5
404
C. Li et al. / Tetrahedron Letters 50 (2009) 5403–5405
HO
HO
OH
O
OH
O
O
HO
HO
HO
HO
O
OH
O
HO
O
O
OH
OH
OH
n OH
Cellulose
OH
OH
O
O
O
O
Hydrolysis HO
Dehydration
OH
HO
OH
Glucose
HMF
2,5-dimethylfuran
Scheme 1. Schematic illustration of the steps for cellulose-to-HMF.
Table 1
HMF production from glucose under various conditionsa
Although an exact catalytic mechanism for the exceptional
effectiveness of CrCl in the conversion of cellulose remains elu-
sive, we would like to offer some insights into the reaction. As
illustrated in Scheme 2, we imagined that CrCl in [C mim]Cl might
form complexes [C mim] [CrCl3+n] (n = 1–3) in a way similar to
LnCl reported by Rogers and co-workers. In the cellulose hydro-
3
Entry
1
Catalyst
Catalyst loading (wt%)
Time (min)
Yieldb (%)
3
4
CrCl
CrCl
CrCl
3
3
3
3.6
3.6
3.6
10
1
60
10
1
91
17
<1
49
c
2
3
4
n
d
15
3
4
2 4
H SO
lysis step, the 1,4-glucosidic bonds were weakened partially be-
nꢁ
a
cause of coordination with [CrCl3+n
]
, resulting in more liable to
Unless otherwise specified, conditions were glucose (100 mg) in [C
1.0 g) under MI at 400 W.
4
mim]Cl
(
water attack to form glucose and oligomers. Then, the complex
promoted rapid mutarotation of the a-anomer of glucose to the
b-one through hydrogen bonds of chloride anions with the hydro-
b
Isolated yields.
Heated with an oil-bath at 100 °C.
c
d
4
Water (1.0 g) was used in lieu of [C mim]Cl as the solvent.
2
xyl groups similar to what was proposed for CrCl in an early
5
b
work. The hemiacetal portion of b-glucopyranose then forms
Cr(III) enolate anion complex leading to isomerization of glucose
to fructose, which would be dehydrated to HMF simultaneously
in the reaction condition (vide ante). There are additional data sup-
porting our speculation. We found that a strong coordinating li-
Table 2
HMF production from cellulose in [C
4
mim]Cl under MIa
b
c
Entry
Substrate
Time (min)
YieldHMF (%)
YieldTRS (%)
1
2
3
4
Avicel
2
2
2
61
53
55
62
17
2.0
16
26
23
20
45
15
9.7
0
Spruce
Sigmacell
a-Cellulose
Avicel
Avicel
Avicel
gand, 2,2 -bipyridine, strongly inhibited the reaction, as HMF
yield dropped from 61% to 1.6% (Table 2, entry 1 vs entry 7). More-
over, under oil-bath heating conditions, both HMF yield and TRS
2
d
e
5
240
240
2
0
d,f
e
yield dropped substantially in the presence of 2,2 -bipyridine (Ta-
6
0
f
e
ble 2, entry 5 vs entry 6). These results suggested that 2,2 -bipyri-
7
1.6
dine restrained both cellulose hydrolysis and glucose dehydration
reactions. Thus, coordination chemistry involving CrCl played a
3
key role not only for glucose dehydration but also for cellulose
hydrolysis.
a
Unless otherwise specified, reaction conditions were cellulose (100 mg) and
O (10 mg) in [C mim]Cl (2.0 g) under 400 W MI.
CrCl
3
ꢀ6H
2
4
b
Isolated yield unless otherwise specified.
c
d
e
f
7
Determined by the DNS method.
Heated by oil-bath at 100 °C.
Zhang and co-workers reported a method for dehydration of
Determined spectrophotometrically at 282 nm.
glucose to HMF and claimed that CrCl
2
was more efficient as a cat-
0
Three equiv of 2,2 -bipyridine (17.5 mg) were introduced thereafter the addi-
5b
alyst than CrCl under conventional heating condition. However,
3
tion of CrCl
3
ꢀ6H
2
O (10 mg).
near identical HMF yields were observed in our system, that is,
when the reaction was heated under MI, implying that the valence
state of Cr is not the determinant in the transformation. Moreover,
Based on current data, the reasons that a combination of
mim]Cl, CrCl , and MI generates the magic power for conversion
of cellulose into HMF may be imagined. First, complete dissolution
of cellulose in [C mim]Cl leaves the cellulose chains accessible to
chemical transformation. Second, [C
[
C
4
3
one might consider to formulate complexes [C
n = 1–4) for CrCl in [C mim]Cl according to Scheme 2. We also
noticed that in Zhang’s work, glucose was introduced into the mix-
ture of CrCl and ionic liquid pretreated at 150 °C for 20 min with
vigorous stirring in a sealed tube because CrCl could not be easily
solubilized in [C mim]Cl. When CrCl was used in our method, in
contrast, stirring at 80 °C for 1 min under ambient pressure en-
sured a solution of CrCl in [C mim]Cl. Thus, it was more practi-
4 n
mim] [CrCl2+n]
(
2
4
4
8
4
mim]Cl has excellent dielec-
2
9
tric properties for transformation of microwave into heat. Under
MI, alternating electric field induces vibrational motion of ions,
and resistance of the reaction mixture to ion flux leads to heat evo-
lution, the higher the concentration of ions, the stronger the heat-
ing effect. Because the reaction in ionic liquids had nearly a pure
ionic circumstance, it was heated up rapidly, volumetrically and
simultaneously by this ‘specific microwave effects’. Therefore, it
circumvented some known problems such as partly overheating
2
4
3
3
4
1
0
1
6
cal. More significantly, for the glucose-to-HMF process, our
method afforded 91% isolated yield whereas a 68% HPLC yield
5
b
was reported in the literature. When cellulose was employed as
the substrate, we also obtained HMF in over 60% isolated yield.
As very high isolated yield for HMF was realized with our meth-
od, it was appealing to envision a more practical process. Noting
that ionic liquids have little vapor pressure, and HMF has a boiling
point of 85–88 °C at 0.01 torr, one can take the advantages of this
system and engineer a continuous reaction-distillation process to
produce HMF using carbohydrates as the feedstock. In this sce-
nario, both ionic liquids and catalyst may be recycled.
11
under conventional oil-bath conditions. Other microwave ef-
1
2
fects, such as lowering activation energy or increasing the pre-
exponential factor in the Arrhenius law due to orientation effect
of polar species in an electromagnetic field,13 might also contrib-
1
7
4
ute. Moreover, [C mim]Cl may also act as a water scavenger by
dilution, circumventing water-associated HMF decomposition as
found in these aqueous systems.14 Thus, a combination of ionic li-
quid and MI might be an interesting condition to try new
chemistry.
In conclusion, we showed multiple beneficiary effects under MI
in ionic liquid for carbohydrate chemistry. For the conversion of