Reaction kinetics and mechanism of acid-catalyzed anomerization of 1-O-acetyl-2,3,5-tri-O-benzoyl-l-ribofuranose

Article abstract of DOI:10.1016/j.carres.2009.02.031

The mechanism of the acid-catalyzed anomerization of 1-O-acetyl-2,3,5-O-benzoyl-α- and -β-l-ribofuranoses in different acetic acid-acetic anhydride mixtures was investigated. The progress of the reactions was followed by NMR spectroscopy and the rate cons

Full text of DOI:10.1016/j.carres.2009.02.031

Carbohydrate Research 344 (2009) 1102–1109
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Reaction kinetics and mechanism of acid-catalyzed anomerization
of 1-O-acetyl-2,3,5-tri-O-benzoyl-L-ribofuranose
Jonas J. Forsman a, Johan Wärnå b, Dmitry Yu. Murzin b, Reko Leino a,
*
a Laboratory of Organic Chemistry, Åbo Akademi University, FI-20500 Åbo, Finland
b Laboratory of Industrial Chemistry and Reaction Engineering, Åbo Akademi University, FI-20500 Åbo, Finland
a r t i c l e i n f o
a b s t r a c t
Article history:
The mechanism of the acid-catalyzed anomerization of 1-O-acetyl-2,3,5-O-benzoyl-a- and -b-L-ribofura-
Received 29 January 2009
Accepted 27 February 2009
Available online 10 March 2009
noses in different acetic acid–acetic anhydride mixtures was investigated. The progress of the reactions
was followed by NMR spectroscopy and the rate constants for the reactions were determined by the use
of a kinetic model. The site of anomeric activation was clarified by the use of 13C-labeled acetic acid and
acetic anhydride, respectively, proving that the anomerization takes place by exocyclic C–O cleavage,
thus ruling out anomerization via acyclic intermediates. The role of the acetyl cation as the catalytically
active species was further verified.
Keywords:
L-Ribose
Anomerization
Ó 2009 Elsevier Ltd. All rights reserved.
13C Labeling
Reaction kinetics
1. Introduction
it relates to fundamental carbohydrate chemistry, has been the sub-
ject of much investigation. However, an ongoing debate exists on
Nucleoside mimics form an emerging class of drugs for antiviral
and anticancer therapies and hence the interest in the use of
whether the anomerization proceeds via endo- or exo-cyclic activa-
tion. Lönnberg and Kulonpää have shown that the hydrolysis of
b-D-ribofuranosides takes place either by opening of the five-mem-
L
-nucleosides as chemotherapeutic agents is growing.1 The some-
what unexpected discovery that some of the non-natural
sides possess better binding properties to enzymes than the
natural -nucleosides has given rise to extensive research within
the fields of nucleobase, as well as sugar modified
-nucleosides.2
L
-nucleo-
bered ring or by formation of a cyclic oxo-carbenium ion, the path
taken depending on the electronegativity of the aglycon group.8 In
the present investigation, we report kinetic data on the anomeriza-
D
L
tion of 1-O-acetyl-2,3,5-tri-O-benzoyl-L-ribofuranose obtained by
A great deal of research has been focused on optimizing yields
and stereospecificities in the glycosylation steps of the nucleoside
synthesis. A number of protocols involving the use of 1-O-acetyl-
NMR spectroscopy and propose a mechanism for the anomeriza-
tion and formation of the acyclic side products in various acetyla-
tion media, including different acetic acid/acetic anhydride
mixtures as catalyzed by sulfuric acid or zinc chloride.
