View Article Online / Journal Homepage / Table of Contents for this issue
Green Chemistry Dynamic Article Links
Cite this: Green Chem., 2012, 14, 2153
www.rsc.org/greenchem
New natural and renewable low transition temperature mixtures (LTTMs):
screening as solvents for lignocellulosic biomass processing
Mara Francisco,* Adriaan van den Bruinhorst and Maaike C. Kroon
Received 1st May 2012, Accepted 20th June 2012
DOI: 10.1039/c2gc35660k
Published on 20 June 2012 on http://pubs.rsc.org | doi:10.1039/C2GC35660K
New natural and biorenewable solvents are prepared in this
work consisting of high melting temperature starting
Downloaded by University of Stellenbosch on 19 March 2013
materials that form low transition temperature mixtures
(LTTMs). The physicochemical behaviour of these new sol-
vents can be tailored by a judicious selection of the constitu-
ents nature and ratio. Their suitability as solvents for
lignocellulosic biomass processing is evaluated.
Deep eutectic solvents (DESs) were presented by Abbott et al.
for the rst time as suitable alternative solvents compared to con-
ventional ionic liquids.1 The main constituents of these eutectic
mixtures are high melting temperature solids that show strong
hydrogen bonding interactions. The resultant combinations often
provide wide liquid range and unusual low transition tempera-
tures. The rst reported DES consisted of urea and choline chlor-
ide and is the most widely investigated. DESs can also be
formed by mixing a quaternary ammonium or phosphonium salt
with a hydrogen bond donor agent, for instance, acids, alcohols,
amines or carbohydrates.24 Most DESs share the promising
solvent characteristics of ILs. They often show low volatility,
wide liquid range, water-compatibility, non-ammability, non-
toxicity, biocompatibility and biodegradability. Furthermore,
they show the ability to customize their physical properties by
choosing the right DES constituents in terms of chemical nature,
relative compositions or water content.5 In addition, they can be
easily prepared from readily available materials at high purities
and low cost compared to ILs, and they can be considered as
environmentally benign solvents. Because of keeping most of
the advantages of ILs but overcoming some of their limitations,
DESs open room to research in multiple applications. For
instance, they have been successfully applied as solvents or cata-
lysts for chemical reactions or biotransformations,6 metal electro-
deposition,7 synthesis of nanoparticles,8 liquid and gas
separations9 and heat transfer uids.10 Still, a limitless variety of
DESs and a wide range of possible applications are to be
discovered.
Separation Technology Group, Department of Chemical Engineering
and Chemistry, Eindhoven University of Technology, Den Dolech 2,
5600 MB Eindhoven, The Netherlands.
E-mail: M.Francisco-Casal@tue.nl; Fax: +31 40 246 3966;
Tel: +31 40 247 4716
Electronic supplementary information (ESI) available: Materials and
methods [Section ESI1]. See DOI: 10.1039/c2gc35660k
This journal is The Royal Society of Chemistry 2012
COMMUNICATION
In a future bio-economy, nding a suitable solvent for ligno-
cellulosic biomass is becoming the Achilless heel of renewable
biofuels processing. Conventional methods for biomass decon-
struction into cellulose, hemicelluloses and lignin bioproducts
often require extreme and expensive techniques, but some ILs
were found to succeed as solvents using milder conditions.11
Nevertheless, IL technologies for large scale application still
show limitations in terms of recoverability and cost. Some inter-
esting studies were found in the use of solvent systems contain-
ing only a minor molar fraction of IL.12 Despite their high
efciency in dissolution of cellulose, new green approaches
towards the replacement of volatile organic solvents were taken
into consideration here.
Inspired by some previous publications,14 DESs came up as a
promising alternative for ILs. In the present work, we explore a
series of new low transition temperature mixtures prepared by
combining some natural and renewable biomaterials presented
in Fig. 1.
The selection of the starting materials was made on the basis
of the available functional groups and the key interactions
involved in lignocellulosic biomass dissolution.13 Recently, new
studies reported the use of amino acid based ILs in the search for
biodegradable and natural alternatives to conventional ILs.14
However, DESs seem to be a cheaper and more promising
alternative. For this reason, a set of representative natural amino
acids with suitable structures and functional groups, some essen-
tial nutrients represented by choline chloride and nicotinic acid,
as well as different natural acids present in fruits and vegetables
were tested as liquid-phase promoters. The preparation of the
new solvents was done by mixing both starting materials in the
solid state, followed by melting them at 60 C for mixtures con-
taining lactic or oxalic acid, or at 130 C for nicotinic and malic
acid mixtures. It is noticeable that the melting temperature was
always found to be lower than the melting point of any of the
starting materials. The higher the temperature and the better the
mixing, the faster the melting was observed. Nevertheless,
higher thermal stability was found for mixtures obtained at lower
temperatures for longer times. Once a clear and transparent
liquid was formed with no evidence of solid particles, the
mixture was cooled down and DSC analysis was carried out to
determine the phase transition temperature of the formed liquids.
Water content was always less than 1 wt%. A more detailed
description of the experimental procedure is included as ESI.
Table 1 shows the hydrogen bond donorhydrogen bond
acceptor combinations, which were resulting in clear liquids
Green Chem., 2012, 14, 21532157 | 2153
 View Article Online

