Carbohydrate Polymers 81 (2010) 668674

Contents lists available at ScienceDirect

Carbohydrate Polymers
journal homepage: www.elsevier.com/locate/carbpol

Dissolution behaviour and solubility of cellulose in NaOH complex solution

Shuai Zhang a,b , Fa-Xue Li a , Jian-yong Yu c, , You-Lo Hsieh b
a
College of Textiles, Donghua University, Shanghai 201620, China
b
Fiber & Polymer Science, University of California, Davis, CA 95616, USA
c
Modern Textile Institute, Donghua University, Shanghai 200051, China

a r t i c l e i n f o a b s t r a c t

Article history: Aqueous mixture of NaOH/urea/thiourea at a 8/8/6.5 composition and pre-cooled at 10 C readily
Received 19 January 2010 dissolved cellulose to produce stable solutions at relatively high concentrations. The exothermic dis-
Received in revised form 13 March 2010 solution process was favored at 2 to 0 C. Aqueous NaOH/urea/thiourea solution as non-derivatizing
Accepted 17 March 2010
solvent broke the intra- and inter-molecular hydrogen bonding of cellulose and prevented the approach
Available online 25 March 2010
toward each other of the cellulose molecules, leading to the good dispersion of cellulose to form solution.
The strength of the solvent network structure as well as the interaction between cellulose and solvent
Keywords:
decreased as a function of increasing solution temperature. In the semi-dilute entangled solutions (>3.5%
Cellulose
Direct dissolution
concentration), the entropy-driven gelation occurred, and the gel temperature dropped with increasing
NaOH complex cellulose contents in the solution. The NaOH/urea/thiourea/H2 O was demonstrated to be the most pow-
Gelation erful solvent among all aqueous NaOH solutions and this novel solvent does not degrade cellulose even
Stability after storage times of up to 1 month.
2010 Elsevier Ltd. All rights reserved.

1. Introduction NH3 /NH4 SCN (Cuculo, Smith, Sangwatanaroj, Stejskal, & Sankar,
1994), N-methylmorpholine-N-oxide monohydrate (NMMO)
Cellulose is the most important skeletal component in plants (Loubinoux & Chaunis, 1987) and ionic liquid (Swatloski, Spear,
and the most abundant renewable material in nature (Dogan & Holbrey, & Rogers, 2002) are limited to laboratory scale applica-
Hilmioglu, 2009; Klemm, Heublein, Fink, & Bohn, 2005). The non- tions due to volatility, toxicity and high cost. Among the solvents
thermal plastic nature and insolubility in most common solvents developed, the NMMO/H2 O system is the most powerful in
remain to be the challenges in processing cellulose for effective attaining exceedingly high concentration solutions and has been
utilization. In fact, dissolution of cellulose without chemical mod- commercialized to produce Tencel or Lyocell bers. However, the
ication or derivatization is difcult to achieve because of the NMMO/H2 O system also has disadvantages of requiring high tem-
rigid long-chain and strongly inter-molecular and intra-molecular perature for dissolution and antioxidants to avoid side reactions of
hydrogen-bonded structure in cellulose (Fink, Weigel, Purz, & solvents, causing degradation of cellulose, and high costs. Thus, it
Ganster, 2001; Zhang, Yang, & Liu, 1999). More common than is not suitable for complete replacement of the viscose technology.
not, cellulose needs to be activated or made accessible to be Kamide, Okajima, Matsui, and Kowsaka (1984) and Isogai and
dissolved, even though these notions are not clearly dened. Tra- Atalla (1998) have systematically studied the solubility of micro-
ditional production of regenerated cellulose bers and lms has crystalline cellulose and steam-exploded cellulose in aqueous
been largely based on the viscose or the cuprammonium technolo- NaOH systems in which native cellulose pulps have very limited sol-
gies, that generate hazardous environmental pollution (Kamide & ubility. Recently, Zhangs group (Cai & Zhang, 2006; Weng, Zhang,
Saito, 1986). Thus, identifying new solvent systems for cellulose Ruan, Shi, & Xu, 2004; Zhang, Ruan, & Gao, 2002; Zhang, Ruan,
processing would help to reduce these environmental problems. & Zhou, 2001; Zhou, Zhang, & Cai, 2004) successfully developed
Some solvents for cellulose, such as cuoxam, cuen, and cadoxen aqueous NaOH solution systems with either urea or thiourea for
as well as lithium chloride/N,N-dimethylacetamide (LiCl/DMAc) cotton linter dissolution. The optimal solubility was found with
contain metal complexes (McCormick, Callais, & Hutchinson either 7/12/81 NaOH/urea/H2 O or 9.5/4.5/86 NaOH/thiourea/H2 O
Jr, 1985). Others including N2 O4 /N,N-dimethylformamide in compositions. Both solvent systems were inexpensive and less
(DMF) (Philipp, Nehls, Wagenknecht, & Schnabelrauch, 1987), toxic, but the precise structure of the solvent complex involved is
still not clearly understood.
Recently, we developed a new complex aqueous solvent that
Corresponding author. Tel.: +86 62378390; fax: +86 62378392. consisted of NaOH, urea and thiourea and was capable of dissolv-
E-mail address: yujy@dhu.edu.cn (J.-y. Yu). ing cellulose quickly when pre-cooled to temperatures between 8

