Effects of anionic structure and lithium salts addition on the dissolution of cellulose in 1-butyl-3-methylimidazolium-based ionic liquid solvent systems

Article abstract of DOI:10.1039/b916882f

Cellulose is the most abundant biorenewable and biodegradable resource on the earth. However, the extent of its application is limited due to its inefficient dissolution in solvents. Thus, the development of new cellulose solvents continues to be an activ

Full text of DOI:10.1039/b916882f

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/greenchem | Green Chemistry
Effects of anionic structure and lithium salts addition on the dissolution of
cellulose in 1-butyl-3-methylimidazolium-based ionic liquid solvent systems†
a
b
b
Airong Xu, Jianji Wang* and Huiyong Wang
Received 17th August 2009, Accepted 30th September 2009
First published as an Advance Article on the web 4th November 2009
DOI: 10.1039/b916882f
Cellulose is the most abundant biorenewable and biodegradable resource on the earth. However,
the extent of its application is limited due to its inefficient dissolution in solvents. Thus, the
development of new cellulose solvents continues to be an active area of investigation. In this work,
a series of ionic liquids (ILs) have been synthesized by coupling the 1-N-butyl-3-methyl-
+
-
COO]-,
imidazolium cation [C
4
mim] with the Brønsted basic anions [CH
3
COO] , [HSCH
2
-
-
-
-
-
-
[
HCOO] , [(C
6
H
5]COO] , [H
2
NCH
2
COO] , [HOCH
2
COO] , [CH
3
CHOHCOO] and [N(CN)
2
] .
The solubilities of microcrystalline cellulose (MCC) in these ionic liquids were determined as a
function of temperature. The effect of the anion structure on the solubility of cellulose has been
1
estimated, and investigated by H NMR and a solvatochromic UV/vis probe. It was found that
the solubility of cellulose increases almost linearly with increasing hydrogen bond accepting ability
of anions in the ionic liquids. At the same time, novel [C
LiAc, LiNO , or LiClO ) solvent systems have been developed by adding 1.0 wt% of lithium salt
into [C mim][CH COO]. It was shown that the addition of lithium salts significantly increased the
4
mim][CH COO]/lithium salt (LiCl, LiBr,
3
3
4
4
3
1
3
solubility of the cellulose. This observation was studied by C NMR spectra, and the results
suggested that the enhanced solubility of cellulose originated from the disruption of the
+
intermolecular hydrogen bond, O(6)H ◊ ◊ ◊ O(3) owing to the interaction of Li with the hydroxyl
oxygen O(3) of cellulose. Furthermore, the cellulose materials regenerated from the ionic liquids
were characterized by scanning electron micrograph, thermogravimetric analysis and Fourier
transform infrared spectroscopy, and the degree of polymerization of the original and regenerated
cellulose materials was also determined. Good thermal stability was found for the regenerated
cellulose. It is expected that the above information is useful for the design of novel ionic liquids
and ionic liquid-based solvent systems for cellulose.
Introduction
the dissolution processes using these solvent systems suffer from
environmental, energy, safety, and other problems. For example,
the viscose process, which is a traditional technology used
in industry, sometimes causes serious environmental pollution
Cellulose is the most abundant biorenewable and biodegradable
resource in the world, with an annual yield of over 1.0 ¥ 10 ton.
1
0
1
Cellulose can be derived from cellulose-rich starting materials
like trees, cotton, and crop wastes. Cellulose and its derivatives
have been widely used in our society in fibers, tissues, paper,
membranes, polymers, paints and medicines. However, due
to its close packing by numerous inter- and intramolecular
hydrogen bonds, cellulose is extremely difficult to dissolve in
water and most of the conventional organic solvents. Therefore,
the main obstacle to the further application of cellulose is the
lack of powerful solvent systems. Until now, only a limited
number of solvent systems have been industrialized. However,
4
and involves derivatization of cellulose products. Although
the N-methylmorpholine-N-oxide (NMMO) process is a direct
dissolution process commercialized for the manufacture of so-
2,3
4
called lyocell fibres, this process also has some problems,
such as the severe fibrillation of the manufactured fibres, a
◦
high dissolution temperature (130 C), and the requirement
of a major investment in safety technology because of its
thermal instability. Other solvent systems investigated include
DMAc/LiCl, DMSO/TBAF, DMF/N
2
O
4
, LiClO
4
·3H
2
O, and
5–8
LiSCN·2H
2
O, where DMAc, DMSO, TBAF, DMF, and
N
2
4
O stand for N,N-dimethylacetamide, dimethyl sulfoxide,
tetrabutyl ammonium fluoride, N,N-dimethylformamide and
nitrous tetroxide, respectively. Such solvent systems still retain
certain drawbacks such as high cost, toxicity, difficulty in
solvent recovery or harsh processing conditions. Therefore, it
is highly desirable to develop less energy consuming, more
environmentally friendly, and highly efficient solvent systems
for cellulose.
a
College of Chemistry and Chemical Engineering, Lanzhou University,
Lanzhou, Gansu, 730000, P. R. China
b
School of Chemistry and Environmental Science, Henan Key Laboratory
of Environmental Pollution Control, Henan Normal University,
Xinxiang, Henan, 453007, P. R. China
1
†
Electronic supplementary information (ESI) available: H NMR data
of the ILs and FTIR spectra of cellulose before and after dissolution.
