Catalytic activity of lanthanide(III) ions for the dehydration of hexose to 5-hydroxymethyl-2-furaldehyde in water

Article abstract of DOI:10.1246/bcsj.74.1145

All of the lanthanide(III) (La³⁺-Lu³⁺) efficiently catalyzed the dehydration of hexose in water at 140°C to produce 5-hydroxymethyl-2-furaldehyde (HMF) without accompanying further hydrolysis to levulinic acid, which is routinely obtained as a by-product in the conventional Bronsted acid-catalyzed reactions. The relationships between the catalytic activity and the ionic radius (r_i) of lanthanides(III) for the dehydration of D-fructose and D-glucose were found to have a double-peak profile maximizing at Pr or Nd (r_i 1.00 A) and Er or Yb (r_i 0.87 A) with a dip at Sm (r_i 0.96 A). Kinetic analyses revealed that the rate-determining step is not the complex formation of lanthanide(III) ion with the saccharide molecule, but a subsequent reaction from the substrate-catalyst complex. Possible reasons why the catalytic activity-ionic radius relationships show a two-peak profile are discussed.

Full text of DOI:10.1246/bcsj.74.1145

c. Jpn.
e Chemical © 2001 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 74, 1145—1150 (2001) 1145
ciety of Ja-
n
e Chemical
ciety of Ja-
n
Catalytic Activity of Lanthanide(III) Ions for the Dehydration of Hexose
to 5-Hydroxymethyl-2-furaldehyde in Water
Lanthanide-Catalyzed Dehydration of Hexose
K. Seri et al.
†
,†
Kei-ichi Seri, Yoshihisa Inoue, and Hitoshi Ishida*
01
4
5
5
0
Interior Design Research Institute, Fukuoka Industrial Technology Center, Ohkawa, Fukuoka 831-0031
Inoue Photochirogenesis Project, ERATO, JST, 4-6-3 Kamishinden Toyonaka, Osaka 560-0085
†
SJA8
0
9-2673
(
Received January 22, 2001)
021
nuary
3
ꢀ
3ꢀ
All of the lanthanide(III) (La –Lu ) efficiently catalyzed the dehydration of hexose in water at 140 ˚C to produce
-hydroxymethyl-2-furaldehyde (HMF) without accompanying further hydrolysis to levulinic acid, which is routinely
01
01
.5
5
obtained as a by-product in the conventional Brønsted acid-catalyzed reactions. The relationships between the catalytic
activity and the ionic radius (r ) of lanthanides(III) for the dehydration of D-fructose and D-glucose were found to have a
double-peak profile maximizing at Pr or Nd (r 1.00 Å) and Er or Yb (r 0.87 Å) with a dip at Sm (r 0.96 Å). Kinetic
i
i
i
i
analyses revealed that the rate-determining step is not the complex formation of lanthanide(III) ion with the saccharide
molecule, but a subsequent reaction from the substrate-catalyst complex. Possible reasons why the catalytic activity–
ionic radius relationships show a two-peak profile are discussed.
Saccharides, including cellulose, have recently attracted
much attention as biomass for substituting petroleum resourc-
es, and their conversion to useful chemicals has actively been
investigated as a promising route to a sustainable, versatile car-
(2)
1
2–6
bon source. The dehydration of hexose, such as D-glucose
4,6–13
and D-fructose,
giving 5-hydroxymethyl-2-furaldehyde
In this paper, we report on the catalytic activities of lan-
thanide(III) ions for the dehydration of D-glucose, D-fructose,
D-galactose, and D-mannose to HMF in water. We elucidate
the relationship between the catalytic activity and the ionic ra-
dius of lanthanide(III) for the dehydration reaction of D-glu-
cose and D-fructose, and further discuss the possible factors
and mechanisms controlling the activity through kinetic analy-
ses.