0144-8617/$ see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carbpol.2010.03.029
 S. Zhang et al. / Carbohydrate Polymers 81 (2010) 668674 669

and 12 C (Jin, Zha & Gu, 2007; Zhang, Li, Yu, & Gu, 2009; Zhang, 10 C with vigorous stirring for 10 min. The solvent and solution
Li, Yu, Fu, & Gu, 2009). With the aim to gain basic understanding of were injected into NMR tubes using syringes and solid-state 13 C
the rapid dissolution of cellulose in this solvent, the present work NMR measurements were carried out using a CP/MAS unit (Bruker
was to further clarify the dissolution behaviour and solubility of AV 400, Switzerland, magnetic eld 9.4 T) at room temperature.
cellulose based on this novel aqueous NaOH/urea/thiourea plat- The refractive index of the cellulose solutions was measured by
form. The behaviour of cellulose solutions among the three aqueous an Abbe refractometer at 20 C. The zero-shear viscosity of the solu-
NaOH/urea/thiourea, NaOH/thiourea and NaOH/urea solvent sys- tions was obtained by measuring the shear viscosity value using a
tems was also compared. rheometer (ARES-RFS, TA Instruments, USA) at varying shear rates
from 0.01 to 100 s1 and extrapolating to zero-shear rate. The gela-
2. Materials and methods tion temperature was also measured on the rheometer from 10 to
100 C.
2.1. Materials The dependence of the viscosity on the storage time in aque-
ous NaOH/urea/thiourea solvent was carried out immediately upon
Cellulose used in this study was cotton linter, with a degree of dissolution of cellulose and when the solution became transparent.
polymerization of 520 and denoted as C520, supplied by Xinxiang The storage time was calculated from when cellulose was placed
Bailu Chemical Fibers Co. Ltd. (Henan, China). All cellulose sam- in the solvent. The viscosities of the cellulose solutions were mea-
ples were shredded into powder to pass 40 mesh and dried under sured with a Ubbelohde capillary viscometer at 25 0.1 C. Relative
vacuum at 70 C for 24 h. All chemicals were of analytical grade viscosity (r ), the ratio of the dynamic viscosity of the solution ()
and used as received. All concentrations of cellulose and solvent to that of the pure solvent (s ), was used. To avoid aggregation of
mixtures were expressed in weight percent unless specied. For cellulose in the solution, a relatively low cellulose concentration of
dual or tri-component solutions, the concentration of each compo- 3 103 g/ml was examined. In addition, the viscosity of the cel-
nent was expressed in number, such as x/y/z NaOH/urea/thiourea lulose in cadoxen at 25 0.1 C was measured. After sealed and
representing x wt% of NaOH, y wt% of urea and z wt% of thiourea. storage at 25 C for 30 days, cellulose solution was taken out and
neutralized with 1 mol/l H2 SO4 to precipitate cellulose. Isolated cel-
2.2. Preparation of cellulose solutions lulose was washed with water and acetone, and vacuum-dried for
24 h. Its viscosity-average molecular weight (M  ) can be calculated
Three aqueous mixtures with the optimal mass ratios of urea, with equation (Brown & Wiskston, 1965):
thiourea and NaOH reported in the literature were used as sol-
vent systems in this study. They were 7/12 NaOH/urea (Cai & [] = 3.85 102 M
0.76 (1)
Zhang, 2006), 9.5/4.5 NaOH/thiourea (Zhou et al., 2004) and 8/8/6.5
NaOH/urea/thiourea (Jin et al., 2007). Two dissolution procedures 3. Results and discussion
were employed. For aqueous NaOH/urea and NaOH/thiourea sys-
tems, the solvents were pre-cooled to 12 to 10 C for 10 min 3.1. Effect of NaOH concentration on solubility
before cellulose was added (referred to as method I). In aqueous
8/8/6.5 NaOH/urea/thiourea, cellulose was dispersed after the sol- The effects of NaOH concentration on cellulose dissolution
vent was pre-cooled to 10 C, followed by vigorously stirring for kinetics was evaluated from 1 to 10% NaOH in 8/6.5 urea/thiourea
3 min. The solution temperature was then maintained between 2 aqueous solutions. The solubility of 7% cellulose improved signif-
to 0 C using a saltice bath and the mixture was vigorously stirred icantly with increasing NaOH concentrations and peaked at 8%
for 7 min to obtain transparent cellulose dope (referred to as method NaOH with 91% solubility (Fig. 1).
II). This is probably due to the concentration-dependent size of
All prepared cellulose dopes were subjected to centrifugation NaOH-water hydrates (Yamashiki et al., 1988). Crystalline cellu-
at 5000 rpm for 10 min at 510 C in order to exclude the minute lose chains are very densely packed, with an inter-sheet distance
amount of undissolved cellulose and to help degasify. The appar- of about 10 and a crystallite diameter of only 10 nm. At low
ent viscosity of each prepared solution was determined at 20 C alkali concentrations, the hydrodynamic diameters of NaOH-water
with a rotational viscometer (NDJ-1, Cany Precision Instruments hydrates may be too large to penetrate into the crystalline region
Co., China).

