Dimethyl phosphorothioate and phosphoroselenoate ionic liquids as solvent media for cellulosic materials

Article abstract of DOI:10.1039/c1gc15407a

A series of novel ionic liquids comprising two asymmetric phosphate-derived anions, namely dimethyl phosphorothioate and dimethyl phosphoroselenoate, and several imidazolium and non-imidazolium-based cations was prepared via a facile synthetic route. Thermal degradation was studied by dynamic thermogravimetric analysis (TGA) revealing a slightly higher stability of the imidazolium ionic liquids and an overall low thermal stability for the phosphoroselenoate salts. Long-term moisture sorption analysis showed correlation with the polarity of the cation and differences in absorption and desorption kinetics. Finally, a Eucalyptus globulus kraft paper grade pulp was dissolved and subsequently regenerated to assess the degradation of the various molecular weight fractions by size exclusion chromatography. In addition, pre-extracted xylan was subjected to the same dissolution procedure to examine the degradation of low-molecular weight components in more detail.

Full text of DOI:10.1039/c1gc15407a

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Green Chemistry
Cite this: Green Chem., 2011, 13, 2507
www.rsc.org/greenchem
PAPER
Dimethyl phosphorothioate and phosphoroselenoate ionic liquids as solvent
media for cellulosic materials†
Michael Hummel,*a Carmen Froschauer,b,c Gerhard Laus,b Thomas Ro¨der,d Holger Kopacka,b
Lauri K. J. Hauru,a Hedda K. Weber,c Herbert Sixtaa and Herwig Schottenbergerb
Received 13th April 2011, Accepted 21st June 2011
DOI: 10.1039/c1gc15407a
A series of novel ionic liquids comprising two asymmetric phosphate-derived anions, namely
dimethyl phosphorothioate and dimethyl phosphoroselenoate, and several imidazolium and
non-imidazolium-based cations was prepared via a facile synthetic route. Thermal degradation
was studied by dynamic thermogravimetric analysis (TGA) revealing a slightly higher stability of
the imidazolium ionic liquids and an overall low thermal stability for the phosphoroselenoate
salts. Long-term moisture sorption analysis showed correlation with the polarity of the cation and
differences in absorption and desorption kinetics. Finally, a Eucalyptus globulus kraft paper grade
pulp was dissolved and subsequently regenerated to assess the degradation of the various
molecular weight fractions by size exclusion chromatography. In addition, pre-extracted xylan was
subjected to the same dissolution procedure to examine the degradation of low-molecular weight
components in more detail.
energy consumption, and the formation of hazardous by-
Introduction
products. Thus, the implementation of non-derivatising solvents
The worldwide increase of energy and material consumption by
should be promoted to pursue a more sustainable way of cellu-
rapidly industrialising nations and the concomitant depletion of
lose processing. Amongst numerous cellulose solvent systems
fossil carbon deposits urge scientists and politicians to develop
like Cuen (cupriethylene diamine),6 Cuoxam (cuprammonium
new strategies to economically exploit renewable biomass.1 This
hydroxide),6 Cadoxen (CdO/ethylenediamine),6 DMAc/LiCl,7
has highlighted the biorefinery of wood and its main constituents
cellulose, hemicelluloses and lignin as alternative sources for
DMSO/TBAF,8,9 phosphoric acid,10 and N-methylmorpholine
N-oxide (NMMO) monohydrate,11 only the latter could be com-
fuel, chemicals and end use products.2 In particular, cellulose,
mercialised so far in the so called Lyocell process.12,13 However,
as the earth’s most abundant biomacromolecule, is still not
this process has several drawbacks, most of them resulting from
used to its full potential.3 One reason, inter alia, is the recalci-
trance of cellulose towards dissolution in conventional organic
the particular chemical structure of NMMO.14 Especially the
incompatibility with redox-active substances and the intrinsic
solvents caused by the numerous intra- and intermolecular
instability of cyclic ethers, necessitating stabilisers to prevent
hydrogen bonds.4,5 Chemical derivatisation of the cellulosic
runaway reactions, initialised the search for alternative solvents.
hydroxyl groups is one possibility to disrupt the interchain
Even though they were discovered in the 1930s, the full poten-
hydrogen bridges and dissolve the modified biopolymer, but is
tial of low melting salts to dissolve cellulose was not recognised
accompanied by the utilisation of toxic chemicals, considerable
for decades.15 Only in 2002, Rogers and co-workers re-discovered
these salts when they introduced 1-butyl-3-methylimidazolium
aDepartment of Forest Products Technology, Aalto University,
Vuorimiehentie 1, 02150, Espoo, Finland.
chloride (BMIM Cl) as direct solvent for microcrystalline
cellulose (MCC).16 By then, intensive research was already
carried out in the field of salts with a melting point below
100 ◦C and the term ionic liquid (IL) had become common
for that type of compounds.
ILs have attracted attention in both academic and industrial
fields because of some unique properties such as exceptionally
low vapor pressure or high thermal stability. Although these
features make them attractive for scale-up processes, it is
doubtful whether the term green solvent is truly justified, as
has been pointed out many times in the past.
E-mail: michael.hummel@aalto.fi; Tel: +358 505124198
bInstitute of General, Inorganic and Theoretical Chemistry, Faculty of
Chemistry and Pharmacy, University of Innsbruck, Innrain 52a, 6020,
Innsbruck, Austria
cCompetence Centre of Wood Composites and Wood Chemistry K-Plus,
Linz, Austria
dLenzing AG, Department of Pulp Research, Lenzing, Austria
† Electronic supplementary information (ESI) available: Experimen-
tal details for O,Se-dimethylphosphoroselenoate compounds; TGA
graphs for dimethylphosphoroselenoate ILs; moisture sorption plots
for dimethylphosphorothioate ILs. See DOI: 10.1039/c1gc15407a
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2507–2517 | 2507
©

View Article Online
Contemporarily investigated ionic liquids comprise bulky
organic cations such as 1-alkyl-3-methylimidazolium, 1-
alkylpyridinium, 1-alkyl-1-methylpyrrolidinium, tetraalkylam-
monium, guanidinium or phosphonium ions.17,18 Due to
their size, reduced symmetry, and conformational flexibil-
ity the respective salts exhibit small lattice enthalpies and
large entropy changes upon liquefaction, thus favouring the
fluid state thermodynamically.19 In combination with weakly-
of potassium O,S-dimethylphosphorothioate on rats have
assessed the anion as non-toxic.46 Preliminary tests have
shown that ILs comprising O,S-dimethylphosphorothioate can
dissolve high molecular weight cellulose without extensive
degradation.
-
Results and discussion
coordinating anions like hexafluorophosphate (PF6 ), tetrafluo-
-
roborate (BF4 ), bis(trifluoromethanesulfonyl)amide (triflimide)
Synthesis of phosphate-derived ILs
or tris(perfluoroalkyl)trifluorophosphate (FAP),20 this effect is
leading to ionic liquids with very low melting points and low
viscosity at room temperature.21 However, these ILs do not dis-
solve cellulose as the anions are weak hydrogen bond acceptors.
High hydrogen bond basicity was claimed by several research
groups to be a prerequisite for ILs to dissolve cellulose.18,22–25
Thus, the first generation of cellulose-dissolving ionic liquids
was composed of chlorides and bromides.16,26,27 But as a matter
of fact, halide salts often exhibit high melting points, high
hygroscopicity, and corrosiveness towards metallic processing
equipment. To overcome these problems, carboxylate-based
ILs were proposed as alternative. Even though some 1-alkyl-
3-methylimidazolium carboxylates28 have low viscosity and
can dissolve cellulose in high concentrations, other problems
such as decarboxylation29 in the case of formates or acetates
induced side reactions30–33 evolved. As another disadvantage,
only the homo-substituted 1,3-dialkylimidazolium carboxylates
are directly accessible34 whereas those with asymmetric cation
have to be synthesised via an additional metathesis step. Ionic
liquids containing phosphates are the third group of molten
salts that exhibit good cellulose-dissolving properties and seem
to be promising candidates to overcome the aforementioned
problems. They offer high thermal and hydrolytic stability,35,36
adequate hydrogen bond basicity23,29,37 and were reported to be
efficient solvents for cellulose37–40 and wood.41 As stated above,
reduced symmetry of the ions leads to lower melting points
and viscosities, respectively. Applying this concept to anions,
Fukaya et al. introduced 1-ethyl-3-methylimidazolium methyl
phosphonate and methyl P-methylphosphonate as solvents for
microcrystalline cellulose (MCC).29,37,42
This principle of reducing the viscosity of a certain ionic
liquid while preserving the ability to dissolve lignocellulosic
material by altering the anion was also pursued in our re-
search group. Recently, we presented the synthesis of 1-butyl-
3-methylimidazolium ethyl methyl phosphate and preliminary
cellulose dissolution tests therein.38
In continuation of these studies, we will herein introduce
ionic liquids with O,S-dimethylphosphorothioate (dmpt) and
O,Se-dimethylphosphoroselenoate (dmpSe) as anions. By re-
placement of one methoxy group oxygen with a chalcogen
atom, the symmetry of the dialkylphosphate anion is re-
duced from C2v to C1, while the hydrogen bond basicity
should remain more or less unaltered, and thus cellulose
dissolution possible. To the best of our knowledge, these
anions have not been implemented into ionic liquids so far,
even though the syntheses of O,O,O-trimethylthiophosphate,
sodium O,S-dimethylphosphorothioate, and respective am-
monium salts were already described in the middle of
the twentieth century.43–45 Moreover, toxicological studies
Imidazolium-based ionic liquids are commonly prepared by
alkylation of the neutral 1-alkylimidazole with alkyl halides
(Menshutkin reaction), followed by a metathesis step where
the halide is substituted by a more desirable anion.3,21 The
respective anion is introduced in the form of the free acid
or a metal salt, causing a phase separation which allows the
removal of the new ionic liquid or the metal halide from the
residual solution, respectively. For some special anions, direct
introduction by alkylation using alkylesters of phosphoric,35,38,47
phosphorous,29 sulfuric,48 sulfurous,49 and carbonic acid50 or
Meerwein’s reagent51 has been reported. Among the aforemen-
tioned alkylesters, alkylphosphates are the most prominent ones.