2,3,5-tri-O-benzoyl-b-L-ribofuranose (2) or the corresponding
sugar halide as the sugar component in glycosylation have been
described. The best known of these are the Wittenburg procedure3
and the silyl-Hilbert–Johnson reaction4 utilizing the sugar bromide
or, alternatively, the latter was performed by using acylated ribose
2. Results and discussion
The anomerization reactions of 1-O-acetyl-2,3,5-tri-O-benzoyl-
in the presence of
a Friedel–Crafts catalyst (Vorbrüggen
a- (1) and b-L-ribofuranoses (2) were carried out at 25 °C in a mix-
conditions).5
ture of acetic acid and acetic anhydride (5:4 v/v) containing 0.75%
sulfuric acid and monitored by 1H NMR spectroscopy. All com-
pounds formed: 1, 2 and the acyclic 1,1,4-tri-O-acetyl-2,3,5-tri-O-
We are currently investigating in detail the synthesis of
ester-protected -ribofuranoses and we became interested in the
fundamental mechanism of anomerization in 1-O-acetylated
-ribofuranoses under standard acetolysis conditions. Whereas
L
benzoyl-L
-ribose hydrate 3 were characterized by 1H and 2D
L
NMR spectroscopic techniques. The structures of all compounds
present in the anomerization reaction studies are shown in Figure
1 (for details, see Supplementary data). Pyranose forms were not
detected in these experiments.
The NMR chemical shifts of the signals arising from the different
compounds present in the reaction mixture clearly deviate from
each other, thus allowing the determination of relative product
concentrations by simple integration of the anomeric signals in
the 1H NMR spectra. As shown in Figure 2, the anomerization
extensive studies on both acid and base catalyzed anomerization
and hydrolysis of various alkyl furanosides have been reported re-
cently,6 only a few investigations have targeted the corresponding
reactions of acylated furanoses regardless of their wide use in syn-
thetic nucleoside chemistry.7 The mechanism of anomerization, as
* Corresponding author. Tel.: +358 2 2154132; fax: +358 2 2154866.
E-mail address: reko.leino@abo.fi (R. Leino).
0008-6215/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2009.02.031

J. J. Forsman et al. / Carbohydrate Research 344 (2009) 1102–1109
1103
Low concentrations of the acyclic ribose hydrate 3 were de-
tected as well, the observation being consistent with earlier reports
on endocyclic cleavage in furanosides.8 The anomerization of acet-
ylated glucopyranosides has been proposed to proceed, under sim-
ilar conditions, via a ring-opening/ring-closing process where the
ring oxygen is acetylated in the acyclic form.9 However, in the
present study, when pure 3 was treated under similar acetylation
conditions as employed for 1 and 2 using the acetylation mixture
containing 0.75 vol % sulfuric acid, none of the furanoses 1 and 2
were detected. This observation is, in fact, in agreement with the
1965 results of Painter who claimed that no ring closing of the fully
acylated acyclic intermediate took place which would require the
C4 ester to act as a nucleophile.10 The fact that no ring closing of
the fully acylated ribose hydrate 3 takes place suggests that the
formation of this product is not in equilibrium with the furanoses
but rather serves as a molecular exit. When the anomerization
reactions were repeated for 1 and 2 with 7.5 vol % of sulfuric acid
a constant accumulation of compound 3 was observed showing
10 mol % of 3 after 96 h. In order to verify whether the anomeriza-
tion takes place via endocyclic or exocyclic cleavage, the reactions
were repeated using acetic acid-1-13C and acetic anhydride-
1,10-13C2. The reactions were followed by quantitative 13C NMR
OAc
O
OBz
O
OBz
R
R
OAc
BzO
BzO
R'
R'
11
: R = OBz; R' = H
12
: R = OBz; R' = H
16: R = H; R' = OBz
17: R = H; R' = OBz
12
13
: R = OAc; R' = H
: R = OAc; R' = H
OBz OAc
OBz OAc
BzO
OAc BzO
OAc
OAc OBz
OAc OBz
3
8
Figure 1. Structures of the compounds studied in the anomerization experiments.
reactions produce equilibrium mixtures of 1 and 2 containing 29%
a and b anomers, respectively, regardless of
whether pure 1 or 2 is used as starting material.
and 71% of the
Figure 2. Time-dependent product distributions in the anomerization of (a) compound 1, and (b) compound 2 in a 5:4 (v/v) AcOH–Ac2O mixture catalyzed by 0.75% of H2SO4.