Table 1 Hydrogen bond donorhydrogen bond acceptor mixtures

resulting in formation of clear DESs (in green), precipitation of solid
particles when cooling down (in yellow) and no melting (in red)

Hydrogen bond donor Molar ratio Hydrogen bond acceptor

Lactic acid Alanine

Lactic acid Alanine
Lactic acid Betaine
Lactic acid Betaine
Lactic acid Choline chloride
Lactic acid Choline chloride
Lactic acid Choline chloride
Published on 20 June 2012 on http://pubs.rsc.org | doi:10.1039/C2GC35660K

Lactic acid Glycine

Lactic acid Glycine
Lactic acid Histidine
Downloaded by University of Stellenbosch on 19 March 2013

Lactic acid Histidine

Lactic acid Proline
Lactic acid Proline
Malic acid Alanine
Malic acid Betaine
Malic acid Choline chloride
Malic acid Choline chloride
Malic acid Choline chloride
Malic acid Glycine
Malic acid Histidine
Malic acid Nicotinic acidb
Malic acid Proline
Oxalic acid (anhydrous) Alanine
Oxalic acid (anhydrous) Betaine
a
Fig. 1 Selected hydrogen bond donors and acceptors for the formation Oxalic acid (anhydrous) ChCl
of new low transition temperature mixtures. Oxalic acid (anhydrous) Glycine
Oxalic acid (anhydrous) Histidine
Oxalic acid (anhydrous) Proline
(in green), liquids at the set temperature but with formation of
solid particles when cooling down (in yellow), and with no evi- Oxalic acid (dihydrate) Alanine
dence of melting for the selected combination or ratio (in red). Oxalic acid (dihydrate) Alanine
The building principles are not easy to generalize. Unlike Oxalic acid (dihydrate) Betaine
normal chemical bonds, hydrogen bonds present different Oxalic acid (dihydrate) Choline chloride
contact distances and binding energies which do not depend Oxalic acid (dihydrate) Glycine
only on the donor and acceptor nature.15 The picture included as Oxalic acid (dihydrate) Histidine
Fig. 2 shows an example of the evolution of the phase transition Oxalic acid (dihydrate) Proline
for the malic acidcholine chloride series reported in Table 1.
Oxalic acid (dihydrate) Nicotinic acid
The proton afnity (PA)/pKa equalization plays a role in
Nicotinic acid ChCl
strengthening the H-bond, so the pKa slide rule was taken into
account in the selection of the H-bonding counterparts.16 The Nicotinic acid Proline
pKa values for the main functional groups are included in Fig. 1. a
Differences found with the phase behavior reported by Abbott et al.1
The acidity of the proton is also responsible for the formation of b
Nicotinic acid has been found to perform better as hydrogen bond
an LTTM instead of an IL. For instance, when lactic acid is com- acceptor counterpart.
bined with choline chloride, a liquid is formed at room tempera-
ture. However, the ionic liquid choline lactate is not produced as
reected in the examples of IR spectra included in ESI. An IL lower eld in 1H-NMR. Representative 1H-NMR spectra are
is a liquid below 100 C, solely consisting of ions. For the IL to included in ESI.
be formed, a stronger base with a higher pKa needs to be facing The colour code in Table 1 evidences that nicotinic acid has
the H-bond donor or a stronger acid needs to be facing the accep- been proved to interact preferentially when facing a hydrogen
tor. Hydrogen bonding can be evidenced as well by the shifts in bond donor counterpart instead of an acceptor.
the representative peaks of the involved bonds in the FTIR Table 2 captures the transition temperatures of the selected
spectra.15 A shift in the resonance signal can also be noticed to mixtures. These values are signicantly lower compared with the

