The effect of the ionic liquid anion in the pretreatment of pine wood chips

Article abstract of DOI:10.1039/b918787a

The effect of the anion of ionic liquids on air-dried pine (Pinus radiata) has been investigated. All ionic liquids used in this study contained the 1-butyl-3-methylimidazolium cation; the anions were trifluoromethanesulfonate, methylsulfate, dimethylphosphate, dicyanamide, chloride and acetate. Using a protocol for assessing the ability to swell small wood blocks (10 × 10 × 5 mm), it was shown that the anion has a profound impact on the ability to promote both swelling and dissolution of biomass. Time course studies showed that viscosity, temperature and water content were also important parameters influencing the swelling process. We used Kamlet-Taft parameters to quantify the solvent polarity of the ionic liquids and found that the anion basicity described by the parameter β correlated with the ability to expand and dissolve pine lignocellulose. It is shown that 1-butyl-3-methylimidazolium dicyanamide dissolves neither cellulose nor lignocellulosic material.

Full text of DOI:10.1039/b918787a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/greenchem | Green Chemistry
The effect of the ionic liquid anion in the pretreatment of pine wood chips†
Agnieszka Brandt,a,b Jason P. Hallett,a David J. Leak,b,c Richard J. Murphyb,c and Tom Welton*a
Received 15th September 2009, Accepted 8th January 2010
First published as an Advance Article on the web 3rd February 2010
DOI: 10.1039/b918787a
The effect of the anion of ionic liquids on air-dried pine (Pinus radiata) has been investigated. All
ionic liquids used in this study contained the 1-butyl-3-methylimidazolium cation; the anions were
trifluoromethanesulfonate, methylsulfate, dimethylphosphate, dicyanamide, chloride and acetate.
Using a protocol for assessing the ability to swell small wood blocks (10 ¥ 10 ¥ 5 mm), it was shown
that the anion has a profound impact on the ability to promote both swelling and dissolution of
biomass. Time course studies showed that viscosity, temperature and water content were also
important parameters influencing the swelling process. We used Kamlet–Taft parameters to
quantify the solvent polarity of the ionic liquids and found that the anion basicity described by the
parameter b correlated with the ability to expand and dissolve pine lignocellulose. It is shown that
1-butyl-3-methylimidazolium dicyanamide dissolves neither cellulose nor lignocellulosic material.
talline polymer, such as N-methylmorpholine-N-oxide (NMO)7
Introduction
or concentrated phosphoric acid.8 However, recently a range
Lignocellulosic biomass has been identified as a suitable
of ionic liquids based on the imidazolium cation have been
feedstock for future large-scale biorefineries.1 Lignocellulose
shown to be effective cellulose solvents,9–14 yielding clear viscous
is widely distributed and can be grown and harvested on a
solutions of cellulose. The dissolved cellulose can subsequently
billion ton scale.2 Fuels and materials derived from biomass
be precipitated by adding water or protic organic solvents such
are potentially “carbon-neutral” or can even help to sequester
as ethanol.9 The crystallinity of the regenerated cellulose is
carbon dioxide. The production cost of lignocellulosic biomass
reduced, which significantly accelerates subsequent hydrolysis
is less than for starch-based substrates, but more extensive
with cellulases.15
pretreatment is required in order to release the carbohydrates
The interaction of solvents with solutes can be described
and other components from the resistant cell wall matrix, adding
by empirical polarity scales. One has been devised by Kamlet
significantly to process complexity and overall costs. Once
and Taft with three polarity parameters. These quantify the
released from the lignocellulose complex, the carbohydrates can
be fermented to liquid fuels or chemicals. Several pretreatment
general dipolarisability (p*), the hydrogen bond acidity (a)
and the hydrogen bond basicity (b). The methodology has
options have been developed.3 None is currently able to provide
been successfully applied to characterise molecular solvents16–19
the optimal economic and environmental performance.4
Recently, attention has been drawn towards the application of
as well as ionic liquids.20 It is possible to correlate Kamlet–
Taft parameters with the rates of chemical reactions, e.g. in
ionic liquids in lignocellulose processing.5,6 Ionic liquids are salts
nucleophilic substitutions21–23 and esterifications.24
that are liquid at or near room temperature. Liquid salts have
It has been observed that for ionic liquids, the anion has an
negligible vapour pressure under ambient conditions and are
impact on the ability to dissolve cellulose.9 NMR studies on
usually non-flammable. By modification of the cations and/or
ionic liquids that dissolved glucose and cellobiose showed that
anions an enormous range of potential ionic liquids is possible,
the anion strongly coordinates to the carbohydrates’ hydroxyl
encompassing a wide variety of solvent properties. Treatment
groups.25
of lignocellulose with certain ionic liquids results in enhanced
After initial studies focused on ionic liquids containing
glucose yields.
chloride as anion,9 Fukaya et al. found cellulose solubility in
Lignocellulose is a complex composite of polymeric carbo-
other 1,3-dialkylimidazolium ionic liquids that contained anions
hydrates (cellulose and hemicelluloses) and lignin. Cellulose
such as formate,11 acetate, phosphate or phosphonate anions,12
is the most abundant component, consisting of 1→4-b linked
and the concomitant measurement of Kamlet–Taft parameters
confirmed that these ionic liquids are characterised by a high b
glucopyranose. Only a few solvents are able to dissolve this crys-
parameter.