Imidazolium dialkylphosphate salts were prepared using tri-
alkylphosphates either via direct alkylation of 1-alkylimidazole
or via reaction of a 1,3-dialkylimidazolium halide with the ester
and concomitant liberation of methyl halide.38,52
Related to those symmetric phosphate esters, O,O,O-
trimethylphosphorothioate is a viscous liquid prepared from
trimethylphosphite and sulfur.43,53 Upon addition of NaI,
methyl iodide is formed and a sequence of intermolecular
dealkylation/alkylation reactions affords the final sodium O,S-
dimethylphosphorothioate.44 An analogous procedure54 was
reported for sodium O,Se-dimethylphosphoroselenoate, the
crystal structure of which was recently published.55 As these
anions lack any symmetry they attracted our interest regarding
their implementation in ILs.
To test the influence of the cation on the dissolution properties
towards cellulose, we prepared a series of ILs comprising
seven imidazole- and three non-imidazole-based compounds,
namely 1-allyl-3-methylimidazolium (1, AMIM), 1-butyl-3-
methylimidazolium (2, BMIM), 1-methyl-3-propylimidazolium
(3, MPIM), 1-benzyl-3-methylimidazolium (4, BnMIM), 1-(2-
hydroxyethyl)-3-methylimidazolium (5, (HOEt)MIM), 1-ethyl-
3-methylimidazolium (6, EMIM), 1-allyl-3-butylimidazolium
(7, ABIM), 1-butyl-3-methylpyridinium (8, BMPy), 1-
methyl-1-propylpyrrolidinium (9, MPPyrr), and 4-ethyl-4-
methylmorpholinium (10, EMMorph) (Scheme 1).
The respective phosphorochalcogenoate salts were pref-
erentially prepared via the metathesis route (Scheme 2)
because sodium O,S-dimethylphosphorothioate and O,Se-
dimethylphosphoroselenoat are air- and shelf-stable solids and
therefore convenient reagents, and many 1,3-dialkylimidazolium
halide salts are meanwhile commercially available in high purity.
Except for the 1-methyl-1-propylpyrrolidinium salts, all
phosphorochalcogenoates were liquids or tended to re-
main in a supercooled state at room temperature. The 4-
methyl-4-ethylmorpholinium ILs, however, were highly viscous
pastes.
2508 | Green Chem., 2011, 13, 2507–2517
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
chloride. Varying the stoichiometry of the starting materials
affected the conversion only marginally.
Incomplete metathesis for ionic liquids has been observed
before and was ascribed to the reduced solubility of the
introduced metal salt in the solvent media.58
In analogy to the aforementioned alternative reaction of im-
idazolium halide and trialkylphosphate, attempts to synthesise
the desired products starting from the respective chloride salts
and O,O,O-trimethylphosphorothioate were made. Heating the
reactant mixture under reflux at different temperatures and
for different times did not afford a pure product. Instead,
new peaks evolved in the NMR spectra indicating gradual de-
composition of O,S-dimethylphosphorothioate. It was reported
earlier that the generated methylchloride reacts again with the
phosphorothioate.45 This was confirmed when the reaction was
carried out in a sealed vessel where the methyl chloride could
not escape from the reaction mixture. Thereafter, no more O,S-
dimethylphosphorothioate was detected in the NMR spectra.
Finally, pure O,S-dimethylphosphorothioate-ILs could be
prepared by promoting the reaction via ultrasonication in an
open reaction vessel. The optimisation of the reaction conditions
was tedious, though, as only exact bath temperature and a nar-
row reaction time window led to a product without impurities.
Besides, O,O,O-trimethylphosphorothioate is a viscous liquid
with a noticeable odour, making it unpleasant to handle.
Halide impurities have been reported to impair the prop-
erties of ionic liquids making them less suitable for cer-
tain applications.59,60 For cellulose dissolution, however, small
quantities of chloride should not have an adverse effect on
the dissolving properties of ionic liquids. Since the exact
composition of the O,S-dimethylphosphorothioate- and O,Se-
dimethylphosphoroselenoate-ILs could be determined easily via
NMR spectroscopy, the halide-containing ILs (chloride content:
1–2 % w/w) were subsequently studied regarding their suitability
as cellulose solvent media.
Scheme 1 Structure of constituent ions of the ionic liquids investigated
herein.
Scheme 2 Preparation of O,S-dimethylphosphorothioate-ILs via anion
metathesis.
1-Ethyl-3-methylimidazolium O,S-dimethylphosphorothio-
ate crystallised after a few days and could therefore be subjected
to single crystal X-ray analysis. The respective ORTEP plot is
depicted in Fig. 1 (crystal data was deposited at the CCDC).56
NMR analysis of the metathesis products, however, showed
a reduced peak intensity for the two methyl groups of the
anion. This could be either attributed to the degradation of
the anion or to incomplete conversion. The latter was confirmed
by ion chromatography which revealed the presence of chloride.
According to the phosphate concentration determined via NMR
integral values and residual amounts of chloride detected by
ion chromatography (IC), the conversion was between 80–85%
(see Experimental section), even after intensifying the reaction
conditions (ultrasonication or refluxing in acetone for several
hours). To exclude the possibility that chloride arises from the
dissolution of sodium chloride in the respective ionic liquid,
all products were subjected to ICP-AES. It was reported that
even dry, hydrophobic ILs can dissolve considerable amounts of
NaCl.57 In our case, virtually no sodium was detected (average
sodium content below 350 ppm/0.035% w/w) confirming that
imperfect anion exchange led to the detected amounts of
Thermal stability
It was shown previously that the thermal stability of
imidazolium-based ionic liquids decreases as the hydrophilicity
of the anion increases.61 Retro-alkylation was proposed as a
possible explanation.62,63 In the process, the anion attacks one
of the side chains in a SN2 reaction. This results in lower
decomposition temperatures when compared to hydrophobic
-
ILs incorporating weakly-coordinating anions like PF6 or
-
N(Tf)2 which decompose at temperatures between 350 and
450 ◦C.
As depicted in Fig. 2a, the imidazolium O,S-dimethyl-
phosphorothioate ILs decompose at around 200–230 ◦C and
are thus slightly less stable than EMIM dimethylphos◦phate
◦
(Td = 289 C),29 MMIM dimethylphosphate (Td = 268 C),40
◦
and EMIM diethylphosphate (Td = 264 C).40 The decrease in
stability might be attributed to the presence of chloride traces.
Nevertheless, th◦ese ILs are in the same stability range as EMIM
OAc (Td = 210 C)64 which is currently intensively investigated
and successfully implemented in both cellulose processing65,66
and wood fractionation.67,68
The non-imidazolium based phosphorothioates showed
degradation at marginally lower temperatures (Fig. 2b). This
Fig. 1 ORTEP-plot of 1-ethyl-3-methylimidazolium O,S-dimethyl-
phosphorothioate (6a).56
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2507–2517 | 2509
©

View Article Online
As expected, the O,Se-dimethylphosphoroselenoate ILs are
less stable than their thio-analogues. The imidazolium-based
ones decompose at 190–200 ◦C (Table 1). This low thermal
stability in combination with the possible release of selenium-
containing degradation products disqualifies this type of ILs
for scale-up applications. Therefore, in this study only the O,S-
dimethylphosphorothioates were further pursued in terms of
biomass processing.
Nevertheless, the O,Se-dimethylphosphoroselenoates bear
potential to probe cellulose-solvent interactions by means of
77Se-NMR.75
Moisture sorption
The dissolving power of ionic liquids is sensitive to moisture.
