-a- and -b-D-
mannopyranosides at different concentrations of sulfuric acid in acetic anhydride–acetic acid mixtures were studied.
The progress of the reactions was followed by gas chromatography, and the rate constants of the partial reactions
were calculated on the basis of the time-dependent product distribution obtained. The mechanisms of the reactions
involved are discussed. The involvement of unstable ionic intermediates is taken into account in the evaluation of the
kinetic results, and simplified and extended models are used in the mathematical treatment of the results. A
fourth-order Runge–Kutta algorithm is used to calculate rate constants. Acetolysis was found to be faster for
mannosides than for glucosides relative to their anomerization. The b-mannopyranoside prefers endocyclic CO-bond
rupture, while in the a anomer the endocyclic and exocyclic cleavages are comparatively rapid. © 2000 Elsevier
Science Ltd. All rights reserved.
Keywords: Acetolysis; Anomerization; Mannopyranosides; Kinetics
1. Introduction
desirable, as it is of fundamental importance
for the chemical manipulation of carbohy-
drates [10]. The identity of the ionic intermedi-
ates involved in acetolysis (cyclic versus
acyclic oxocarbenium ion) is particularly
relevant.
The most commonly used medium for ace-
tolysis is a mixture of acetic anhydride and
acetic acid containing a catalytic amount of
sulfuric acid. In such solutions, the sugars are
released from polysaccharides as an anomeric
mixture of peracetates. Similarly, methyl gly-
copyranoside peracetates have been shown to
yield anomeric mixtures of fully acetylated
glycopyranoses, while glycofuranosides are
Acetolysis is one of the degradation meth-
ods used to elucidate the structure of oligo-
and polysaccharides [1–5]. The specificity of
this approach is clearly seen in the fragmenta-
tion of polysaccharides containing (1“6)
linkages, which are extremely labile to acetol-
ysis but relatively resistant to acid hydrolysis
[6–9]. Thus, a detailed understanding of the
influence of the sugar configuration on the
kinetics and mechanism of acetolysis appears
* Corresponding author. Fax: +48-58-3410357.
0008-6215/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(99)00308-0

J. Kaczmarek et al. / Carbohydrate Research 325 (2000) 16–29
17
acetic acid to yield an anomeric mixture of
glucose penta-acetate, or an acyclic 2,3,4,5,6-
converted under the same conditions to either
acyclic aldehyde acetates or to furanoid ac-
etates [11]. We have shown previously [12]
that the acetolysis of ethyl 2,3,4,6-tetra-O-
penta-O-acetyl-1-O-ethyl- -glucitol-1-C-yl-
D
ium ion is generated and converted via the
diastereoisomeric acyclic ethyl acetals of per-
acetyl-a- and -b- -glucopyranosides is accom-
D
O-acetyl-aldehydo-
D
-glucose and traces of
panied by rather rapid anomerization.
Together with the time-dependent product dis-
tributions, this observation suggests that ace-
tolysis proceeds by one of two parallel routes.
Either a cyclic glucopyranosyl oxocarbenium
ion is formed and subsequently trapped by
anomeric -glucofuranose peracetates. The re-
D
cyclization of the intermediate oxocarbenium
ion is fast, compared with its reaction with an
external nucleophile, i.e., acetic acid.
We now report on a similar kinetic analysis
of the acetolysis of methyl 2,3,4,6-tetra-O-
acetyl-a- and -b- -mannopyranosides with an
D
extended model of the reaction set that in-
cludes ionic species.
2. Results and discussion
The acetolysis reactions of the anomeric
methyl mannopyranoside tetra-acetates (1
and 2) were carried out in a 1:1 acetic anhy-
dride–acetic acid mixture at three different
concentrations of sulfuric acid (0.25, 0.5 and
1%), followed by neutralization with sodium
acetate, and analyzed by gas chromatogra-
phy (GC). Both mixtures were observed to
contain seven components, which were iden-
tified by spiking with authentic samples:
mannosides 1 and 2, the anomeric
mannopyranose penta-acetates and 5,
equimolar diastereoisomeric (1R,1S)-1,2,3,4,-
5,6-hexa-O-acetyl- -mannose methyl hemiac-
D
-
4
D
etals 3R and 3S (further labelled as 3 in the
text or 3a and 3b in Fig. 1 and Scheme 3)
and aldehydo-D-mannose-hepta-acetate
6
(Fig. 1). Tables 1 and 2 present the time-de-
pendent product distributions of 1–6 for the
acetolysis of anomeric methyl per-O-acetyl-
D-mannopyranosides. The plots of the time-
dependent distributions of 1–6 obtained by
quantitative GC with either 1 or 2 as start-
ing material are illustrated in Figs. 2 and 3
(only for a 0.25% sulfuric acid content). A
similar seven-component mixture was ob-
tained elsewhere from 1 in an acetic anhy-
dride–boron trifluoride medium [17].
