Accepted Manuscript Title: True molecular solutions of natural cellulose in the binary ionic liquid-containing solvent mixtures Author: Dmitry M. Rein Rafail Khalfin Noemi Szekely Yachin Cohen PII: DOI: Reference:
S0144-8617(14)00536-0 http://dx.doi.org/doi:10.1016/j.carbpol.2014.05.059 CARP 8926
To appear in: Received date: Revised date: Accepted date:
27-1-2014 23-4-2014 22-5-2014
Please cite this article as: Rein, D. M., Khalfin, R., Szekely, N., and Cohen, Y.,True molecular solutions of natural cellulose in the binary ionic liquid-containing solvent mixtures, Carbohydrate Polymers (2014), http://dx.doi.org/10.1016/j.carbpol.2014.05.059 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Highlights
1 •
Cellulose dissolves molecularly in mixtures of organic solvents and ionic liquid.
3
•
Scattering measurements and TEM imaging showed insignificant molecular aggregation.
4
•
Highly concentrated cellulose/IL solutions are prepared by co-solvent evaporation.
5
•
Facile cellulose dissolution correlates with higher conductivity of solvent mixture.
cr
ip t
2
6
us
7 8
an
9
Ac ce p
te
d
M
10
1
Page 1 of 34
True molecular solutions of natural cellulose in the binary
10
ionic liquid‐containing solvent mixtures
12
Dmitry M. Rein a,*, Rafail Khalfin a, Noemi Szekely b, Yachin Cohen a
ip t
11
13
a
14
b
15
Garching Germany
Faculty of Chemical Engineering, Technion – Israel Institute of Technology, Haifa 32000, Israel;
cr
Jülich Centre for Neutron Science, Forschungszentrum Jülich GmbH Outstation at MLZ, 85747
us
16
Abstract
17
Evidence is presented for the first time of true molecular dissolution of cellulose in binary mixtures of common polar organic solvents with ionic liquid. Cryogenic transmission electron microscopy, small‐angle neutron‐, x‐ray‐ and static light scattering were used to investigate the structure of cellulose solutions in mixture of dimethyl formamide and 1‐ethyl‐3‐methylimidazolium acetate. Structural information on the dissolved chains (average molecular weight ~5∙104 g/mol; gyration radius ~36 nm, persistence length ~4.5 nm), indicate the absence of significant aggregation of the dissolved chains and the calculated value of the second virial coefficient ~2.45∙10‐2 mol∙ml/g2 indicates that this solvent system is a good solvent for cellulose. More facile dissolution of cellulose could be achieved in solvent mixtures that exhibit the highest electrical conductivity. Highly concentrated cellulose solution in pure ionic liquid (27 wt. %) prepared according to novel method, utilizing the rapid evaporation of a volatile co‐solvent in binary solvent mixtures at superheated conditions, shows insignificant cellulose molecular aggregation.
38
___________________________________________________________________
39 40 41 42
*Corresponding author. Tel.: +972 4 8292113; fax: +972 4 829 5672. E‐mail: [email protected]
Ac ce p
te
d
M
an
18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
Keywords: cellulose true molecular solution; amorphous amphiphilic cellulose; binary solvent mixture; small‐angle neutron‐, x‐ray‐ diffraction
Authors’ e‐mails: R. Khalfin: [email protected]; N. Szekely: n.szekely@fz‐juelich.de; Y. Cohen: [email protected]
2
Page 2 of 34
43
1.
Introduction One major disadvantage of natural cellulose, regarding its industrial
45
application, is that due to its crystal fibril structure and the significant presence of
46
intermolecular hydrogen bonding and hydrophobic interactions (Medronho,
47
Romano, Miguel, Stigsson & Lindman, 2012), it is insoluble in common solvents.
48
Nevertheless after years of research many specific cellulose solvents were found,
49
some of which are used now in the manufacture of commercial cellulose products
50
(viscose and cuprammonium fibers, Lyocell®, etc.).
us
cr
ip t
44
Derivatives of cellulose (Hunt, Newman, Scheraga, & Flory, 1956) or its
52
metallo‐complexes (Burchard & Klufers, 1994) are formed by an interaction
53
between cellulose and the corresponding solvent components. In such solutions
54
the macromolecules of cellulose are claimed to be completely separated from each
55
other, and their dimension in the dissolved state is defined by its gyration radius
56
(Rg), which is close to the size of a single macromolecule. For example, in cadmium
57
complexing Cd‐tris (2 aminoethyl) amine solution, the size of cellulose particles is
58
equal to about 20 nm (Saalwachter et al., 2000; Schulz, Seger, & Buchard, 2000).