Small amounts of water impair the dissolution of biopolymers
and provoke their precipitation, respectively. Hence, the water
content of the solvent system has to be known and controlled
throughout the whole process line.
Various techniques can be found in the literature to determine
the moisture content of ILs and study the time-dependent water
uptake, respectively. A near-infrared spectrometry study has
shown that the kinetics of water absorption of ILs exposed to
humidity cannot be described by simple models because physical
properties, namely polarity, viscosity, and surface tension of ILs
are dependent on the concentration of gradually incorporated
water.76
Vitz et al. presented a more practical approach to water sorp-
tion analysis.40,77 In a so-called dynamic vapor sorption (DVS)
measurement, a sample is subjected to an atmosphere of varying
humidity, and the sample response measured gravimetrically. We
have used this technique to characterise the dmpt-ILs in terms
of moisture uptake. By means of a high-throughput analyser,
all samples could be measured simultaneously, providing good
comparability of the results. Thus, the tested ILs were kept at
25 ◦C and exposed to controlled humidity which was increased
stepwise until a relative humidity (r.H.) of 90% was reached
and then decreased correspondingly. The weight increase was
recorded as a function of time at a constant humidity level for
at least 120 min and no longer than 48 h or until each sample
had reached the equilibrium state, i.e. the weight change was less
than 0.01% after 30 min.
Fig. 3 presents the moisture uptake in terms of relative
weight gain as a function of time for BMIM dmpt (2a). The
graph shows that gradually absorbed water accelerates further
moisture uptake. This might be attributed to a decrease of
viscosity and hence an increase of diffusion mobility as stated
above.76 However, during the measurement interval at 70% r.H.,
a sudden rise in water absorption could be observed, as seen by
a bend in the curve after approximately 260 h. This is when the
sample shows dissemination due to a concomitant increase of
Fig. 2 Thermogravimetric analysis (TGA) of imidazolium (top) and
non-imidazolium dmpt-ILs.
coincides with other studies on homologous series of a certain
anion combined with various cationic structures.69–71 Each non-
imidazolium cationic moiety investigated herein has intrinsic
features which can destabilise the structure upon thermal
impact. Pyridinium seems to be slightly more sensitive to
retroalkylation,71,72 while pyrrolidinium is reported to show
pronounced Hofmann elimination.73 The ether bond in the
morpholinium ring is responsible for the lower decomposition
temperature, although an ether group in one of the alkyl
substituents at the nitrogen atom would have a bigger impact
on the thermal stability.69,74
For the sake of comparison, the onset temperatures are
summarised in Table 1. However, it should be noted that these
values do not reflect precisely the degradation behaviour of
the ionic liquids. Weight loss might be observed already at
lower temperatures as depicted by the TGA graphs (Fig. 2) and
could be detected even more sensitively via a recently reported
potentiometric titration approach.63
Table 1 TGA-onset temperaturesa of O,S-dimethylphosphorothioate (dmpt) and O,Se-dimethylphosphoroselenoate (dmpSe) ionic liquids
AMIM
BMIM
MPIM
BnMIM
(HOEt)MIM
ABIM
BMPy
MPPyrr
EMMorph
dmpt
dmpSe
232
192
230
208
227
196
220
195
227
195
228
204
218
215
226
201
208
191
a Of decomposition (in ◦C).
2510 | Green Chem., 2011, 13, 2507–2517
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Fig. 3 Moisture absorption and desorption of BMIM dmpt (2a). Left: weight change as a function of time upon stepwise increase and decrease of
relative humidity (r.H.); right: sorption isotherm, the ¥ indicates that the system did not reach the equilibrium state within the measurement interval.
sample volume and surface tension during water uptake.78,79 The
accompanied increase in effective surface facilitates subsequent
water absorption markedly. During the last step, at 90% r.H.,
the moisture content is then almost doubled although the
finally measured value would have been even higher, but
the system could not fully equilibrate within the interval of
48 h.
Upon subsequent decrease of humidity, the water release is
slightly slower than the uptake, resulting in a hysteresis when
the data is plotted as a sorption isotherm (Fig. 3; ESI, Fig.
S2–S9†). This is in contrast to the findings of Vitz et al., where
they observed equal absorption and desorption behaviour. The
discrepancy might originate from differences in the final r.H.
level and measurement interval times.
All absorption isotherms of the imidazolium-based ILs are
summarised in Fig. 4 where the relative amount of absorbed
water is represented as molecules of water per ion pair of ionic
liquid. The comparison demonstrates that the water uptake
decreases with longer and bulkier alkyl/aryl-residues. ABIM
dmpt (7a) shows the lowest water affinity. This is, again, in
accordance with the findings of Vitz et al. for imidazolium
chlorides77 and corresponds to the trend seen in the literature
for viscosity and melting points.
At the first sight it might be astonishing that the water
uptake of (HOEt)MIM dmpt (5a) is slightly lower than for the
alkylmethylimidazolium ILs. However, this can be attributed to
a hydroxyl group promoted cation-anion interaction leading to
a reduction of the IL’s polarity.80
Fig. 4 Absorption isotherms of imidazolium dmpt ILs (for clarity the
desorption data was omitted). Note: at 90% r.H. none of the depicted
ILs reached the equilibrium state within the measurement interval, hence
the depicted data points do not represent the equilibrated values.
Dissolution and regeneration of cellulose
There are plenty of reports in the literature describing the dis-
solution of cellulosic material in ionic liquids. But only recently
the respective research was systemised to such an extent that
the published data became somehow comparable. For example,
the information about purity and moisture content81 of the ionic
liquid was insufficient precluding repeatability of the dissolution
experiments in terms of maximum solute concentration or
degree of biopolymer degradation. Besides, microcrystalline
The final water content values of all ILs at 90% r.H. are
summarised in Table 2.
Table 2 Water uptake of O,S-dimethylphosphorothioate-ILs at 90% r.H.
Ionic liquid
wu (%)a
nH O/nIL
Ionic liquid
wu (%)a
nH O/nIL
2
2
AMIM (1a)
145.6
133.7
151.6
86.0
21.4
20.8
22.4
15.0
19.6
ABIM (7a)
77.3
124.8
155.1
148.1
13.1
20.2
23.2
22.3
BMIM (2a)
BMPy (8a)
MPIM (3a)
BnMIM (4a)
(HOEt)MIM (5a)
MPPyrr (9a)
EMMorph (10a)
131.7
a Water uptake.
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2507–2517 | 2511
©

View Article Online
cellulose (MCC) with a low degree of polymerisation (DP) is
often used as model solute for dissolution trials. But when
aiming for products with high structural performance upon
regeneration, such as fibres and films, one has to deal with
high molecular weight cellulose like cotton linters or wood pulp.
Results obtained with MCC studies do not necessarily hold
true when dissolving high DP material. Dissolution kinetics and
maximum solute load can differ markedly.
To process cellulose, knowledge about the rheological char-
acteristics of the solution (dope) is a prerequisite.66 The visco-
elastic properties of the dope, however, cannot be assigned to a
single value such as the average DP only. Instead, the molecular
weight distribution (MWD) of the solute has to be determined.
Therefore, degradation of the solute during dissolution has to be
assessed by subsequent regeneration and MWD measurement82
of the precipitate to investigate possible unequal degradation of
low and high molecular weight fractions.
Fig. 5 Molecular weight (MW) distribution of Ence kraft pulp
before dissolution and after regeneration, respectively, in the indicated
ionic liquid. The inset shows the Gaussian functions simulating the
hemicelluloses and cellulose parts of the pulp regenerated from MPIM
dmpt (3a).
We have chosen a Eucalyptus globulus kraft paper-grade pulp
from Ence, Spain, to test the dissolution properties of the O,S-
dimethylphosphorothioate-containing ILs.
Since 1-methyl-1-propylpyrrolidinium and 1-ethyl-3-methyl-
imidazolium dmpt were solids and 4-methyl-4-ethylmor-
pholinium dmpt a highly viscous paste at room temperature,
these three ILs were excluded from the following cellulose
dissolving experiments.
The dissolution studies were conducted in sealed vessels under
a dry nitrogen atmosphere. The respective water free ILs were
pre-heated to 100 ◦C and 0.5 % w/w pulp was added.
Especially AMIM Cl and BMIM Cl required propyl gallate
as stabiliser to avoid excessive degradation of the pulp.82 Thus,
dissolution experiments with chloride-free AMIM and BMIM
dmpt, prepared via the alternative synthesis, were conducted
to extract the influence of the residual chloride. However,
the resulting MWD-curves did not deviate from the previous
ones.
To examine the degradation in the hemicellulose region quan-
titatively, the hemicellulose and cellulose part were separated
by fitting the experimental data with a linear combination
of several Gaussian functions. The inset in Fig. 5 shows the
calculated molecular weight distribution functions in MPIM
dmpt (3a) representing the hemicellulose and cellulose fraction,
respectively. The functions permitted to assess the degradation
of the hemicelluloses in terms of average molecular mass (Table
3). AMIM dmpt (1a), BMIM dmpt (2a), and (HOEt)MIM
dmpt (5a) caused severe degradation leading to reduced number
average molecular weight and higher polydispersity whereas the
regenerated pulp from MPIM dmpt (3a), BnMIM dmpt (4a),
and BMPy dmpt (8a) was de facto unaltered.