Ionic liquids have also been used to dissolve lignocellulosic
aDepartment of Chemistry, Imperial College London, London, SW7
2AZ, UK. E-mail: t.welton@imperial.ac.uk
material. So far, most literature reports only used ground or
milled lignocellulose. The partial dissolution of wood shavings
in 1-butyl-3-methylimidazolium chloride {[C4C1im]Cl} was first
described by Rogers and co-workers,5 with the dissolution of
wood sawdust and thermomechanical wood pulp in 1-butyl-3-
methylimidazolium chloride and in 1-allyl-3-methylimidazolium
bThe Porter Alliance, Imperial College London, London, SW7 2AZ, UK
cDivision of Biology, Imperial College London, London, SW7 2AZ, UK
† Electronic supplementary information (ESI) available: Swelling and
dissolution of air-dried pine wood chips in [C4C1im]Cl at 120 ◦C;
synthesis of ionic liquids. See DOI: 10.1039/b918787a
672 | Green Chem., 2010, 12, 672–679
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Table 1 Ionic liquids used in this study and their Kamlet–Taft parameters. Values with standard deviation were measured in triplicate. Measurements
were taken at 25 ◦C unless stated otherwise.
Ionic liquid
Abbreviation
a
b
p*
1-Butyl-3-methylimidazolium trifluoromethanesulfonate
[C4C1im][OTf]
[C4C1im][OTf]
0.634 (0.008)
0.6320
0.483 (0.008)
0.4620
0.974 (0.002)
1.0120
1-Butyl-3-methylimidazolium dicyanamide
1-Butyl-3-methylimidazolium methylsulfate
1-Butyl-3-methylimidazolium chloride
[C4C1im][N(CN)2]
[C4C1im][MeSO4]
0.543 (0.002)
0.545 (0.006)
0.49
0.596 (0.004)
0.672 (0.009)
0.83
1.052 (0.003)
1.046 (0.002)
1.03
[C4C1im]Cl, 75 ◦
[C4C1im]Cl
C
0.4411
0.8411
1.1411
1-Butyl-3-methylimidazolium dimethylphosphate
[C4C1im][Me2PO4]
[C4C1im][Me2PO4], wet, 8 wt%
[C2C1im][Me2PO4]
[C4C1im][MeCO2]
[C4C1im][MeCO2]
0.452 (0.003)
0.44
1.118 (0.012)
1.00
0.970 (0.006)
0.97
1-Ethyl-3-methylimidazolium dimethylphosphate
1-Butyl-3-methylimidazolium acetate
0.5112
1.0012
1.0612
0.470 (0.012)
1.201 (0.017)
0.971 (0.007)
0.4329
1.0529
1.0429
chloride being reported shortly afterwards.6 Lee et al. reported
partial delignification of 5 wt% maple wood flour with 1-ethyl-3-
methylimidazolium acetate {[C2C1im][MeCO2]},26 while the ob-
served wood solubility was very low (less than 0.5%). An attempt
was made to link solvent properties of 1,3-dialkylimidazolium
ionic liquids with the impact on lignocellulose, e.g. the extraction
of lignin with the Hildebrand solubility parameter. Wood flour
solubility was compared with b parameters taken from the
literature. A clear correlation could not be confirmed due to
lack of data. In contrast, Sun et al. were able to observe almost
complete dissolution of 5 wt% southern yellow pine and red oak
particles in [C2C1im][MeCO2].27 The wood could be fractionated
into a carbohydrate-rich fraction with reduced lignin content
(75% of original lignin content) and pure lignin after the treat-
ment. When catalytic amounts of acid were added to 1-butyl-
3-methylimidazolium chloride {[C4C1im]Cl} the carbohydrates
were hydrolysed to oligomers and monomers.28 The sugars were
partially transformed to degradation products such as furfurals,
while the insoluble fraction comprised mainly lignin.
Herein, we investigate the behaviour of chip-sized wood
specimens when brought into contact with a range of ionic
liquids that contain the 1-butyl-3-methylimidazolium cation. We
use Kamlet–Taft parameters to explain the exceptional solubility
of cellulose and lignocellulose in some of these ionic liquids. We
present evidence that the amount of water present in the biomass
plays an important role in the swelling/dissolution process of
lignocellulose.
on the ionic liquid–wood interaction. The ionic liquids were all
dried under vacuum; the moisture content of the individual ionic
liquids is listed in the experimental section.
We set out to monitor the behaviour of chips of pine in
ionic liquids at three different temperatures, 60 ◦C, 90 ◦C
and 120 ◦C. The initial observation was that, under the
conditions used, none of the ionic liquids investigated was able
to completely disintegrate the wood specimens. Instead a more-
or-less pronounced change of dimensions was observed. Hence
we decided to measure the specimens’ size changes.
Wood is an anisotropic and microstructured material. It is
made up of elongated cell channels that are several micrometers
long and are surrounded by lignified cell walls. The majority
of cells are orientated in the axial direction. There are few
longitudinal cells, restricting swelling in the radial direction.
