9660
J. Phys. Chem. B 2008, 112, 9660–9662
Abnormal Viscosity Increment Observed for an Ionic Liquid by Dissolving Lithium Chloride Akihiko Takada,* Kenta Imaichi, Tomoyasu Kagawa, and Yoshiaki Takahashi Institute for Materials Chemistry and Engineering, and Department of Molecular and Materials Sciences, Kyushu UniVersity, 6-1 Kasuga-koen, Kasuga, Fukuoka 816-8580 Japan ReceiVed: January 22, 2008; ReVised Manuscript ReceiVed: April 22, 2008
The effect of adding an inorganic salt, lithium chloride, and water on the viscosity of an ionic liquid, 1-nbutyl-3-methylimidazolium chloride (BmimCl), was investigated by shear stress measurements with a rheometer. The shear rate dependence of the viscosity showed shear thinning behavior, which implies that some structure should exist in the liquid and the structure should change at high shear rates. Addition of LiCl enhances the viscosity of BmimCl. The logarithmic value of zero-shear viscosity, η0, of BmimCl increases linearly and largely with increasing added salt content. The increasing rate of the viscosity by addition of LiCl was about 10 times larger than that for an aqueous solution of LiCl. When water is added into BmimCl, viscosity decreased. The increasing rate of the viscosity by addition of LiCl for BmimCl with about 5 wt % of water was almost the same as that for BmimCl without addition of water. Introduction In the past decade or so, a category of new organic liquids called room-temperature ionic liquids, or simply ionic liquids, has been attracting many researchers in a wide variety of academic and industrial fields because of their unique properties.1 It is reported that ionic liquids can dissolve some materials which are difficult to dissolve in conventional solvents. Among them, one of the striking reports is that 1-n-butyl-3-methylimidazolium chloride (BmimCl) easily dissolves cellulose.2 Solution properties of cellulose are interesting and important but not wellstudied yet, especially for relatively high concentrations, since there was no solvent which can easily dissolve cellulose. Thus, cellulose/BmimCl may be a suitable system to advance the study of the solution properties of cellulose over a wide range of concentrations. The very low or almost no vapor pressure of BmimCl, which is also a unique characteristics of ionic liquids, is appropriate for rheological measurements. Therefore, we have started to study the viscoelastic properties of cellulose solutions in BmimCl. As a first step of the study, we examined a few different dissolution conditions including added salt effects to obtain uniform solutions with relatively high concentrations. During the examination, we found that the viscosity of BmimCl increases more than 1 order of magnitude when LiCl is dissolved. Such viscosity increment by dissolution of simple inorganic salts into solvents or mixed solvents has been never reported, as far as we know. In this article, we report the preliminary results of shear viscosity measurements for BmimCl at different LiCl and H2O concentrations. The effect of addition of LiBr is also slightly shown to compare the effect of LiCl. Strictly speaking, the tested system is a three component system, LiCl/H2O/BmimCl, since it is very difficult to completely remove H2O from BmimCl. The melting temperature of pure BmimCl is about 68 °C,3 but the residual water drastically lowers the melting point and glass transition tem* Corresponding author. Phone and Fax: +81-92-583-8821. E-mail:
[email protected].