59
60
bringing the cellulose macromolecules into the aqueous solution by forming the
61
hydrogen‐bonded complex networks. The mean size of the self‐assembled solvent‐
62
cellulose clusters in these solvents is 60‐160 nm (Luo, & Zhang, 2013).
63
64
on complex formation or derivatization of cellulose. Direct solvents transfer the
65
polymer into solution by the mechanism of physical solvation. Relatively large
66
dimensions of aggregates have been evaluated for cellulose dissolved in direct
67
solvents. For example, diluted solutions of cellulose in N‐methylmorpholine‐N‐
Ac ce p
te
d
M
an
51
The fast cellulose dissolution could be realized in the alkali–urea solvents,
Dissolution of cellulose in the so called “direct solvents” does not rely either
3
Page 3 of 34
68
oxide (NMMO) exhibit an Rg value of about 200 nm (Arndt, Morgenstern, & Roder,
69
2002; Drechsler, Radosta, & Vorwerg, 2000; Roder, & Morgenstern, 1999), which
70
considerably exceeds that of the isolated macromolecule.
71
72
condition of a true molecular solution of cellulose is not achieved.
ip t
Thus, it may be assumed that with alkali–urea and direct solvents the
Ionic liquids (IL) are a unique class of direct solvents for cellulose, which
74
have attracted significant interest due to their excellent solvation characteristics,
75
low toxicity and volatility leading to facile recovery (Mizumo, Marwanta, Matsumi
76
& Ohno, 2004; Swatloski, Spear, Holbrey, & Rogers, 2002; Wang, Gurau, & Rogers,
77
2012). However, even in one of the most efficient IL solvent, 1‐ethyl‐3‐
78
methylimidazolium acetate (EMIMAc) (Cheng et al., 2012; Kyllonen et al., 2013;
79
Rinaldi, 2011; Zavrel, Bross, Funke, Buchs, & Spiess, 2009) there is evidence for
80
aggregates with Rg about 150 nm, much higher than that of a single cellulose
81
macromolecule of the same molecular weight (Kuzmina, Sashina, Troshenkowa, &
82
Wawro, 2010) , presumably due to undissolved solvated cellulose crystals. It was
83
shown that, according to rheological measurements, the thermodynamic
84
properties of EMIMAc correspond to the properties of a θ solvent for cellulose
85
(Gericke, Schlufter, Liebert, Heinze, & Budtova, 2009; Sescousse, Le, Ries, &
86
Budtova, 2010).
Ac ce p
te
d
M
an
us
cr
73
In recent years it was revealed that mixtures of polar organic solvents with
87 88
a small fraction of IL can instantaneously dissolve cellulose even at high
89
concentration (Gericke, Liebert, El Seoud, & Heinze, 2011; Luo & Neogi, 2009;
90
Rinaldi, 2011; Xu, Zhang, Zhao & Wang, 2013). However, a reasonable explanation
91
of this effect is still lacking. According to one explanation, solvation effects such as
92
the chemical interactions due to the specific composition of the solute and solvent
93
play a major role. These can be due to acid‐base or polar interactions, or an ability 4
Page 4 of 34
to form cyclic arrangements of solvent and cellulose during dissolution (Remsing,
95
Liu, Sergeyev, & Moyna, 2008; Pinkert, Marsh, & Pang, 2010). According to another
96
explanation, the enhanced cellulose dissolution in co‐solvent/IL mixtures is caused
97
by the increased concentration of the “free” anions from the dissociated IL owing
98
to the preferential solvation of the cations by the aprotic polar solvents. (Xu,
99
Zhang, Zhao & Wang, 2013) A third explanation considers possible electric
100
interactions between cellulose chains, which are sufficiently weakened in the co‐
101
solvent/IL environment with high relative dielectric permittivity (the dielectric
102
constant for the imidazolium‐based IL it is about 10 (Wakai, Oleinikova, Ott, &
103
Weingartner, 2005)) (Stana‐Kleinschek, & Ribitsch, 1998). It was noted that, basing
104
on rheological measurements, EMIMAc/dimethyl sulfoxide mixture is more like a
105
θ solvent for cellulose, and the thermodynamic properties of EMIMAc/dimethyl
106
sulfoxide mixtures are similar with those of ILs. The conformation of cellulose in
107
ILs would not be changed with the addition of dimethyl sulfoxide not only in the
108
dilute cellulose concentrations but also in the entanglement concentrations (Lv et
109
al., 2012).