2154 | Green Chem., 2012, 14, 21532157 This journal is The Royal Society of Chemistry 2012
 View Article Online
Published on 20 June 2012 on http://pubs.rsc.org | doi:10.1039/C2GC35660K

Fig. 2 Malic acidcholine chloride mixtures showing the phase tran-

sition for different hydrogen bond donorhydrogen bond acceptor molar
ratios at room temperature.
Downloaded by University of Stellenbosch on 19 March 2013

Table 2 Selection of solvents for screening of biopolymer solubilities,

including glass transition temperatures (C)

Hydrogen bond
Name Hydrogen bond donor Ratio acceptor Tg

LA9:1 Lactic acid 9:1 Alanine 59.31

LB2:1 Lactic acid 2:1 Betaine 46.86
LC1.3:1 Lactic acid 1.3 : 1 Choline chloride 76.75
LC2:1 Lactic acid 2:1 Choline chloride 77.73
LC5:1 Lactic acid 5:1 Choline chloride 69.23
LC10:1 Lactic acid 10 : 1 Choline chloride 66.3
LG9:1 Lactic acid 9:1 Glycine 54.51
LH9:1 Lactic acid 9:1 Histidine 39.22
LP2:1 Lactic acid 2:1 Proline 36.69
MA1:1 Malic acid 1:1 Alanine 42.64
MB1:1 Malic acid 1:1 Betaine 20.01
MC1:1 Malic acid 1:1 Choline chloride 56.48 Fig. 3 DSC curves for some representative mixtures described in
MG1:1 Malic acid 1:1 Glycine 34.08 Table 2 screened for lignin, cellulose and starch solubility.
MP1:1 Malic acid 1:1 Proline 13.64
MP1:2 Malic acid 1:2 Proline 15.51
MP1:3 Malic acid 1:3 Proline 44.38
MH2:1 Malic acid 2:1 Histidine
MN9:1 Malic acid 9:1 Nicotinic acid
OB1:1 Oxalic acid dihydrate 1:1 Betaine 17.19 The importance of this screening lies in the evaluation of the
OP1:1 Oxalic acid dihydrate 1:1 Proline 42.91 potential ability of LTTMs to deconstruct the lignocellulosic
OC1:1 Oxalic acid dihydrate 1:1 Choline chloride 40.17 biomass structure. High selectivity is desirable for separating
OG3:1 Oxalic acid dihydrate 3:1 Glycine
ON9:1 Oxalic acid dihydrate 9:1 Nicotinic acid
lignin from cellulose and hemicellulose, and high solubility
OH9:1 Oxalic acid dihydrate 9:1 Histidine leads to efcient hydrolysis. Two different approaches can be
OCA1:1 Oxalic acid anhydrous 1:1 Choline chloride 46.06 considered for the hydrolysis with these solvents: catalytic or
OPA1:1 Oxalic acid anhydrous 1:1 Proline 14.45 enzymatic hydrolysis. For catalytic hydrolysis, LTTMs are likely
to act as solvents as well as catalysts or co-catalysts, considering
their acid character.17
They can also be designed to be an enzyme-tolerant
melting point of the starting materials. Some representative DSC medium18 which allows the development of a one-pot process
curves can be found in Fig. 3 showing unusual low glass tran- where both deconstruction and enzymatic hydrolysis occur.
sition temperatures. Achieving good solubilities is a must for catalytic hydrolysis
Because for the selected mixtures no melting point was found, of cellulose or hemicellulose, while delignication and decrystal-
but most of the mixtures showed glass transitions instead, we lization are desirable for enzymes to perform better. In all cases,
named them LTTMs instead of DESs. the biorenewable and natural character of the solvent constituents
Biomass processing faces two main challenges for a better is of most importance.
exploitation of lignocellulosic biomass: the recalcitrant nature of For further steps in the process, the recoverability of the
lignocellulosic biopolymers, making them difcult to dissolve, solvent is considered. Solvents that are able to form hydrogen
and the efcient hydrolysis into sugars or high-valuable pro- bonding are very likely to phase separate when adding a non-
ducts. In both cases, the solvent plays a crucial role. In this hydrogen bonding solvent. As an example, acetone was proved
context, 26 new mixtures, listed in Table 2, are screened as sol- to work as an anti-solvent for solvent recoverability. The exper-
vents for lignin, cellulose and starch. imental procedure is described in ESI.