The cells are inter-connected via pits and perforation plates,
creating long channels that enable transport of nutrients and
water between roots and leaves. Wood from temperate regions
forms annual rings (growth rings) while growing, due to seasonal
change of growth conditions.
Due to the known anisotropy of wood expansion, the
specimens were cut in a specific orientation to obtain consistent
results (Fig. 1). The three dimensions were defined as: a
tangential (parallel to growth ring), b radial (perpendicular to
growth ring) and c axial (longitudinal).
Results and discussion
The ionic liquids that are reported to affect lignocellulose
and its components often contain dialkylimidazolium cations.
Based on this, we screened a selected range of ionic liquids
(Table 1), containing the popular 1-butyl-3-methylimidazolium
{[C4C1im]+} cation and a variety of anions. We included anions
that are known to dissolve cellulose, lignocellulose and/or
lignin. All selected ionic liquids are miscible with water.
Fig. 1 Pine chip with growth rings. Definition of lengths a (tangential),
b (radial) and c (axial) relative to growth rings.
Kamlet–Taft parameters
The pine wood was used in air-dried form and thus contained
a substantial amount of water. We decided to use air-dried wood
because rigorous drying requires additional energy input, which
is undesirable for an energy-efficient pretreatment process. We
were also interested in the impact that this water might have
The abilities of the ionic liquids investigated to interact with
solutes were measured using a set of three solvatochromic
dyes. In accordance with previous measurements the dipolar-
ity/polarisability expressed through p* is high for all ionic
liquids. The a value and the b value, which express the ability
This journal is
The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 672–679 | 673
©

View Article Online
to donate and accept hydrogen bonds, respectively, depend on
the ionic liquid composition.20 The cation is responsible for the
a value,20 and in 1,3-dialkylimidazolium ionic liquids it largely
arises due to the acidity of the proton that is attached to the
carbon between the two ring nitrogen atoms. Our measurements
show that the a values were between 0.45 and 0.64, a range that
is typical of 1,3-dialkylimidazolium ionic liquids.20 The anion
is the most important determinant for b.20 In our study, the
b parameter varied between 0.5 and 1.2, which reflects the
range of anions. [C4C1im][OTf] was at the lower end of the
scale, while [C4C1im][Me2PO4] and [C4C1im][MeCO2] exhibited
very high b values. This means that the acetate anion is a
considerably stronger hydrogen bond acceptor than the chloride
or the dimethylphosphate (Table 1).
and [C4C1im]Cl exceeded 8%, correlating well with their high
b values of 1.12, 1.20 and 0.83. The swelling in these ionic
liquids was temperature-dependent. [C4C1im][MeCO2] swelled
the specimens by 15% at 90 C and more than 20% at 120 C
(Fig. 2b and c).
◦
◦
It is noteworthy that we observed a significant decrease of the
a value when b was high. This has been noticed previously for
other ionic liquids and ascribed to the strong hydrogen bonding
of the anion to the cation in these cases.20 This results in reduced
hydrogen bond donation from the cation to the dye probe.
For [C4C1im][MeCO2] and [C4C1im][Me2PO4], we found
significantly higher b values than previously reported in the
literature (Table 1). We suspected that the moisture content
of the ionic liquids might influence the Kamlet–Taft values,
as these ionic liquids are known to be very hygroscopic. We
deliberately left a [C4C1im][Me2PO4] sample to equilibrate with
air and determined the Kamlet–Taft values. We found that,
while a and p* remain relatively unaffected by the water, b
was significantly lower. This might explain the lower b values
found in the literature.
Swelling and dissolution of pine wood
Pine, a softwood, was chosen as the substrate because it is a
widely cultivated, fast-growing tree. Softwood is also regarded
as particularly recalcitrant to many pretreatment methods, due
to its high lignin content.4
The specimens with a mass of 0.15–0.20 g were immersed
in 1.5–2.0 g of ionic liquid (10% by weight). The ionic liquid
was added under vacuum. This preparation method enabled the
rapid filling of all the wood voids in the specimens, maximising
surface contact of the ionic liquid which resulted in a more
uniform swelling across all orientations and more consistent
results.
Correlation of anion basicity and the impact on wood specimens
Distinct differences were observed in the maximal swelling of
the wood specimen, depending in which ionic liquid it had
been placed. Expansion in the tangential direction (a in Fig. 1)
was chosen as the principal indicator for swelling, because the
swelling was most pronounced in this direction (Fig. 7).
The ionic liquid [C4C1im][OTf] with a b value of 0.49 was a
very poor swelling agent at all three temperatures, inducing some
slow swelling only at 120 ◦C (Fig. 2). The ionic liquids with
midrange b values, [C4C1im][N(CN)2] and [C4C1im][MeSO4]
(0.60 and 0.69 respectively), were able to swell the wood up to
8%. This exceeds the swelling ability of water, which expanded
air-dried pine chips by 5% at 60 ◦C and 90 ◦C. The swelling in
these ionic liquids was not affected by temperature. The swelling
in the ionic liquids [C4C1im][Me2PO4] and [C4C1im][MeCO2]
Fig. 2 Swelling in the tangential direction of pine specimens at different
temperatures. Each time point was averaged over three samples, except
for some of the [C4C1im][MeCO2] time points. Only one specimen
survived to the end of the experiment; the other two split in half between
the 2 h and 15 h measurements.