Fig. 1. Gas chromatograms of the product mixtures after (a)
5 min, and (c) 24 h acetolysis of methyl 2,3,4,6-tetra-O-acetyl-
As can be seen from Figs. 2 and 3, the
acetolysis of 1 and 2 is always accompanied
b-
D
-mannopyranoside 2, and (b) 1 h acetolysis of methyl
-mannopyranoside 1, in a 1:1 acetic
2,3,4,6-tetra-O-acetyl-a-
D
by mutual anomerization. As the a- -man-
D
anhydride–acetic acid mixture, containing 1% sulfuric acid, at
25 °C. For the structures of the compounds, see Scheme 3.
nopyranoside is thermodynamically much

Table 1
Time-dependent product distribution for the acetolysis of methyl 2,3,4,6-tetra-O-acetyl-a-
D
-mannopyranoside 1 in a 1:1 acetic acid–acetic anhydride mixture containing 0.25,
0.5, and 1% sulfuric acid, at 25 °C a
Time (h)
Concentration of the catalyst
0.25%
0.5%
1%
Compounds (%)
Compounds (%)
Compounds (%)
/
1
2
3 b
4
5
6
1
2
3 b
4
5
6
1
2
3 b
4
5
6
0
100
0
0
0
0
0
100
0
0
0
0
0
100
0
0
0
0
0
0.0833
0.1667
0.3333
0.5
1
2
3
6
97.65
0.67
1.13
1.50
1.64
2.09
2.08
1.80
1.47
0.80
0.22
1.47
2.34
3.67
4.82
8.97
15.56
21.44
38.02
62.13
83.40
91.73
0.21
0.41
0.61
0.61
1.41
2.08
2.54
4.34
5.98
7.87
7.19
96.17
1.03
1.17
1.33
1.79
2.20
1.82
1.13
0.75
0.21
2.54
3.12
6.06
9.35
16.01
31.25
44.53
66.13
86.54
92.39
92.55
0.26
0.50
0.97
1.39
2.77
2.75
4.30
6.00
6.47
6.52
6.79
90.65
1.73
1.98
1.88
1.83
1.36
0.68
0.66
0.25
6.80
0.82
1.46
1.96
2.04
4.27
5.79
7.04
6.73
6.66
6.18
6.19
96.12
94.22
92.93
87.45
80.16
74.09
56.00
30.76
8.07
95.21
91.64
87.40
78.94
64.01
49.82
26.76
6.31
84.93
78.79
73.48
52.59
27.04
13.74
2.63
11.63
17.31
22.49
41.61
66.10
78.17
89.88
92.61
93.09
92.79
0.06
0.16
0.17
0.39
0.39
0.51
0.30
0.28
0.36
0.07
0.08
0.17
0.22
0.36
0.42
0.39
0.38
0.08
0.12
0.13
0.17
0.33
0.35
0.29
325
0
12
24
48
0.05
0.08
0.28
0.43
0.45
0.66
1
0.09
0.21
0.62
0.58
a For the structures of the compounds, see Scheme 3.
b 3 is the sum of 3R+3S.

Table 2
Time-dependent product distribution for the acetolysis of methyl 2,3,4,6-tetra-O-acetyl-b-
D
-mannopyranoside 2 in a mixture of 1:1 acetic acid–acetic anhydride containing
0.25, 0.5, and 1% sulfuric acid, at 25 °C a
Time (h)
Concentration of the catalyst
0.25%
0.5%
1%
Compounds (%)
Compounds (%)
Compounds (%)
1
2
3 b
4
5
6
1
2
3 b
4
5
6
1
2
3 b
4
5
6
/
0
0
100
0
0
0
0
0
100
0
0
0
0
0
100
0
0
0
0
0.0833
0.1667
0.3333
0.5
1
2
3
6
27.27
44.59
55.77
67.76
82.69
81.76
75.43
59.43
33.72
10.09
1.96
71.64
0.22
0.25
0.26
0.34
0.42
0.48
0.53
0.66
0.74
0.85
0.75
0.87
1.11
1.79
2.54
6.45
13.01
19.58
34.27
58.59
81.02
90.29
50.76
65.36
76.42
81.25
78.33
67.09
51.64
27.90
8.00
47.78
0.15
0.28
0.32
0.42
0.50
0.63
0.87
0.93
0.94
0.93
0.54
1.12
1.76
3.83
6.26
16.50
27.34
42.09
65.34
84.17
91.18
92.28
0.19
0.34
0.44
0.92
2.10
3.34
4.27
5.03
6.42
6.63
6.70
75.84
80.82
79.76
79.39
54.48
27.04
16.00
1.90
18.68
0.50
0.56
0.56
0.63
0.67
0.76
0.90
0.91
0.71
0.53
0.35
4.37
7.49
0.61
1.03
1.60
1.76
3.97
5.85
6.08
6.83
6.73
6.50
6.52
53.80
41.96
29.00
9.48
3.09
2.08
1.56
1.01
0.41
0.25
0.22
0.36
0.96
1.66
2.38
4.08
5.76
7.32
6.58
32.26
18.99
11.15
2.57
1.60
1.13
0.80
0.27
0.13
10.10
3.46
3.33
2.12
0.78
0.49
0.26
0.14
0.09
0.08
14.62
14.89
38.76
65.57
76.53
89.84
90.96
91.87
91.83
325
)
0.26
0.92
1.01
1.22
12
24
48
0.18
0.31
0.42
0.20
0.26
0.48
0.54
1
0.87
a For the structures of the compounds, see Scheme 3.
b 3 is the sum of 3R+3S.