Ac ce p
te
d
M
an
us
cr
ip t
94
The decisive influence of the cellulose amphiphilicity and hydrophobic
110 111
interactions on the processes of its dissolving currently is being actively discussed
112
in the scientific literature (Glasser et al., 2012; Medronho et al., 2012). In previous
113
studies we described a novel phenomenon: outstanding amphiphilic properties
114
inherent to cellulose molecules dissolved in co‐solvent/IL binary mixtures and to
115
cellulose in the amorphous state in hydrogels regenerated from these solutions, as
116
evident in their ability to form a stabilizing coating for oil‐in‐water (O/W) and
117
water‐in‐oil (W/O) emulsions (Rein, Khalfin, & Cohen, 2012). Similar properties
118
were also found in cellulose regenerated from the phosphoric acid (Jia et al., 2013).
119
We had presented the hypothesis that this manifestation of cellulose’s amphiphilic 5
Page 5 of 34
character is made possible by its molecular dissolution in the solvent mixture prior
121
to the emulsification process (Rein et al., 2012). The amorphous cellulose hydrogel,
122
received on the basis of these free‐molecular‐chain solutions, also holds noticeable
123
chain mobility and pronounced amphiphilic properties (Rein et al., 2012), which is
124
consistent with the previously detected phenomena that even in the crystalline
125
state, in disordered structural regions, plasticized by sorbed water, there is
126
evidence for the increased mobility of cellulose chain segments. (Bradley & Carr,
127
1976; Froix & Nelson, 1975), a thorough analysis of the molecular state of cellulose
128
in solution in ionic liquids and their binary solvent mixtures using light scattering
129
and transmission electron microscopy methods is lacking. For this purpose we
130
seek to utilize a combination of scattering techniques, static light scattering (SLS),
131
small‐angle x‐ray and neutron scattering (SAXS and SANS) with cryogenic
132
transmission electron microscopy (cryo‐TEM).
133 134
2. Materials and methods
135
136
(given by the supplier) ~ 4.6∙104 g/mol (degree of polymerization 295 and particle
137
size in the range of 20 ‐ 160 μm), and ionic liquid EMIMAc of 90% purity, were
138
obtained from Sigma–Aldrich. EMIMAc and cellulose were dried in a vacuum
139
oven at 60 °C and 0.26 kPa for at least 24 h. Chloroform, dichloromethane (DCM),
140
dimethyl formamide (DMF), deuterated DMF (DMF‐d7), acetonitrile and
141
propylene carbonate were purchased from Sigma–Aldrich. These chemicals were
142
used without additional purification. Unless otherwise stated, the weight
143
percentages of the components were calculated based on the total composition of
144
the final mixture.
145
te
d
M
an
us
cr
ip t
120
Ac ce p
Microcrystalline cellulose powder Avicel® with nominal molecular weight
6
Page 6 of 34
146
2.1 Dissolution of cellulose Dissolution of cellulose was realized in two stages so as to reduce the
148
aggregation (Rein et al., 2012; Rinaldi, 2011). First, cellulose was dispersed in a
149
specific volume of the co‐solvent. Next, the required volume of EMIMAc was
150
added into the suspension, and the mixture was heated to 90 °C using magnetic
151
stirring during about 10 min. Subsequently, the obtained solution was cooled to
152
room temperature and was not further treated.
154
2.2 Equipment and characterization methods
us
153
cr
ip t
147
The morphology of the solutions was investigated by cryogenic
156
transmission electron microscope (cryo‐TEM) T12G2 (FEI Co., Netherlands).