2.3. Experimental design

Solubility of cellulose in aqueous NaOH/thiourea, NaOH/urea

and NaOH/urea/thiourea systems was studied at 17.3% (w/w) cel-
lulose concentrations. Dissolution of cellulose was observed by
the disappearance of the highly crystalline cotton linter using a
polarized light microscope (PLM, Olympus RX51-P, Japan). In the
aqueous NaOH/urea/thiourea system, the effects of NaOH on the
dissolution of cellulose were observed at 10 C using 110% NaOH
at a constant 8% urea and 6.5% thiourea composition. The effects of
solvent temperatures from 15 to 15 C on the appearance of the
solutions was recorded by photography. The solubility (Sa ) char-
acterized by the quantity of cellulose dissolved was expressed as
Sa = [W1 /(W1 + W2 )], where W1 was the weight of the dissolved cel-
lulose and W2 was the combined weight of the gel and insoluble
cellulose isolated by centrifugation.
The solvent for 13 C NMR measurements, i.e., 8/8/6.5
Fig. 1. Dependence of cellulose solubility values on NaOH in aqueous 8%/6.5%
NaOH/urea/thiourea in D2 O, was prepared in a 5-mL tube. A urea/thiourea. The insets are the PLM images of cellulose solutions with specic
7 wt% cellulose solution was prepared by dispersing cellulose at NaOH contents.
670 S. Zhang et al. / Carbohydrate Polymers 81 (2010) 668674

of cellulose. It is also possible that there is not enough NaOH to

dissolve cellulose at very low NaOH concentrations. At NaOH con-
centrations higher than 1820%, on the other hand, NaOH prefers
to stay close to the cellulose chains, forming Na/cellulose crystals
as in the mercerization process. In that case, cellulose will not be
dissolved either.
It has been well recognized that NaOH play a leading role during
the dissolution of cellulose in NaOH/urea/thiourea aqueous solu-
tion, that is, NaOH can penetrate not only between crystallites but
also into the crystallites to destroy inter- and intra-hydrogen bonds
between cellulose molecules. Urea and thiourea, however, function
as hydrogen bond donor and receptor to prevent the re-association
of cellulose molecular chains (Zhou & Zhang, 2000). Since none of
the individual component of this solvent system can dissolve native
cellulose, a synergic interaction of NaOH, urea and thiourea was
thought to exist during the dissolving of cellulose (Jin et al., 2007).
At 8/6.5 urea/thiourea composition, the synergic effect was optimal
at 8% NaOH at 10 C. At the optimal 8/8/6.5 NaOH urea/thiourea Fig. 3. Arrhenius plots of temperature dependence of Sa value for 6% cellulose solu-
tion in the range from 10 to 10 C.
ratio, NaOH-water hydrates can penetrate the amorphous area to
solvate cellulose then destruct the neighboring crystalline regions.
molecules (Hinteraction ); and thirdly, the mixing of the solvation
3.2. Effect of pre-cool temperature on cellulose solubility of the polymer molecules to give an innitely diluted solution
(Hmixing ). The only endothermic term in the equation is Hfusion
The dissolving ability of solvent systems has been shown to associated with breaking the hydrogen bonds in the crystalline
improve with decreasing temperatures (Cai & Zhang, 2006; Kamide regions. All other terms relate to interactions between cellulose
et al., 1984; Zhang et al., 2002; Zhang et al., 1999; Zhou et al., 2004). hydroxyl groups and the solvent system and are exothermic. The
Cellulose has shown to be highly swollen and even soluble in sol- overall process of cellulose dissolution is exothermic and is favored
vents with 610% NaOH at temperatures from 15 to 4 C. Kunze by lower temperature.
and Fink (2005) also stated NaOH to be similarly active at decreased The change in energy of cellulose from solid to solution can be
temperatures as at increased concentrations at room temperature. illustrated quantitatively by the apparent activation energy Ea,s . The
As seen in Fig. 2, the solubility of cellulose in aqueous relationship between Sa and temperature can be described by the
NaOH/urea/thiourea solution was enhanced with decreasing sol- Arrhenius equation (Su & Puls, 1999):
vent temperatures, reaching the highest solubility at 10 C.
Ea,s
The more favorable dissolution at lower temperatures has been ln Sa = ln A (3)
RT
analyzed by the total enthalpy of cellulose dissolution as given by
(Marson & El Seoud, 1999; Wang & Deng, 2009): where Ea,s is the apparent activation energy of dissolution (kJ/mol),
R is the molar gas constant (0.008314 kJ mol1 K1 ), T is the abso-
HDissolution = Hfusion + Htransition + Hinteraction + Hmixing lute temperature (K), and A is a pre-exponential factor. The Ea,s
(2) could be obtained from the slope of ln Sa vs. 1/T plot using linear
least-squares analysis (Fig. 3). Interestingly, the calculated Ea value
In that, the dissolution of crystalline cellulose into molecular dis- was 49.77 kJ/mol in the 10 to 10 C range, as a result of the disso-
persion is divided into several stages: rstly, the transition of the lution at low temperature. The negative apparent activation energy
solid polymer to a hypothetical, highly elastic liquid state which in our system suggests that the soluble form of cellulose has a lower
corresponds to disintegration of the crystalline regions (Hfusion ) enthalpy than the solid cellulose, and the dissolution of cellulose in
and transition of the amorphous regions from a vitreous to a highly aqueous NaOH/urea/thiourea solution pre-cooled to low tempera-
elastic state (Htransition ); secondly, the solvation of the polymer ture may be described as an entropy-driven process as previously
proposed (Cai et al., 2007).
In addition, temperature appeared to play an important role in
affecting the solvent properties. As the temperature lowered, the
network of the solvent hydrates became stronger, probably due to
the increasing hydrogen bonding strength. Take NaOH hydrates for
example, the Na+ and OH ions are surrounded by a cage of water
molecules because of ionization equilibrium and hydration. That is,
the Na+ ion exists as [Na(H2 O)m ]+ , and the OH ion as [OH(H2 O)n ]
at sub-zero temperatures (Cdric, Tatiana, Patrick, & Olivier, 2001).
At room temperature, the fast exchange between bulk and coor-
dinated water molecules cannot sustain such [Na(H2 O)m ]+ and
[OH(H2 O)n ] structures. When cellulose is dissolved at low temper-
atures, this strong network of solvent hydrates cleave mainly the
closely chain packed cellulose chains through the formation of new
hydrogen bonds between cellulose and NaOH hydrates. Meanwhile,
urea and thiourea associate with the cellulose hydroxyl groups and
prevent the aggregation of cellulose chains through the inter- and
intra-molecular hydrogen bonds association. This behaviour is sim-
Fig. 2. Dependence of cellulose dissolution on solvent temperature. The insets ilar to the cuprammonium solvation of cellulose, in which ammonia
present the state of the solvent under different temperature. molecules coordinated to copper are hydrogen-bonded to cellulose
 S. Zhang et al. / Carbohydrate Polymers 81 (2010) 668674 671