1104
J. J. Forsman et al. / Carbohydrate Research 344 (2009) 1102–1109
spectroscopy. By carrying out the reaction from 1, a slow decrease
the b-arabinofuranose 7 followed a reaction pattern similar to that
observed for 1 to give a mixture of the labeled compounds 9 and 10
in 74% and 26% proportions, respectively, results provided in Sup-
plementary data (Figure S5). Due to minor overlapping of the 13C
NMR methyl carbon signals of the acetyl groups, the accuracy of
the analytical results in the arabinose study was not as high as that
in the ribose case. When the reaction time was prolonged some
in the concentration of starting material and in the concurrent for-
mation of the 13C-labeled
a and b anomers 4 and 5 was observed as
presented in Supplementary data (Fig. S4). When, on the other
hand, compound 2 was used as the starting material, an immediate
exchange of the acetoxy group was observed followed by subse-
quent anomerization of the 13C-labeled b-anomer 5 to provide
the equilibrium mixture of 4 and 5 (Fig. 3a).
formation of the corresponding acylated acyclic L-arabinose hy-
Neither of the reactions showed any formation of the other
unlabeled anomer, clearly suggesting anomerization via exocyclic
cleavage. The difference in the reaction rates between 1 and 2
may be explained by the trans-1,2-configuration in the b-anomer,
which evidently gives rise to neighboring group participation from
the benzoyl group at C2 thus resulting in fast exchange of the 1-O-
acetyl group11. In order to verify the role of neighboring group
participation, we synthesized the analogous 1-O-acetyl-2,3,5-tri-
drate (8) was observed.
Earlier, Lindberg has studied the action of strong acids on acet-
ylated glucosides.9 He proposed that the source of catalytic activity
in a solution of sulfuric acid in acetic acid/acetic anhydride is the
acetyl cation formed upon reaction of the anhydride with sulfuric
acid. We aimed to verify this hypothesis by comparison of our re-
sults obtained from the sulfuric acid-catalyzed reactions with
those of anomerization experiments catalyzed by zinc chloride in
acetic acid–acetic anhydride. Acetic anhydride and zinc chloride
are used in Friedel–Crafts acylation reactions where the Lewis acid
supposedly reacts with the anhydride to form the active acetyl cat-
ion. The anomerization solution was prepared by dissolving 13% by
weight of zinc chloride in a mixture of acetic acid and acetic anhy-
dride (3:7, v/v). The reactions were monitored by 1H and 13C NMR
spectroscopy in order to clarify the mechanism. The anomerization
of 1 and 2 under these conditions followed the same pattern as
observed for the sulfuric acid-catalyzed reactions. It appears that
O-benzoyl-a- (6) and b- (7) L-arabinofuranoses, the structures of
which are shown in Figure 1. Similar experiments as carried out
earlier for the ribose-based compounds were then repeated using
6 and 7 producing the expected reaction pattern: The acylated
a-arabinofuranose 6 possessing 1,2-trans configuration underwent
immediate cleavage and exchange of the anomeric acetyl group
similar to the analogous b-ribofuranose to form 13C-labeled 1-O-
acetyl
a-L-arabinofuranose (9) which then anomerized to give
the corresponding b-anomer 10 (Fig. 3b). In an analogous fashion,
Figure 3. Anomerization and acetyl exchange of (a) acylated b-
and modeled results are presented as symbols and as solid lines, respectively. (a) Carbonyl carbon labeled in acetyl group. (b) Methyl carbon labeled in acetyl group.