See DOI: 10.1039/b916882f
2
68 | Green Chem., 2010, 12, 268–275
This journal is © The Royal Society of Chemistry 2010

View Article Online
Ionic liquids (ILs) are a novel class of green solvents. They
have many distinct advantages such as chemical and thermal
stability, non-flammability, non-detectable vapor pressures and
Table 1 Solubility of microcrystalline cellulose in the ILs at different
temperatures
Solubility/wt%
9–11
chemical tunabilities,
and promise widespread application
◦
◦
◦
◦
12
Entry IL
40 C 50 C
60 C
70 C
in industry. In 2002, Rogers and coworkers reported, for the
first time, that cellulose could be efficiently dissolved in 1-N-
1
2
3
4
5
6
7
8
[C
[C
[C
4
4
4
mim][CH
mim][HSCH
mim][HCOO]
3
COO]
11.5
<1.0 9.5
7.5 8.5
12.5
13.0
12.5
9.0
<1.0
8.0
15.5
13.5
12.5
12.0
12.0
10.5
9.5
butyl-3-methylimidazolium chloride [C
4
mim]Cl, whereas the ILs
2
COO]
-
]- anions were poor solvents.
13
containing [BF
4
] and [PF
6
C
3
5/37
[C mim][(C H )COO]
<1.0 <1.0
4
4
4
6
5
and
the high solubility of cellulose in [C
to the formation of hydrogen bonding between the hydroxyl
Cl NMR relaxation measurements demonstrated that
[C
[C
mim][H
2
NCH
2
COO]
<1.0 2.0
4
mim]Cl was attributable
a
mim][HOCH
2
COO]
—
—
—
7.5
9.0
a
a
a
[C mim][CH CHOHCOO]
4
3
—
8.0
13,14
a
a
a
protons of cellulose and the chloride anion of the IL.
Since
[C
4
mim][N(CN)
2
]
—
—
—
then, a series of researches has been conducted in this field.
a
Insoluble at the given temperature.
15
Zhang et al. reported that 1-N-allyl-3-methylimidazolium
chloride [Amim]Cl was a powerful solvent for cellulose, in which
microcrystalline cellulose could be dissolved quickly without
activation or pretreatment. They also indicated that 1-N-ethyl-
Results and discussion
Influence of the ionic liquids’ anion structure on the solubility of
cellulose
3
-methylimidazolium acetate [C
2
mim][CH COO], which has
3
a lower melting point and viscosity, showed a much higher
16
17,18
ability to dissolve cellulose. Ohno et al.
dialkylimidazolium formates, [RR¢im][HCOO], and N-ethyl-
N¢-methylimidazolium alkylphosphates, [C mim][(MeO) PO ],
found that N,N¢-
The solubility values of microcrystalline cellulose in the studied
ILs are shown in Table 1 as a function of temperature. It can
be seen that the structure of the anions significantly affects
the solubility of the cellulose. Among the investigated ILs,
2
2
2
displayed superior solubility for a variety of polysaccharides,
including cellulose. In the extended dissolution studies of
[C
dissolution of cellulose. In contrast, cellulose is insoluble in
[C mim][N(CN) ]. This indicates that the anions of the ILs play
a key role in the disruption of the inter- and intramolecular hy-
drogen bonds of cellulose. Generally, the solubility of cellulose in
4
mim][CH
3
COO] is the most efficient ionic liquid for the
19
Schubert et. al. for cellulose in imidazolium-based ionic
liquids, an odd–even effect was found for different alkyl side-
chain lengths of the imidazolium chlorides, and 1-ethyl-3-
methylimidazolium diethyl phosphate was found to be best
suited to the dissolution of cellulose. It was also reported that
wood and other biomasses such as bagasse and straw can be
4
2
these ILs decreases in the order: [C
[HSCH COO] > [C mim][HCOO] > [C
[C mim][H NCH COO] > [C
[CH CHOHCOO] > [C
4
mim][CH
mim][(C
mim][HOCH COO] > [C
]. It is interesting to
3
COO] > [C
)COO] >
mim]-
4
mim]-
2
4
4
6
H
5
dissolved, at least partially, in ILs such as [Amim]Cl, [C
4
mim]Cl
4
2
2
4
2
4
20–24
and [C
2
mim][CH
3
COO].
These encouraging results provide
3
4
mim][N(CN)
2
possibilities for the dissolution of cellulose and biomass by
ILs instead of conventional organic solvents, as well as for
solutions to the problems described above in the solvent systems
of cellulose.
note that replacing H in the CH
liquid [C mim][CH COO] with electron-withdrawing groups
such as OH, SH, NH
3
COO- anion of the ionic
4
3
2
or CH
3
OH leads to decreased solubility.
-
Considering the fact that replacement of H in the CH COO
3
Although previous investigations revealed some important
aspects for the dissolution of cellulose in ILs, very few studies
have been conducted to explore the influence of the structure of
ILs on the dissolution performance of cellulose. Considering the
fact that the dissolution of cellulose is significantly determined
anion by an electron-withdrawing group would decrease the
-
ability of CH
2
XCOO (X = OH, SH, NH
2
and CH OH) to
3
form hydrogen bonds with the hydroxyl protons of cellulose,
the decreased solubility observed is not difficult to understand.
This suggests that the hydrogen bond accepting ability of the
anions strongly dominates the ILs’ capacity for the dissolution of
cellulose. In addition, it is noted from Table 1 that the solubility
of cellulose in these ILs is enhanced with increasing temperature.
13,14
by the nature of the anions in ILs,
we prepared a series
of ionic liquids with a fixed cationic backbone and a fixed
alkyl chain length, i.e. the 1-butyl-3-methylimidazolium cation
+
[C
4
mim] , but with varied anion structures. This allowed us to
In the case of [C
4
mim][CH
3
COO], the solubility of cellulose at
◦
◦
examine the effect of the ILs’ anion structure on the solubility
of cellulose, as well as the related dissolution mechanism. In
order to further enhance the dissolution of cellulose, several
novel solvent systems have been developed by the addition of
70 C (15.5 wt%) is higher than that at 40 C (11.5 wt%) by about
35%. This is an indication that hydrogen bonds of the cellulose
were also partially disrupted by the increased temperature.
The b parameter, introduced by Kamlet and Taft, has been
established as a measure of the hydrogen bond accepting ability
1
.0 wt% of a lithium salt (LiAc, LiCl, LiBr, LiClO
into [C mim][CH COO]. The dissolution of cellulose in the ILs
and the [C mim][CH COO]/lithium salt solvent systems was
4
, or LiNO )
3
25–27
4
3
of anions of ionic liquids.