This journal is The Royal Society of Chemistry 2012 Green Chem., 2012, 14, 21532157 | 2155
 View Article Online

Table 3 Solubility values in wt% for lignin, starch and cellulose for Depending on the selected donor, the trend in solubilities can be
the selected LTTMs inverted. For instance, lignin solubility shows the trend LB > LH
Name Ttest (C) Lignin Starch Cellulose > LC > LG > LA > LP for the lactic acid sequence, while malic
or oxalic acid follows MP > MC > MG MH MN MB
LP2:1 60 7.56 0.00 0.00 and OC > OP > OB > OG OH ON respectively.
LB2:1 60 12.03 0.00 0.00 Even though cellulose solubility was found to be very poor or
LC1.3:1 60 4.55 0.00 0.00
LC2:1 60 5.38 0.00 0.00 negligible for most of the studied solvents, there was a noticeable
LC5:1 60 7.77 0.00 0.00 change in cellulose crystallinity in most cases, obtaining a turbid
LC10:1 60 11.82 0.13 0.00 liquid after stirring overnight. In the case of lactic acidcholine
LH9:1 60 11.88 0.13 0.00
LG9:1 60 8.77 0.00 0.00
chloride mixtures for example, the cellulose bers were not dis-
LA9:1 60 8.47 0.26 0.00 solved, but for the mixtures containing proline, only a new
MA1:1 100 1.75 0.59 0.11 turbid liquid phase was formed with no evidence of solid
MB1:1 100 0.00 0.81 0.00 particles.
Published on 20 June 2012 on http://pubs.rsc.org | doi:10.1039/C2GC35660K

MC1:1 100 3.40 7.10 0.00

MG1:1 100 1.46 7.65 0.14 Preliminary experiments were done to test the solubility of
MP1:1 100 0.00 0.00 0.00 real wheat straw biomass samples. The results are promising,
MP1:2 100 6.09 0.32 0.24 although the experiments are in an early stage. For the rst
Downloaded by University of Stellenbosch on 19 March 2013

MP1:3 100 14.90 5.90 0.78 experiment, wheat straw biomass was suspended in the LC2:1
MH2:1 85 0.00 0.00 0.00
MN9:1 85 0.00 0.00 0.00 LTTM. The LTTM coloured upon stirring the biomass suspen-
OB1:1 60 0.66 0.00 0.00 sion overnight, although biomass particles or bers were still
OP1:1 60 1.25 0.00 0.00 present (Fig. 4 and 5). After three washing cycles with ethanol,
OC1:1 60 3.62 2.50 0.00
OG3:1 85 0.28 0.00 0.00 90.5 wt% of the non-dissolved biomass could be recovered.
ON9:1 60 0.00 2.83 0.00
OH9:1 60 0.00 0.00 0.25
OCA1:1 60 0.00 0.15 0.00
OPA1:1 60 0.00 0.15 0.15