20
J. Kaczmarek et al. / Carbohydrate Research 325 (2000) 16–29
Fig. 2. Time-dependent product distributions for the acetolysis of methyl 2,3,4,6-tetra-O-acetyl-a-D-mannopyranoside 1, in a 1:1
acetic anhydride–acetic acid mixture, containing 0.25% sulfuric acid, at 25 °C. For the structures of the compounds, see Scheme
3; e1–e6 correspond to experimental points, t1–t6 correspond to theoretical curves. Upper curves concern the product
distributions appearing in the mixture at low concentration.
more stable than its b anomer [13], this
anomerization is more readily detectable when
the less stable b anomer 2 is used as starting
material. But as the anomerization is actually
faster than the acetolysis, an equilibrium mix-
ture of 1 and 2 containing 97.6 and 2.4% of
the a and b anomer, respectively, is produced.
The same equilibrium mixture is also obtained
on using the a anomer 1 as the starting
material.
and 2 is shown in Scheme 1. This depicts the
smallest amount of reactions needed to de-
scribe the kinetics of the two anomers. The
least-squares method was employed for the
conversion ratios (ca. 0–60%), while the first-
order rate constants for the anomerization
and acetolysis of 1 and 2 were calculated as in
Ref. [12]. The rate constants obtained for the
partial reactions involved are listed in Table 3.
At this step of evaluation, the following con-
clusions may be drawn for 2 as substrate: (i)
anomerization of methyl 2,3,4,6-tetra-O-
A preliminary, simplified approach to the
reactions taking place during acetolysis of 1

J. Kaczmarek et al. / Carbohydrate Research 325 (2000) 16–29
21
acetyl-a-
D
-mannopyranoside 1 to its b anomer
the partial reactions increase with catalyst
concentration; these relations are obvious for
anomers 1 and 2, and even more for the b
anomer; (v) acetolysis is faster for mannosides
than for glucosides, relative to their anomer-
ization [12].
proceeds about four times as fast, and that of
the b anomer to the a anomer about 40 times
as fast, as the acetolysis of the anomeric mix-
ture 1 and 2 to anomeric mannose penta-ac-
etates 4 and 5, or acyclic acetyl methyl acetals
3R and 3S; (ii) formation of the anomeric
mixture of mannose penta-acetates 4 and 5 is
about 30 times as fast as that of acetyl methyl
acetals 3R and 3S; (iii) acetolysis of the acetyl
methyl acetals yields only the acyclic hepta-ac-
etate 6; (iv) the first-order rate constants for
As discussed previously [12], the anomeriza-
tion of ethyl 2,3,4,6-tetra-O-acetyl-a- and -b-
D
-glucopyranosides proceeds by the formation
of cyclic 2,3,4,6-tetra-O-acetyl-
osylium 7 and acyclic 2,3,4,5,6-penta-O-acet-
yl-1-O-methyl- -glucitol-1-C-ylium ions.
D
-glucopyran-
D
8
Fig. 3. Time-dependent product distributions for the acetolysis of methyl 2,3,4,6-tetra-O-acetyl-b-D-mannopyranoside 2, in a 1:1
acetic anhydride–acetic acid mixture, containing 0.25% sulfuric acid, at 25 °C. For the structures of the compounds, see Scheme
3; e1–e6 correspond to experimental points, t1–t6 correspond to theoretical curves. Upper curves concern the product
distributions appearing in the mixture at low concentration.

22
J. Kaczmarek et al. / Carbohydrate Research 325 (2000) 16–29
Scheme 1.
Therefore, a more extensive model of the ki-
netics and mechanism must be employed,
which includes the ion formation (Scheme 2).
As in the simplified approach, compounds 1
and 2 are quite readily interconvertible
through the acyclic ion 8, but acetolysis takes
place via a different ionic intermediate. Both 1
and 2 were observed to yield 1,2,3,4,6-penta-
Scheme 2.
off in the form of a methyl acetate molecule.
As 1 and 2 both produce 4 and 5 in the same
7:1 ratio, the oxocarbenium ion of 7 appears
to be generated before the attack of acetic
acid on the anomeric carbon (Scheme 2).