Unfortunately, most of the b
4
1
3
parameters for the ILs investigated in the present work are
not available in the literature. Therefore, we determined these
parameters and the results are given in Table 2. It is evident that
the ILs with different anion structures exhibit a marked variation
in the b parameter. As shown in Fig. 1, the solubility values of
cellulose increase almost linearly with increasing b parameter
of the ILs. This indicates that the solubility of the cellulose
strongly depends on the hydrogen bond-accepting ability of the
13
studied by H and C NMR spectroscopies and solvatochromic
UV/vis probe. In addition, the regenerated cellulose materials
were investigated by means of scanning electron microscopy
(
SEM), thermogravimetric analysis (TGA), and Fourier trans-
form infrared (FT-IR), and the dissolution temperature and
dissolution time effects on the degree of polymerization of the
regenerated cellulose were also discussed.
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 268–275 | 269

View Article Online
Table 2 The b parameter values for the ILs investigated
Table 3 Solubility of microcrystalline cellulose in [C
lithium salt solvent systems
4
mim][CH
3
COO]/
Entry
IL
b
Solubility (wt%) at different
temperatures
1
2
3
4
5
6
7
8
[C
[C
[C
[C
[C
[C
[C
[C
4
mim][CH
mim][HSCH
mim][HCOO]
mim][(C )COO]
mim][H NCH COO]
mim][HOCH COO]
mim][CH CHOHCOO]
mim][N(CN)
3
COO]
1.161
1.032
1.008
0.987
1.096
0.967
0.964
0.621
4
4
4
4
4
4
4
2
COO]
◦
◦
◦
◦
◦
Solvent system
40 C 50 C 60 C 70 C 80 C
6
H
5
[C
[C
[C
[C
[C
[C
4
4
4
4
4
4
mim][CH
mim][CH
mim][CH
mim][CH
mim][CH
mim][CH
3
3
3
3
3
3
COO]
11.5 12.5
12.0 16.0
12.0 14.5
11.5 13.5
12.0 15.0
12.0 15.0
13.0
17.5
16.0
15.0
17.5
18.5
15.5
19.0
18.5
18.0
20.0
19.5
19.0
20.0
20.0
19.5
21.0
21.0
2
2
COO]/LiAc
COO]/LiCl
COO]/LiBr
COO]/LiClO
2
3
2
]
4
COO]/LiNO
3
Fig. 1 Linear correlation between solubility of microcrystalline cellu-
◦
lose at 70 C and b Parameter of the ILs investigated.
Fig. 2 Linear correlation between solubility of microcrystalline cel-
◦
1
anions of the ILs. The stronger the hydrogen bond accepting
ability of the anions, the higher the solubility of cellulose in
lulose at 70 C and the H NMR chemical shifts of the proton in
the 2-position of the imidazolium ring measured in DMSO-d at a
6
-1
the IL. For example, the anion of [C
4
mim][CH
3
COO] has the
concentration of 1.0 mol kg .
strongest hydrogen bond accepting ability (highest b value) of
the ILs investigated, and solubility of the cellulose is found to
the cellulose is enhanced significantly after the addition of
a small amount of lithium salt into [C mim][CH COO]. As
4
3
be the highest in [C
4
mim][CH COO].
3
29
1
reported in the literature, some lithium salt hydrates have the
capacity to dissolve cellulose through the hydroxyl groups of the
It is known that the H NMR chemical shift (d) of the
proton in the 2-position of the imidazolium ring can be used
as a measure of the capacity of the anion of an IL to form
+
cellulose taking part in the coordination of Li . Compared to
28
LiX·nH
2
O, the lack of water molecules in LiX leads to more
hydrogen bonds, which is crucial to the breaking of the inter-
17,18
free coordination sites at the lithium cation. These sites can
be occupied by other coordinating groups, e.g. the hydroxyl
groups of cellulose, resulting in the breaking of some inter-
and intramolecular hydrogen bonds of cellulose. This is a
possible reason why higher solubility was observed for cellulose
and intramolecular hydrogen bonds of the cellulose chains,
1
and thus to the dissolution of cellulose. Therefore, H NMR
chemical shifts of the proton in the 2-position of the imidazolium
ring have been determined for the ILs under study, and the
correlation between the solubility of cellulose and the chemical
shift of this proton is shown in Fig. 2. As expected, the solubility
of cellulose increases almost linearly with the chemical shift of
the proton in the 2-position of the imidazolium ring. Hence,
the solubility of cellulose in a certain IL can be predicted by
using this linear correlation. According to our result, the ILs
with a d value smaller than 9.5 are not suitable solvents for the
dissolution of cellulose.
in [C
mim][CH
In order to further examine how the addition of lithium
4
mim][CH
3
COO]/lithium salt solvent systems than in
[
C
4
3
COO] alone.