Cellulose and lignin are the two most abundant renewable

polymers in lignocellulosic biomass, and starch is chosen as the
representative polysaccharide for this study. For a selection of 26
solvents, listed in Table 3, the solubility of these biopolymers
was determined by using the cloud point method. Vials contain-
ing 2 g of solvent were placed in an oil bath. The selected temp-
erature (Ttest) was set constant for the whole experiment.
Consecutive additions of 0.21 mg of solute were done under
vigorous stirring to ensure good contact between phases. Once
the turbidity or the presence of particles was noticeable by the
cloud point method, the samples were equilibrated for at least
24 h to check that the turbidity had not disappeared. Zero solubi-
lity was considered when the addition of 0.1 wt% of solute
showed turbidity. The dissolution temperature was set as 60 C
for less viscous mixtures and as 100 C for the ones showing
higher viscosity. Fig. 4 Biomass processing with LC2:1 (top) and MP1:3 (bottom) at
different stages: (A) wheat straw raw biomass samples, (B) after pretreat-
From Table 3 it can be observed that a high selectivity for the
ment with LTTM overnight, (C) after centrifuging and (D) after washing
separation of lignin from a mixture of lignin and cellulose was
with ethanol and centrifuging.
found. Furthermore, very different solubility values were
obtained for the different combinations. Choline chloridelactic
acid mixtures show high solubility for lignin, while cellulose
was found to be immiscible with the whole series. Lignin solubi-
lity shows in this case a clear trend when increasing the acid
content (LC1.3:1 < LC1:1 < LC3:1 < LC5:1 < LC9:1).
However, the opposite trend was found when comparing with
the malic acidproline series (MP1:1 MP1:2 < MP1:3). The
malic acid combinations, in general, were found to show much
higher solubilities for starch and lower solubilities for lignin
when comparing with other hydrogen bond donors. But for the
former series, an increase in solubility of cellulose (also of lignin
and starch) was found when increasing the proline ratio. The role Fig. 5 Wheat straw biomass samples after pretreatment with and
of the hydrogen bond acceptor opens room to further discussion. LC2:1 at 60 C (I) MP1:3 at 85 C (II) overnight.

2156 | Green Chem., 2012, 14, 21532157 This journal is The Royal Society of Chemistry 2012