L-ribofuranose (2) and (b) acylated a-L
-arabinofuranose (6) as studied by 13C NMR spectroscopy. Experimental

J. J. Forsman et al. / Carbohydrate Research 344 (2009) 1102–1109
1105
Figure 4. Time-dependent product distribution in the anomerization experiments of (a) 1 and (b) 2 in 0.75 vol % solution of H2SO4 in AcOH. Experimental and modeled
results are presented as symbols and as solid lines, respectively.
the change from Brønsted to Lewis acid does not influence the
mechanism of anomerization, providing further evidence for the
role of the acetyl cation as the catalytically active species, in accor-
dance with the Lindberg hypothesis. The only observable differ-
ence between the results from the two sets of experiments
(Lewis vs Brønsted acid) is that in the reactions catalyzed by zinc
chloride a smaller amount of the acyclic product was formed. Nev-
ertheless, these observations disagree with the results of Mont-
gomery and coworkers,12 who reported that the treatment of
a
-anomer 1 was still observed, but the main reaction consisted of
the formation of two new compounds 1,2-di-O-acetyl-3,5-di-O-
benzoyl- - (11) and b- (12) -ribofuranoses, characterized by a
set of 2D NMR experiments. When starting the reaction from the
-anomer 1, anomerization to form the b-anomer 2 was first ob-
a
L
a
served, followed by the subsequent formation of compounds 11
and 12 (Fig. 4).
Due to severe overlap of the 13C NMR signals from the acetyl
groups in compounds 11 and 12 and the anomeric acetyl group
of compound 2, much information on the exchange of the anomer-
ic acetoxy groups could not be extracted from the experiments
with 13C-labeled reagents. The only confident conclusion from
these 13C NMR experiments was that no immediate exchange of
the unlabeled acetyl group of compound 2 with a labeled one, as
seen in the earlier experiments, takes place.
methyl 2,3,4-tri-O-acetyl-b-D-arabinopyranoside under similar
conditions only gave acyclic acetyl methyl acetals without any for-
mation of furanoses or pyranoses. To further shed light on the
mechanism of anomerization in ribofuranoses 1 and 2, we re-
moved the acetic acid from the zinc chloride-catalyzed system
and repeated the experiments. Again, the anomerization of both
1 and 2 followed the earlier observed route to reach the equilib-
rium. However, in this case, we did not detect any formation of
compound 3 at all. This observation further supports the sugges-
tion that the anomerizations of 1 and 2, and the formation of the
acyclic ribose hydrate 3, follow different mechanistical pathways.
In addition, a set of experiments catalyzed by sulfuric acid in acetic
acid in the absence of added acetic anhydride were carried out in
order to verify whether the removal of the source of the catalyti-
cally active acetyl cation would influence the reaction mechanism.
For these experiments, a 0.75 vol % solution of sulfuric acid in ace-
tic acid was prepared. The course of the reaction changed dramat-
ically upon removing the acetic anhydride from the system: By use
of the b-anomer 2 as starting material, a slow anomerization to the
Based on these results, we propose
a
mechanism for
-ribofuranose,
theanomerizationof1-O-acetyl-2,3,5-tri-O-benzoyl-
L
wheretheacetylcationformedintheacetylationmediaactivatesthe
anomericacetylgroupforexoC–Ocleavage,aspresentedinScheme1.
The fact that no change in the overall reaction path is observed
as long as acetic anhydride, the source of the acetyl cation, is pres-
ent, is in good agreement with the earlier results of Lindberg.9 Fur-
thermore, the results from the 13C-labeled experiments clearly rule
out any anomerization via acyclic intermediates, as this would re-
sult in the formation of the opposite anomer while retaining the
original unlabeled acetoxy group. The immediate exchange of the
acetoxy groups of b-ribofuranose and
evidences the formation of a cis-fused dioxonium ion intermediate
a-arabinofuranose further

1106
J. J. Forsman et al. / Carbohydrate Research 344 (2009) 1102–1109
OBz
BzO
O
OBz
BzO
O
O
*
O
O
O
OBz
AcOH
OBz
*
5
4
AcOH
AcOH
Ph
+
O
OBz
O
OBz
BzO
+
O
O
OBz
OBz
-Ac2O
-Ac2O
-Ac2O
OBz
BzO
OBz
BzO
O
O
O
+
O
O
O
O
O
+
OBz
OBz
Ac+
Ac+
OBz
BzO
AcO
OBz
BzO
O
O
AcO
OBz
OBz
1
2
H+
H+
OBz
BzO
AcO
OBz
BzO
O
O
AcO
+
+
OBz
H
OBz
BzO
H
OBz
BzO
+
OBz OAc
Ac2O
OAc
AcO
HO
OAc OBz
OBz
3
Scheme 1. Proposed mechanism for the acid-catalyzed anomerization of 1-O-acetyl-2,3,5-tri-O-benzoyl-
a
- (1) and b-
L-ribofuranose (2).