The kinetics of the above processes, which
involve the formation of cyclic and acyclic
ions, can be described by a set of first-order
differential equations (Eqs. (1)–(8)), where kij
are the pseudo-first-order rate constants indi-
cated in Scheme 2, and xi denotes the time-de-
pendent molar fractions of compounds 1, 2, 3,
4, 5 and 6, and intermediates 7 and 8. The
values kij were determined by a fourth-order
Runge–Kutta algorithm [18] to give the best
O-acetyl-a- and -b- -mannopyranose in a
D
constant 7:1 concentration ratio. It is espe-
cially worthwhile recording that this ratio re-
mains unchanged during the acetolysis of 2, in
spite of the fact that 2 simultaneously under-
goes anomerization to 1. Thus, it appears
reasonable to assume that the acetolysis of 1
and 2 proceeds via a common intermediate,
i.e., the cyclic mannosyl oxocarbenium ion (7
in Scheme 2), which is produced when the
acetylium ion of the 1-methoxy group is split
Table 3
First-order rate constants for the partial reactions involved in the acid-catalyzed acetolysis of methyl 2,3,4,6-tetra-O-acetyl-a- (1)
and -b-
D
-mannopyranoside (2) in a 1:1 acetic anhydride–acetic acid mixture, containing 0.25, 0.5, and 1% sulfuric acid, at 25 °C a
Rate constant 106 Acetolysis of 1
Rate constant 106 Acetolysis of 2
k (s−1
k (s−1
)
)
Concentration of the catalyst
0.25% 0.5% 1%
25930 40950
Concentration of the catalyst
0.25%
60940
0.5%
1%
k12
k21
k13
k14
k15
k36
1109200 k12
53009900 k21
2509150
6009350
11009100
0.1790.05
2490.5
14509300
0.2890.1
5691
9309100 23009300
60009600
5.092.5
152.794.5
1592
0.9290.4
18095
1692
k23
k24
k25
k36
2.090.8
20.690.5
3.090.5
0.790.6
3.091.2
48.790.7
6.090.5
0.490.5
3.490.7
891
0.0590.07
0.0690.08
1.291.7
0.791.0
a For the rate constants, see Scheme 1.

J. Kaczmarek et al. / Carbohydrate Research 325 (2000) 16–29
23
dx
dt
least-squares fit between the calculated
and experimental values of xi. The results
obtained are set out in Table 4, and show
good agreement between the calculated and
experimental time-dependent distributions of
compounds 1, 2, 3, 4, 5 and 6 (see also Figs. 2
and 3).
8=k18x1+k28x2 −k81x8−k82x8−k83x8
(8)
The rate constants listed in Table 4 allow
the following inferences to be made: (i)
Methyl 2,3,4,6-tetra-O-acetyl-b- -mannopyr-
D
anoside 2 forms the cyclic mannopyranosyl
oxocarbenium ion 7 (Scheme 3) slightly faster
than its anomer 1 (cf. k27 and k17); in other
words, the b anomer undergoes acetolysis
faster than its counterpart does. The anomeric
distribution of per-O-acetyl mannopyranoses
is due to thermodynamic factors. (ii) The ring
opening of the two substrates proceeds in a
similar fashion, though much more distinctly:
b anomer 2 forms the acyclic oxocarbenium
ion 8 (Scheme 3) 30 times faster than the a
anomer (cf. k28 and k18). Accordingly, the b
anomer prefers endocyclic CO-bond rupture,
while in the a anomer the endocyclic and
exocyclic cleavages are comparatively rapid.
(iii) Acetic-acid attack on the mannopyranosyl
oxocarbenium ion 7 from the a face is ten
times faster than from the b face (cf. k74 and
dx
1 =k81x8 −k18x1−k17x1
2 =k82x8 −k28x2−k27x2
3 =k83x8 −k36x3+k63x6
4 =k74x7 −k47x4
(1)
(2)
(3)
(4)
(5)
(6)
dt
dx
dt
dx
dt
dx
dt
dx
dt
5 =k75x7 −k57x5
dx
dt
6 =k36x3 −k63x6
dx
dt
7 =k17x1 +k27x2+k47x4+k57x5−k74x7
k75). The equilibrium mixture of anomeric
D
-
−k75x7
(7)
mannopyranose penta-acetates contains 92%
Table 4
First-order rate constants for the partial reactions involved in the acid-catalysed acetolysis of methyl 2,3,4,6-tetra-O-acetyl-a- (1)
and -b-
-mannopyranoside (2) in a 1:1 acetic anhydride–acetic acid mixture, containing 0.25, 0.5, and 1% sulfuric acid, at 25 °C a
D
Rate constant 106 k (s−1
)
Acetolysis of 1
Acetolysis of 2
Concentration of the catalyst
Concentration of the catalyst
0.25%
0.5%
1%
0.25%
0.5%
1%
417910
19,2009400
11,7009700
k18
k81
k28
k82
k74
k47
k75
k57
k17
k27
k83
k36
k63
4893
26,7009300
18009100
11195
23,1009600
34709500
417910
15,60091000 27,2009800
11,7009700 16009100
4793
11395
25,6009300
36709400
26,4009500
19,40092000 13,90092000 27,50091000 24,40091000 20,50092000
27,50091000 3,200091500 33,60092000 27,00091000 28,00092000 27,00092000
0
0
0
0
0
0
2940980
2.290.5
2691
8192
8292
29209100
1.490.7
5692
306950
7893
28609150
1.690.9
14695
15809100
5195
29609100
3.090.5
2591
2592
11998
692
25009150
1.790.7
5492
81915
10993
592
21809200
0.891
14295
3069100
8395
3.890.5
0.090.5
491
1693
1993
091
0.692
795
592
492
S6i=1 Z2i (%)
1.7
3.5
4.4
1.3
1.8
1.5
a For the rate constants, see Scheme 2.