157
Vitrified samples for cryo‐TEM (Talmon, 1996) were prepared in a controlled
158
environment vitrification system (CEVS). (Bellare, Davis, Scriven, & Talmon, 1988)
M
an
155
SLS measurements were carried out with a goniometer BI‐200SM
160
(Brookhaven Instruments Corp., USA) using an argon‐ion laser (wavelength 514.5
161
nm). Data obtained at several concentrations (0.4 – 4 wt. %) and scattering angles
162
(30 – 150°) were evaluated by the common Zimm plot yielding the root‐mean‐
163
square Rg, weight‐average molar mass Mw and the second osmotic virial coefficient
164
A2 (Debye, 1944; Zimm, 1945). The indices of the solvent refraction no (no=1.44 for
165
the investigated DMF/EMIMAc mixture with molar ratio 9:1) and the mass
166
concentration (c) increment of the solution refraction index dn/dc (for investigated
167
DMF/EMIMAc mixture dn/dc=0.061±0.06 cm3/g) were measured by a differential
168
refractometer BI‐DNDC (Brookhaven Instruments Corp., USA).
Ac ce p
te
d
159
169
Electrical conductivity of solvent mixtures was measured with a Precision
170
Impedance Analyzer Agilent 4294A (Agilent Technologies) at 0.5 V and 10 kHz.
171
Each sample was analyzed in triplicate and the mean values were plotted (average
172
measurements error about 12%). 7
Page 7 of 34
X‐ray diffractometry of the samples was performed using a small/wide‐
174
angle diffractometer (Molecular Metrology SAXS/WAXS system) equipped with a
175
sealed microfocus tube (MicroMax‐002 + S) emitting CuKα radiation, two Göbel
176
mirrors, three‐pinhole slits, and generator that is powered at 45 kV and 0.9 mA.
177
The scattering patterns were recorded by a two‐dimensional position sensitive
178
wire detector positioned 150 cm behind the sample. The solutions were sealed in
179
thin‐walled glass capillaries about 2 mm in diameter and 0.01 mm wall thickness,
180
and measured under vacuum at the ambient temperature. The scattered intensity
181
I(q) was recorded at the interval 0.07
S0144-8617(14)00536-0 http://dx.doi.org/doi:10.1016/j.carbpol.2014.05.059 CARP 8926
To appear in: Received date: Revised date: Accepted date:
27-1-2014 23-4-2014 22-5-2014
Please cite this article as: Rein, D. M., Khalfin, R., Szekely, N., and Cohen, Y.,True molecular solutions of natural cellulose in the binary ionic liquid-containing solvent mixtures, Carbohydrate Polymers (2014), http://dx.doi.org/10.1016/j.carbpol.2014.05.059 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Highlights
1 •
Cellulose dissolves molecularly in mixtures of organic solvents and ionic liquid.
3
•
Scattering measurements and TEM imaging showed insignificant molecular aggregation.
4
•
Highly concentrated cellulose/IL solutions are prepared by co-solvent evaporation.
5
•
Facile cellulose dissolution correlates with higher conductivity of solvent mixture.
cr
ip t
2
6
us
7 8
an
9
Ac ce p
te
d
M
10
1
Page 1 of 34
True molecular solutions of natural cellulose in the binary
10
ionic liquid‐containing solvent mixtures
12
Dmitry M. Rein a,*, Rafail Khalfin a, Noemi Szekely b, Yachin Cohen a
ip t
11
13
a
14
b
15
Garching Germany
Faculty of Chemical Engineering, Technion – Israel Institute of Technology, Haifa 32000, Israel;
cr
Jülich Centre for Neutron Science, Forschungszentrum Jülich GmbH Outstation at MLZ, 85747
us
16
Abstract
17
Evidence is presented for the first time of true molecular dissolution of cellulose in binary mixtures of common polar organic solvents with ionic liquid. Cryogenic transmission electron microscopy, small‐angle neutron‐, x‐ray‐ and static light scattering were used to investigate the structure of cellulose solutions in mixture of dimethyl formamide and 1‐ethyl‐3‐methylimidazolium acetate. Structural information on the dissolved chains (average molecular weight ~5∙104 g/mol; gyration radius ~36 nm, persistence length ~4.5 nm), indicate the absence of significant aggregation of the dissolved chains and the calculated value of the second virial coefficient ~2.45∙10‐2 mol∙ml/g2 indicates that this solvent system is a good solvent for cellulose. More facile dissolution of cellulose could be achieved in solvent mixtures that exhibit the highest electrical conductivity. Highly concentrated cellulose solution in pure ionic liquid (27 wt. %) prepared according to novel method, utilizing the rapid evaporation of a volatile co‐solvent in binary solvent mixtures at superheated conditions, shows insignificant cellulose molecular aggregation.