via neighboring group participation. The retention of stereochem-
istry in the fast acetoxy exchange observed for the b-anomer 2 is
by zinc chloride. These observations indicate that the ring-opening
reaction is catalyzed by protonation of the ring oxygen followed by
subsequent acetylation to form 3, which in turn is not in equilib-
rium with 1 and 2. From the results obtained here, it can also be
not observed for the
a-anomer 1 possessing a 1,2-cis configuration.
Formation of the acyclic 1,1,4-tri-O-acetyl-2,3,5-tri-O-benzoyl-
L-ri-
bose hydrate (3), detected in the reactions catalyzed by sulfuric
acid or carried out in the presence of a Brønsted acid, did not take
place in acetic anhydride when the anomerization was catalyzed
seen that the
3 than the b-anomer 2. The observed difference toward ring open-
ing between the - and b-anomers is in good agreement with the
a-anomer 1 is more prone to form the acyclic hydrate
a

J. J. Forsman et al. / Carbohydrate Research 344 (2009) 1102–1109
1107
OBz
BzO
+
OBz OAc
OAc
AcOH
BzO
AcO
HO
OAc OBz
OBz
3
OBz
BzO
AcO
OBz
BzO
O
O
AcO
OBz
OBz
1
2
OBz
O
+
O
O
OBz
OBz
AcO
AcO
OBz
AcO
O
O
AcO
OBz
OBz
12
11
Scheme 2. Proposed mechanism for the formation of 1,2-di-O-acetyl-3,5-di-O-benzoyl-a-12 and b-L-ribofuranose 13.
previously published results, although the anomeric effects used to
explain this type of differences are not as significant in furanoses as
they are in pyranoses.13 The formation of compounds 11 and 12 in
the absence of acetic anhydride can be explained by acetyl group-
assisted replacement of the benzoyl group at C2 (Scheme 2).
As evidenced by the experimental results, it is the b-anomer 2
that is converted to the diacetylated b-anomer 12 with subsequent
anomerization to 11. As the reaction proceeds with retention of
stereochemistry at C2, it is reasonable to suggest that the 1-O-acet-
yl group at C1 participates in the replacement of the benzoyl group
at C2. The cyclic oxonium ion formed is then attacked by acetic
acid at C2 to form the 1,2-trans-diacetylated compound 12.
Reaction schemes representing the smallest amount of reac-
tions required to describe the anomerizations of 1 and 2 and the
subsequent formation of the acyclic hydrate 3, as well as the for-
mation and anomerization of the labeled compounds 4 and 5 are
depicted in Schemes 3 and 4, respectively.
Kinetic calculations of the anomerization reaction presented in
Scheme 3 were based on the 7.5% sulfuric acid containing mixture
of acetic acid and acetic anhydride (5:4 v/v). When lower concen-
trations of sulfuric acid were used, the amount of 3 formed in the
reaction was so low that the accuracy of the integration of the pro-
ton spectra was not sufficient for kinetic calculations. Kinetic mod-
eling showed that the reaction rate r2–3 is equal to zero which
further supports that compound 3 is formed only from the
mer 1. The equilibrium constant for the reaction between 1 and 2
was determined from the experimental data as K2–1 = 0.43 while
a
-ano-
3
k1-3
k2-3
k1-2
k2-1
1
2
Scheme 3.
1
2
5
k1-5
k2-4
k1-4
k2-5
k4-5
k5-4
4
Scheme 4.