(
HMF), has been known as a primary process for converting
saccharides to a variety of chemicals (Eq. 1). This reaction is
2
–4,6–9,11,12,14,15
catalyzed by various acids. Thus, Brønsted acids
have been widely used, while the use of Lewis acids
5
,7,10,13
is
quite limited in number, probably because they are active as
catalysts only in organic solvents. Although it is advantageous
that the Brønsted acid-catalyzed dehydration proceeds in aque-
ous solution, the initial product HMF often suffers further hy-
dration to levulinic acid under the same reaction condition (Eq.
Results and Discussion
2
), and therefore control of the dehydration/hydration process-
Catalytic Conversion of Monosaccharides to HMF by
Lanthanide(III) Ions. The dehydration of D-glucose, D-
fructose, D-galactose, and D-mannose in the presence of a cata-
lytic amount of lanthanide(III) salt was carried out in water at
140 ˚C in an autoclave. The major product, HMF, was isolated
1
es is difficult in general. In our recent study in search of novel
catalysts for converting saccharides to chemicals,
for the first time that lanthanide(III) ions can efficiently cata-
lyze the conversion (dehydration) of saccharides to HMF in or-
ganic solvents, such as DMSO and DMF; e.g. D-fructose is
converted to HMF in > 90% yield at 100 ˚C in 3.5 h. More
interestingly, this catalyzed dehydration proceeds even in
aqueous solutions.
1
6,17
we found
by extraction with benzene, and identified by NMR and mass
1
7
16
spectroscopy.
The yields of HMF, obtained after heating
ꢁ
3
aqueous solutions of hexoses (0.30 mol dm ) and lan-
thanide(III) salts (2.0 mmol dm ) for 1 h in water at 140 ˚C,
1
6
ꢁ3
are listed in Table 1. Although practically no HMF was pro-
duced in the absence of the lanthanide(III) catalysts, except for
the D-fructose case, a catalytic amount of lanthanide(III) dra-
matically enhanced the rate of HMF production in all cases, as
can be seen from Table 1. It is interesting to note that the for-
mation rate of HMF increases with increasing atomic number,
or decreasing ionic radius, of lanthanide(III). Among the hex-
oses used, D-fructose is most efficiently dehydrated to HMF
both in the presence and in the absence of lanthanide(III) ions,
giving the highest product yields and catalytic enhancements.
(
1)

1
146 Bull. Chem. Soc. Jpn., 74, No. 6 (2001)
Lanthanide-Catalyzed Dehydration of Hexose
Table 1.
Yields (in µmol) of 5-Hydroxymethyl-2-furaldehyde (HMF) Produced in the Ln(III)-Catalyzed Reaction of Hexose
3
in Water (15 cm ) at 140 ˚C for 1 h
a)
Catalyst
b)
Substrate
None
198
trace
trace
trace
LaCl
707
125
152
180
3
NdCl
741
219
164
182
3
EuCl
788
252
234
210
3
DyCl
773
333
255
243
3
YbCl
831
363
264
222
3
D-Fructose
D-Glucose
D-Mannose
D-Galactose
ꢁ
3
ꢁ3
a) 2.0 mmol dm . b) 0.30 mol dm
.
Fig. 1.
DyCl
consumed substrate (square).
HMF and substrate concentrations employed in the dehydration reactions of (a) D-fructose and (b) D-glucose catalyzed by
ꢁ
3
3
(2.0 mmol dm ) at 140 ˚C in water: the amounts of substrate (closed circle) and HMF (open circle) and the yield based on
This is consistent with the observations reported for other acid-
catalyzed dehydration reactions that ketohexoses, such as D-
fructose, afford HMF in better yields than aldohexoses, such as
8
,10,18
D-glucose, D-galactose, and D-mannose.
The time profiles of the dehydration of D-fructose and D-
glucose catalyzed by DyCl in water at 140 ˚C are shown in
3
Fig. 1, where the amounts of substrate and product, as well as
the yield of HMF based on the consumed substrate, are plotted
as functions of the reaction time. As can be seen from Fig. 1,
both D-fructose and D-glucose are smoothly dehydrated to give
HMF catalytically and the product yields based on conversion
are higher than 95%, at least in the initial stages of the reaction
Scheme 1.