Fig. 5. 13 C NMR spectra of (a) aqueous NaOH/urea/thiourea solvent and (b) 6% cel-
Fig. 4. Dependence of cellulose solubility values on dissolving methods. The insets lulose solution. (c) The solid-state CP/MAS 13 C NMR spectrum of raw cellulose.
present the PLM pictures of cellulose solution under different dissolving method.

as blockers. When the temperature is raised, the hydrogen bonds aids, the solvent appeared to lose its dissolution ability with the
are weaker, and the network of hydrates is gradually destroyed elevation in temperature and the cellulose solution would easily
(Egal, Budtova, & Navard, 2007). suffer gelation with prolonged stirring.
It also could be seen from inset images in Fig. 2 that crystal pre- To improve dissolution of cellulose in aqueous
cipitation appeared during pre-cooling. It was interesting that the NaOH/urea/thiourea, the optimal dissolution temperature of
long, needle-like crystal precipitated out of the solvent at the opti- 10 C observed previously was combined with the most stable
mal pre-cool temperature (10 C) was tested to be thiourea, not solution at 0 C reported by others (Jin et al., 2007) as method
urea, NaOH or their mixtures. The amount of thiourea precipitation II. 6% cellulose was dispersed in solvent pre-cooled to 10 C,
and the ice formation increase with the further decreases in tem- followed by vigorous stirring for 3 min. After 3 min, most of the
perature. Consequently, the solubility of cellulose also decreases, cellulose was dissolved while the temperature of the solution
because the concentration and the composition ratio of thiourea in increased to approximately 0 C. Further dissolution at 2 to 0 C
the solvent is changed signicantly the concentration and the com- for another 7 min reached a high 96% dissolution or Sa = 0.96. The
position ratio of thiourea in the solvent is changed signicantly. In solution was more transparent and rather easy to lter compared
fact, the entire solvent would become solid ice-block at 15 C and with solution prepared at room temperature or by method I. At a
could not be used to dissolve cellulose. The temperature at which higher cellulose concentration of 7%, continuing dissolution at 2
thiourea precipitated out was also found to be higher in any other to 0 C also increased the solubility to 0.91 from 0.72 by method
NaOH/urea/thiourea solvent composition. It can be concluded that I. Therefore, method II was proven to be the more efcient and
crystal separating out is not only caused by decreased solubility of time-saving approach to dissolve cellulose and is very suitable for
thiourea but also inuenced by interaction between compositions preparing more concentrated cellulose solutions.
of solvent. This interaction is being studied in detail to be reported
later. 3.5. Structure of the cellulose solution