674 | Green Chem., 2010, 12, 672–679
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Several specimens incubated in [C4C1im][MeCO2] at 90 ◦C or
120 ◦C cracked along the growth ring. We suggest that the strong
swelling exerted considerable tension on the tissue structure of
the wood, which ultimately led to breaking.
Lignocellulosic components, particularly the carbohydrate
polymers, are rich in hydroxyl groups. Ionic liquid anions that
can strongly coordinate these hydroxyl groups (high b) are able
to weaken the hydrogen-bonding interactions in the wood matrix
and allow the chips to expand. Ionic liquids with low hydrogen-
bond basicity, such as [C4C1im][OTf], are not able to break the
hydrogen bonds in the matrix and therefore do not expand it at
all or only minimally, as observed for [C4C1im][OTf] at 120 ◦C
(Fig. 2c). Since these ionic liquids cannot dissolve the cellulose
fibres, the wooden ultrastructure is preserved and no extraction
is observed.
Fig. 3 displays the relationship between b and a, as well as the
b and c dimensions. The greatest dependency on b is observed for
the dimensions a and c, while length b was hardly susceptible to
variations in b. We ascribe the different behaviour of b compared
to a to the presence of radial pores (ray cells), which restrict
swelling in length b. Length a was more susceptible to expansion
due to the lack of radial cells. The data show that the swelling
in dimension a can be correlated with the b values of the ionic
liquids (Fig. 3a). Therefore we conclude that, when the cation
is 1-butyl-3-methylimidazolium, an anion with high hydrogen-
basicity allows stronger expansion of the wood matrix.
For the c dimension, we also observed a dependency on
b. The less basic ionic liquids caused expansion of c, with a
slight b dependency. For the ionic liquids [C4C1im][Me2PO4]
and [C4C1im][MeCO2], a reduction in c over time was observed,
which we ascribe to dissolution of cell wall material.
Rate of swelling
At 60 ◦C, the lowest temperature investigated, different rates for
the initial swelling were observed. The swelling was particularly
slow with [C4C1im][MeSO4] and [C4C1im][Me2PO4] (Fig. 2). In
contrast, the ionic liquid [C4C1im][N(CN)2] acted quickly on
lignocellulose even at this relatively low temperature.
We ascribe this behaviour to the relatively high viscosity of
ionic liquids, compared to molecular solvents. [C4C1im][N(CN)2]
has the lowest viscosity at room temperature of the ionic liquids
tested (29.3 cP at 298 K30). Viscosity is temperature-dependent
and, in the case of ionic liquids, the relationship follows the
Vogel–Tammann–Fulcher (VTF) equation.31 The viscosity is
greatest close to the melting point and decreases exponentially
with increasing temperature. 120 ◦C is more than 100 ◦C
above the melting point for the ionic liquids shown in Fig. 2.
Therefore viscosity differences between different ionic liquids at
Fig. 3 Correlation between the relative cube dimensions a, b and c and
Kamlet Taft parameter b. The pine chips were pretreated at 120 ◦C. A
regression line is shown for the 15 h data points (a/a0 = 0.154b + 0.978,
p = 0.054). Error bars have been omitted for clarity.
◦
◦
90 C will be small, and even less at 120 C. This explains, at
least partially, why the observed swelling rates are similar for
[C4C1im][N(CN)2], [C4C1im][MeSO4], [C4C1im][Me2PO4] and
[C4C1im][MeCO2] at 90 ◦C and 120 ◦C.
the presence of hot ionic liquid. For dimension a, swelling is the
dominant process, as discussed earlier. [C4C1im][MeCO2] was
exceptionally capable of expanding the a dimension.
Comparison of [C4C1im][Me2PO4] and [C4C1im][MeCO2]
We observed dissolution of pine cubes in the ionic liquids
[C4C1im][MeCO2] (Fig. 4a) and [C4C1im]Cl (see ESI), and
to a lesser extent in [C4C1im][Me2PO4] (Fig. 4b). The three
dimensions (a, b and c in Fig. 1) were affected differently by
The profile for [C4C1im][Me2PO4] was similar to that of
[C4C1im][MeCO2], but swelling and dissolution were much less
pronounced (Fig. 4b). An explanation could be the lower b value
of [C4C1im][Me2PO4]. It is possible that the size of the anion may
This journal is
The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 672–679 | 675
©

View Article Online
liquid. All ionic liquids are more or less hygroscopic. For a given
cation the amount of water that can be taken up by the ionic
liquid increases with increasing hydrogen bond basicity of the
anion.33 We found that [C4C1im][Me2PO4] can take up more than
24 wt% (>400 mol%) water if allowed to equilibrate with the air
in the laboratory. [C4C1im][MeCO2] could take up 17 wt% (>200
mol%) water. Therefore, the ionic liquids needed to be dried
thoroughly before use in order to allow comparison. Our drying
procedure brought the water content to below 0.3 wt%. For
example, Karl Fisher titration of dried [C4C1im][Me2PO4] and
[C4C1im][MeCO2] showed that 1.5 ml of ionic liquid contained
2.1 mg and 2.5 mg water, respectively. The pine chips used
in this study contained 9 wt% water compared to the oven-
dry weight. A typical wood specimen of 150 mg therefore
contained 14 mg water. Hence, in our experiments the air-
dried biomass introduced the greater part of the water into the
system.