1
3
salts increases the solubility of cellulose, C NMR measure-
ments of cellulose were carried out in solutions of [C mim]-
COO]/LiCl
4
[
(
CH
3
COO]/cellulose (9.0 wt%) and [C
4
mim][CH
3
◦
13
1.0 wt%)/cellulose (9.0 wt%) at 90 C. The C NMR spectra
1
3
are shown in Fig. 3, and the C NMR chemical shift data
are presented in Table 4. It can be seen from Table 4 that
the addition of LiCl produces a downfield shift (an increase
in chemical shift) for C3 and upfield shifts (decrease in
chemical shift) for the other C atoms. Based on the results
Lithium salt-promoted dissolution of cellulose in
[C
4
mim][CH
3
COO]
The solubility of cellulose in various [C
4
mim][CH COO]/
3
30,31
lithium salt solvent systems was investigated at different
temperatures. Table 3 summarizes the results obtained for
microcrystalline cellulose. It can be seen that solubility of
reported in literature,
the hydrogen bonds of cellulose are
mainly the intramolecular hydrogen bonds, O(3)H ◊ ◊ ◊ O(5) and
O(2)H ◊ ◊ ◊ O(6), present on both sides of the cellulose chain,
2
70 | Green Chem., 2010, 12, 268–275
This journal is © The Royal Society of Chemistry 2010

View Article Online
13
Table 4
in [C mim][CH
1.0 wt%)/cellulose (9.0 wt%) systems
C NMR chemical shifts (d (ppm) relative to TMS) of cellulose
4
3
COO]/cellulose(9.0 wt%) and [C mim][CH COO]/LiCl
4
3
(
d (ppm)
Solvent system
C(1)
C(2)
C(3)
C(4)
C(5)
C(6)
[
[
C
C
4
4
mim][CH
mim][CH
3
COO] 103.165 74.188 75.468 79.488 76.242 61.063
COO]/ 103.146 74.182 75.478 79.467 76.189 61.034
3
LiCl
the same time, disruption of this intermolecular hydrogen
bond prompts further dissolution of the cellulose. However,
the chemical shifts for the other C atoms of cellulose behave
differently. This may be interpreted as the interactions of the
-
Cl anion of LiCl with the hydroxyl proton of cellulose in
the [C
4
mim][CH COO]/LiCl/cellulose system, which leads to
3
an increase of electron cloud density of the C atoms and an
1
3
upfield shift of their C NMR signals. The dissolution mech-
anism of cellulose in [C mim][CH COO]/LiCl reported here is
4
3
somewhat similar to that in N,N-dimethylacetamide/LiCl, and
the breaking of intermolecular hydrogen bonds results from
the direct interaction of lithium with the cellulosic hydroxyl
32
oxygen. The difference is that [C
4
mim][CH
3
COO] in the
[
C
4
mim][CH COO]/LiCl solvent system can dissolve cellulose,
3
but N,N-dimethylacetamide in N,N-dimethylacetamide/LiCl
cannot.
Compared to other solvent systems for cellulose, the
[
C
4
mim][CH COO]/lithium salt solvent systems reported in the
3
present work have the advantages of higher solubility and lower
◦
dissolution temperature. For example, at 40 C, the solubility
1
3
of cellulose in [C mim][CH COO]/LiX (X = Cl, NO , ClO ,
Fig. 3
C NMR spectra of microcrystalline cellulose in IL and
4
3
3
4
◦
Ac) systems could be as high as 12.0 wt%, whereas only
IL/lithium salt systems at 90 C: (A) [C
4
mim][CH
3
COO]; (B)
18
[C
4
mim][CH
3
COO]/LiCl.
8.0 wt% of cellulose was dissolved in [C
2
mim][(MeO)HPO ].
2
◦
At 60 C, 18.5 wt% of cellulose can be dissolved in
and the intermolecular hydrogen bond, O(6)H ◊ ◊ ◊ O(3), which
builds a bridge between two neighboring molecules (Fig. 4).
[C
has been dissolved in [C
cellulose cannot be dissolved in [Amim]Cl at temperatures below
70 C, and the temperature needs to be maintained at 130 C
for the dissolution of 10–15 wt% of cellulose in the NMMO
4
mim][CH
3
COO]/LiNO
3
, but only 10.0 wt% of cellulose
17
2
mim][HCOO]. On the other hand,
After the addition of LiCl into the [C
4
mim][CH COO]/cellulose
3
◦
15
◦
system, the intermolecular hydrogen bond, O(6)H ◊ ◊ ◊ O(3), was
+
broken owing to the interaction of Li with the hydroxyl oxygen
O(3) of cellulose, and thus the electron cloud density of C3
bonded to the hydroxyl oxygen O(3) of cellulose decreased.
Consequently, the C NMR signal of C3 shifts downfield,
and the chemical shift of C3 increases correspondingly. At
system. In addition, our ILs and [C
4
mim][CH COO]/lithium
3
salt solvent systems can be used for the direct dissolution of
cellulose without any pretreatment or activation. In contrast,
in most of the conventional solvent systems for cellulose such
1
3
Fig. 4 Schematic structure and numbering of cellulose.
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 268–275 | 271

View Article Online
Fig. 5 SEM images of the free and fracture surfaces for the regenerated cellulose films from [C
4
mim][CH
3
COO]/LiCl/cellulose solution: (A) free
surface at 10 000¥ magnification; (B), (C) fracture surface at 3000¥ magnification and 10 000¥ magnification, respectively.
as DMAc/LiCl and NaOH/H
the original cellulose is often necessary.
2
O, pretreatment or activation of
suggests that the macromolecular structure of cellulose has not
been destroyed.
33,34
Encouragingly, it is also found that absorbent cotton with a
degree of polymerization as high as 1586 could be dissolved in
TGA curves for both the original cellulose and the regenerated
cellulose are shown in Fig. 6. It is noted that the two TGA curves
nearly overlapped at lower temperatures, and rapid decomposi-
the [C
4
mim][CH
3
COO]/LiCl system in high concentrations (up
◦
◦
to 21 wt%) at 90 C. We have tried to increase the content of
tion occurred in the temperature range from 313 to 351 C for
◦
the lithium salts in [C
4
mim][CH
3
COO] to enhance the cellulose
the original cellulose and from 301 to 341 C for the regenerated
solubility further. However, we failed to do this because of the
limited solubility of the lithium salts in the ILs.
cellulose. The regenerated sample exhibits a slightly lower onset
◦
temperature (301 C) for the decomposition compared to the
◦
original cellulose (313 C), and gives a slightly higher char
yield (nonvolatile carbonaceous material) on pyrolysis, indicated
by the slightly higher residual mass after the decomposition
step. This result indicates that the cellulose regenerated from
Structure and properties of the regenerated cellulose
The cellulose regenerated from [C
4
mim][CH COO]/lithium salt
3
solvent systems (see Experimental section) was characterized by
SEM, TGA, and FTIR spectroscopy. The degree of polymer-
ization of the original and the regenerated cellulose materials
was also determined, and the dissolution temperature and the
dissolution time effects were examined.
the [C
4
mim][CH COO]/lithium salts solvent systems has good
3
thermal stability. This is significantly different from the cellulose
12,15
regenerated from the reported solvent systems
in which a
large difference in the decomposition temperature was observed
between the regenerated and the original materials.