38
___________________________________________________________________
39 40 41 42
*Corresponding author. Tel.: +972 4 8292113; fax: +972 4 829 5672. E‐mail: [email protected]
Ac ce p
te
d
M
an
18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
Keywords: cellulose true molecular solution; amorphous amphiphilic cellulose; binary solvent mixture; small‐angle neutron‐, x‐ray‐ diffraction
Authors’ e‐mails: R. Khalfin: [email protected]; N. Szekely: n.szekely@fz‐juelich.de; Y. Cohen: [email protected]
2
Page 2 of 34
43
1.
Introduction One major disadvantage of natural cellulose, regarding its industrial
45
application, is that due to its crystal fibril structure and the significant presence of
46
intermolecular hydrogen bonding and hydrophobic interactions (Medronho,
47
Romano, Miguel, Stigsson & Lindman, 2012), it is insoluble in common solvents.
48
Nevertheless after years of research many specific cellulose solvents were found,
49
some of which are used now in the manufacture of commercial cellulose products
50
(viscose and cuprammonium fibers, Lyocell®, etc.).
us
cr
ip t
44
Derivatives of cellulose (Hunt, Newman, Scheraga, & Flory, 1956) or its
52
metallo‐complexes (Burchard & Klufers, 1994) are formed by an interaction
53
between cellulose and the corresponding solvent components. In such solutions
54
the macromolecules of cellulose are claimed to be completely separated from each
55
other, and their dimension in the dissolved state is defined by its gyration radius
56
(Rg), which is close to the size of a single macromolecule. For example, in cadmium
57
complexing Cd‐tris (2 aminoethyl) amine solution, the size of cellulose particles is
58
equal to about 20 nm (Saalwachter et al., 2000; Schulz, Seger, & Buchard, 2000).
59
60
bringing the cellulose macromolecules into the aqueous solution by forming the
61
hydrogen‐bonded complex networks. The mean size of the self‐assembled solvent‐
62
cellulose clusters in these solvents is 60‐160 nm (Luo, & Zhang, 2013).
63
64
on complex formation or derivatization of cellulose. Direct solvents transfer the
65
polymer into solution by the mechanism of physical solvation. Relatively large
66
dimensions of aggregates have been evaluated for cellulose dissolved in direct
67
solvents. For example, diluted solutions of cellulose in N‐methylmorpholine‐N‐
Ac ce p
te
d
M
an
51
The fast cellulose dissolution could be realized in the alkali–urea solvents,
Dissolution of cellulose in the so called “direct solvents” does not rely either
3
Page 3 of 34
68
oxide (NMMO) exhibit an Rg value of about 200 nm (Arndt, Morgenstern, & Roder,
69
2002; Drechsler, Radosta, & Vorwerg, 2000; Roder, & Morgenstern, 1999), which
70
considerably exceeds that of the isolated macromolecule.
71
72
condition of a true molecular solution of cellulose is not achieved.