1108
J. J. Forsman et al. / Carbohydrate Research 344 (2009) 1102–1109
Table 1
dent on the choice of type of acid employed (Brønsted vs Lewis
acid), providing evidence for the role of the acetyl cation as the cat-
alytically active species. Also, the dramatical change of reaction
mechanism when acetic anhydride was removed from the system
supports this hypothesis. Based on the experiments using 13C-la-
beled acetic acid and acetic anhydride it can be deduced that the
anomerization occurs via exocyclic C–O cleavage. The role of the
neighboring benzoyl group at C2 in the cleavage, and the introduc-
tion of the anomeric acetyl group are significant as it becomes evi-
dent by inspecting Figure 3. The results herein contribute to the
elucidation of fundamental reaction mechanisms in organic chem-
istry in general and in carbohydrate chemistry in particular, and
may be utilized in further development of stereospecific glycosyl-
ation reactions in nucleoside synthesis.
First order rate constants for the reactions involved in the H2SO4 catalyzed
anomerization of 1 and 2 carried out at 25 °C in different AcOH/Ac2O mixtures
Reaction conditionsa
k (hꢀ1
Reaction conditionsb
Reaction conditionsc
k (hꢀ1
)
k (hꢀ1
)
)
k1–2
k2–1
k1–3
k2–3
33.21 1.33
14.28 0.57
0.0037 0.00006
0d
k1–4
k1–5
k2–4
k2–5
k4–5
k5–4
0d
k1–2
k2–1
k1–3
k2–3
0.06 0.0014
0.034 0.0021
0.002 0.00056
0d
2.26 0.022
0d
89.4 0.12
2.35 0.0023
0.94 0.0014
k2–12
k12–11
0.058 0.0011
0.012 0.00059
a
The reaction was carried out in AcOH/Ac2O (5:4 v/v) containing 7.5% H2SO4.
The reaction was carried out in 13C-labeled AcOH/Ac2O (5:4 v/v) containing
b
0.75% H2SO4.
c
The reaction was carried out in AcOH containing 0.75% H2SO4.
Rate constants set to 0 based on kinetic modeling.
d
4. Experimental
the rate constants k1–2, k2–1, and k1–3 were estimated. The initial
rates when starting from the -anomer 1 or the b-anomer 2 are gi-
4.1. General experimental details
a
ven in Supplementary data (Table S5). The rate constants obtained
for the partial reactions involved have been listed in Table 1.
The experimental data combined with the kinetic modeling
show that, in the acetyl exchange experiments performed with
13C-labeled acetic acid and acetic anhydride, the reaction rates
r1–4 and r2–4 are equal to zero (Scheme 4). This further indicates
that regardless of whether the reaction is started from the
Synthesis of all starting compounds and detailed kinetic data
are provided in Supplementary data. The anomerization experi-
ments were carried out in sealed NMR tubes inside the magnet
thermostated to 25 °C by a Bruker variable temperature unit. The
NMR spectra were recorded using a Bruker Avance 600 MHz spec-
trometer equipped with a 5-mm inverse z-axis fg probe operating
at 600.13 MHz for 1H and at 150.92 MHz for 13C. 1H NMR spectra
were acquired with single-pulse excitation, 45° flip angle, pulse re-
cycle time of 3.6 s, and with spectral widths of 12 kHz consisting of
64 k data points. The quantitative 13C NMR spectra were recorded
with single-pulse excitation, 90° flip angle, pulse recycle time of
10 s, and with spectral widths of 3 kHz consisting of 64 k data
points. Inverse-gated decoupling techniques were applied in order
to avoid NOE.
a-anomer 1 or the b-anomer 2, the starting material is always first
converted into the labeled b-anomer which subsequently
5
anomerizes to reach the equilibrium between compounds 4 and
5. The equilibrium constant for the reaction between 4 and 5
was determined from the experimental data as K5–4 = 0.399.