Reaction scheme for acid-catalyzed dehydra-
12
tion of D-fructose or D-glucose, proposed by Kuster et al. ;
the thick solid line (arrow) (i) denotes the lanthanide cata-
lyzed path, the thin solid line (iii) indicates the non-cata-
lyzed path, and the dashed lines ((ii) and (iv)) denote the
steps which were not observed in the lanthanide-catalyzed
reactions.
(
< 15 min). However, the material balance gradually decreases
after 20 min in both cases, reaching the ultimate product yield
1
9
of ca. 40% attained at 120 min.
In the elucidation of possible mechanism(s) for the gradual-
ly decreasing HMF yield based on the consumed saccharide,
we first examine the mechanism of a Brønsted acid-catalyzed
thanide ions, because the substrate consumption matches to the
HMF production, and no appreciable side reaction is seen at
the initial stages of the reaction. In order to evaluate the possi-
ble intervention of the indirect path (iii), HMF was subjected
to catalytic dehydration under the same conditions. Thus,
1
2
reaction of hexoses proposed by Kuster et al. (Scheme 1).
There are four successive/competing processes in the Scheme:
i.e. (i) production of HMF from hexose, (ii) direct formation of
char, a polymeric material insoluble in both water and organic
solvents, (iii) indirect formation of char via HMF, and (iv) fur-
ther hydration of HMF to levulinic acid. In the present case
using the lanthanides(III) as catalysts, path (i) is obviously the
major process throughout the reaction. The direct path to char
ꢁ
3
aqueous solutions of HMF (300 mmol dm ) were heated to
ꢁ
3
140 ˚C in the presence and absence of LaCl (2.0 mmol dm ).
3
After 120 min of heating, the amounts of recovered HMF were
ꢁ
3
219 and 205 mmol dm in the presence and absence of the
lanthanide catalyst, respectively. Although ca. 30% of HMF
was certainly consumed in both cases, the presence of the cata-
lyst did not affect the rate of consumption. These results clear-
(
ii) from saccharides does not appear to be catalyzed by lan-

K. Seri et al.
Bull. Chem. Soc. Jpn., 74, No. 6 (2001) 1147
2
1
ly indicate that the non-catalytic, indirect char formation, path
diaminetetraacetic acid (EDTA), 1,2-diaminocyclohexane-
2
2
(
iii), occurs in the dehydration of saccharides, consuming the
N,N,N′,N′-tetraacetic acid (DCTA), and diethylenetriamine-
23
product HMF. Significantly, levulinic acid was not detected in
the reaction mixture, even after 120 min, which rules out the
possible decomposition of HMF through path (iv) in the
present dehydration catalyzed by lanthanides. In summary, we
have revealed that the lanthanide-catalyzed dehydration of
hexoses preferentially affords HMF in > 95% yield at the ini-
tial stages, but unfortunately the product HMF is not stable un-
der the reaction conditions and suffers non-catalytic secondary
char formation without accompanying the decomposition to le-
vulinic acid. Since the material balance is excellent and no se-
rious decomposition of HMF to char is observed during the
initial stages of the reaction, all of the following discussion is
made based on the initial rates.
N,N,N′,N′′,N′′-pentaacetic acid (DTPA), display sudden
changes at Gd. However, the break in the rate of dehydration
occurs not at Gd, but at Sm, in the present cases. This rather
unusual break at Sm may be related to a change in the hydra-
2
4
tion number. Hydration energy calculations indicate that lan-
thanide(III) ions lighter than Sm are most stabilized as nonahy-
drates, while heavier ones are stable as octahydrates. Indeed,
the hydration numbers of lanthanide(III) ions in water are re-
ported to be 9 for La to Nd, 8 or 9 for Nd to Tb, and 8 for Tb to
2
5–27
Lu.
Hence, it is not unreasonable to attribute the Sm-break
observed for the catalytic activity to the hydration number
change somewhere between Nd and Tb. We further discuss the
plausibility of this hypothesis later.