perature of BmimCl so that it is in the liquid state even below room temperature. Experimental Section BmimCl was synthesized and purified by following the method reported in literatures and further purified by absorbing impurities to charcoal.1,3 Identification of BmimCl was carried out by elemental analysis and 1H NMR. Neither the raw materials nor the solvent was detected. BmimCl was dried in a vacuum oven at around 100 °C for 3 h before use. Anhydrous LiCl (>99.0%) and BrCl (>99.0%) purchased from Kishida Chemical Co. Ltd. were used as received. Weighed amounts of BmimCl, LiCl, BrCl, and deionized H2O were mixed under nitrogen atmosphere and stirred for a few hours at 70 °C in a sealed vessel to obtain uniform samples. The amount of H2O was analyzed by Karl Fischer titration. Concentrations of LiCl and H2O, CLiCl and Cw, respectively, are presented by mol kg-1, since the precise density of the tested system is unknown. Steady shear viscosities of the solutions were measured by an Anton Paar MCR300 rheometer. A cone-plate geometry with 25 mm diameter and 2° cone angle was used. The measurements were carried out at 25.0 ( 0.5 °C and 80.0 ( 0.5 °C under dry nitrogen atmosphere. Results and Discussion Figure 1 shows double-logarithmic plots of shear viscosity, η(γ˙ ), against shear rate, γ˙ , for different CLiCl in BmimCl containing 0.44 mol kg-1 H2O. For all the samples, η(γ˙ ) became constant in a wide range of γ˙ but showed shear thinning behavior at the higher end of γ˙ . The data were reproducible in repeated measurements even without any rest time. Zero-shear viscosity, η0, for each sample was determined as an average value neglecting the data in the shear thinning region. Figure 2 shows double-logarithmic plots of reduced shear viscosity, η(γ˙ )/η0, against reduced shear rate, γ˙ /γ˙ *, where γ˙ * is determined to superpose all data on one curve. In this case, γ˙ * is determined as η(γ˙ *) ) 0.8η0 in order to make the
10.1021/jp800633x CCC: $40.75 2008 American Chemical Society Published on Web 07/23/2008
Viscosity of the BmimCl/LiCl System
Figure 1. Double-logarithmic plots of shear viscosity η(γ˙ ) vs shear rate γ˙ for LiCl solutions with different CLiCl in BmimCl containing 0.44 and 2.2 mol kg-1 H2O. η(γ˙ ) for samples with different water contents are also plotted.
Figure 2. Double-logarithmic plots of reduced shear viscosity, η(γ˙ )/ η0, against reduced shear rate, γ˙ /γ˙ *, where γ˙ * is determined as η(γ˙ *) ) 0.8η0.
estimation easy. For samples whose thinning phenomenon is very weak, the superposition is performed manually. It is clear that all the data compose a single master curve, which denotes that the degree of shear thinning behavior is the same for all the samples irrespective of CLiCl and Cw in the tested range of CLiCl and Cw and that the shear thinning mechanism should be essentially same for all samples. The shear thinning behavior observed in Figures 1 and 2 implies that some structure should exist in the system and the structure is deformable or breakable at high shear rates. In imidazolium-based ionic liquids, the existence of an interaction between the halogen anion and the hydrogen of imidazolium cation was known.4–6 The association of ions or local structure of liquid is reported,7–11 which should be brought about by the interaction of ions. The existence of long-range ordering in the imidazolium-based ionic liquids with halogen counteranions is proposed by Hamaguchi and co-workers,12–17 based on Raman spectroscopy, wide-angle X-ray scattering, and neutron diffraction measurements. Such a long-range ordering could provide the liquid the shear thinning phenomenon. In this stage, we could not conclude what type of structure exists in the system. However, it is expected that the structure in the ionic liquid is distorted by the shear flow deformation and the structure is recovered with a certain relaxation time. When the shear rate is low enough to recover the structure, the shear deformation is applied on the recovered structure. Consequently, its viscosity does not depend on the shear rate in such a low shear region. On the other hand, when the shear rate becomes higher than the rate of the relaxation process for the liquid structure, the shear deformation is applied on the broken or nonrecovered structure, which brings about the shear thinning phenomenon. From the applicability of the superposition of all data on one curve, the essential structure and the essential mechanism of the shear thinning should not be changed by addition of LiCl and water.
J. Phys. Chem. B, Vol. 112, No. 32, 2008 9661
Figure 3. CLiCl dependence of logarithmic value of zero-shear viscosity for BmimCl with different water contents.
Figure 4. Relative viscosity η0/η0(CLiCl)0) for BmimCl semilogarithmically plotted against CLiCl at 25 °C. η0/η0(CLiCl)00) vs CLiCl at 80 °C (b) and η0/η0(CLiCl)0) vs CLiBr at 25 °C (/) are also plotted. Besides them, the CLiCl dependence of η/η(CLiCl)0) for water at 25 °C is also shown as the dashed line (ref 21).