To investigate the effect of dmpt-ILs on hemicelluloses in more
detail, additional dissolution studies with pure xylan isolated via
cold caustic extraction from the Ence kraft pulp were performed
under the same conditions used for the initial Ence pulp. The
low DP of the solute allowed the determination of the saturation
point because the viscosity rise upon concentration increase was
less steep than for the pulp. Thus, homogenisation via magnetic
stirring was possible throughout the whole concentration range.
The dissolved amounts of xylan are summarised in Table 3.
BMPy dmpt (8a) could dissolve 7 % w/w of xylan within 2 h,
BnMIM dmpt (4a) only 1 % w/w. Despite different dissolution
capacities, all tested ILs degraded the solute to the same extent
which differs from the findings with the Ence pulp.
The solute load was kept low as the viscosity rises tremen-
dously with increased concentration, preventing efficient stirring
with the magnetic stir bar. Even at such low concentration,
the viscosity increased noticeable and in some cases complete
dissolution could not be achieved within 2 h. ABIM dmpt (7a)
did not dissolve the pulp at all. The other ILs, however, dissolved
the solute typically within 1–4 h. Subsequently, the material
was regenerated and washed with methanol to facilitate the
drying of the resulting fibrous material. Control experiments
using water as anti-solvent showed no differences in terms of
molecular weight distribution of the isolated material. The mass
balance showed complete regeneration of the cellulosic material.
Thus, the ionic liquid could be reused after evaporation of the
anti-solvent. The initial and regenerated pulps were subjected
to GPC-analysis to determine possible degradation of the single
weight fractions (Fig. 5).
The MWD-curve of the dissolving pulp in Fig. 5 shows a
bimodal distribution. The shoulder in the high molecular weight
region is possibly an artifact which might result from aggregate
formation owing to non-monodisperse solution state and was
therefore not accounted for in the interpretation of the results.
In principle, the high DP cellulose fractions were not degraded
during dissolution in the ILs. Only BMIM dmpt (2a) showed
a significant cleavage of the pulp’s glycosidic bonds leading
to a displacement of the whole curve towards the left. The
lower molecular weight parts (hemicelluloses-region), however,
are substantially more affected by most of the ILs. The curves
of AMIM dmpt (1a), BnMIM dmpt (4a), and (HOEt)MIM
dmpt (5a) showed the most pronounced deviation from the
original pulp-MWD. In an earlier study we observed similar
degradation when a series of chloride-based ILs was tested.
Nevertheless, the dmpt-ILs seem suitable to dissolve high DP
material without extensive degradation and thus turn them into
candidates for cellulose processing like continuous extrusion or
stationary film casting.
2512 | Green Chem., 2011, 13, 2507–2517
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Table 3 Average molecular masses of hemicelluloses in the original and regenerated pulp/xylan. Numbers in parentheses represent the percental
values
Hemicelluloses in Ence pulp
Mna/kg mol-1 Mwb/kg mol-1 Mz /kg mol-1 PDId
Extracted xylan
cmax (%) Mna/kg mol-1 Mwb/kg mol-1 Mz /kg mol-1 PDId
c
e
c
Original
28.5
23.9
19.3
27.9
29.8
38.2
33.8
27.7
37.8
48.4
45.4 (94)
37.1 (77)
1.34
1.41 (105) 4
1.44 (107) 3
9
18.6
17.2
16.8
18.0
n/d
18.2
17.3
39.5
44.3
44.2
48.1
n/d
46.7
40.8
2.06
AMIM (1a)
BMIM (2a)
MPIM (3a)
BnMIM (4a)
(HOEt)MIM (5a) 28.3
BMPy (8a) 26.9
(84)
(68)
(98)
(89)
(73)
(99)
5.7
5.6
6.1
n/d
5.7
5.9
(63)
(62)
(68)
(93)
(90)
(97)
(112) 3.02 (146)
(112) 3.00 (145)
(122) 2.96 (143)
n/d
(118) 3.19 (155)
(103) 2.91 (141)
48.8 (101) 1.36 (101) 4
(104) 39.8
(104) 51.1 (106) 1.34 (100) 1
(104) 52.6 (109) 1.4 (104) 2
(99)
(95)
39.6
37.3
(63)
(66)
(98)
(93)
(98)
50.3 (104) 1.38 (103) 7
a Number-average, b weight-average, c and z-average molecular mass. d Polydispersity index. e Maximum solute concentration in terms of weight.
˚
radiation, l = 0.71073 A). The structure was solved using direct
methods and refined by full-matrix least-squares techniques.
Chloride was determined using a Dionex ICS-1500 ion
Experimental
Materials
R
chromatography system with an Ionpackꢀ AS9-HC analytical
Quaternised N-compounds were purchased as chlorides
or methylcarbonates from IoLiTec (Germany) and used
as received. O,O,O-Trimethylphosphorothioate and O,O,O-
trimethylphosphoroselenoate and their sodium salts were pre-
pared according to literature procedures.43,45,53,54 Eucalyptus
globulus kraft pulp from Ence, Spain (817 ml g-1; R10 = 87.93;
R18 = 91.54) and cold caustic extracted xylan (81.0% xylan; 4.6%
glucan) were donated from Wood K Plus, Austria. All other
chemicals were purchased from Sigma-Aldrich and VWR Int.
and used without further purification.
column.
Sodium content was determined with a Horiba Jobin Yvon
ACTIVA spectrometer.
General synthetic procedure
Sodium O,S-dimethylphosphorothioate [32586-90-6] (ca. 20 g)
and an equimolar amount of the respective cation as its chloride
salt were placed in a round bottom flask and suspended in
250 ml of acetonitrile by stirring at room temperature for
15 min. Afterwards, the suspension was treated with ultrasound
for 1–2 h and stirred again at room temperature for 24–48 h.
Upon stirring, sodium chloride starts to precipitate, which
was then filtered by means of a sintered glass frit with a Celite
layer. The solid matter was washed 2 times with 15 ml of
acetonitrile. The solvent was stripped off the filtrate by means of
a rotary evaporator and the residue finally dried in high vacuum
(rotary vane oil pump) at 80 ◦C. Karl Fischer titration showed a
residual water content of 0.2% (w/w) or lower. The condensed
acetonitrile could be reused. The final products were stored
under an atmosphere of nitrogen to avoid moisture uptake from
the air.
Instrumentation
TGA was carried out with a TGA7 system (Perkin-Elmer,
Norwalk, CT, USA) using 2 to 10 mg of sample in a platinum
pan (50 ml). Dry nitrogen was used as a purge gas (sample purge:
20 mL min-1, balance purge: 40 mL min-1) and the general heat-
ing rate was 5 K min-1. The weight calibration was performed
with a 100-mg Class M calibration standard. The temperature
signal was calibrated with ferromagnetic materials (Alumel and
Ni, Curie point standards, Perkin-Elmer).
NMR spectra were recorded using a Bruker Avance DPX
300 spectrometer.
1-Allyl-3-methylimidazolium O,S-dimethylphosphorothioate
IR spectra were obtained with a Nicolet 5700 FT spectrometer
in ATR mode.
1a. 1-Allyl-3-methylimidazolium
chloride
[65039-10-3]
(40.00 g, 252.2 mmol), sodium O,S-dimethylphosphorothioate
(41.38 g, 252.2 mmol). Yield: 57.76 g, nD20: 1.5307, density:
1.212 g ml-1.
n
max/cm-1: 2926, 2833, 1645, 1564, 1426, 1313, 1239, 1169,
Density measurements were conducted at 20 ◦C using a 5 ml
pycnometer (according to Gay-Lussac).
Refractive index was determined with an Abbe refractometer.
Sorption analyses were performed on a Sorption Analyzer
SPS11-10. Resolution: 10 g/100 g (dual range); temperature
range: +10 ◦C to +60 ◦C; humidity range: 0% r.H. to 98% r.H.;
gas supply: compressed air/nitrogen, dried, oil-free; validation:
DKD Calibration Certificate with “min weight” certification;
salt validation test with humidity standards.
Molecular mass distribution of cellulose and hemi-cellulose
samples was determined by gel permeation chromatography
(size exclusion chromatography) using DMAc/LiCl as eluent.
The method was described in detail earlier.83
1069, 1044, 943, 739, 675, 627, 558.
1H NMR (300 MHz, DMSO-d6): d 1.93 (2.3H, d, 11.0 Hz,
P(SCH3)), 3.31 (2.4H, d, J = 12.2 Hz, P(OCH3)), 3.88 (3H, s,
NCH3), 4.88 (2H, d, J = 5.8 Hz, NCH2CH CH2), 5.28 (2H,
m, NCH2CH CH2), 6.03 (1H, m, NCH2CH CH2), 7.82 (1H,
s, NCHCHN), 7.84 (1H, s, NCHCHN), 9.51 (1H, s, NCHN)
ppm.