Catalytic Activity and Ionic Radius of Lanthanide(III).
In order to elucidate the relationships between the catalytic ac-
tivity and the ionic radius of lanthanide(III), the initial rates of
catalyzed dehydration were measured for D-fructose and D-
glucose in the presence of various lanthanides(III) under com-
parable conditions. The initial rates are plotted against the ion-
ic radius of lanthanides in Fig. 2, where the net rates for D-
fructose were calculated by subtracting the blank rates (ob-
tained in the absence of catalyst) from the observed initial rates
Kinetic Analysis. To effect catalytic dehydration, lan-
thanide(III) ions have to interact with the saccharide sub-
strates. In order to elucidate the lanthanide(III)–substrate in-
teraction, we first performed NMR measurements with
2
mixtures of several lanthanide ions and hexoses in D O. How-
ever, no appreciable spectral changes were observed in the
shape and chemical shifts of saccharide proton signals. Previ-
2
0
2
8
29
ous NMR and calorimetric studies also indicate that the in-
teractions between lanthanide(III) ions and neutral saccharides
are extremely weak.
in the presence of catalysts. Although the reaction rate (v
shows a global increasing tendency with decreasing ionic radi-
us (r ) from La to Lu, closer examinations of the relationships
obtained for both D-fructose and D-glucose exhibit the charac-
teristic two-peak profiles, where the v value maximizes at Pr
or Nd (r 1.00 Å) and Er or Yb (r 0.87 Å) with a drop at Sm (r
.96 Å).
It is well documented that several properties of lan-
0
)
i
0
i
i
i
0
Scheme 2.
lanthanide(III)-catalyzed dehydration of saccharides,
where K
Michaelis–Menten mechanism applied to the
thanide(III) ions are not simple, but discontinuous, functions
of the ionic radius, often displaying a break somewhere in the
middle of the lanthanide series, most at Gd. For instance, the
stability constants of the lanthanide(III) complexes of ethylen-
d 1
ꢂ (kꢁ1 ꢀ kcat)/k .
In order to gain insights into the nature of lanthanide(III)–
saccharide complexes and the subsequent reaction rates, we
performed a kinetic analysis based on the Michaelis–Menten-
type mechanism, shown in Scheme 2, and the Michaelis–
Menten Eqs. 3 and 4. The initial rates of dehydration were
measured for D-fructose and D-glucose at varying concentra-
ꢁ
3
tions of 0.10–0.50 mol dm in the presence of LaCl
3 3
, NdCl ,
EuCl , DyCl , or YbCl at a fixed concentration of 2.0 mmol
3
3
3
3
ꢁ
dm . Plots of the inverse reaction rates against the reciprocal
substrate concentrations gave the good straight lines shown in
Fig. 3. From the slopes and intercepts of the regression lines,
d
the dissociation constants (K ) and the catalytic reaction rate
constants (kcat) were calculated (see experimental section for
ꢁ
3
d
detail). The K values obtained are 0.2–0.3 mol dm for D-
ꢁ
3
fructose and 0.2–0.4 mol dm for D-glucose. These large dis-
sociation constants mean weak interaction between the lan-
thanide(III) ions and the saccharides, which is in good agree-
28,29
ment with the results of NMR and calorimetric studies.
1
K
d
1
1
ꢂ
ꢂ
•
ꢀ
(3)
(4)
3
ꢁ1
ν
0
k
cat[Cat]
0
[Sub]
kcat[Cat]
0
Fig. 2.
s
Relationship between the initial rate (in dm mol
) and the ionic radius of lanthanides(III) for HMF pro-
duction from D-fructose (closed circle) and D-glucose
ꢁ
1
k
ꢁ1 ꢀ kcat
K
d
ꢁ
3
k1
(
open circle); the catalyst concentration: 2.0 mmol dm .

1
148 Bull. Chem. Soc. Jpn., 74, No. 6 (2001)
Lanthanide-Catalyzed Dehydration of Hexose
ꢁ
1
ꢁ1
Fig. 3.