In order to test this hypothesis, we conducted an experiment
with thoroughly dried wood specimens and handled the samples
under a nitrogen atmosphere at all times. The pine cubes were
dried under high vacuum at room temperature and immersed in
the dried ionic liquids. The pretreatment temperature was set to
120 ◦C.
Interestingly, the performance of [C4C1im][MeCO2] improved
greatly under these conditions, resulting in softening and disin-
tegration of the pine specimen at 120 ◦C in less than 15 h (Fig. 5).
When exposing the [C4C1im][MeCO2] treated wood to air, the
structure re-hardened. We hypothesise that this was due to the
uptake of atmospheric moisture. However, dry pine specimens
treated with [C4C1im][Me2PO4] and [C4C1im]Cl remained intact
and no softening was observed.
We suggest that the differences in moisture content of ionic liq-
uids and/or biomass is a likely explanation for the contradicting
observations for wood solubility in [C2C1im][MeCO2] reported
in the literature,26,27 and that researchers must carefully monitor
the water content in their systems when using ionic liquids for
lignocellulose treatment.
Fig. 4 Swelling profile of a pine chip in [C4C1im][MeCO2] and in
[C4C1im][Me2PO4] at 120 ◦C. Only one out of three replicates in
[C4C1im][MeCO2] remained intact during the treatment.
also play a role, a smaller ion being able to diffuse faster within
the cell wall matrix.
The observed water-sensitivity is also of particular signifi-
cance for the practical application of ionic liquids in biomass
processing, as the feedstock will, under normal circumstances,
have a proportion of bound water (8 to 15% compared to
oven-dry biomass) and, when freshly harvested, appreciable
quantities of free water. An industrial process that is based on
dissolving lignocellulose with a dry ionic liquid will have to deal
with additional energy requirements for thoroughly drying the
substrate as well as the ionic liquid.
The choice of ionic liquid anion in combination with water
content may have a profound impact on the dissolution process.
This is shown in Fig. 6. Pine flour, either air-dried or vacuum-
dried, was immersed in [C4C1im][MeCO2] and [C4C1im]Cl
and incubated at 90 ◦C under a nitrogen atmosphere. The
[C4C1im][MeCO2] was slightly yellow, while the [C4C1im]Cl
was a white powder at room temperature. [C4C1im]Cl is very
hygroscopic, attracting large amounts of water until it becomes
a liquid, if it is not stored well sealed and under a dry atmosphere.
We observed that the [C4C1im][MeCO2] liquor became more
coloured when vacuum-dried pine powder was used. Conversely,
the liquor of [C4C1im]Cl with vacuum-dried pine remained
colourless during the 15 h treatment. The [C4C1im]Cl resolidified
The exact amount of wood material that was solubilised
cannot be estimated from this profile, because swelling and
dissolution take place simultaneously. While reduction of the
specimen’s size is probably due to solubilisation, the swelling
caused by continued disintegration of the cell wall structure
counteracts this effect. Interestingly, the dissolution of air-dried
wood specimens under the chosen conditions was incomplete
even after 2 days. This could be due to saturation of the ionic
liquid with solutes. However, this is inconsistent with a linear
decrease of length c, suggesting that the solubility limit had not
been reached in our experiment. While dissolution of particles
up to 1 mm appears to give good yields,27 the yield from larger
wood chips seems to be impaired. Slow dissolution of chip-sized
lignocellulose has been previously reported.5,6 This suggests that
particle size is an important factor for lignocellulose dissolution
in ionic liquids.
The influence of water
We hypothesised that the lignocellulose dissolution may be
inhibited by water, as noted by others for cellulose dissolution.32
Moisture can be introduced by the biomass and/or by the ionic
676 | Green Chem., 2010, 12, 672–679
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
also reported that 1-benzyl-3-methylimidazolium dicyanamide
{[BzC1im][N(CN)2]} dissolved 2 wt% of Southern Pine thermo-
mechanical pulp.6 However, doubts have already been raised by
others.36 We investigated this claim by adding 2% dry cellulose
to rigorously dried [C4C1im][N(CN)2]. No dissolution was
observed either at room temperature or at 100 ◦C. We speculated
that chloride ions, a common contaminant of dicyanamide
ionic liquids, could bring about cellulose dissolution, as it is
known that [C4C1im]Cl is a potent cellulose solvent. Therefore,
increasing amounts [C4C1im]Cl were added. However, a mix-
ture of [C4C1im][N(CN)2] and [C4C1im]Cl containing 7 mol%
[C4C1im]Cl did not dissolve 2 wt% cellulose. Also, 20 mol% and
30 mol% solutions were not able to dissolve cellulose.
Fig.
5
Dry pine specimens treated at 120 ◦C for 15
h with
[C4C1im][Me2PO4] (left), [C4C1im][MeCO2] (middle) and [C4C1im]Cl
(right).
A b value of 0.60 was measured for the pure ionic liquid
(Table 1). This is significantly lower than the b values of other
known cellulose-dissolving ionic liquids, such as [C4C1im]Cl,
[C4C1im][Me2PO4] or [C4C1im][MeCO2]. We therefore conclude
that the dicyanamide anion is not capable of forming sufficiently
strong hydrogen bonds to disrupt the hydrogen bonds in
cellulose.