24
J. Kaczmarek et al. / Carbohydrate Research 325 (2000) 16–29
Scheme 3.

J. Kaczmarek et al. / Carbohydrate Research 325 (2000) 16–29
25
kinetic study indicates that acetolysis and
anomerization depend on the respective at-
tack of the acetylium ion (CH3CO+) on the
glycosidic or heterocyclic oxygen, as was
suggested earlier. These results and those re-
ported recently for ethyl glucoside tetra-ac-
etates corroborate the early suggestion of
Lindberg [20,21] that anomerization takes
place via an acyclic intermediate. In fact,
acyclic glycosyl oxocarbenium ions are
formed more readily than cyclic glycopyra-
nosyl oxocarbenium ions, the relative rates
of these two reactions being markedly de-
pendent on the sugar moiety configuration.
It should, however, be noted that acetolysis,
i.e., the cleavage of the glycosidic bond, does
not take place via the acyclic oxocarbenium
ion, but by the formation of a cyclic oxocar-
benium ion. Obviously, very rapid recycliza-
tion keeps the concentration of the acyclic
ion at such a low level that acetolysis via
this intermediate cannot compete with that
via the cyclic oxocarbenium ion.
Calculation of the rate constants.—The re-
actions involved in the kinetics of the acetol-
ysis reactions are summarized in Scheme 2.
The rates of these reactions are described by
a set of first-order differential equations
(Eqs. (1)–(8)). The rate constants kij were
determined by the solution for the set that
gives the best least-squares fit between the
calculated and experimental values of xi. If
all the constants kij are known, Eqs. (1)–(8)
can be solved by numerical integration. The
values of kij were initially guessed, and the
fourth-order Runge–Kutta algorithm [18]
was applied to solve the equation system.
The values of kij were then optimized to give
the best fit. For the best fit of the experi-
mental and theoretical values, the relative
differences li, Eq. (9), and the mean-squared
relative difference Z2i , Eq. (10), were used,
where xeij and xtij are the respective experi-
mental and theoretical values for the ith
compound and jth point in time.
of the a anomer 4 and 8% of the b anomer
5. The acyclic 2,3,4,5,6-penta-O-acetyl-1-O-
methyl- -mannitol-1-C-ylium ion 8 is cy-
D
clized to methyl a and b mannosides (1 and
2) (cf. k81 and k82). In contrast to ring open-
ing, reclosure is not accelerated by increasing
the concentration of sulfuric acid. (iv) The
attack of acetic acid on the acyclic oxocarbe-
nium ion 8, giving diastereoisomeric (1R,1S)-
1,2,3,4,5,6-hexa-O-acetyl- -mannose methyl
D
hemiacetals (3R, 3S), is more than 200 times
slower than the cyclization of 8 (cf. k83, k81
and k82). The rate of this reaction is also
independent of the sulfuric acid concentra-
tion. (v) The subsequent acetolysis of 3R
and 3S to 1,1,2,3,4,5,6-hepta-O-acetyl-alde-
hydo- -mannose 6 is acid catalyzed and is
D
slower than the formation of 3R and 3S
from 8 (cf. k36 and k83). In accordance with
the above, we suggest that the mechanism of
the acetolysis of methyl per-O-acetyl man-
nosides is as in Scheme 3, which takes into
account not only all the stable products (1–
6) but also the ionic intermediates 7 and 8.
The acid-catalyzed acetolysis of anomeric
methyl 2,3,4,6-tetra-O-acetyl- -mannopyran-
D
osides 1 and 2 is quite similar to that de-
scribed previously [12] for anomeric ethyl
2,3,4,6-tetra-O-acetyl- -glucopyranosides.
D
The acetolysis of all the glycosides men-
tioned is accompanied by a markedly faster
anomerization of the starting material. This
anomerization is faster with mannosides than
with glucosides: the conversion of cis-1,2-
mannoside (b) to its more stable trans coun-
terpart (a) is four times as fast as the
corresponding isomerization (a“b) of glu-
cosides. The acetolysis rates differ even more
strikingly: the anomeric mixture of man-
nosides 1 and 2 is converted to anomeric
penta-acetates up to 30 times as fast as
the anomeric glucosides. Both glycosides
produce 1,1,2,3,4,5,6-hepta-O-acetyl-aldehydo-
hexose only as a minor side product.