ip t
Thus, it may be assumed that with alkali–urea and direct solvents the
Ionic liquids (IL) are a unique class of direct solvents for cellulose, which
74
have attracted significant interest due to their excellent solvation characteristics,
75
low toxicity and volatility leading to facile recovery (Mizumo, Marwanta, Matsumi
76
& Ohno, 2004; Swatloski, Spear, Holbrey, & Rogers, 2002; Wang, Gurau, & Rogers,
77
2012). However, even in one of the most efficient IL solvent, 1‐ethyl‐3‐
78
methylimidazolium acetate (EMIMAc) (Cheng et al., 2012; Kyllonen et al., 2013;
79
Rinaldi, 2011; Zavrel, Bross, Funke, Buchs, & Spiess, 2009) there is evidence for
80
aggregates with Rg about 150 nm, much higher than that of a single cellulose
81
macromolecule of the same molecular weight (Kuzmina, Sashina, Troshenkowa, &
82
Wawro, 2010) , presumably due to undissolved solvated cellulose crystals. It was
83
shown that, according to rheological measurements, the thermodynamic
84
properties of EMIMAc correspond to the properties of a θ solvent for cellulose
85
(Gericke, Schlufter, Liebert, Heinze, & Budtova, 2009; Sescousse, Le, Ries, &
86
Budtova, 2010).
Ac ce p
te
d
M
an
us
cr
73
In recent years it was revealed that mixtures of polar organic solvents with
87 88
a small fraction of IL can instantaneously dissolve cellulose even at high
89
concentration (Gericke, Liebert, El Seoud, & Heinze, 2011; Luo & Neogi, 2009;
90
Rinaldi, 2011; Xu, Zhang, Zhao & Wang, 2013). However, a reasonable explanation
91
of this effect is still lacking. According to one explanation, solvation effects such as
92
the chemical interactions due to the specific composition of the solute and solvent
93
play a major role. These can be due to acid‐base or polar interactions, or an ability 4
Page 4 of 34
to form cyclic arrangements of solvent and cellulose during dissolution (Remsing,
95
Liu, Sergeyev, & Moyna, 2008; Pinkert, Marsh, & Pang, 2010). According to another
96
explanation, the enhanced cellulose dissolution in co‐solvent/IL mixtures is caused
97
by the increased concentration of the “free” anions from the dissociated IL owing
98
to the preferential solvation of the cations by the aprotic polar solvents. (Xu,
99
Zhang, Zhao & Wang, 2013) A third explanation considers possible electric
100
interactions between cellulose chains, which are sufficiently weakened in the co‐
101
solvent/IL environment with high relative dielectric permittivity (the dielectric
102
constant for the imidazolium‐based IL it is about 10 (Wakai, Oleinikova, Ott, &
103
Weingartner, 2005)) (Stana‐Kleinschek, & Ribitsch, 1998). It was noted that, basing
104
on rheological measurements, EMIMAc/dimethyl sulfoxide mixture is more like a
105
θ solvent for cellulose, and the thermodynamic properties of EMIMAc/dimethyl
106
sulfoxide mixtures are similar with those of ILs. The conformation of cellulose in
107
ILs would not be changed with the addition of dimethyl sulfoxide not only in the
108
dilute cellulose concentrations but also in the entanglement concentrations (Lv et
109
al., 2012).
Ac ce p
te
d
M
an
us
cr
ip t
94
The decisive influence of the cellulose amphiphilicity and hydrophobic
110 111
interactions on the processes of its dissolving currently is being actively discussed
112
in the scientific literature (Glasser et al., 2012; Medronho et al., 2012). In previous
113
studies we described a novel phenomenon: outstanding amphiphilic properties
114
inherent to cellulose molecules dissolved in co‐solvent/IL binary mixtures and to
115
cellulose in the amorphous state in hydrogels regenerated from these solutions, as
116
evident in their ability to form a stabilizing coating for oil‐in‐water (O/W) and
117
water‐in‐oil (W/O) emulsions (Rein, Khalfin, & Cohen, 2012). Similar properties
118
were also found in cellulose regenerated from the phosphoric acid (Jia et al., 2013).
119
We had presented the hypothesis that this manifestation of cellulose’s amphiphilic 5
Page 5 of 34
character is made possible by its molecular dissolution in the solvent mixture prior
121
to the emulsification process (Rein et al., 2012). The amorphous cellulose hydrogel,
122
received on the basis of these free‐molecular‐chain solutions, also holds noticeable
123
chain mobility and pronounced amphiphilic properties (Rein et al., 2012), which is
124
consistent with the previously detected phenomena that even in the crystalline
125
state, in disordered structural regions, plasticized by sorbed water, there is
126
evidence for the increased mobility of cellulose chain segments. (Bradley & Carr,
127
1976; Froix & Nelson, 1975), a thorough analysis of the molecular state of cellulose
128
in solution in ionic liquids and their binary solvent mixtures using light scattering
129
and transmission electron microscopy methods is lacking. For this purpose we
130
seek to utilize a combination of scattering techniques, static light scattering (SLS),
131
small‐angle x‐ray and neutron scattering (SAXS and SANS) with cryogenic
132
transmission electron microscopy (cryo‐TEM).