Fig. 6 IL solutions after treating pine flour at 90 ◦C. From left to right:
vacuum-dried and air-dried flour in [C4C1im][MeCO2] and vacuum-
dried and air-dried flour in [C4C1im]Cl.
upon cooling, confirming that water uptake had been prevented
during the treatment.
Conclusions
This observation suggests that a certain amount of water
seems to be necessary to promote lignin solubilisation in
[C4C1im]Cl, while this does not seem to be true if the anion
is acetate. Miyafuji et al. report that, at lower temperatures
(<100 ◦C), the carbohydrate fraction of Japanese beech is
preferentially solubilized in [C2C1im]Cl.34 Other researchers have
noted that high lignocellulose loading influences the ratio of
carbohydrate/lignin solubilisation.27 Using these findings and
our own, there might be interesting opportunities for selective
extraction of the cellulose fraction from biomass. The influence
of the water content on the yield remains to be investigated.
In agreement with Zavrel et al.,37 we have found that ionic liquids
that dissolve cellulose also dissolve lignocellulose. When paired
with the 1-butyl-3-methylimidazolium cation, the hydrogen-
bond basicity of the anion, expressed by the Kamlet–Taft
parameter b, can be correlated with the swelling and dissolution
of lignocellulose. Ionic liquids with b >0.80 exhibited swelling,
followed by reduction of chip size. The swelling and dissolution
in these ionic liquids is temperature-dependent, being better at
◦
temperatures larger than 100 C. Swelling of wood blocks was
most pronounced in the tangential direction, while dissolution
seems to take place mainly in the axial direction of the wood
blocks. Out of the investigated ionic liquids, [C4C1im][MeCO2]
was identified as most effective swelling and dissolution agent,
in agreement with its high b value. The dissolution of air-dried
wood chips was slow and incomplete after 2 days at 120 ◦C. We
have also shown that the amount of water present within the
system is important for the swelling and dissolution of biomass.
Our results cast doubt that ionic liquids containing the
dicyanamide anion can dissolve significant amounts of wood
as well as cellulose. A significant amount of chloride impurities
does not improve the solubility of cellulose.
The inability of [C4C1im][N(CN)2] to dissolve cellulose
We could not find evidence for dissolution of significant
amounts of wood into [C4C1im][N(CN)2] (Fig. 7). Earlier reports
claim that imidazolium-based ionic liquids with dicyanamide
anions can dissolve cellulose in large amounts.35 It was
Experimental
Materials
Cellulose (fibrous, long) was obtained from Sigma. 1-
Methylimidazole (Acros Organics) and 1-butylimidazole
(Aldrich) were distilled under vacuum from potassium hy-
droxide. 1-Chlorobutane and 1-bromobutane were distilled
over P2O5. Dimethylsulfate (Aldrich) was distilled over
CaO. Trimethylphosphate (Aldrich), silver acetate, sodium di-
cyanamide and silver nitrate (Aldrich) were used as received.
Ethyl acetate and acetonitrile were dried over CaH2. Toluene
was distilled over sodium (with benzophenone as moisture
indicator).
Fig. 7 Swelling profile of pine specimens in [C4C1im][N(CN)2] at
120 ◦C.
This journal is
The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 672–679 | 677
©

View Article Online
All ionic liquids were prepared using literature procedures
(see ESI).24,38,39 1-Butyl-3-methylimidazolium dicyanamide and
1-butyl-3-methylimidazolium acetate were synthesised using 1-
butyl-3-methylimidazolium bromide24 and the silver salts of the
respective anions.40
within 60 h. We could not detect any decomposition products in
the respective 1H NMR spectrum.
Dissolution of pine flour
Pine wood was ground for 5 min in a water-cooled bench-top
grinding machine. Part of the flour was dried under vacuum at
50 ◦C for 5 h. The dried pine flour was transferred into the glove
box for storage. 30 mg of wet or dried flour was mixed with 1.0 g
of [C4C1im][MeCO2] or [C4C1im]Cl. The flasks were kept under
nitrogen atmosphere at all times. The mixtures were heated to
90 ◦C while stirring.
Preparation of wood chips
Pine (Pinus radiata) sapwood was obtained from a 15 year
old tree harvested at Silwood Park, Ascot, in June 2008. All
wood was taken from material sawn from the main trunk.
Pine sapwood specimens of dimensions 10 mm ¥ 10 mm ¥
5 mm were prepared. Specimens were stored air-dried at normal
temperature and room humidity in the laboratory. The moisture
content of the pine wood was 9% on oven-dry basis.
Measurement of Kamlet–Taft parameters
Stock solutions of the Kamlet–Taft dyes (Reichhardt’s dye, 4-
nitroaniline and N,N-diethyl-4-nitroaniline) were prepared in
dichloromethane. The solution was transferred by syringe into
a cuvette of 1 mm thickness under nitrogen atmosphere. The
solvent was evaporated, the vacuum-dried ionic liquid added
and the solution carefully mixed with shaking or a syringe
needle. A spectrum was recorded using a PC-controlled Perkin
Elmer Lambda 2 spectrometer with thermostatted sample
holder. The measurements were conducted at 25 ◦C, ex◦cept for
[C4C1im]Cl, for which the spectra were recorded at 75 C. The
maximum of the absorption peak was determined using Origin
7. A Gaussian function was fitted to a 40 nm wide region with
the peak maximum in the centre.