SEM images of the regenerated cellulose films from
[C
4
mim][CH COO]/LiCl solution are shown in Fig. 5. It can be
3
seen that the free surface and fracture surface of the dried films
display homogeneous structures from the interior to the surface,
indicating a dense architecture, in which no undissolved original
cellulose can be detected. The structure of the regenerated
1
5
cellulose is quite similar to that reported by Zhang et al.
35
and Zhang et al., but different from the porous structure
36
regenerated from aqueous NaOH/thiourea solution.
FTIR spectra of the original and the regenerated cellulose
materials are shown in Fig. S1 (see ESI†). It can be seen that all
the spectra are quite similar and no new peaks are observed in
the regenerated sample. This indicates that no chemical reaction
takes place during the cellulose dissolution and regeneration
-
1
processes. The absorption band at 1427 cm in the regenerated
cellulose was assigned to the CH scissoring vibration. This band
was weakened and shifted to a lower wavenumber compared to
2
Fig. 6 Thermal decomposition profiles of the original cellulose and
the regenerated cellulose materials: (A) the original cellulose; (B) the
cellulose regenerated from [C mim][CH COO]/LiCl/cellulose after 1 h
-
1
the peak at 1431 cm for the original cellulose, suggesting the
3
5
destruction of an intra-molecular hydrogen bond involving O6.
4
3
-
1
◦
A new shoulder at 990 cm was observed in the regenerated
cellulose, which could be assigned to a C–O stretching vibration
in the amorphous region. The O–H vibration in the regenerated
of dissolution at 60 C.
37
The effects of the dissolution temperature and dissolution
time on the degree of polymerization (DP) of the regenerated cel-
lulose are shown in Fig. 7 and 8, respectively. In the temperature
-
1
cellulose shifts to a higher wavenumber (3427 cm ), indicating
38,39
the breaking of hydrogen bonds to some extent.
The ab-
-
1
◦
sorption bands in the range of 1164–1061 cm are assigned to
the C–O–C stretching of the original cellulose. The presence
of such bands in the absorption of the regenerated cellulose
range from 40 to 80 C, the DP values (from 215 to 209) of the
40
regenerated cellulose are slightly smaller than that of the original
◦
cellulose (DP = 229). Only at 100 C was a slight degradation of
2
72 | Green Chem., 2010, 12, 268–275
This journal is © The Royal Society of Chemistry 2010

View Article Online
the anion and the addition of lithium salts significantly affected
the solubility of cellulose. The hydrogen bond accepting ability
of the anions of the ILs, as characterized by the chemical shift
values of the proton in the 2-position of the imidazolium ring,
and the b parameter of the ILs, are closely linked to the solubility
-
of cellulose. Thus, replacement of H in the CH
mim][CH
OH, SH, NH
3
COO anion of
[
C
4
3
COO] by an electron- withdrawing group such as
or CH OH leads to decreased solubility. The
2
3
nature the enhanced dissolution achieved with lithium salts is
+
suggested to result from the interaction of Li with the hydroxyl
oxygen O(3) of the cellulose, which disrupts the intermolecular
hydrogen bond O(6)H ◊ ◊ ◊ O(3). Such information is expected
to be useful for the design of novel ILs and IL-based solvent
systems for cellulose.
Cellulose can be regenerated from the ILs and the [C
4
mim]-
[
CH COO]/lithium salt solvent systems by precipitation with
3
Fig. 7 Degree of polymerization for the cellulose regenerated from
mim][CH COO]/LiCl/cellulose after 1 h of dissolution at different
temperatures.
[C
4
3
water. The regenerated cellulose was studied by SEM, TGA,
FTIR and degree of polymerization measurements. It was found
that no chemical reaction took place between cellulose and
the ILs or the [C
4
mim][CH COO]/lithium salt systems during
3
the dissolution and precipitation processes. The regenerated
cellulose exhibited a quite similar thermal stability to the original
cellulose. Therefore, the [C
4
mim][CH COO]/Li salt solvent sys-
3
tems reported here are expected to provide a possible platform
for the homogeneous processing of cellulose and the production
of advanced cellulose materials under mild conditions. However,
it should be pointed out that the recovery and reuse of the
ILs and the lithium salts are the key issue for the practical
application of the new solvent systems mentioned above.
Although both of them can be recovered by simple evaporation
from their aqueous solutions and then dried under vacuum,
this method is energy consuming and thus expensive. Therefore,
alternative low-cost methods for the recovery of the ILs and the
lithium salts from the aqueous solutions need to be developed.
Fig. 8 Degree of polymerization for the original cellulose and the
◦
cellulose regenerated from [C
4
mim][CH
3
COO]/LiCl/cellulose at 40 C
Experimental
as a function of the dissolution times: (᭹) the original cellulose; (ꢀ) the
regenerated cellulose.
Materials
Absorbent cotton and 1-methylimidazole (99%) were pur-
chased from Shanghai Chem. Co.; 1-chlorobutane (>99.0%),
cellulose observed, with the DP value of the regenerated cellulose
being 181. It is worth noting that at 40 C, the DP value of the
◦
1
-bromobutane (98.0%), mercaptoacetic acid (>97.0%), gly-
colic acid (98.0%), glycine (>99.0%), microcrystalline cellulose
MCC) and anion exchange resin (Ambersep 900 OH) were from
regenerated cellulose hardly changed, even if the dissolution
time was prolonged to as much as 5 h. This result indicates
◦
(
that no cellulose degradation took place at 40 C during 5 h
Alfa Aesar; acetic acid (>99.5%), formic acid (>88.0%), ben-
zoic acid (>99.5%)and lactic acid (85.0–90.0%) were obtained
from Shanghai Shiyi Chem. Reagent Co. Ltd.; N,N-diethyl-
4-nitroaniline (97.0%) was purchased from Nanjing Chemlin
Chemical Industrial Co. Ltd.; deuterated DMSO (DMSO-
of dissolution in the [C
systems. Obviously, the [C
systems of cellulose are superior to some of the reported systems
4
mim][CH
3
COO]/lithium salt solvent
4
mim][CH
3
COO]/lithium salt solvent
15,41
in which cellulose degradation definitely occurred.