3.3. Effect of stirring rate on cellulose solubility Fig. 5 shows the 13 C NMR spectrum of cellulose in deuter-
ated 8/8/6.5 NaOH/urea/thiourea. Five major peaks were sharp
The stirring rate has also been found to be of great importance and readily identied in the spectra as cellulose I. The chemical
to prepare solutions of higher cellulose concentrations. Cellulose shifts at 104.1 ppm, 79.5 ppm, 75.8 ppm, 74.3 ppm, and 61.4 ppm
solubility increased with increasing stirring rates. It was found that for C1, C4, C3/C5, C2 and C6, respectively. This indicates that the
certain amount of cellulose could be dissolved in the pre-cool sol- NaOH/urea/thiourea aqueous solution is indeed a good solvent
vent even without any stirring. for cellulose (Cai, Liu, & Zhang, 2006). The C-4 chemical shift for
It indicated a rapid dissolution of cellulose in aqueous raw cellulose (Fig. 5c) is located at 88.9 ppm, which is assigned
NaOH/urea/thiourea solution. However, the remained cellulose to intra-molecular hydrogen bonds (C-3H O-5 ) (Zhang et al.,
cannot be dissolved completely without stirring or under low rate 2002). It is obvious that the chemical shift of the C-4 peak shifts
stirring. The increasing shear stress from high stirring rates is upeld compared with that for raw cellulose, which indicates
thought to help to improve solvent diffusion and break off partially that the intra-molecular hydrogen bonds of cellulose have been
dissolved chains from the solid surfaces to accelerate dissolution. destroyed. The chemical shifts of these peaks are very similar to
those for some good solvents of cellulose, such as cadoxen (Nehls,
3.4. Effect of dissolution method on cellulose solubility Wagenknech, Philipp, & Stscherbina, 1994), DMAc/LiCl (Hattori,
Koga, Shimaya, & Saito, 1998), NMMO/DMSO (Gagnaire, Mancier,
The solubility of cellulose under the two dissolution methods & Vincendon, 1980), NaSCN/D2 O (McCormick et al., 1985) and
was studied over time. In dissolution method I, cellulose was dis- BMIMCl (Moulthrop, Swatloski, Moyna, & Rogers, 2005). More-
persed in pre-cool solvent and stirred at room temperature. Nearly over, no new peaks appeared in the 13 C NMR spectrum, indicating
80% of the cellulose dissolved in 3 min and the solubility increased the absence of derivatization of cellulose. Therefore, the aqueous
slightly to 85% after 10 min (Fig. 4). However, dissolution drasti- 8/8/6.5 NaOH/urea/thiourea dissolve cellulose directly rather than
cally reduced to 77% and 65% after 15 and 20 min, respectively. The through a cellulose derivative.
undissolved cellulose solids appeared to be surrounded by more One of the earliest techniques used to study polymer dissolu-
viscous solution. With vigorous stirring without further cooling tion was refractometry (Ueberreiter, 1968). This technique is based
672 S. Zhang et al. / Carbohydrate Polymers 81 (2010) 668674

Fig. 6. The dependence of refractive index (n) and zero-shear viscosity (0 ) on Fig. 7. Phase diagram of gel formation in the cellulose solution. The insets present
cellulose concentration (C) at 20 C. the sol and gel states of cellulose solutions.