In the reactions where the anomerization was studied in the ab-
sence of acetic anhydride (Fig. 4), compound 3 was formed in small
quantities when starting from 1. When, on the other hand, 2 was
used as the starting material, compound 3 was not detected at
all. That r2–3 in Scheme 5 equals to zero was also shown by the ki-
netic modeling which is in good agreement with all other results
on the formation of compound 3. The equilibrium constant for
the anomerization of 1 and 2 in the absence of acetic anhydride
is given in Supplementary data (Table S6).
In the modeling, the reaction r12–11 was considered to be irre-
versible as no equilibrium between the compounds 11 and 12
was formed during the time period when the reaction was followed.
For this reason we could not determine the equilibrium constant
K12–11, although the eventual equilibrium between compounds
11 and 12 would probably have been established if the reaction
was allowed to continue for a longer time period.
The kinetics of the anomerization process was described with a
first order reaction kinetics model. The reactor was described with
a batch reactor model.
dci
¼ ri
dt
For the parameter estimation the following objective function was
minimized:
X X
2
Q ¼
ðci;t;exp ꢀ ci;t;modelÞ wi;t
t
i
where ci,t,exp and ci,t,model are the experimentally recorded concen-
trations and the concentrations predicted by the model, respec-
tively. The weight factor w was set to 1 for all experimental
points. The software Modest was used to estimate the rate con-
stants and to solve the reactor mass balances, the software mini-
mizes the objective function with the Levenberg–Marquardt
method, and solves the ODEs describing the reactor model by the
backward difference method.14 The fit of the model to experimental
data of the reactions is presented in Figures 3 and 4 and in Supple-
mentary data (Figs. S7–S14). As can be seen from the figures, the
model can describe the experimental data very well. The estimated
rate constants, kinetic models, and reactor component mass bal-
ances are listed in Tables S1–S4 (Supplementary data). The esti-
mated kinetic constants are well identified (parameter sensitivity
analysis plots in Supplementary data) and all the parameter errors
are low.
3. Summary and conclusions
To summarize, new mechanistic and kinetic data on the ano-
merization of acylated L-ribofuranoses under different acetolysis
conditions have been obtained thus contributing to the under-
standing of fundamental phenomena in carbohydrate chemistry.
Our results show that the mechanism of anomerization is indepen-
3
k1-3
k2-3
k1-2
k2-1
1
2
k2-12
12
4.2. General procedures for the anomerization reactions in
AcOH–Ac2O mixture with H2SO4
k12-11
11
Compounds 1–3 and 6–7 (20 mg) were dissolved in 550
ll of
CD2Cl2. To this solution was added 63 l of a 5:4 (v/v) mixture of
l
Scheme 5.

J. J. Forsman et al. / Carbohydrate Research 344 (2009) 1102–1109
1109
AcOH and Ac2O containing 0.75 or 7.5 vol % H2SO4. 1H NMR spectra
were recorded at different time intervals and the molar concentra-
tions of the products were determined from the integral ratios.
determination of the reaction rates. The experiments were per-
formed following the general procedures given above.
Acknowledgments
4.3. General procedure for the anomerization in AcOH–Ac2O
catalyzed by ZnCl2
Financial support to J.F. from the Finnish National Glycoscience
Graduate School is gratefully acknowledged. The authors also wish
to thank Professor Harri Lönnberg and Dr. Patrik Eklund for fruitful
discussions. Mr. Markku Reunanen is thanked for recording the
EIMS and HRMS analysis.
Compounds 1 and 2 (20 mg) were dissolved in 550
ll of CD2Cl2.
A solution of ZnCl2 (20 mg) in AcOH (37 l) and Ac2O (85
l
ll) was
prepared. The reaction was started by mixing the two solutions
in the NMR tube. 1H NMR spectra were recorded at different time
intervals and the molar concentrations of the products were deter-
mined from the integral ratios.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.carres.2009.02.031.
4.4. General procedure for the anomerization in Ac2O catalyzed
by ZnCl2
References
Compounds 1 and 2 (20 mg) were dissolved in 550
To this solution was added a solution of ZnCl2 (5 mg) in Ac2O
(100
l). 1H NMR spectra were recorded at different time intervals
and the molar concentrations of the products were determined
from the integral ratios.
ll of CD2Cl2.