All of
these facts suggest that our new solvent systems are excellent
solvents for cellulose, and they could provide a new platform for
the homogeneous processing of cellulose and the production of
advanced cellulose materials under mild conditions.
d ) used for NMR samples was purchased from Qingdao
6
Weibo Tenglong Technol. Co. Ltd. The above materials were
used as received. 1,1,1-Trichloroethane (95.0%) was obtained
from Shanghai Shanpu Chem. Co. Ltd., and distilled before
use. 4-Nitroaniline (98.0%) from Alfa Aesar was dissolved in
methanol, and the mixture was stirred and then filtered to
remove insoluble impurities. The methanol in the filtrate was
Conclusions
1
13
In this work, solubility, H and C NMR and solvatochromic
UV/vis probe measurements have been performed to study the
effects of the structure of the anion in selected ILs and the effect
evaporated, and 4-nitroaniline was dried under vacuum for 12 h
◦
at 40 C. LiCl·3H
2
O, LiBr·2H
2
O, LiAc·2H
2
O, LiClO
4
·3H
2
O and
of adding lithium salts to [C
4
mim][CH
3
COO] on the solubility
LiNO
3
·3H
2
O were obtained from Shanghai Chem. Co. and dried
of microcrystalline cellulose. It was shown that the structure of
under vacuum before use to remove the crystallization water.
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 268–275 | 273

View Article Online
Synthesis of 1-butyl-3-methylimidazolium bromide [C
and 1-butyl-3-methylimidazolium chloride [C mim]Cl
4
mim]Br
lithium salt solvent systems. These solvent mixtures were
prepared immediately before use to minimize moisture
uptake.
4
[
C mim]Br was prepared and purified by using the proce-
4
42
dure described in the literature. Briefly, the reaction of 1-
methylimidazole with an excess of 1-bromobutane was per-
Dissolution of cellulose in the ILs and
[
C
4
mim][CH COO]/lithium salt solvent systems
3
◦
formed in 1,1,1-trichloroethane at 70 C for 48 h. The solid
product of [C
4
mim]Br was washed with 1,1,1-trichloroethane
In a typical dissolution experiment, microcrystalline cellulose or
absorbent cotton sample was added into a 20 mL colorimetric
tube which contained 2.0 g of the dried IL or the IL/lithium
salt, and the tube was sealed with parafilm. The tube was then
immersed in an oil bath (DF-101S, Gongyi Yingyu Instrument
Factory); the instability of the bath temperature was estimated to
and then filtered. The residual solvent was removed by rotary
evaporation, and the resulting product was dried under vacuum
for 24 h. [C
4
mim]Cl was prepared similarly, but ethyl acetate was
43
used instead of 1,1,1-trichloroethane.
◦
Synthesis of 1-butyl-3-methylimidazolium acetate
be ± 0.5 C. The mixture was heated at a given temperature and
[
[
[
C
C
C
4
4
4
mim][CH
3
COO], 1-butyl-3-methylimidazolium formate
stirred under argon atmosphere. Additional cellulose was added
until the solution became optically clear under a polarization
microscope (Nanjing Jiangnan Novel Optics Co. Ltd.). When
cellulose became saturated, judged to be the point where
cellulose could not be dissolved further within 1 h, its solubility
(expressed in g per 100 g of IL) at the given temperature
could be calculated from the amount of solvent and cellulose
mim][HCOO], 1-butyl-3- methylimidazolium lactate
mim][CH
3
CHOHCOO], 1-butyl-3-methylimidazolium
mim][HOCH COO], 1-butyl-3-methylimidazolium
mim][HSCH COO], and
-butyl-3-methylimidazolium benzoate [C mim][(C
glycollate [C
thioglycollate [C
1
4
2
4
2
4
6
5
H )COO]
An aqueous solution of [C mim]Br was allowed to pass
through a column filled with anion exchange resin to obtain
4
added. For each IL or [C
4
mim][CH COO]/lithium salt solvent,
3
the solubility values of cellulose were measured at different
17
[C
4
mim][OH]. The aqueous [C
4
mim][OH] solution was then
◦
temperatures with 10 C intervals.
neutralized with equal molar acetic acid. After removing water
by evaporation under reduced pressure, the viscous liquid
1
13
Measurements of H NMR and C NMR spectra
[C
4
mim][CH
3
COO] was thoroughly washed with diethyl ether,
1
◦
H NMR spectra of the ILs were collected at room tem-
and finally dried under vacuum for 72 h at 70 C. The other ionic
liquids were prepared by a similar process.
perature on a Bruker Avance-400 NMR spectrometer operat-
1
3
ing at 400.13 MHz. Measurements of C NMR spectra for
the cellulose in [C mim][CH COO]/cellulose (9.0 wt%) and
Synthesis of 1-butyl-3-methylimidazolium aminoethanic acid salt
4
3
[
C
4
mim][CH
3
COO]/LiCl (1.0 wt%)/cellulose (9.0 wt%) were
[C
4
mim][H
2
NCH
2
COO]
◦
performed on a Bruker DMX 300 spectrometer at 90 C. Chem-
ical shifts were given in ppm downfield from TMS.
This IL was prepared by a similar procedure to that reported by
Ohno et al. Aqueous [C mim][OH] solution was thoroughly
4
4
4
mixed with a slight excess of aqueous amino acid solution.