on polymer concentration dependence of refractive index dur- be a result of altered solvent components caused by chemical reac-
ing dissolution. The zero-shear viscosity (0 ) is also always used tions among NaOH, urea and thiourea at elevated temperatures (Cai
to investigate the structure properties of polymer solution. Fig. 6 & Zhang, 2006), losing the ability to prevent self-association of cel-
shows the relations between n and 0 of solution and cellulose lulose macromolecules. The random self-association of cellulose
concentration (C). It is interesting to nd that the transitions of chains having exposed hydroxyl in the aqueous solution promotes
refractive index and zero-shear viscosity curves occurred at almost the entanglement networks at elevated temperatures.
the same cellulose content of 3.5%. Thermodynamically, gelation occurs only when the free energy
This 3.5% concentration was therefore used to divide the change (G) is negative at a given constant temperature T, or
semi-dilute unentangled regime below from the semi-dilute entan- G = H TS, where H and S are the changes of enthalpy
gled regime above. The n and 0 values rise in different rates and entropy, respectively. Gelation is a thermodynamic process
with increasing concentrations in both regions. In the C < 3.5% as it involves energy exchanges between the gel system and the
semi-dilute unentangled regime, n and 0 increase linearly with surroundings, resulting in the solgel transition of the system. It
increasing concentrations. The thermodynamics of cellulose chains should be taken into consideration both the process of the cellu-
in this region are different from that of linear exible polymers lose molecular chains self-associated in the aqueous solution and
as can be described by the Rouse model (Heo & Larson, 2005). the action of small molecules including water, thiourea, urea and
In the semi-dilute entangled regime (C > 3.5%), n and 0 increase some ions. When cellulose is dissolved in an aqueous solution, the
more sharply which indicates some aggregation of cellulose chains small solvent molecules interact with cellulose molecules to give
caused by increasing chance of collision among cellulose molecules. some degree of order in the solution. When the cellulose solu-
This concentration dependence is consistent with that observed tion is heated to gel, the heat provides the activation energy to
for cellulose solutions in ionic liquids and the MCC solutions in destroy the interaction between the solvent molecules and cel-
LiCl/DMAc (Kuang, Zhao, Niu, Zhang, & Wang, 2008; Tamai, Aono, lulose. The exposed hydroxyls form random junctions between
Tatsumi, & Matsumoto, 2003). cellulose molecules from self-association, forming gel network.
Meanwhile, the solvent molecules were released and move ran-
3.6. Gelation behaviour of cellulose solution domly. The changes in entropy and enthalpy for the system are
dened as:
Cellulose dissolved in NaOH gels easily and is not practical for
processing into membranes or bers. Moreover, thermal gelation H = Hself-association + Hdemixing (4)
of cellulose in aqueous NaOH/thiourea and NaOH/urea solutions S = Sentanglement + Sdemixing (5)
has also been reported. Such gelation was believed to be due to
self-aggregation of the cellulose chains in the solution at elevated where Hself-association is from self-association of the hydroxyl
temperatures (Cai & Zhang, 2006; Weng et al., 2004). groups in gelation and is exothermic (Hself-association < 0);
The thermally induced gelation of cellulose solution in aque- Hdemixing stands for destroying the interaction between the
ous NaOH/urea/thiourea solvent was investigated by rheometry. small molecules and cellulose and is endothermic (Hdemixing > 0);
Fig. 7 displays the phase diagram of gel formation in the pre- Hentanglement refers to the aggregation of cellulose molecule chains
pared cellulose solutions. Gelation of solutions was observed with and should be negative (Sentanglement < 0); while the release of
increasing temperatures from 38 to 81 C. Below 3.5% cellulose con- small molecules from the chain gives an increase in the entropy
centration, thermally induced gelation behaviour was not observed of the system (Sdemixing > 0).
from 20 to 90 C. This suggests that the extent of cellulose chains The gelation occurred at a high temperature with the endother-
aggregation is very limited in relatively dilute or semi-dilute unen- mic heat H > 0 (the absolute value of Hdemixing is greater than
tangled solutions. The dropping temperatures of gelation (Tgel ) with that of Hself-association ), as conrmed by the preliminary results
increasing cellulose concentrations above 3.5% can be explained by of differential scanning calorimetric analysis, which revealed a rel-
the progressively increased number of junction zones between the atively broad endothermic peak. If the gelation process can take
cellulose chains in the semi-dilute entangled solution. However, place, the entropy (S) should be more positive to meet the G < 0
the gel formed could not be re-dissolved even when the temper- requirement. As we noticed that, compared with the energy needed
ature was cooled to 10 C, the optimal solvent temperature to in the destruction of the cellulosesolvent interaction structure,
dissolve cellulose. Thus, the irreversible gelation from heating may the energy for the formation of self-association (its intrinsic ten-
 S. Zhang et al. / Carbohydrate Polymers 81 (2010) 668674 673

unentangled to entangled regions at 3.5% concentration. Whereas

gelation was absent in the semi-dilute unentangled concentration
range from 20 to 90 C, thermally induced gelation was observed
at lower temperatures at cellulose concentrations above 3.5%. This
unique gelation process was thought to be an entropy-driven pro-
cess. Moreover, this aqueous NaOH/urea/thiourea system exhibits
higher solubility capacity for cellulose when compared with
either aqueous NaOH/thiourea or NaOH/urea systems and the as-
prepared cellulose solution is rather stable. These ndings conrm
a direct and efcient process for preparing stable cellulose solu-
tions for lm and ber processing, a promising substitution for
the viscose technology that involves hazardous by-products. This
dissolution approach offers a convenient, environmentally friendly
process platform and potential new strategies for expanding appli-
cations or chemical modication of cellulose, the natures most
abundant polymer.

Fig. 8. Comparison of the solubility values and apparent viscosity (a ) of cellulose
in different NaOH complex solvent systems. References