The acetolysis of glycosides has been stud-
ied by several research teams since the late
1940s, and alternative mechanisms via either
cyclic or acyclic intermediates [10,19–23]
have been suggested for the acetolysis and/or
accompanying anomerization. The present
12
12
li = % (xeij −xtij)/ % xeij
(9)
j=1
j=1
12
12
Z2i = % (xeij −xtij)2/ % xe2ij
(10)
j=1
j=1

26
J. Kaczmarek et al. / Carbohydrate Research 325 (2000) 16–29
Fig. 4. Theoretical time-dependent ion intermediate 7 (above) and 8 (below) distributions for the acetolysis of methyl
2,3,4,6-tetra-O-acetyl-a- -mannopyranoside 1, in a 1:1 acetic anhydride–acetic acid mixture, containing 0.25, 0.5, and 1% sulfuric
acid, at 25 °C. For the structures of the ion intermediates, see Scheme 3.
D
The final step involves minimizing the
squared value Z2i , not only for each product
of the reaction mixture but also for the total
sum of all observable compounds (i=
1, 2, 3, …6, Eq. (11)).
3. Experimental
General procedures.—The purity of the
compounds prepared, 1, 2, 4, 5, 3R, 3S and
6, was assessed by GC on a 60 m Rtx-225
capillary column (Restec Co., USA) under
isothermal conditions (215 °C) using flame-
ionization detection. The intermediates and
products of acetolysis were identified by co-
injection with authentic samples. The AcOH
and Ac2O used in the experiments were dis-
6
TSum= % Z2i
(11)
i=1
The total sum TSum was B5% for all the
products studied (see Table 4), where Z62
contains about 70% of TSum.
tilled prior to use. H and 13C NMR spectra
1
were recorded using a 400 MHz spectrome-
ter (Varian-Mercury, USA). The specific ro-
tations were measured using a Jasco J-20
spectropolarimeter (Japan).
The theoretical concentrations of ion inter-
mediates (7 and 8) of the acetolysis for dif-
ferent catalyst concentrations and different
reaction times are presented in Figs. 4 and
5. For a 0.25% catalyst concentration, inter-
mediates 7 and 8 exist during the whole time
of acetolysis of anomers 1 and 2. For higher
catalyst concentrations, the intermediate con-
centrations increase. The relation between
the concentrations of intermediates and the
catalyst concentration in the case of methyl
Methyl 2,3,4,6-tetra-O-acetyl-h- (1) and
-i- -mannopyranoside (2).—A mixture of
D
methyl mannosides was obtained by reflux-
ing -mannose in MeOH in the presence of
D
a strong cation-exchange resin [24]. The
products were separated on a 50×2.7 cm
ion-exchange column (Dowex 1×8, 200–400
mesh, OH− form). The products obtained
were identified, following silylation, by GC.
Methyl a- and -b-
acetylated with a 1:1 Ac2O–pyridine mix-
per-O-acetyl-b- -mannopyranoside is more
D
distinct, and the concentrations of the for-
mer achieve a maximum in shorter times.
D
-mannopyranosides were

J. Kaczmarek et al. / Carbohydrate Research 325 (2000) 16–29
27
ture, 1.5 h, 90 °C, in order to obtain 1 and 2,
5.06 (m) H-5, 4.22 (m) and 4.03 (m), H-6, 3.41
and 3.40 (2 s, 6 H, 1R,1S CH3OꢀHC-1-
OCOCH3), 2.12 and 2.10 (2 s, 6 H, 1R,1S
CH3OCOꢀHC-1-OCH3), 2.09–2.04 (10 s, 30
H, (1R,1S)×5×CH3CO); 13C NMR: l
170.2–169.1 (12 peaks, CꢁO), 95.8 and 94.6
C-1, 68.1 and 68.0 C-2, 67.2 and 67.3 C-3,
67.3 and 67.4 C-4, 67.7, 67.5 C-5, 61.7, 61.7
C-6; 57.6, 57.4 (CꢀOCH3), 20.8–20.6 (12
peaks CH3CꢁO). Anal. Calcd for C19H28O13:
C, 49.14; H, 6.03. Found: C, 49.00; H, 6.02.
respectively.
1,2,3,4,6-Penta-O-acetyl-h-
D
- (4) and -i- -
D
mannopyranose (5).—Compounds 4 and 5
were prepared as described by Hudson and
Dale [15].