After 12 h of reaction under cooling, water in the mixture was
evaporated at 50 C, and a slightly viscous liquid was obtained.
Measurements of the b parameter for the ILs
◦
The b parameter of the ILs, introduced as a measure of the
hydrogen bond accepting ability of a solvent, was determined by
After addition of acetonitrile–methanol (9 : 1, v/v), the mixture
was stirred and then filtered to remove the excess amino acid.
The organic solvents in the filtrate were evaporated and the
18
following the procedure reported by Ohno et al. To calculate
the b parameter, two different dyes, 4-nitroaniline (NA) and
N,N-diethyl-4-nitroaniline (DENA) were employed. In a dry
box, a given amount of the dried IL and a concentrated solution
of NA or DENA in dry methanol were added into a vial
and mixed homogeneously. The methanol was then carefully
◦
product was dried under vacuum for 72 h at 80 C.
Synthesis of 1-butyl-3-methylimidazolium dicyanamide
[C
4
mim][N(CN)
2
]
◦
removed by drying under vacuum at 40 C for 12 h. To avoid
This IL was synthesized and purified by using the procedure
aggregation of the dyes, concentrations of NA or DENA in the
ILs were selected to give an absorbance between 0.15 and 0.30.
The IL containing NA or DENA was added into a quartz cell
in a dry box and the cell was capped and sealed. The visible
spectra of the ILs containing NA or DENA were recorded
on a TU-1810 ultraviolet-visible spectrophotometer at 25.0 ±
45
described by Sheldon and coworkers. Briefly, to a solution
of [C mim]Cl in acetone, Na[N(CN) ]was added. The mixture
4
2
was stirred for 24 h and the solid NaCl was removed by
filtration. After removing acetone by evaporation under reduced
pressure, the viscous liquid [C
4
mim][N(CN)
2
] was washed with
◦
◦
dichloromethane, and dried under vacuum for 24 h at 60 C.
0.1 C. The wavelength at the maximum absorption (lmax) was
1
The H NMR data of the above ILs are summarized in the
determined and used to calculate the b values by the following
18
supporting information. They were in good agreement with
equations:
42,43,45
those available in literature.
Preparation of [C mim][CH
.0% (w/w) of LiCl, LiBr, LiAc, LiNO
-
4
n
(NA) = 1/(lmax (NA) ¥ 10 )
(1)
(2)
(3)
4
3
COO]/lithium salt solvent systems
-
4
n
(DENA)) = 1/(lmax (DENA) ¥ 10 )
1
3
or LiClO
4
was dissolved
COO]/
in [C mim][CH COO] to prepared the [C
4
3
4
mim][CH
3
b = (1.035n(NA) + 2.64 - n(DENA))/2.80
2
74 | Green Chem., 2010, 12, 268–275
This journal is © The Royal Society of Chemistry 2010

View Article Online
Recovery and recycling of [C
systems
4
mim][CH
3
COO]/Li salt solvent
3 F. T. Moutos, L. E. Freed and F. Guilak, Nat. Mater., 2007, 6, 162.
4
F. Hermanutz, F. G a¨ hr, E. Uerdingen, F. Meister and B. Kosan,
Macromol. Symp., 2008, 262, 23.
5
6
7
C. L. McCormick and T. R. Dawsey, Macromolecules, 1990, 23, 3606.
S. Fischer, W. Voigt and K. Fischer, Cellulose, 1999, 6, 213.
T. Heinze and T. Liebert, Prog. Polym. Sci., 2001, 26, 1689.
After the complete dissolution of cellulose in the
mim][CH COO]/Li salt solvent systems, cellulose can
be regenerated and the [C mim][CH COO]/Li salt solvent
mixtures can be recovered by addition of water or ethanol. In a
typical recovery trial, 2.0 g of [C mim][CH COO]/LiCl solvent
and 5.0 wt% cellulose solution were used. The cellulose solution
was poured into a 100 mL beaker containing 10 mL of water.
The beaker was sealed with preservative film and the mixture
was stirred for 30 min at ambient temperature. The precipitated
cellulose was separated by filtration through a ceramic funnel
with Nylon filter paper on a B u¨ chner flask under vacuum.
The cellulose was washed four times to ensure that all the
[C
4
3
4
3
8 Z. Wang, T. Yokoyama, H. M. Chang and Y. Matsumoto, J. Agric.
Food Chem., 2009, 57, 6167.
T. Welton, Chem. Rev., 1999, 99, 2071.
9
4
3
1
1
0 P. Wasserscheid and W. Keim, Angew. Chem., Int. Ed., 2000, 39, 3772.
1 J. Dupont, R. F. de Souza and P. A. Z. Suarez, Chem. Rev., 2002,
102, 3667.
2 R. P. Swatloski, S. K. Spear, D. John, J. D. Holbrey and R. D. Rogers,
J. Am. Chem. Soc., 2002, 124, 4974.
3 J. S. Moulthrop, R. P. Swatloski, G. Moyna and R. D. Rogers, Chem.
Commun., 2005, 1557.
14 R. C. Remsing, R. P. Swatloski, R. D. Rogers and G. Moyna, Chem.
Commun., 2006, 1271.
1
1
1
5 H. Zhang, J. Wu, J. Zhang and J. S. He, Macromolecules, 2005, 38,
272.
16 Y. Cao, J. Wu, J. Zhang, H. Q. Li, Y. Zhang and J. S. He, Chem.
[C
4
mim][CH COO]/LiCl solvent had been washed out. The
3
8
filtrates were combined in a round bottomed flask, and water
was removed by rotatory evaporation under reduced pressure.
Eng. J., 2009, 147, 13.
1
1
1
2
7 Y. Fukaya, A. Sugimoto and H. Ohno, Biomacromolecules, 2006, 7,
The resulting [C
dried under vacuum for 24 h at 70 C, and then could be used in
the next dissolution process. In each dissolution–recovery cycle,
4
mim][CH
3
COO]/LiCl solvent mixture was
3
295.