1. (a) Mathé, C.; Gosselin, G. Antiviral Res. 2006, 71, 276–281; (b) Gumina, G.;
Song, G.-Y.; Chu, C. K. FEMS Microbiol. Lett. 2001, 202, 9–15.
2. (a) Sabini, E.; Hazra, S.; Konrad, M.; Lavie, A. J. Med. Chem. 2007, 50, 3004–3014;
(b) Hocek, M.; Šilhár, P.; Shih, I.; Mabery, E.; Mackman, R. Bioorg. Med. Chem.
Lett. 2006, 16, 5290–5293; (c) Herbal, K.; Kitteringham, J.; Voyle, M.;
Whitehead, A. J. Tetrahedron Lett. 2005, 46, 2961–2964; (d) Shi, J.; Du, J.; Ma,
T.; Pankiewicz, K. W.; Patterson, S. E.; Tharnish, P. M.; McBrayer, T. R.; Stuyver,
L. J.; Otto, M. J.; Chu, C. K.; Schinazi, R. F.; Watanabe, K. A. Bioorg. Med. Chem.
2005, 13, 1641–1652.
l
4.5. General procedure for the anomerization in AcOH
catalyzed by H2SO4
3. Wittenburg, E. Chem. Ber. 1968, 101, 1095–1114.
4. (a) Hilbert, G. E.; Johnson, T. B. J. Am. Chem. Soc. 1930, 52, 4489–4494; (b)
Kotick, M. P.; Szantay, C.; Bardos, T. J. J. Org. Chem. 1969, 34, 3806–3813.
5. Niedballa, U.; Vorbrüggen, H. J. Org. Chem. 1974, 39, 3654–3660.
6. (a) Nowacki, A.; Smiataczowa, K.; Kasprzykowska, R.; Dmochowska, B.;
Wis´niewski, A. Carbohydr. Res. 2002, 337, 265–272; (b) Garegg, P. J.;
Johansson, K.-J.; Konradsson, P.; Lindberg, B. J. Carbohydr. Chem. 1999, 18,
31–40; (c) Snyder, J. R.; Serianni, A. S. Carbohydr. Res. 1988, 184, 13–25.
7. Zhang, P.; Dong, Z. E.; Cleary, T. P. Org. Process Res. Dev. 2005, 9, 583–592.
8. Lönnberg, H.; Kulonpää, A. Acta Chem. Scand. A 1977, 31, 306–312.
9. Lindberg, B. Acta Chem. Scand. 1949, 3, 1153–1169.
Compounds 1 and 2 (20 mg) were dissolved in 550 ll of CD2Cl2.
Next, 63 ll of a 0.75 vol % solution of H2SO4 in AcOH was added to
initiate the reaction. 1H NMR spectra were recorded at different
time intervals and the molar concentrations of the products were
determined from the integral ratios.
4.6. General procedures for the quantitative 13C experiments
10. Painter, E. P. Chem. Ind. (London) 1965, 1380–1381.
11. Štimac, A.; Kobe, J. Carbohydr. Res. 2000, 324, 149–160.
12. Montgomery, E. M.; Hann, R. M.; Hudson, C. S. J. Am. Chem. Soc. 1937, 59, 1124–
1129.
13. (a) McPhail, D. R.; Lee, J. R.; Fraser-Reid, B. J. Am. Chem. Soc. 1992, 114, 1905–
1906; (b) Praly, J.-P.; Lemieux, R. U. Can. J. Chem. 1987, 65, 213–223.
14. Haario, H. Modest 6.0—A User’s Guide; ProfMath: Helsinki, 2001.
For all reactions with compounds 1–3 acetic acid-1-13C and ace-
tic anhydride-1,10-13C2 were used, while with compounds 6 and 7
acetic acid-2-13C and acetic anhydride-2,20-13C2 were used to ob-
tain the best possible separation of the acetyl resonances for the