133 134
2. Materials and methods
135
136
(given by the supplier) ~ 4.6∙104 g/mol (degree of polymerization 295 and particle
137
size in the range of 20 ‐ 160 μm), and ionic liquid EMIMAc of 90% purity, were
138
obtained from Sigma–Aldrich. EMIMAc and cellulose were dried in a vacuum
139
oven at 60 °C and 0.26 kPa for at least 24 h. Chloroform, dichloromethane (DCM),
140
dimethyl formamide (DMF), deuterated DMF (DMF‐d7), acetonitrile and
141
propylene carbonate were purchased from Sigma–Aldrich. These chemicals were
142
used without additional purification. Unless otherwise stated, the weight
143
percentages of the components were calculated based on the total composition of
144
the final mixture.
145
te
d
M
an
us
cr
ip t
120
Ac ce p
Microcrystalline cellulose powder Avicel® with nominal molecular weight
6
Page 6 of 34
146
2.1 Dissolution of cellulose Dissolution of cellulose was realized in two stages so as to reduce the
148
aggregation (Rein et al., 2012; Rinaldi, 2011). First, cellulose was dispersed in a
149
specific volume of the co‐solvent. Next, the required volume of EMIMAc was
150
added into the suspension, and the mixture was heated to 90 °C using magnetic
151
stirring during about 10 min. Subsequently, the obtained solution was cooled to
152
room temperature and was not further treated.
154
2.2 Equipment and characterization methods
us
153
cr
ip t
147
The morphology of the solutions was investigated by cryogenic
156
transmission electron microscope (cryo‐TEM) T12G2 (FEI Co., Netherlands).
157
Vitrified samples for cryo‐TEM (Talmon, 1996) were prepared in a controlled
158
environment vitrification system (CEVS). (Bellare, Davis, Scriven, & Talmon, 1988)
M
an
155
SLS measurements were carried out with a goniometer BI‐200SM
160
(Brookhaven Instruments Corp., USA) using an argon‐ion laser (wavelength 514.5
161
nm). Data obtained at several concentrations (0.4 – 4 wt. %) and scattering angles
162
(30 – 150°) were evaluated by the common Zimm plot yielding the root‐mean‐
163
square Rg, weight‐average molar mass Mw and the second osmotic virial coefficient
164
A2 (Debye, 1944; Zimm, 1945). The indices of the solvent refraction no (no=1.44 for
165
the investigated DMF/EMIMAc mixture with molar ratio 9:1) and the mass
166
concentration (c) increment of the solution refraction index dn/dc (for investigated
167
DMF/EMIMAc mixture dn/dc=0.061±0.06 cm3/g) were measured by a differential
168
refractometer BI‐DNDC (Brookhaven Instruments Corp., USA).
Ac ce p
te
d
159
169
Electrical conductivity of solvent mixtures was measured with a Precision
170
Impedance Analyzer Agilent 4294A (Agilent Technologies) at 0.5 V and 10 kHz.
171
Each sample was analyzed in triplicate and the mean values were plotted (average
172
measurements error about 12%). 7
Page 7 of 34
X‐ray diffractometry of the samples was performed using a small/wide‐
174
angle diffractometer (Molecular Metrology SAXS/WAXS system) equipped with a
175
sealed microfocus tube (MicroMax‐002 + S) emitting CuKα radiation, two Göbel
176
mirrors, three‐pinhole slits, and generator that is powered at 45 kV and 0.9 mA.
177
The scattering patterns were recorded by a two‐dimensional position sensitive
178
wire detector positioned 150 cm behind the sample. The solutions were sealed in
179
thin‐walled glass capillaries about 2 mm in diameter and 0.01 mm wall thickness,
180
and measured under vacuum at the ambient temperature. The scattered intensity
181
I(q) was recorded at the interval 0.07