Plots of v
0
vs [S] for dehydration of (a) D-fructose and (b) D-glucose catalyzed by lanthanide(III) ions (2.0 mmol
, (C) SmCl , (D) DyCl , and (E) LuCl
ꢁ
3
dm ): (A) LaCl , (B) NdCl
3
3
3
3
3
.
Fig. 4.
d
Kinetic parameters (K (open circle) and kcat (closed circle)) as functions of the ionic radii of lanthanides(III) for catalytic
dehydration of (a) D-fructose and (b) D-glucose.
The K
d
and kcat values are plotted against the ionic radius of
),
d
stant (K ) with increasing atomic number of lanthanide(III),
lanthanide(III) in Fig. 4. The dissociation constants (K
d
the lanthanide(III)–saccharide interaction becomes stronger as
the ionic radius decreases and the surface charge increases.
This tendency clearly indicates that the complexation of lan-
thanide(III) ion with hexose is primarily driven by the electro-
static interaction. Similarly, the catalytic reaction rate constant
(kcat) also increases with decreasing ionic radius, accompany-
ing a break at Eu, probably due to a change in the hydration
number. The enhanced reactivities observed for the heavier
lanthanides are also rationalized by the increased electrostatic
interactions. It is therefore evident that the nature and struc-
ture of the lanthanide-saccharide complex are the keys to this
lanthanide(III)-catalyzed dehydration of saccharides in aque-
ous solution. However, such complexes are not stable in neu-
tral aqueous solution, and the structures of such complexes
have rarely been elucidated in detail. It has been suggested
that the lanthanide(III) complex of saccharide is stabilized
when three adjacent hydroxy groups on the pyranose ring are
which monotonically decrease with decreasing ionic radius,
are fairly large and no significant differences in the magnitude
or tendency were observed between the two substrates. Not-
withstanding the large dissociation constants, more than 50%
of the lanthanide ion existing in the system can form a cata-
lyst–substrate complex under the reaction condition, as the
substrate concentration is very high (0.30 mol dm ). This
means that changes in the K values do not significantly affect
the initial reaction rate, and that the rate-determining step is
not the complex formation process, but a subsequent reaction
from the complex. In contrast, the kcat value differs for each
substrate and shows a sudden jump at Eu in both cases. Thus,
the consistently faster reaction rates (v ) for D-fructose and the
0
discontinuous change around Sm in the catalytic activity are
reasonably attributed to the kcat value.
Interaction and Reaction Mechanisms. As can be rec-
ognized from the monotonically decreasing dissociation con-
ꢁ
3
d
2
8,30
in the axial, equatorial, axial conformations.
However, D-

K. Seri et al.
Bull. Chem. Soc. Jpn., 74, No. 6 (2001) 1149
scarcely been elucidated so far because such complexes are
hardly isolated due to the very weak interaction. The lan-
thanide-catalyzed dehydration of saccharides in organic sol-
vents is currently under investigation for significantly different
catalysis profiles. Further application of the present catalytic
reaction to a process that directly transforms biomass, such as
oligosaccharides, starch and cellulose, to some useful chemi-
cals is also a future task.
Experimental
Materials. Lanthanide(III) chlorides and saccharides of the
highest available grade were purchased from Wako Pure Chemical
Industries, Ltd. and used without further purification.
3
Catalytic Reactions. An aqueous solution (15 cm ) of
ꢁ
3
monosaccharides (0.30 mol dm ) and lanthanide(III) chloride
ꢁ
3
(
1.0–5.0 mmol dm ) was charged in a glass tube (25 mmφ ꢃ 165
Fig. 5.
Schematic drawing of the proposed lanthanide(III)
mm) which was placed in an autoclave (Taiatsu Techno Corpora-
tion, TVS-N2 (100 cm )), and the whole system was heated to 140
3
ion–D-glucose complexes.