Swelling experiments
Pyrex culture tubes (Ø = 25 mm) with screw caps and Teflon
linings were used as reaction vessels. The weight of dry wood
specimens chips was determined using a precision balance
( 0.0001 g). Lengths were measured with a calliper with an
accuracy of 0.01 mm. Experiments were performed in triplicate
to account for sample heterogeneity. The ionic liquids were
thoroughly dried before use at 40 ◦C under vacuum and
overnight, except for [C4C1im][MeCO2] and [C4C1im][Me2PO4],
which were dried at 50 ◦C and 100 ◦C, respectively. The residual
water content was 200 ppm for [C4C1im][OTf], 400 ppm for
[C4C1im][MeSO4], 3100 ppm for [C4C1im][N(CN)2], 1200 ppm
for [C4C1im][Me2PO4] and 1600 ppm for [C4C1im][MeCO2],
according to Karl Fisher titration. A Schlenk flask was used as a
soaking apparatus. The culture tube containing a wood sample
was inserted into the Schlenk flask and the Schlenk sealed with
a rubber septum. The flask was evacuated for 60 s and 1.5 ml
ionic liquid injected through the septum with a syringe. The
vacuum was applied until the chip appeared to be fully soaked.
The tube was tightly sealed and the sample kept at 4 ◦C up for
to 5 h until preparation of the remaining samples was finished.
Samples were heated in an oven to the desired temperature. For
measurements they were taken out and left to cool. The timing
was stopped while samples were outside the oven. The wood
specimens were removed from the ionic liquid and their surface
wiped dry with a paper towel. The lengths were determined and
samples placed in the glass tubes as quickly as possible.
Acknowledgements
We are grateful to the Porter Institute, Imperial College London,
UK, for financial support of the PhD studentship for Agnieszka
Brandt. The authors would like to thank Dr Mike Ray for
assistance and discussion in preparation of the wood materials.
References
1 A. J. Ragauskas, C. K. Williams, B. H. Davison, G. Britovsek, J.
Cairney, C. A. Eckert, W. J. Frederick, J. P. Hallett, D. J. Leak, C. L.
Liotta, J. R. Mielenz, R. Murphy, R. Templer and T. Tschaplinski,
Science, 2006, 311, 484–489.
2 R. D. Perlack, L. L. Wright, A. F. Turhollow, R. L. Graham, B. J.
Stokes and D. C. Erbach, Biomass as Feedstock for a Bioenergy
and Bioproducts Industry: The Technical Feasibility of a Billion-Ton
Annual Supply, U.S. Department of Energy and U.S. Department of
Agriculture, 2005.
3 N. Mosier, C. Wyman, B. Dale, R. Elander, Y. Y. Lee, M. Holtzapple
and M. Ladisch, Bioresour. Technol., 2005, 96, 673–686.
4 P. Kumar, D. M. Barrett, M. J. Delwiche and P. Stroeve, Ind. Eng.
Chem. Res., 2009, 48, 3713–3729.
5 D. A. Fort, R. C. Remsing, R. P. Swatloski, P. Moyna, G. Moyna
and R. D. Rogers, Green Chem., 2007, 9, 63–69.
6 I. Kilpelainen, H. Xie, A. King, M. Granstrom, S. Heikkinen and
D. S. Argyropoulos, J. Agric. Food Chem., 2007, 55, 9142–9148.
7 C. H. Kuo and C. K. Lee, Bioresour. Technol., 2009, 100, 866–871.
8 Y. H. P. Zhang, S. Y. Ding, J. R. Mielenz, J. B. Cui, R. T. Elander, M.
Laser, M. E. Himmel, J. R. McMillan and L. R. Lynd, Biotechnol.
Bioeng., 2007, 97, 214–223.
9 R. P. Swatloski, S. K. Spear, J. D. Holbrey and R. D. Rogers, J. Am.
Chem. Soc., 2002, 124, 4974–4975.
10 J. Wu, J. Zhang, H. Zhang, J. S. He, Q. Ren and M. Guo,
Biomacromolecules, 2004, 5, 266–268.
For swelling under anhydrous conditions the pine specimens
(10 ¥ 10 ¥ 5 mm) were dried under reduced pressure for 5 days,
resulting in a mass loss of 9%. Oven-drying resulted in a
comparable mass loss. The samples were handled in a glove
box or under nitrogen atmosphere at all times. 2 g of [C4C1im]Cl
or 1.5 ml of dry [C4C1im][Me2PO4] and [C4C1im][MeCO2] were
transferred into Schlenk flasks, the pine specimens added and
the Schlenk flask evacuated to allow filling of the wood pores.
The samples were heated to 120 ◦C for 15 h.