13C NMR (75 MHz, DMSO-d6):
d
12.9 (d, J =
3.6 Hz, P(SCH3)), 35.7 (NCH3), 50.7 (NCH2CH CH2), 51.4
(d, J = 5.8 Hz, P(OCH3)), 120.1 (NCH2CH CH2), 122.3
(NCH2CH2CH3), 123.8 (NCHCHN), 131.9 (NCH2CH CH2),
137.0 (NCHN) ppm.
Crystal structure analyses were performed on a Stoe IPDS-II
image plate diffractometer (graphite-monochromated Mo-Ka
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2507–2517 | 2513
©

View Article Online
1-Butyl-3-methylimidazolium O,S-dimethylphosphorothioate
2a. 1-Butyl-3-methylimiodazolium chloride [79917-90-1]
(17.56 g, 89.2 mmol), sodium O,S-dimethylphosphorothioate
(18.59 g, 89.2 mmol). Yield: 26.83 g; colourless oil. nD20: 1.5098;
density: 1.162 g ml-1.
n
max/cm-1: 3145, 3094, 2927, 2837, 1567, 1441, 1339, 1313,
1226, 1167, 1063, 1041, 871, 747, 652, 623, 562.
1H NMR (300 MHz, DMSO-d6): d 1.93 (2.5H, d, J = 11.3 Hz,
P(SCH3)), 3.33 (2.6H, d, J = 12.3 Hz, P(OCH3)), 3.68 (2H, t,
J = 4.5 Hz, NCH2CH2OH), 3.86 (3H, s, NCH3), 4.23 (2H, t, J =
4.5 Hz, NCH2CH2OH), 7.73 (1H, s, NCHCHN), 7.77 (1H, s,
NCHCHN), 9.28 (1H, s, NCHN) ppm.
n
max/cm-1: 3054, 2958, 2931, 2872, 1568, 1463, 1381, 1337,
1313, 1237, 1172, 1070, 1045, 740, 655, 626, 560.
1H NMR (300 MHz, DMSO-d6): d 0.84 (3H, t, J =
7.4 Hz, (CH2)3CH3), 1.21 (2H, m, CH2CH2CH2CH3), 1.74 (2H,
m, CH2CH2CH2CH3), 1.92 (2.3H, d, J = 11.2 Hz, P(SCH3)),
3.31 (2.2H, d, J = 12.1 Hz, P(OCH3)), 3.87 (3H, s, NCH3), 4.19
(2H, t, J = 7.2 Hz, NCH2-Pr), 7.85 (1H, s, NCHCHN), 7.94
(1H, s, NCHCHN), 9.64 (1H, s, NCHN) ppm.
13C NMR (75 MHz, DMSO-d6): d 12.7 (d, J = 3.7 Hz,
P(SCH3)), 35.6 (NCH3), 51.6 (m, NCH2CH2OH and P(OCH3)),
59.4 (NCH2CH2OH), 122.7 (NCHCHN), 123.3 (NCHCHN),
137.0 (NCHN) ppm.
1-Ethyl-3-methylimidazolium O,S-dimethylphosphorothioate
6a. 1-Ethyl-3-methylimiodazolium chloride [65039-09-0] (1.13
g, 7.71 mmol), sodium O,S-dimethylphosphorothioate (1.27 g,
7.71 mmol). Yield: 1.71 g; mp: 39 ◦C.
13C NMR (75 MHz, DMSO-d6): d 12.8 (d, J = 3.7 Hz,
P(SCH3)), 13.3 ((CH2)3CH3), 18.8 (CH2CH2CH2CH3), 31.5
(CH2CH2CH2CH3), 35.6 (NCH3), 48.4 (NCH2-Pr), 51.4 (d, J =
6.5Hz, P(OCH3)), 122.3(NCHCHN), 123.6(NCHCHN), 137.0
(NCHN) ppm.
n
max/cm-1: 3061, 2974, 2928, 2829, 1578, 1568, 1467, 1430,
1342, 1234, 1172, 1068, 1040, 965, 840, 737, 710, 623, 558, 485,
439.
1H NMR (300 MHz, DMSO-d6): d 1,40 (3H, t, J = 7.3 Hz,
NCH2CH3), 1.93 (2.6H, d, J = 11.1 Hz, P(SCH3)), 3.32 (2.6H,
d, J = 12.1 Hz, P(OCH3)), 3.87 (3H, s, NCH3), 4.22 (2H, t,
J = 7.3 Hz, NCH2CH3), 7.80 (1H, s, NCHCHN), 7.90 (1H, s,
NCHCHN), 9.53 (1H, s, NCHN) ppm.
1-Methyl-3-propylimidazolium O,S-dimethylphosphorothioate
3a. 1-Methyl-3-propylimidazolium chloride [79917-89-8]
(20.00 g, 124.5 mmol), sodium O,S-dimethylphosphorothioate
(20.43 g, 124.5 mmol). Yield: 27.78 g, nD20: 1.5175, density:
1.188 g ml-1.
13C NMR (75 MHz, DMSO-d6): d 12.8 (d, J = 3.5 Hz,
P(SCH3)), 15.1 (NCH2CH3), 35.5 (NCH3), 44.0 (NCH2CH3),
51.4 (d, J = 6.4 Hz, P(OCH3)), 122.0 (NCHCHN), 123.5
(NCHCHN), 136.7 (NCHN) ppm.
n
max/cm-1: 3058, 2964, 2936, 2876, 1569, 1462, 1386, 1313,
1238, 1173, 1069, 1044, 741, 654, 625, 560.
1H NMR (300 MHz, DMSO-d6): d 0.80 (3H, t, J = 7.4 Hz,
NCH2CH2CH3), 1.78 (2H, m, NCH2CH2CH3), 1.93 (2.2H, d,
J = 11.3 Hz, P(SCH3)), 3.33 (2.3H, d, J = 12.2 Hz, P(OCH3)),
3.87 (3H, s, NCH3), 4.15 (2H, t, J = 7.1 Hz, NCH2CH2CH3),
7.82 (1H, s, NCHCHN), 7.90 (1H, s, NCHCHN), 9.56 (1H, s,
NCHN) ppm.
1-Allyl-3-butylimidazolium O,S-dimethylphosphorothioate 7a.
1-Allyl-3-butylimidazolium chloride [887276-30-4] (7.40 g,
36.8 mmol) and sodium O,S-dimethylphosphorothioate
(6.05 g, 36.8 mmol). Yield: 10.17 g, nD20: 1.5169, density: 1.132
g ml-1.
13C NMR (75 MHz, DMSO-d6): d 10.4 (NCH2CH2CH3),
12.8 (d, J = 3.7 Hz, P(SCH3)), 22.9 (NCH2CH2CH3), 35.6
(NCH3), 50.1 (NCH2CH2CH3), 51.4 (d, J = 6.5 Hz, P(OCH3)),
122.3(NCHCHN), 123.6 (NCHCHN), 137.0 (NCHN) ppm.
n
max/cm-1: 3054, 2958, 2928, 2873, 1645, 1562, 1463, 1313,
1241, 1168, 1070, 1047, 941, 739, 644, 559.
1H NMR (300 MHz, DMSO-d6): d 0.86 (3H, t, J = 7.4 Hz,
(CH2)3CH3), 1.23 (2H, m, CH2CH2CH2CH3), 1.76 (2H, m,
CH2CH2CH2CH3), 1.93 (2.2H, d, J = 11.2 Hz, P(SCH3)), 3.31
(2.2H, d, J = 12.2 Hz, P(OCH3)), 4.21 (2H, t, J = 7.2 Hz,
NCH2-Pr), 4.89 (2H, d, 5.9 Hz, NCH2CH CH2), 5.29 (2H,
m, NCH2CH CH2), 6.06 (1H, m, NCH2CH CH2), 7.84 (1H,
s, NCHCHN), 7.94 (1H, s, NCHCHN), 9.63 (1H, s, NCHN)
ppm.
1-Benzyl-3-methylimidazolium O,S-dimethylphosphorothioate
4a. 1-Benzyl-3-methylimidazolium chloride [36443-80-8]
(20.00 g, 95.84 mmol), sodium O,S-dimethylphosphorothioate
(15.73 g, 95.84 mmol). Yield: 27.35 g; nD20: 1.5685; density:
1.244 g ml-1.
n
max/cm-1: 2926, 2832, 1561, 1497, 1455, 1313, 1236, 1163,
1069, 1044, 742, 719, 699, 663, 626, 559.