˚
C under a nitrogen atmosphere at a 10 bar pressure. The temper-
glucose does not exactly meet this condition. Hence, we spec-
ulate that the conformationally flexible 6-hydroxy group of D-
glucose, as well as the 1-axial- and 2-equatorial-hydroxys, co-
ordinate to the lanthanide ion to form the lanthanide–glucose
complex shown in Fig. 5.
ature was kept constant at 140 ˚C for a given period of time by a
heater equipped with a digital thermometer (CT-700S). After
cooling the autoclave to room temperature, the resultant mixture
was subjected to a gas chromatographic analysis on a Shimadzu
GC-9A instrument equipped with an FID detector and a glass col-
umn (3.2 mmφ ꢃ 1.6 m) packed with polyethylene glycol 20 M
Although the detailed mechanism of the dehydration pro-
cess is not clear for the lanthanide-catalyzed dehydration of
saccharides, the Brønsted acid-catalyzed dehydration is be-
lieved to be initiated by removal of the 2-hydroxy of D-glucose
(
2
10% (w/w) supported on acid-washed Chromosorb W); N was
used as a carrier gas. The chromatograms were recorded and inte-
grated by a Shimadzu C-R6A integrator. The decrease in the sub-
strate was analyzed by HPLC, using a Waters LC Module 1 plus
equipped with a differential refractometer, Waters 410. A Sugar-
Pak Ca column (7.8 mmφ ꢃ 300 mm) was used with a guard col-
umn (6 mmφ ꢃ 50 mm). Degassed water was used as the eluent.
The chromatographs were monitored and recorded by using soft-
ware, Waters Chromatograph Manager, Millennium 2010J, work-
ing on a computer, Digital VENTRIUS 4100.
10
or D-fructose. In this context, it seems reasonable to con-
clude that the markedly enhanced reactivities upon lanthanide-
catalyzed dehydration with Dy–Lu (Figs. 2 and 4) are correlat-
ed to the stronger and closer contacts with the less-hydrated,
smaller-sized, heavier lanthanide ions.
Conclusion
Kinetic Analysis. The initial rates (v
0
) of the dehydration of
ꢁ
3
The catalytic activity of lanthanide(III) ions for the dehydra-
tion of hexoses, particularly D-glucose and D-fructose, yielding
HMF has been investigated. The lanthanide(III) ions are ex-
cellent Lewis acid catalysts giving HMF without accompany-
ing an undesirable further decomposition to levulinic acid, a
process which is often observed upon dehydration with con-
ventional Brønsted acids. The relationship between the cata-
lytic activity and the ionic radius displays the characteristic
two-peak profile maximizing at Pr (or Nd) and Er (or Yb) with
a break at Sm. A kinetic analysis using the Michaelis–Menten
model clearly demonstrated that the rate-determining step is
not complex formation, but a subsequent dehydration reaction.
D-fructose and D-glucose (0.10–0.50 mol dm ) catalyzed by lan-
ꢁ
3
thanide(III) ions (2.0 mmol dm ) were analyzed by using the
Michaelis–Menten model (Scheme 2). In the D-fructose case, the
net rates were corrected for the appreciable dehydration which oc-
curred in the absence of lanthanide, by subtracting the blank rate
from the observed rates. According to Eqs. 3 and 4, the inverse
initial rates were plotted against the corresponding inverse sub-
strate concentrations, as shown in Fig. 3. The dissociation con-
d
stants (K ) and catalytic reaction rate constants (kcat) were calcu-
lated from the slopes and y-intercepts of the plots.
The present work was supported by the Basic Research
Project Fund for New Technology Creation from the Depart-
ment of Commerce and Industry, Fukuoka Prefecture.
d
Although the binding ability (1/K ) of the lanthanide(III) ions
towards the saccharides monotonically increases with decreas-
ing ionic radius as a consequence of the smaller catalyst–sub-
strate distance in the complex, the reactivity (kcat) shows an
abrupt jump between Eu and Dy for both D-glucose and D-
fructose. Thus, the characteristic two-peak profile observed
References
1
E. Sjöström, “Wood Chemistry, Fundamentals and Appli-
0
for the reaction rate (v ) is attributable to the change in the hy-
cations,” 2nd ed, Academic Press, New York (1993).
dration number of lanthanide ions around Sm.