Ionic liquid stability
Thermogravimetric analysis (TGA) did not show no-
ticeable decomposition for [C4C1im]Cl, [C4C1im][N(CN)2],
[C4C1im][MeSO4] and [C4C1im][Me2PO4] at 120 ◦C over a period
of 60 h. For [C4C1im][MeCO2], we observed a 12% weight loss
11 Y. Fukaya, A. Sugimoto and H. Ohno, Biomacromolecules, 2006, 7,
3295–3297.
678 | Green Chem., 2010, 12, 672–679
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
12 Y. Fukaya, K. Hayashi, M. Wada and H. Ohno, Green Chem., 2008,
10, 44–46.
13 S. Kohler, T. Liebert, M. Schobitz, J. Schaller, F. Meister, W. Gunther
and T. Heinze, Macromol. Rapid Commun., 2007, 28, 2311–2317.
14 H. Zhao, G. A. Baker, Z. Y. Song, O. Olubajo, T. Crittle and D.
Peters, Green Chem., 2008, 10, 696–705.
27 N. Sun, M. Rahman, Y. Qin, M. L. Maxim, H. Rodriguez and R. D.
Rogers, Green Chem., 2009, 11, 646–655.
28 C. Sievers, M. B. Valenzuela-Olarte, T. Marzialetti, D. Musin, P. K.
Agrawal and C. W. Jones, Ind. Eng. Chem. Res., 2009, 48, 1277–1286.
29 Y. S. Wu, T. Sasaki, K. Kazushi, T. Seo and K. Sakurai, J. Phys.
Chem. B, 2008, 112, 7530–7536.
15 H. Zhao, C. I. L. Jones, G. A. Baker, S. Xia, O. Olubajo and V. N.
Person, J. Biotechnol., 2009, 139, 47–54.
16 M. J. Kamlet, J. L. M. Abboud, M. H. Abraham and R. W. Taft,
J. Org. Chem., 1983, 48, 2877–2887.
30 Y. Yoshida, O. Baba, C. Larriba and G. Saito, J. Phys. Chem. B, 2007,
111, 12204–12210.
31 Ionic Liquids in Synthesis, ed. P. Wasserscheid and T. Welton, Wiley-
VCH, 2007.
17 M. J. Kamlet, J. L. Abboud and R. W. Taft, J. Am. Chem. Soc., 1977,
99, 6027–6038.
32 M. Mazza, D. A. Catana, C. Vaca-Garcia and C. Cecutti, Cellulose,
2009, 16, 207–215.
33 L. Cammarata, S. G. Kazarian, P. A. Salter and T. Welton, Phys.
Chem. Chem. Phys., 2001, 3, 5192–5200.
18 M. J. Kamlet and R. W. Taft, J. Am. Chem. Soc., 1976, 98,
377–383.
19 R. W. Taft and M. J. Kamlet, J. Am. Chem. Soc., 1976, 98, 2886–2894.
20 L. Crowhurst, P. R. Mawdsley, J. M. Perez-Arlandis, P. A. Salter and
T. Welton, Phys. Chem. Chem. Phys., 2003, 5, 2790–2794.
21 G. Ranieri, J. P. Hallett and T. Welton, Ind. Eng. Chem. Res., 2008,
47, 638–644.
22 J. P. Hallett, C. L. Liotta, G. Ranieri and T. Welton, J. Org. Chem.,
2009, 74, 1864–1868.
34 H. Miyafuji, K. Miyata, S. Saka, F. Ueda and M. Mori, in 58th
Annual Meeting of the Japan Wood Research Society, Springer,
Tsukuba, Tokyo, 2008, pp. 215–219.
35 Q. B. Liu, M. H. A. Janssen, F. van Rantwijk and R. A. Sheldon,
Green Chem., 2005, 7, 39–42.
36 G. Laus, G. Bentivoglio, H. Schottenberger, V. Kahlenberg, H.
Kopacka, T. Ro¨der and H. Sixta, Lenzinger Berichte, 2005, 84, 71–85.
37 M. Zavrel, D. Bross, M. Funke, J. Buchs and A. C. Spiess, Bioresour.
Technol., 2009, 100, 2580–2587.
23 L. Crowhurst, R. Falcone, N. L. Lancaster, V. Llopis-Mestre and T.
Welton, J. Org. Chem., 2006, 71, 8847–8853.
24 T. P. Wells, J. P. Hallett, C. K. Williams and T. Welton, J. Org. Chem.,
2008, 73, 5585–5588.
25 R. C. Remsing, G. Hernandez, R. P. Swatloski, W. W. Massefski,
R. D. Rogers and G. Moyna, J. Phys. Chem. B, 2008, 112, 11071–
11078.
38 E. Kuhlmann, S. Himmler, H. Giebelhaus and P. Wasserscheid, Green
Chem., 2007, 9, 233–242.
39 J. D. Holbrey, W. M. Reichert, R. P. Swatloski, G. A. Broker, W. R.
Pitner, K. R. Seddon and R. D. Rogers, Green Chem., 2002, 4, 407–
413.
26 S. H. Lee, T. V. Doherty, R. J. Linhardt and J. S. Dordick, Biotechnol.
40 D. Gerhard, S. C. Alpaslan, H. J. Gores, M. Uerdingen and P.
Wasserscheid, Chem. Commun., 2005, 5080–5082.
Bioeng., 2009, 102, 1368–1376.
This journal is
The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 672–679 | 679
©