13C NMR (75 MHz, DMSO-d6):
d
12.9 (d, J =
1H NMR (300 MHz, DMSO-d6): d 1.95 (2.3H, d, J = 11.1 Hz,
P(SCH3)), 3.34 (2.4H, d, J = 12.1 Hz, P(OCH3)), 3.86 (3H, s,
NCH3), 5.49 (2H, s, NCH2Ph), 7.36 (3H, m, CHo,p), 7.50 (2H,
m, CHm), 7.82 (1H, s, NCHCHN), 7.95 (1H, s, NCHCHN),
9.71 (1H, s, NCHN) ppm.
3.7 Hz, P(SCH3)), 13.3 ((CH2)3CH3), 18.8 (CH2CH2CH2CH3),
31.3 (CH2CH2CH2CH3), 48.5 (CH2CH2CH2CH3), 50.7
(NCH2CH CH2), 51.4 (d, J = 6.4 Hz, P(OCH3)), 120.0
(NCH2CH CH2), 122.5 (NCH2CH2CH3), 122.6 (NCHCHN),
131.9 (NCH2CH CH2), 136.5 (NCHN) ppm.
13C NMR (75 MHz, DMSO-d6): d 12.9 (d, J = 3.6 Hz,
P(SCH3)), 35.8 (NCH3), 51.4 (d, J = 23.8 Hz, P(OCH3)), 51.8
(NCH2Ph), 122.3 (NCHCHN), 124.0 (NCHCHN), 128.4 (Cm),
129.0 (Co,p), 135.0 (Ci), 136.9 (NCHN) ppm.
1-Butyl-3-methylpyridinium O,S-dimethylphosphorothioate
8a. 1-Butyl-3-methylpyridinium chloride [125652-55-3] (19.12
g, 103.0 mmol), sodium O,S-dimethylphosphorothioate (16.90
g, 103.0 mmol). Yield: 26.45 g; dark oil. nD20: 1.5370; density:
1.146 g ml-1.
1-(2-Hydroxyethyl)-3-methylimidazolium O,S-dimethylphos-
phorothioate
5a. 1-(2-Hydroxyethyl)-3-methylimidazolium
n
max/cm-1: 2958, 2929, 2878, 1633, 1505, 1465, 1312, 1242,
chloride [61755-34-8] (20.00 g, 123.0 mmol), sodium O,S-
dimethylphosphorothioate (20.18 g, 123.0 mmol). Yield: 28.30
g, nD20: 1.5291, density: 1.276 g ml-1.
1158, 1071, 1046, 960, 817, 738, 691, 560.
1H NMR (300 MHz, DMSO-d6): d 0.88 (3H, t, J = 7.4 Hz,
(CH2)3CH3), 1.27 (2H, m, CH2CH2CH2CH3), 1.86 (2H, m,
2514 | Green Chem., 2011, 13, 2507–2517
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
CH2CH2CH2CH3), 1.93 (2.2H, d, J = 11.1 Hz, P(SCH3)), 2.50
(3H, s, CCH3), 3.31 (2.2H, d, J = 12.2 Hz, P(OCH3)), 4.63 (2H,
t, J = 7.9 Hz, NCH2-Pr), 8.06 (1H, t, J = 7.0 Hz, NCHCHCHC),
8.47 (1H, d, J = 7.9 Hz, NCHCHCHC), 9.10 (1H, d, J = 6 Hz,
NCHCHCHC), 9.22 (1H, s, NCHC) ppm.
ultrasonicated immediately for 2 h at 46–48 ◦C. (Note: the flask
has to be open to allow the methyl chloride to escape and
should therefore always be kept in a fume hood). Subsequently
the mixture was subjected to high vacuum (rotary vane oil
pump) to strip off excessive O,O,O-trimethylthiophosphate
along with impurities like water. The product was stored under
an atmosphere of nitrogen to avoid moisture uptake from the
air. Yield: 8.23 g (91.7% of theory), nD20: 1.5122, density: 1.19 g
ml-1.
13C NMR (75 MHz, DMSO-d6):
d
12.9 (d, J =
3.7 Hz, P(SCH3)), 13.3 ((CH2)3CH3), 17.8 (CCH3), 18.8
(CH2CH2CH2CH3), 32.8 (CH2CH2CH2CH3), 51.4 (d, J =
6.3 Hz, P(OCH3)), 60.3 (NCH2-Pr), 127.4 (NCHCHCHC),
138.8 (NCHC), 142.2 (NCHCHCHC), 144.5 (NCHCHCHC),
145.7 (NCHC) ppm.
n
max/cm-1: 2958, 2931, 2873, 1570, 1464, 1312, 1238, 1174,
1070, 1043, 737, 655, 626, 557.
1H NMR (300 MHz, DMSO-d6): d 0.84 (3H, broad s,
(CH2)3CH3), 1.21 (2H, m, CH2CH2CH2CH3), 1.74 (2H, m,
CH2CH2CH2CH3), 1.93 (3H, d, J = 12.5 Hz, P(SCH3)), 3.33
(3H, d, J = 13.5 Hz, P(OCH3)), 3.88 (3H, s, NCH3), 4.19 (2H,
t, J = 6.7 Hz, NCH2-Pr), 7.85 (1H, s, NCHCHN), 7.95 (1H, s,
NCHCHN), 9.65 (1H, s, NCHN) ppm.
1-Methyl-1-propylpyrrolidinium O,S-dimethylphosphorothio-
ate
9a. 1-Methyl-1-propylpyrrolidinium
chloride
[528818-82-8] (18.23 g, 111.4 mmol), sodium O,S-
dimethylphosphorothioate (18.28 g, 111.4 mmol); Primarily, a
light yellow oily liquid was obtained, which solidified slowly at
room temperature (solid after 3 days). Upon re-dissolving the
crude product in acetone, residual amounts of sodium chloride
could be removed by centrifugation. Acetone was stripped
off by means of rotary vane vacuum pump at 90 ◦C giving
13C NMR (75 MHz, DMSO-d6): d 12.8 (d, J = 3.6 Hz,
P(SCH3)), 13.2 ((CH2)3CH3, 18.8 (CH2CH2CH2CH3), 31.5
(CH2CH2CH2CH3), 35.6 (NCH3), 48.4 (NCH2-Pr), 51.4 (d, J =
6.4Hz, P(OCH3)), 122.4(NCHCHN), 123.6(NCHCHN), 137.2
(NCHN) ppm.
1-methyl-1-propylpyrrolidinium chloride as light yellow oil
20
which solidified again after several days. Yield: 19.0 g g; nD
:
1.5078 (as subcooled melt); mp: 30–36 ◦C.
1-Allyl-3-methylimidazolium O,S-dimethyphosphorothioate.
n
max/cm-1: 2963, 2878, 2830, 1459, 1311, 1249, 1072, 1046,
A
round bottom flask was charged with 1-allyl-3-
1006, 975, 943, 905, 889, 735, 557.
methylimidazolium chloride (9.65 g, 60.8 mmol) and O,O,O-
trimethylthiophosphate (10.00 g, 64.0 mmol). Reactions con-
1H NMR (300 MHz, DMSO-d6): d 0.88 (3H, t, J = 7.3 Hz,
NCH2CH2CH3), 1.69 (2H, m, NCH2CH2CH3), 1.91 (2.5H, t,
J = 11 Hz, P(SCH3)), 2.05 (4H, s, CH2), 3.00 (3H, s, NCH3),
3.28 (2.6H, d, P(OCH3)), 3.34 (2H, m, NCH2CH2CH3), 3.49
(4H, d, J = 7.3 Hz, NCH2) ppm.
ditions and work-up procedure were analogous to the above
20
described BMIM dmpt. Yield: 14.78 g (91.9% of theory), nD
1.5288, density: 1.24 g ml-1.
n
:
max/cm-1: 2926, 2833, 1564, 1426, 1313, 1239, 1171, 1069,
13C NMR (75 MHz, DMSO-d6): d 10.7 (NCH2CH2CH3), 12.8
(d, J = 3.9 Hz, P(SCH3)), 16.7 (NCH2CH2CH3), 21.0 (2C, CH2),
47.3 (NCH3), 51.5 (d, J = 6.6 Hz, P(OCH3)), 63.2 (2C, NCH2),
64.2 (NCH2) ppm.
1043, 939, 739, 676, 628, 558.
1H NMR (300 MHz, DMSO-d6): d 1.93 (3H, d, J = 11.2 Hz,
P(SCH3)), 3.33 (3H, d, J = 12.2 Hz, P(OCH3)), 3.89 (3H, s,
NCH3), 4.90 (2H, d, J = 5.8 Hz, NCH2CH CH2), 5.28 (2H, m,
NCH2CH CH2), 6.01 (1H, m, NCH2CH CH2), 7.88 (1H, s,
NCHCHN), 7.90 (1H, s, NCHCHN), 9.63 (1H, s, NCHN) ppm.