Brown, W., & Wiskston, R. (1965). Viscosity-molecular weight relationship for cel-
dency) is much lower (Weng et al., 2004), so the gelation process is lulose incadoxen and hydrodynamic interpretation. European Polymer Journal,
an endothermic process (H > 0) on the whole. For the same rea- 1, 110.
Cai, J., Liu, Y. T., & Zhang, L. N. (2006). Dilute solution properties of cellulose in
son, Sdemixing should be more signicant than Sentanglement and
LiOH/urea aqueous system. Journal of Polymer Science Part B: Polymer Physics, 44,
it should exhibit as a positive value. It can be concluded that the 30933101.
gelation of cellulose in the aqueous solution upon heating is an Cai, J., & Zhang, L. N. (2005). Rapid dissolution of cellulose in LiOH/urea and
NaOH/urea aqueous solutions. Macromolecular Bioscience, 5, 539548.
entropy-driven process.
Cai, J., & Zhang, L. N. (2006). Unique gelation behavior of cellulose in NaOH/urea
aqueous solution. Biomacromolecules, 7, 183189.
3.7. Comparison of the dissolution power of varied NaOH Cai, J., Zhang, L. N., Chang, C. Y., Cheng, G. Z., Chen, X. M., & Chu, B. (2007). Hydrogen-
aqueous systems and the stability of the prepared solution bond-induced inclusion complex in aqueous cellulose/LiOH/urea solution at low
temperature. ChemPhysChem, 8, 15721579.
Cdric, R., Tatiana, B., Patrick, N., & Olivier, B. (2001). Structure of cellulosesoda
In comparing solubility of cellulose in the different aqueous solutions at low temperatures. Biomacromolecules, 2, 687693.
NaOH solutions (Fig. 8), the aqueous NaOH/urea/thiourea system Cuculo, J. A., Smith, C. B., Sangwatanaroj, U., Stejskal, E. O., & Sankar, S. S. (1994).
A study on the mechanism of dissolution of the cellulose/NH3 /NH4 SCN system.
was obviously most powerful. The apparent viscosity values among Journal of Polymer Science Part A: Polymer Chemistry, 32, 229239.
the three prepared cellulose solutions appeared to be more distinc- Dogan, H., & Hilmioglu, N. D. (2009). Dissolution of cellulose with NMMO by
tively different at higher cellulose concentrations. Moreover, the microwave heating. Carbohydrate Polymers, 75, 9094.
Egal, M., Budtova, T., & Navard, P. (2007). Structure of aqueous solutions of micro-
aqueous NaOH/urea/thiourea solution is capable of dissolving cel- crystalline cellulose/sodium hydroxide below 0 C and the limit of cellulose
lulose with higher molecule weights (M  as high as 2.3 105 , Jin et dissolution. Biomacromolecules, 8, 22822287.
al., 2007) than any other aqueous NaOH systems (Cai & Zhang, 2005; Fink, H. P., Weigel, P., Purz, H. J., & Ganster, J. (2001). Structure formation of regener-
ated cellulose materials from NMMO-solutions. Progress in Polymer Science, 26,
Zhou et al., 2004). This is another evidence to support the better
14731524.
dissolving capacity of the aqueous NaOH/urea/thiourea solution. Gagnaire, D., Mancier, D., & Vincendon, M. (1980). Cellulose organic solutions: A
The stability of cellulose solutions is essential for wet-spinning nuclear magnetic resonance investigation. Journal of Polymer Science Part A:
Polymer Chemistry, 18, 1325.
and thus industrial applications. The relative viscosity (r ) of cellu-
Hattori, M., Koga, T., Shimaya, Y., & Saito, M. (1998). Aqueous calcium thio-cyanate
lose solutions exhibited a 15% reduction after being stored under solutions as a cellulose solvent structure and interactions with cellulose. Polymer
a sealed condition at 25 C for 30 days with most reduction occur- Journal, 30, 4348.
ring within the rst 10 days. The cellulose precipitated from the Heo, Y., & Larson, R. G. (2005). The scaling of zero-shear viscosities of semidilute
polymer solutions with concentration. Journal of Rheology, 49, 11171128.
30-day stored solution showed the M  decreased from 10.2 104 Isogai, A., & Atalla, R. H. (1998). Dissolution of cellulose in aqueous NaOH solutions.
to 8.83 104 , indicating slight degradation of cellulose occur during Cellulose, 5, 309319.
dissolution and storage periods. Therefore, the cellulose solution in Jin, H. J., Zha, C. X., & Gu, L. X. (2007). Direct dissolution of cellulose in
NaOH/thiourea/urea aqueous solution. Carbohydrate Research, 342, 851858.
NaOH/urea/thiourea aqueous system is considered stable at room Kamide, K., Okajima, K., Matsui, T., & Kowsaka, K. (1984). Study on the solubility
temperature for practical processing purposes. of cellulose in aqueous alkali solution by deuteration IR and carbon-13 NMR.
Polymer Journal, 16, 857866.
Kamide, K., & Saito, M. (1986). Light scattering and viscometric study of cellulose in
4. Conclusions aqueous lithium hydroxide. Polymer Journal, 18, 569579.
Klemm, D., Heublein, B., Fink, H.-P., & Bohn, A. (2005). Cellulose: Fascinating biopoly-
The rapid dissolution behaviour and solubility of cellulose in mers and sustainable raw material. Angewandte Chemie International Edition, 44,
33583393.
aqueous NaOH/urea/thiourea solution at a 8/8/6.5 mass ratio com- Kuang, Q. L., Zhao, J. C., Niu, Y. H., Zhang, J., & Wang, Z. G. (2008). Celluloses in an
position and pre-cooled to 10 C was aided by vigorous stirring for ionic liquid the rheological properties of the solutions spanning the dilute and
3 min and further dissolution at 2 to 0 C for 710 min evaluated semidilute regimes. Journal of Physical Chemistry B, 112, 1023410240.
Kunze, J., & Fink, H. P. (2005). Structural changes and activation of cellulose by caustic
by 13 C NMR, optical microscopy, solubility as well as viscometry.
soda solution with urea. Macromolecular Symposia, 223, 175188.
The low temperature, thought to be driven by entropy hydrate Loubinoux, D., & Chaunis, S. (1987). An experimental approach to spinning new
formation kinetics, played an important role in creating the sta- cellulose bers with N-methylmorpholine oxide as a solvent. Textile Research
Journal, 57, 6165.
ble hydrogen-bonded networks between cellulose, NaOH hydrate,
Marson, G. A., & El Seoud, O. A. (1999). Cellulose dissolution in lithium chloride/N,N-
urea, thiourea and water clusters, leading to the dissolution of dimethylacetamide solvent system: Relevance of kinetics of decrystallization
cellulose. The low temperature condition for dissolution of cellu- to cellulose derivatization under homogeneous solution conditions. Journal of
lose avoids the evaporation of the chemical agents and qualies Polymer Science Part A: Polymer Chemistry, 37, 37383744.
McCormick, C. L., Callais, P. A., & Hutchinson, B. H., Jr. (1985). Solution studies of
this aqueous 8/8/6.5 NaOH/urea/thiourea system as a green sol- cellulose in lithium chloride and N,N-dimethylacetamide. Macromolecules, 18,
vent. The cellulose solutions exhibit a transition from semi-dilute 23942401.
674 S. Zhang et al. / Carbohydrate Polymers 81 (2010) 668674