[(1R)-
and
(1S)-]-1,2,3,4,5,6-Hexa-O-
acetyl- -mannose methyl hemiacetals 3R and
D
3S.—The procedure of Keith et al. [14] to
obtain acetal ester from dimethyl acetal was
adapted for the synthesis of compounds 3R
and 3S. 2,3,4,5,6-Penta-O-acetyl-D-mannose
1,1,2,3,4,5,6-Hepta-O-acetyl-aldehydo- -
D
dimethyl acetal (4.34 g, 9.95 mmol) was
treated with Ac2O (12.6 mL) in the presence
of Dowex 50W-X4 (1 g, 100–200 mesh, H+
form). The mixture was kept at 65 °C for 1.5
h, after which time the resin was filtered off
and the soln concentrated to an oil, under
reduced pressure. The crude, yellow syrup was
purified with a mixture of diethyl ether and
n-hexane. The oily product was separated
from the solvents and stored under diminished
pressure at 60 °C until constant weight to
yield 3.37 g (73%). [h]2D5+0.4° (c 1, CHCl3);
1H NMR data (CDCl3): l 5.77 (d) and 5.63
(d) H-1, 5.20 (dd) and 5.11 (dd) H-2, 5.48 (dd)
and 5.60 (dd) H-3, 5.41 (dd) 5.53 (dd) H-4,
mannose (6).—This compound was prepared
according to Lerner’s procedure for obtaining
the hepta-acetate of
D
-galactose [16]. 2,3,4,5,6-
Penta-O-acetyl-
D
-mannose diethyl dithioac-
etal [25] (2.0 g, 4.03 mmol) was stored with an
acetylating mixture (45.8 mL) of 4:40:1.8
acetic anhydride–acetic acid–sulfuric acid, for
72 h at room temperature (rt). The reaction
mixture was poured onto crushed ice and
mixed for 1 h. The product was extracted
several times with CHCl3 and this CHCl3 soln
was rinsed alternately with water, a satd soln
of NaHCO3, and water, then finally dried over
anhyd Na2SO4. The solvent was evaporated,
and the residue dissolved in EtOH and
Fig. 5. Theoretical time-dependent ion intermediate 7 (above) and 8 (below) distributions for the acetolysis of methyl
2,3,4,6-tetra-O-acetyl-b- -mannopyranoside 2, in a 1:1 acetic anhydride–acetic acid mixture, containing 0.25, 0.5, and 1% sulfuric
acid, at 25 °C. For the structures of the ion intermediates, see Scheme 3.
D

28
J. Kaczmarek et al. / Carbohydrate Research 325 (2000) 16–29
The molar percentages of the reaction mix-
ture were obtained from the ratios of the
experimental data and molar response factor.
One of the two GC peaks of the diaster-
eoisomeric [(1R)- and (1S)-]-1,2,3,4,5,6-hexa-
refluxed with charcoal. The suspension was
filtered and the supernatant concentrated to a
small volume and left at rt. The product was
recrystallized from a diethyl ether–n-hexane
mixture to yield 0.88 g of 6 (44.4%) as a solid;
1
mp 116–118 °C. [h]2D5+1.5° (c 1, CHCl3); H
O-acetyl- -mannose methyl hemiacetals 3R
D
and 3S, corresponding to compound 3a in
NMR data (CDCl3): l 6.78 (d) H-1, 5.25 (dd)
H-2, 5.51 (dd) H-3, 5.44 (dd) H-4, 5.05 (o)
H-5, 4.21 (q) and 4.04 (q) H-6, 2.13 and 2.12
(2 s, 6 H, CH3OCOꢀHC-1-OCOCH3), 2.10–
2.01 (5 s, 15 H, 5×CH3CO); 13C NMR: l
170.1–167.9 (six peaks CꢁO), 85.9 C-1, 67.0
C-2, 66.7 C-3, 67.0 C-4, 67.6 C-5, 61.7 C-6;
20.8–20.5 (six peaks CH3CꢁO). Anal. Calcd
for C20H28O14: C, 48.78; H, 5.69. Found: C,
48.91; H, 5.69.
Fig. 1, was overlapped by a large signal of
1,2,3,4,6-penta-O-acetyl-a- -mannopyranose
D
4. Assuming both hemiacetals to have equal
areas, the total area of acetals was calculated
by multiplying the area of 3b by two. In turn,
the area of per-O-acetylated a- -mannopyr-
D
anose was calculated by subtracting the area
of 3b from that of 4+3a.
Kinetic measurements.—Acetolysis was car-
ried out in sealed tubes immersed in a water
bath thermostatted at 25 °C. The tubes, each
containing 1 mL of 28 M soln of the starting
material in a 1:1 AcOH–Ac2O mixture con-
taining 1, 0.5 or 0.25% H2SO4, were removed
at suitable intervals. An aliquot of 0.1 mL was
withdrawn, neutralized with a satd soln of
NaOAc in AcOH, and centrifuged. The peaks
of the chromatograms were assigned by spik-
ing with standard samples.
Acknowledgements
Financial support from the Polish State
Committee for Scientific Research (grant DS/
8200-4-0085 of the University of Gdan´sk) is
gratefully acknowledged.