13C NMR (75 MHz, DMSO-d6): d 12.8 (d, J = 3.7 Hz,
P(SCH3)), 35.7 (NCH3), 50.6 (NCH2CH CH2), 51.5 (d,
J = 6.1 Hz, P(OCH3)), 120.0 (NCH2CH = CH2), 122.4
(NCH2CH2CH3), 122.8 (NCHCHN), 132.0 (NCH2CH CH2),
137.3 (NCHN) ppm.
4-Ethyl-4-methylmorpholinium O,S-dimethylphosphorothio-
ate 10a. 4-Ethyl-4-methylmorpholinum chloride [6343-87-9]
(7.80 g, 47.1 mmol), sodium O,S-dimethylphosphorothioate
(7.73 g, 47.1 mmol). Yield: 11.87 g
n
max/cm-1: 2981, 2932, 2840, 2359, 2338, 1708, 1651, 1461,
1435, 1362, 1316, 1235, 1182, 1131, 1108, 1070, 1041, 952, 883,
761, 700, 616, 566.
1H NMR (300 MHz, DMSO-d6): d 1.23 (3H, t, J = 7.0 Hz,
NCH2CH3), 1.92 (2.2H, d, J = 11.1 Hz, P(SCH3)), 3.10 (3H,
s, NCH3), 3.29 (2.5H, d, J = 12.1 Hz, P(OCH3)), 3.39 (4H, m,
N(CH2CH2)2O), 3.52 (q, 2H, J = 7.2 Hz, NCH2CH3), 3.89 (4H,
m, N(CH2CH2)2O) ppm.
Dissolution studies
A Schlenk tube equipped with a magnetic stir bar was charged
with 1–2 g of ionic liquid and heated to 90 ◦C by means of
an oil bath. To remove possibly absorbed moisture the ILs were
subjected to high vacuum for several hours (rotary vane vacuum
pump). Subsequently, the temperature was increased to 100 ◦C
and the pre-dried pulp or xylan were added under nitrogen
purge. The dissolution state was checked with a microscope. To
determine the maximum solute concentration, more biopolymer
or IL was added gradually.
13C NMR (75 MHz, DMSO-d6): d 6.9 (NCH2CH3), 12.8
(d, J = 4.1 Hz, P(SCH3)), 45.2 (NCH3), 51.4 (d, J = 6.6 Hz,
P(OCH3)), 58.4 (2C, N(CH2CH2)2O), 59.3 (NCH2CH3), 59.8
(2C, N(CH2CH2)2O) ppm.
The analytical data of the respective O,Se-dimethyl-
phosphoroselenoate compounds can be found in the ESI.†
Regeneration was carried out by adding 5 ml of dry methanol
(£50 ppm H2O) and ultrasonication of the resulting mixture. The
precipitated fibres were isolated by means of a glass frit, washed
three times with methanol (10 ml each), dried and subjected to
GPC analysis.
Alternative syntheses
1-Butyl-3-methylimidazolium O,S-dimethylphosphorothioate..
A
round bottom flask was charged with 1-butyl-3-
methylimidazolium chloride (5.59 g, 32.0 mmol) and an excess
of O,O,O-trimethylthiophosphate (6.00 g, 38.4 mmol) and then
Both methanol and IL could be reused.
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2507–2517 | 2515
©

View Article Online
18 O. A. El Seoud, A. Koschella, L. C. Fidale, S. Dorn and T. Heinze,
Biomacromolecules, 2007, 8, 2629–2647.
Conclusions
19 I. Krossing, J. M. Slattery, C. Daguenet, P. J. Dyson, A. Oleinikova
and H. Weinga¨rtner, J. Am. Chem. Soc., 2006, 128, 13427–13434.
20 N. V. Ignat’ev, U. Welz-Biermann, A. Kucheryna, G. Bissky and H.
Willner, J. Fluorine Chem., 2005, 126, 1150–1159.
21 T. Welton, Chem. Rev., 1999, 99, 2071–2083.
22 L. Crowhurst, P. R. Mawdsley, J. M. Perez-Arlandis, P. A. Salter and
T. Welton, Phys. Chem. Chem. Phys., 2003, 5, 2790–2794.
23 A. Brandt, J. P. Hallett, D. J. Leak, R. J. Murphy and T. Welton,
Green Chem., 2010, 12, 672–679.
24 T. V. Doherty, M. Mora-Pale, S. E. Foley, R. J. Linhardt and J. S.
Dordick, Green Chem., 2010, 12, 1967–1975.
25 A. Xu, J. Wang and H. Wang, Green Chem., 2010, 12, 268–275.
26 H. Zhang, J. Wu, J. Zhang and J. He, Macromolecules, 2005, 38,
8272–8277.
The introduction of new ionic liquids is usually based on
novel cationic systems. Herein, pre-described anions were imple-
mented for the first time to generate a set of cellulose dissolving
ionic liquids. However, there was no clear trend showing a
reduced viscosity due to the anions’ asymmetry, as we expected
beforehand. Elevated temperatures are required to dissolve
biomass in ionic liquids within a reasonable time frame. There-
fore, assessment of the solvent’s thermal stability is a prerequisite
to pre-select potential biopolymer solvents. In our case, this
disqualified the utilization of O,Se-dimethylphosphoroselenoate
containing ionic liquids for further dissolution studies. Struc-
tural similarity of both cationic and anionic moieties does
not necessarily lead to comparable properties. Especially the
degradation of polymeric solutes during dissolution in an ionic
liquid and subsequent regeneration has to be studied case by
case. Although so far mainly the anion is stated to interact
with cellulosic material and thus promoting dissolution, a
pronounced dependence of chain degradation on the cation was
observed. Particularly high DP material with a multi-modal
molecular weight distribution can be changed significantly
which has to be considered when aiming for products with high
structural performance upon regeneration.
27 C. Cuissinat, P. Navard and T. Heinze, Carbohydr. Polym., 2008, 72,
590–596.
28 Y. Fukaya, A. Sugimoto and H. Ohno, Biomacromolecules, 2006, 7,
3295–3297.
29 Y. Fukaya, K. Hayashi, M. Wada and H. Ohno, Green Chem., 2008,
10, 44–46.
30 G. Ebner, S. Schiehser, A. Potthast and T. Rosenau, Tetrahedron Lett.,
2008, 49, 7322–7324.
31 V. K. Aggarwal, I. Emme and A. Mereu, Chem. Commun., 2002,
1612–1613.
32 S. Ko¨hler, T. Liebert, M. Scho¨bitz, J. Schaller, F. Meister, W. Gu¨nther
and T. Heinze, Macromol. Rapid Commun., 2007, 28, 2311–2317.
33 T. Heinze, S. Dorn, M. Scho¨bitz, T. Liebert, S. Ko¨hler and F. Meister,
Macromol. Symp., 2008, 262, 8–22.
34 J. Zimmermann, B. Ondruschka and A. Stark, Org. Process Res. Dev.,
2010, 14, 1102–1109.
35 E. Kuhlmann, S. Himmler, H. Giebelhaus and P. Wasserscheid, Green
Chem., 2007, 9, 233–242.
Acknowledgements
36 J.-y. Wang, F.-Y. Zhao, R.-j. Liu and Y.-q. Hu, J. Chem. Thermodyn.,
2011, 43, 47–50.
Financial support is gratefully acknowledged by M.H. and
L.K.J.H. (Forestcluster Ltd., Tekes) and C.F. (Wood K plus).
Further, we would like to thank M. Noisternig for advice on
moisture sorption analyses.
37 M. Abe, Y. Fukaya and H. Ohno, Green Chem., 2010, 12, 1274–1280.
38 M. Hummel, G. Laus, A. Schwa¨rzler, G. Bentivoglio, E. Rubatscher,
H. Kopacka, K. Wurst, V. Kahlenberg, T. Gelbrich, J. Griesser
Ulrich, T. Ro¨der, K. Weber Hedda, H. Schottenberger and H.
Sixta, in Cellulose Solvents: For Analysis, Shaping and Chemical
Modification, ACS Symposium Series, ed. T. Liebert, T. J. Heinze
and K. J. Edgar, American Chemical Society, 2010, vol. 1033, pp.
229–259.
39 J. Vitz, T. Erdmenger and S. Schubert Ulrich, in Cellulose Solvents:
For Analysis, Shaping and Chemical Modification, ACS Symposium
Series, ed. T. Liebert, T. J. Heinze and K. J. Edgar, American
Chemical Society, 2010, vol. 1033, pp. 299–317.
Notes and references
1 A. J. Ragauskas, C. K. Williams, B. H. Davison, G. Britovsek, J.
Cairney, C. A. Eckert, W. J. Frederick, Jr., 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 L. da Costa Sousa, S. P. S. Chundawat, V. Balan and B. E. Dale, Curr.
Opin. Biotechnol., 2009, 20, 339–347.
3 A. Pinkert, K. N. Marsh, S. Pang and M. P. Staiger, Chem. Rev.,
2009, 109, 6712–6728.
4 A. S. Gross and J.-W. Chu, J. Phys. Chem. B, 2010, 114, 13333–13341.
5 B. Lindman, G. Karlstro¨m and L. Stigsson, J. Mol. Liq., 2010, 156,
76–81.