◦
8 Y. Fukaya, K. Hayashi, M. Wadab and H. Ohno, Green Chem., 2008,
10, 44.
9 J. Vitz, T. Erdmenger, C. Haensch and U. S. Schubert, Green Chem.,
the recovery percentage of [C
4
mim][CH COO]/LiCl solvent
3
2
009, 11, 417.
is approximately 99.2 wt%, and the dissolving capacity of the
recovered solvent for cellulose is equivalent to the original
solvent.
0 D. A. Fort, R. C. Remsing, R. P. Swatloski, P. Moyna, G. Moyna
and R. D. Rogers, Green Chem., 2007, 9, 63.
21 I. Kilpel a¨ inen, H. Xie, A. King, M. Granstrom, S. Heikkinen and
D. S. Argyropoulos, J. Agric. Food Chem., 2007, 55, 9142.
2
2
2 M. Zavrel, D. Bross, M. Funke, J. B u¨ chs and A. C. Spiess, Bioresour.
Technol., 2009, 100, 2580.
Characterization of the regenerated cellulose
3 N. Sun, M. Rahman, Y. Qin, M. L. Maxim, H. Rod r´ iguez and R. D.
Rogers, Green Chem., 2009, 11, 646.
Fourier transform infrared (FTIR) spectra were recorded on
a Necolet Nexus spectrometer with KBr pellets. A total of 16
scans were taken for each sample at a resolution of 2 cm .
Scanning electron micrographs were taken with a JEOL JSM-
24 S. D. Zhu, J. Chem. Technol. Biotechnol., 2008, 83, 777.
-
1
25 A. Oehlke, K. Hofmann and S. Spange, New J. Chem., 2006, 30,
5
33.
2
6 C. Chiappe and D. J. Pieraccini, J. Phys. Org. Chem., 2005, 18, 275.
27 J. G. Huddlestone, G. A. Broker, H. D. Willauer and R. D. Rogers,
6
390LV scanning electron microscope. The regenerated cellulose
films in the dry state were frozen in liquid nitrogen, immediately
snapped, and then dried under vacuum. The free surface (side in
direct contact with the coagulant) and the fracture surface of the
films were sputtered with gold, and then photographed. TGA
was carried out with a NETZSCH STA 449 C thermal analyser
using alumina crucibles. The sample mass was ca. 10–15 mg per
measurement. The measurements were carried out under flowing
ACS Symp. Ser., 2002, 818, 270.
2
2
3
8 R. Lungwitz and S. Spange, New J. Chem., 2008, 32, 392.
9 E. Brendler, S. Fischer and H. Leipner, Cellulose, 2001, 8, 283.
0 B. Hinterstoisser, M. Akerholm and L. Salm e´ n, Biomacromolecules,
2003, 4, 1232.
˚
3
1 Y. Sun, L. Lin, C. S. Pang, H. B. Deng, H. Peng, J. Z. Li, B. H. He
and S. J. Liu, Energy Fuels, 2007, 21, 2386.
3
2 B. Morgensrern, H. W. Kammer and W. Berger, Acta Polym., 1992,
4
3, 356.
◦
-1
N
2
at a heating rate of 10 C min . The viscosity-average degree
33 C. L. McCormick, P. A. Callais and B. H. Hutchinson Jr., Macro-
molecules, 1985, 18, 2394.
of polymerization (DP) for the original and regenerated cellulose
materials was determined by using an Ubbelohde viscometer in
cupriethylenediamine hydroxide solutions.
3
4 D. Ishii, D. Tatsumi and T. Matsumoto, Biomacromolecules, 2003, 4,
238.
35 L. N. Zhang, D. Ruan and J. P. Zhou, Ind. Eng. Chem. Res., 2001,
0, 5923.
1
4
3
3
3
6 L. N. Zhang, D. Ruan and S. J. Gao, J. Polym. Sci., Part B: Polym.
Phys., 2002, 40, 1521.
7 H. G. Higgins, C. M. Stewart and K. J. Harrington, J. Polym. Sci.,
1961, 51, 59.
8 S. M. Zhou, K. Tashiro, T. Hongo, H. Shirataki, C. Yamane and T.
Ii, Macromolecules, 2001, 34, 1274.
9 Y. Kataoka and T. Kondo, Macromolecules, 1998, 31, 760.
0 L. J. Qu, Y. Zhang, J. Q. Wang and D. L. Chi, J. QingDao, University,
2008, 23, 44.
Acknowledgements
This work was supported financially by the National High
Technology Research and Development Program (863 Program,
No.2007AA05Z454), the National Natural Science Foundation
of China (No.20873036) and the Innovation Scientists and
Technicians Troop Construction Projects of Henan Province
3
4
4
4
4
1 A. Potthast, T. Rosenau, H. Sixta and P. Kosmaa, Tetrahedron Lett.,
(
No.084200510015).
2
002, 43, 7757.
2 P. Bonhote, D. Ana-Paula, N. Papageorgiou, K. Kalyanasundaram
and M. Gratzel, Inorg. Chem., 1996, 35, 1168.
3 J. G. Huddleston, H. D. Willauer, R. P. Swatloski, A. E. Visser and
R. D. Rogers, Chem. Commun., 1998, 1765.
References
1
D. Klemm, B. Heublein, H. P. Fink and A. Bohn, Angew. Chem., Int.
Ed., 2005, 44, 3358.
44 K. Fukumoto, M. Yoshizawa and H. Ohno, J. Am. Chem. Soc., 2005,
2
K. J. Edgar, C. M. Buchanan, J. S. Debenham, P. A. Rundquist, B. D.
Seiler, M. C. Shelton and D. Tindall, Prog. Polym. Sci., 2001, 26,
127, 2398.
45 Q. B. Liu, M. H. A. Janssen, F. V. Rantwijk and R. A. Sheldon, Green
Chem., 2005, 7, 39.
1
605.
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 268–275 | 275