Direct separation of the suspended and dissolved biomass was
done by centrifuging the LC2:1 with biomass without washing.
Fig. 4 shows the processed material after each one of the
described steps.
After washing, 81.6% and 2.0 wt% of the added biomass was
recovered from the separated precipitate and supernatant respect-
ively, which means that 2 wt% of the biomass can be considered
as being dissolved and composed entirely of lignin.
A second experiment was done following the same procedure
for MP1:3. In this case, much less biomass particles can be
noticed after the pretreatment (Fig. 4 and 5). This result is con-
sistent with the higher values reported for the solubilities of
cellulose and starch. Lignin is soluble in higher extension, as
Published on 20 June 2012 on http://pubs.rsc.org | doi:10.1039/C2GC35660K
reected in the colour change of the solvent.
Conclusions
Downloaded by University of Stellenbosch on 19 March 2013
In this communication, we proved that novel renewable liquid
solvents can be prepared by mixing suitable hydrogen bond
donoracceptor combinations with high melting points. The
resultant liquid upon mixing can be a versatile solvent for mul-
tiple applications. Their physicochemical properties can be
tailored by choosing the nature and ratio of the constituents.
The novel LTTMs were tested as solvents for biopolymers
(lignin, cellulose and starch). Most of the selected combinations
show high lignin solubility and very poor or negligible cellulose
solubility. Therefore, the selectivities for the separation of lignin/
cellulose are very high.
Lactic acidcholine chloride mixtures showed no solubility of
cellulose but lignin solubility increased with the acid ratio. For
the series containing malic acidproline much higher solubilities
for starch and cellulose were found compared to the other pairs.
Moreover, for this former pair, solubilities were proved to
increase with the proline ratio.
Preliminary experiments of real biomass pretreatment were
consistent with the screening on solubilities of the different bio-
polymers. The solubility of the raw biomass in the selected
LTTMs is very low. However, it was possible to dissolve part of
the lignin, reected in the change in colour of the solvent.
Further experiments need to be carried out for optimization of
This journal is The Royal Society of Chemistry 2012
View Article Online
the pretreatment process. Deeper studies on hydrogen bond inter-
actions are needed to establish the building principles for the for-
mation of LTTMs. These principles would also allow the
improvement of predicting the physical properties and the
biomass solubilities.
Notes and references
1 A. P. Abbott, D. Boothby, G. Capper, D. L. Davies and R. K. Rasheed,
J. Am. Chem. Soc., 2004, 126, 9142.
2 A. P. Abbott, G. Capper, D. L. Davies, R. K. Rasheed and V. Tambyrajah,
Chem. Commun., 2003, 70.
3 Y. H. Choi, J. V. Spronsen, Y. Dai, M. Verberne, F. Hollmann, I. W. C.
E. Arends, G.-J. Witkamp and R. Verpoorte, Plant Physiol., 2011, 156,
1701.
4 M. A. Kareem, F. S. Mjalli, M. A. Hashim and I. M. AlNashef, J. Chem.
Eng. Data, 2010, 55, 4632.
5 A. Abbott, R. C. Harris, K. S. Ryder, C. DAgostino, L. F. Gladdenb and
M. D. Mantleb, Green Chem., 2011, 13, 82.
6 P. Domnguez de Mara and Z. Maugeri, Curr. Opin. Chem. Biol., 2011,
15, 220.
7 G. Saravanan and S. Mohan, J. Alloys Compd., 2012, 522, 162.
8 H.-G. Liao, Y.-X. Jiang, Z.-Y. Zhou, S.-P. Chen and S.-G. Sun, Angew.
Chem., Int. Ed., 2008, 47, 9100.
9 (a) M. A. Kareem, F. S. Mjalli, M. Ali Hashim and I. M. AlNashef, Fluid
Phase Equilib., 2012, 314, 52; (b) A. P. Abbott, G. Capper, D. L. Davies,
R. K. Rasheed and P. Shikotra, Inorg. Chem., 2005, 44, 6497; (c) X. Li,
M. Hou, B. Han, X. Wang and L. Zou, J. Chem. Eng. Data, 2008, 53,
548.
10 R. B. Leron and M.-H. Li, Thermochim. Acta, 2012, 530, 52.
11 N. Sun, H. Rodrguez, M. Rahman and R. D. Rogers, Chem. Commun.,
2011, 47, 1405.
12 R. Rinaldi, Chem. Commun., 2011, 47, 51513.
13 (a) R. C. Remsing, R. P. Swatloski, R. D. Rogers and G. Moyna, Chem.
Commun., 2006, 1271; (b) H. Wang, G. Gurau and R. Rogers, Chem.
Soc. Rev., 2012, 41, 1519.
14 (a) K. Ohira, Y. Abe, M. Kawatsura, K. Suzuki, M. Mizuno, Y. Amano
and T. Itoh, ChemSusChem, 2012, 5, 388; (b) Q.-P. Liu, X.-D. Hou, N. Li
and M.-H. Zong, Green Chem., 2012, 14, 304.
15 G. R. Desiraju and T. Steiner, The Weak Hydrogen Bond in Structural
Chemistry and Biology, Oxford University Press Inc., New York, 1999.
16 P. Gilli, L. Pretto, V. Bertolasi and G. Gilli, Acc. Chem. Res., 2009, 42,
33.
17 S. Hu, X. Zhang, Y. Zhou, J. Song, H. Fan and B. Han, Green Chem.,
2009, 11, 873.
18 J. Gorke, F. Srienc and R. Kazlauskas, Biotechnol. Bioprocess Eng.,
2010, 15, 4053.
Green Chem., 2012, 14, 21532157 | 2157