2
M. L. Mednick, J. Org. Chem., 27, 398 (1962).
This paper not only reports on the novel catalytic activity of
lanthanide(III) ions, but also provides information through a
kinetic analysis about the reactivity of lanthanide(III) complex
with D-fructose or D-glucose in aqueous solutions, which has
3
D. W. Harris and M. S. Feather, J. Org. Chem., 39, 724
(1974).
4
5
C. J. Moye, Aust. J. Chem., 19, 2317 (1966).
T. G. Bonner, E. J. Bourne, and M. Ruszkiewicz, J. Chem.

1
150 Bull. Chem. Soc. Jpn., 74, No. 6 (2001)
Lanthanide-Catalyzed Dehydration of Hexose
Soc., 1960, 787.
hydrolysis at temperatures lower than 140 ˚C, but the reaction be-
came significantly slow. Hence, we employed the present reaction
condition.
6
B. F. M. Kuster and H. M. G. Temmink, Carbohydr. Res.,
5
4, 185 (1977).
7
C. J. Moye and Z. S. Krizeminski, Aust. J. Chem., 16, 258
20 The ion radii used in this study are for hexahydrates: R. D.
Shannon, Acta Crystallogr., Sect. A, 32, 751(1976).
21 E. J. Wheelwright, F. H. Spedding, and G. Schwarzenbach,
J. Am. Chem. Soc., 75, 4196 (1953).
22 T. Moeller, J. Chem. Educ., 47, 417 (1970).
23 T. Moeller, D. F. Martin, L. C. Thompson, R. Ferrus, G. R.
Feistel, and W. J. Randall, Chem. Rev., 65, 1 (1965).
24 K. Miyakawa, Y. Kaizu, and H. Kobayashi, J. Chem. Soc.,
Faraday Trans., 84, 1517 (1988).
25 A. Habenshuss and F. H. Spedding, J. Chem. Phys., 70,
2798 (1979).
26 A. Habenshuss and F. H. Spedding, J. Chem. Phys., 70,
3758 (1979).
(
1962).
8
B. F. M. Kuster and H. S. Van der Baan, Carbohydr. Res.,
5
4, 165 (1977).
P. E. Shaw, J. H. Tatum, and R. E. Berry, Carbohydr. Res.,
, 266 (1967).
H. H. Szmant and D. D. Chundury, J. Chem. Tech. Biotech-
nol., 31, 135 (1981).
B. F. M. Kuster and H. M. G. Temmink, Carbohydr. Res.,
4, 266 (1977).
9
5
1
0
1
1
5
1
1
1
1
2
3
4
5
B. F. M. Kuster, Carbohydr. Res., 54, 177 (1977).
L. R. G. N. Poncini, Carbohydr. Res., 87, 209 (1980).
F. H. Newth, Adv. Carbohydr. Chem., 6, 83 (1951).
W. N. Haworth and W. G. M. Jones, J. Chem. Soc., 1944,
27 A. Habenshuss and F. H. Spedding, J. Chem. Phys., 73, 442
(1980).
6
67.
1
6
H. Ishida and K. Seri, J. Mol. Catal. A: Chem., 112, L163
28 S. J. Angyal, D. Greeves, L. Littlemore, and V. A. Pickles,
Aust. J. Chem., 29, 1231 (1976).
(
1996).
17
18
19
K. Seri, Y. Inoue, and H. Ishida, Chem. Lett., 2000, 22.
E. F. L. J. Anet, Adv. Carbohydr. Chem., 19, 181 (1964).
In an effort to suppress the formation of by-products and to
29 N. Morel-Desrosiers, C. Lhermet, and J. P. Morel, J. Chem.
Soc., Faraday Trans., 89, 1223 (1993).
30 S. J. Angyal, Aust. J. Chem., 53, 567 (2000).
improve the product yield based on conversion, we performed the