Moulthrop, J. S., Swatloski, R. P., Moyna, G., & Rogers, R. D. (2005). High-resolution Yamashiki, T., Kamide, K., Okajima, K., Kowsaka, K., Matsui, T., & Fukase, H. (1988).
13
C NMR studies of cellulose and cellulose oligomers in ionic liquid solutions. Some characteristic features of dilute aqueous alkali solutions of specic alkali
Chemical Communications, 12, 15571559. concentration (2.5 mol/l) which possess maximum solubility power against cel-
Nehls, I., Wagenknecht, W., Philipp, B., & Stscherbina, D. (1994). Characterization lulose. Polymer Journal, 20, 447457.
of cellulose and cellulose derivatives in solution by high resolution 13 C-NMR Zhang, L. N., Ruan, D., & Gao, S. J. (2002). Dissolution and regeneration of cellulose
spectroscopy. Progress in Polymer Science, 19, 2978. in NaOH/thiourea aqueous solution. Journal of Polymer Science Part B: Polymer
Philipp, B., Nehls, I., Wagenknecht, W., & Schnabelrauch, M. (1987). Carbon-13 NMR Physics, 40, 15211529.
spectroscopic study of the homogeneous sulfation of cellulose and xylan in the Zhang, L. N., Ruan, D., & Zhou, J. P. (2001). Structure and properties of
dinitrogen tetroxideDMF system. Carbohydrate Research, 164, 107116. regenerated cellulose lms prepared from cotton linters in NaOH/urea
Su, C., & Puls, R. W. (1999). Kinetics of trichloroethene reduction by zerovalent aqueous solution. Industrial & Engineering Chemistry Research, 40, 5923
iron and tin: Pretreatment effect, apparent activation energy, and intermediate 5928.
products. Environmental Science & Technology, 33, 163168. Zhang, L. N., Yang, G., & Liu, H. (1999). Recent progress in studies of cellulose in
Swatloski, R. P., Spear, S. K., Holbrey, J. D., & Rogers, R. D. (2002). Dissolution of cellu- China. Cellulose Communications, 6, 8996.
lose with ionic liquids. Journal of the American Chemical Society, 124, 49744975. Zhang, S., Li, F. X., Yu, J. Y., Fu, C. F., & Gu, L. X. (2009). Directly preparation of a mem-
Tamai, N., Aono, H., Tatsumi, D., & Matsumoto, T. (2003). Differences in brane from unsubstituted cellulose in NaOH complex solution. Iranian Polymer
rheological properties of solutions of plant and bacterial cellulose in LiCl/N,N- Journal, 18, 767776.
dimethylacetamide. Journal of the Society of Rheology, Japan, 31, 119130. Zhang, S., Li, F. X., Yu, J. Y., & Gu, L. X. (2009). Novel bers prepared from cel-
Ueberreiter, K. (1968). The solution process. In J Crank, & G. S. Park (Eds.), Diffusion lulose in NaOH/thiourea/urea aqueous solution. Fibers and Polymers, 10, 34
in polymers (pp. 219257). New York, NY: Academic Press. 39.
Wang, Y., & Deng, Y. L. (2009). The kinetics of cellulose dissolution in sodium Zhou, J. P., & Zhang, L. N. (2000). Solubility of cellulose in NaOH/urea aqueous solu-
hydroxide solution at low temperatures. Biotechnology and Bioengineering, 102, tion. Polymer Journal, 32, 866870.
13981405. Zhou, J. P., Zhang, L. N., & Cai, J. (2004). Behavior of cellulose in NaOH/urea aqueous
Weng, L. H., Zhang, L. N., Ruan, D., Shi, L. H., & Xu, J. (2004). Thermal gelation of solution characterized by light scattering and viscometry. Journal of Polymer
cellulose in a NaOH/thiourea aqueous solution. Langmuir, 20, 20862093. Science Part B: Polymer Physics, 42, 347353.