6 A. L. Horvath, J. Phys. Chem. Ref. Data, 2006, 35, 77–92.
7 C. M. Hansen and A. Bjo¨rkman, Holzforschung, 1998, 52, 335–344.
8 S. Ko¨hler and T. Heinze, Macromol. Biosci., 2007, 7, 307–314.
9 T. Heinze, R. Dicke, A. Koschella, A. H. Kull, E.-A. Klohr and W.
Koch, Macromol. Chem. Phys., 2000, 201, 627–631.
10 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.
40 J. Vitz, T. Erdmenger, C. Haensch and U. S. Schubert, Green Chem.,
2009, 11, 417–424.
41 H. Xie, I. Kilpela¨inen, A. King, T. Leskinen, P. Ja¨rvi and S.
Argyropoulos Dimitris, in Cellulose Solvents: For Analysis, Shaping
and Chemical Modification, ACS Symposium Series, ed. T. Liebert,
T. J. Heinze and K. J. Edgar, American Chemical Society, 2010, vol.
1033, pp. 343–363.
42 Y. Fukaya, K. Hayashi, S. Kim Seung and H. Ohno, in Cellulose
Solvents: For Analysis, Shaping and Chemical Modification, ACS
Symposium Series, ed. T. Liebert, T. J. Heinze and K. J. Edgar,
American Chemical Society, 2010, vol. 1033, pp. 55–66.
43 G. Mastrantonio and C. O. D. Ve´dova, J. Mol. Struct., 2001, 561,
161–174.
44 G. Hilgetag and H. Teichmann, J. Prakt. Chem., 1959, 8, 90–96.
45 G. Hilgetag and H. Teichmann, J. Prakt. Chem., 1959, 8, 97–
103.
11 H. P. Fink, P. Weigel, H. J. Purz and J. Ganster, Prog. Polym. Sci.,
2001, 26, 1473–1524.
46 R. D. Verschoyle, E. Reiner, E. Bailey and W. N. Aldridge, Arch.
12 T. Heinze and T. Liebert, Prog. Polym. Sci., 2001, 26, 1689–1762.
13 T. Liebert, in Cellulose Solvents: For Analysis, Shaping and Chemical
Modification, ACS Symposium Series, ed. T. Liebert, T. J. Heinze and
K. J. Edgar, American Chemical Society, vol. 1033, pp. 3–54.
14 T. Rosenau, A. Potthast, H. Sixta and P. Kosma, Prog. Polym. Sci.,
2001, 26, 1763–1837.
15 C. Graenacher, US Pat., 1,943,176, 1934.
16 R. P. Swatloski, S. K. Spear, J. D. Holbrey and R. D. Rogers, J. Am.
Chem. Soc., 2002, 124, 4974–4975.
Toxicol., 1982, 49, 293–301.
47 C. J. Bradaric, A. Downard, C. Kennedy, A. J. Robertson and Y.
Zhou, Green Chem., 2003, 5, 143–152.
48 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.
49 L. Szarvas, T. Rohde, L. Wittenbecher and K. Massonne, US Pat.,
20080033209 A1, 2008.
50 J. D. Holbrey, R. D. Rogers, S. S. Shukla and C. D. Wilfred, Green
Chem., 2010, 12, 407–413.
17 G. Laus, G. Bentivoglio, H. Schottenberger, V. Kahlenberg, H.
Kopacka, T. Ro¨der and H. Sixta, Lenzinger Ber., 2005, 84, 71–85.
2516 | Green Chem., 2011, 13, 2507–2517
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
51 M. Egashira, Y. Yamamoto, T. Fukutake, N. Yoshimoto and M.
Morita, J. Fluorine Chem., 2006, 127, 1261–1264.
52 N. Ignatyev, U. Welz-Biermann, A. Kucheryna and H. Willner, WO
Pat., 2006063655 A1, 2006.
53 N. Farschtschi, D. G. Gorenstein, T. Fanni and K. Taira, Phosphorus,
Sulfur Silicon Relat. Elem., 1990, 47, 93–104.
54 P. Pistschimuka, J. Prakt. Chem., 1911, 84, 746–760.
55 C. Froschauer, C, Himmer, K. Wurst, G. Laus and H. Schottenberger,
Z. Kristallogr. - New Cryst. Struct., 2010, 225, 569–570.
56 M. Hummel, H. Sixta, C. Froschauer, H. K. Weber, V. Kahlenberg,
G. Laus and H. Schottenberger, Private communication to the
Cambridge Structural Database, deposition number CCDC 772705,
2010.
57 C. H. C. Janssen, M. C. Kroon, S. J. Metz, J. van Spronsen and G.-J.
Witkamp, J. Chem. Eng. Data, 2010, 55, 3391–3394.
58 J. H. Davis, C. M. Gordon, C. Hilgers and P. Wasserscheid, in Ionic
Liquids in Synthesis, ed. P. Wasserscheid and T. Welton, Wiley-VCH
Verlag GmbH & Co. KGaA, 2003, pp. 7–40.
68 N. Sun, X. Jiang, M. L. Maxim, A. Metlen and R. D. Rogers,
ChemSusChem, 2011, 4, 65–73.
69 M. J. Monteiro, F. F. Camilo, M. C. C. Ribeiro and R. M. Torresi, J.
Phys. Chem. B, 2010, 114, 12488–12494.
70 S. Bulut, P. Klose, M.-M. Huang, H. Weinga¨rtner, P. J. Dyson, G.
Laurenczy, C. Friedrich, J. Menz, K. Ku¨mmerer and I. Krossing,
Chem.–Eur. J., 2010, 16, 13139–13154.
71 D. Bejan, N. Ignat’ev and H. Willner, J. Fluorine Chem., 2010, 131,
325–332.
72 B. Tong, Q.-S. Liu, Z.-C. Tan and U. Welz-Biermann, J. Phys. Chem.
A, 2010, 114, 3782–3787.
73 R. E. Del Sesto, T. M. McCleskey, C. Macomber, K. C. Ott, A. T.
Koppisch, G. A. Baker and A. K. Burrell, Thermochim. Acta, 2009,
491, 118–120.
74 J. Kim, R. P. Singh and J. n. M. Shreeve, Inorg. Chem., 2004, 43,
2960–2966.
75 R. C. Remsing, R. P. Swatloski, R. D. Rogers and G. Moyna, Chem.
Commun., 2006, 1271–1273.
76 C. D. Tran, S. H. De Paoli Lacerda and D. Oliveira, Appl. Spectrosc.,
2003, 57, 152–156.
59 K. R. Seddon, A. Stark and M.-J. Torres, Pure Appl. Chem., 2000,
72, 2275–2287.
60 M. Fremantle, An introduction to ionic liquids, RSC, Cambridge,
2010.
77 T. Erdmenger, J. Vitz, F. Wiesbrock and U. S. Schubert, J. Mater.
Chem., 2008, 18, 5267–5273.
61 J. G. Huddleston, A. E. Visser, W. M. Reichert, H. D. Willauer, G.
A. Broker and R. D. Rogers, Green Chem., 2001, 3, 156–164.
62 Y. Hao, J. Peng, S. Hu, J. Li and M. Zhai, Thermochim. Acta, 2010,
501, 78–83.
78 J. W. Russo and M. M. Hoffmann, J. Chem. Eng. Data, 2010, 55,
5900–5905.
79 J.-y. Wang, F.-Y. Zhao, R.-j. Liu and Y.-q. Hu, J. Chem. Thermodyn.,
2011, 43, 47–50.
80 S. Zhang, X. Qi, X. Ma, L. Lu and Y. Deng, J. Phys. Chem. B, 2010,
114, 3912–3920.
63 N. Meine, F. Benedito and R. Rinaldi, Green Chem., 2010, 12, 1711–
1714.
64 H. Zhao, G. A. Baker, Z. Song, O. Olubajo, T. Crittle and D. Peters,
Green Chem., 2008, 10, 696–705.
65 F. Hermanutz, F. Ga¨hr, E. Uerdingen, F. Meister and B. Kosan,
Macromol. Symp., 2008, 262, 23–27.
66 B. Kosan, C. Michels and F. Meister, Cellulose, 2007, 15, 59–66.
67 N. Sun, M. Rahman, Y. Qin, M. L. Maxim, H. Rodriguez and R. D.
Rogers, Green Chem., 2009, 11, 646–655.
81 M. E. Zakrzewska, E. Bogel-Łukasik and R. Bogel-Łukasik, Energy
Fuels, 2010, 24, 737–745.
82 G. Bentivoglio, T. Ro¨der, M. Fasching, M. Buchberger, H.
Schottenberger and H. Sixta, Lenzinger Ber., 2006, 86, 154–
161.
83 N. Schelosky, T. Ro¨der and T. Baldinger, Das Papier, 1999, 53, 728–
738.
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2507–2517 | 2517
©