References
[1] R.D. Guthrie, J.F. McCarthy, Ad6. Carbohydr. Chem.,
22 (1967) 11–23.
To determine the molar response factors for
the seven compounds in the reaction mixtures,
a standard mixture of methyl 2,3,4,6-tetra-
[2] B. Lindberg, J. Lo¨nngren, S. Svensson, Ad6. Carbo-
hydr. Chem. Biochem., 31 (1975) 185–240.
[3] A.F. Bochkov, GE Zaikov, Chemistry of the O-Glyco-
sidic Bond, Pergamon, Oxford, 1979, pp. 177–201.
[4] C.J. Biermann, Ad6. Carbohydr. Chem. Biochem., 46
(1988) 251–271.
[5] G.O. Aspinall, in G.O. Aspinall (Ed.), The Polysaccha-
rides, Vol. 1, Academic Press, New York, 1982, pp.
35–131.
[6] J.K.N. Jones, W.H. Nicholson, J. Chem. Soc., 27
(1958).
[7] K. Matsuda, H. Watanabe, K. Fujimoto, K. Aso, Na-
ture, 191 (1961) 278.
[8] I.J. Goldstein, W.J. Whelan, J. Chem. Soc., 170 (1961).
[9] C.E. Ballou, W.C. Raschke, Science, 184 (1974) 127–
134.
[10] R.U. Lemieux, Ad6. Carbohydr. Chem., 9 (1954) 1–57.
[11] E.M. Montgomery, R.M. Hann, C.S. Hudson, J. Am.
Chem. Soc., 59 (1937) 1124–1129.
[12] J. Kaczmarek, M. Preyss, H. Lo¨nnberg, J. Szafranek,
Carbohydr. Res., 279 (1995) 107–116.
[13] B. Capon, Chem. Re6., 69 (1969) 407–498.
[14] D.D. Keith, R. Yang, J.A. Tortora, M. Weigele, J.
Org. Chem., 43 (1978) 3713–3716.
O-acetyl-a-
penta-O-acetyl-b-
mixture of diastereoisomers [(1R)- and (1S)-]-
1,2,3,4,5,6-hexa-O-acetyl- -mannose methyl
hemiacetals 3R and 3S, and 1,1,2,3,4,5,6-
hepta-O-acetyl-aldehydo- -mannose dis-
D
-mannopyranoside 1, 1,2,3,4,6-
D
-mannopyranose 5, the
D
D
6
solved in CH2Cl2 was injected into the gas
chromatograph. The peak areas are expressed
relative to the peak area of methyl 2,3,4,6-te-
tra-O-acetyl-a- -mannopyranoside 1, which
D
was used as an internal standard. The relative
molar response factors for the compounds
under consideration were: methyl 2,3,4,6-te-
tra-O-acetyl-a-1 and -b-
2, 1.00; 1,2,3,4,6-penta-O-acetyl-a-
-b- -mannopyranose 5, 1.26; [(1R)- and
(1S)-]-1,2,3,4,5,6-hexa-O-acetyl- -mannose
methyl hemiacetals 3R and 3S, 1.69;
D-mannopyranoside
D
- 4 and
D
D
[15] C.S. Hudson, J.K. Dale, J. Am. Chem. Soc., 37 (1915)
1264–1268.
[16] L.M. Lerner, Carbohydr. Res., 282 (1996) 189–192.
[17] F.W. Lichtenthaler, J. Breunig, W. Fischer, Tetrahe-
dron Lett., 30 (1971) 2825–2828.
1,1,2,3,4,5,6-hepta-O-acetyl-aldehydo-
nose 6, 1.65.
D
-man-

J. Kaczmarek et al. / Carbohydrate Research 325 (2000) 16–29
29
[18] (a) W. Kutta, Z. Math. u. Phys., 46 (1901) 435–453. (b)
C. Runge, Math. Ann., 46 (1895) 167–174. (c) L. Gollatz,
Numerische Behandlung 6on Differentialgleichungen, sec-
ond ed., Springer, Berlin, 1955.
[19] D.R. McPhail, J.R. Lee, B. Fraser-Reid, J. Am. Chem.
Soc., 114 (1992) 1905–1906.
[22] L. Rosenfeld, C.E. Ballou, Carbohydr. Res., 32 (1974)
287–298.
[23] F.W. Lichtenthaler, G. Bambach, Carbohydr. Res., 68
(1979) 305–312.
[24] J.E. Cadotte, F. Smith, D. Spriestersbach, J. Am. Chem.
Soc., 74 (1952) 1501–1504.
[20] B. Lindberg, Acta Chem. Scand., 3 (1949) 1153–1169.
[21] B. Lindberg, Acta Chem. Scand., 3 (1949) 1350–1354.
[25] M.L. Wolfrom, A. Thompson, Methods Carbohydr.
Chem., 2 (1963) 427–430.