We tried to carry out small-angle X-ray scattering measurement on our samples. However, because of the large absorption of X-rays by halogen atoms, we could not obtain enough large scattering intensity to analyze. We also tried to carry out smallangle neutron scattering measurement and preliminarily obtained data for partially deuterized BmimCl with 0.34 mol kg-1 water. The data showed that, in the scattering vector q range higher than around 0.1 nm-1, scattering intensity gradually increased with decreasing q, which implies a certain large structure should exist in the ionic liquid. In this stage, we could not analyze the data in detail because the measurement time was not long and the measurement q range was limited. Figure 3 shows the CLiCl dependence of η0 at four different Cw. The logarithmic value of viscosity for all samples seems to linearly increase with increasing CLiCl. η0 at high CLiCl becomes more than 10 times larger than those without LiCl. When water content increased, η0 decreased, but the dependency on CLiCl seems to be unchanged. Figure 4 shows the reduced plots of η0/η0(CLiCl)0) against CLiCl. All the data in Figure 3 are located on a straight line even though some points are scattered. This figure implies that, in this water content range, the logarithmic value of viscosity linearly increases with increasing CLiCl and the essential mechanism of the drastic viscosity change should be the same for all the samples irrespective of water content in the tested region. In this figure, the CLiCl dependence of η/0η0(CLiCl)0) for water is also shown as a dashed line. When the viscosity increment for water with that for BmimCl is compared, it is clearly shown that the viscosity increment for BmimCl is drastically large. Takamuku and co-workers reported that phase separation is induced when some salts are added to miscible solutions, even for well-miscible combinations such as H2O and ethanol.18,19 Our ionic liquid system is also a ternary system LiCl/H2O/
9662 J. Phys. Chem. B, Vol. 112, No. 32, 2008 BminCl containing LiCl salt. We checked that the phase separation did not appear for all samples under no shear condition but did not check it under shear flows. Therefore, we can take phase separation as one of the possibilities for the cause of the shear thinning behavior. However, we think that the possibility should be very low because phase separation normally does not give such a large viscosity change. Phase separation gives at most the same order of viscosity change as the viscosity of the matrix phase.20 One of the possibilities of the mechanism of the drastic viscosity change is that addition of LiCl may change the number and/or size of the large-scale association structure in the ionic liquid. Another possibility is that addition of LiCl may change the binding constant, which changes energy to break the association structure and/or the relaxation time of the association structure. These factors should change the dissipation rate of energy applied on the sample as work, in other words, change the viscosity. In this stage, it is difficult to conclude which is the main factor for the drastic increase of viscosity by addition of the salt. Addition of H2O decreased the viscosity as opposite to addition of LiCl. It is also difficult to explain the role of H2O in this stage. H2O may give an effect opposite to LiCl. These viscosity measurements were carried out at 25 °C. BmimCl has a melting point around 68 °C. In our sample, BmimCl contains a small amount of H2O, which makes the melting point lower. However, the system may be under a supercooling state. We also measure the viscosity change with CLiCl at 80 °C, which is shown in Figure 4. The increment of viscosity is not steeper than that of 25 °C. However, it still has a very strong CLiCl dependence. The difference of CLiCl dependence between 25 and 80 °C comes from the different temperature dependence of the viscosity for different CLiCl. Preliminarily we think that the difference of the temperature dependence comes from a change in the glass transition temperature by CLiCl. We tried with a few other inorganic salts such as NaCl and KCl. These salts did not dissolve into BmimCl. We also tried to mix aqueous solutions of the above salts with BmimCl, but the salts precipitated at low water contents. However, LiBr could be dissolved in BmimCl. We also measured the viscosity of the LiBr solution of BmimCl, which is shown in Figure 4. The relative viscosity also drastically increased with increasing LiBr concentration, CLiBr. These results shows that Li+ plays an important role of dissolution of salts in the ionic liquid. As is well-known, Li+ has a small radius and its strong interaction with solvent, which gives the good solubility of Li salts. As for the effect on the viscosity, both Cl- and Br- ions should play a similar role in the liquid.
Takada et al. Several other kind of ionic liquids, which were given by Kohei Chemical Company as codes of IL-P14, IL-A2, and ILA4 for the pyridinium type and aliphatic type of ionic liquids, were also tested, but none of them dissolved inorganic salts. Therefore, we tentatively conclude that the abnormal viscosity increment reported in this article is not common for ionic liquids but is a unique property for LiCl/H2O/BmimCl or LiBr/H2O/ BmimCl systems. Studies on other imidazolium ionic liquids will be carried out. More systematic studies of this system including a few fundamental physical properties are now in progress, and the results will be reported in the near future. References and Notes (1) Ion Ekitai (Ionic Liquids) II; Ohno, H., Ed.; CMC Shuppan: Tokyo, 2006. (2) Swatloski, R. P.; Spear, S. K.; Holbrey, J. D.; Rogers, R. D. J. Am. Chem. Soc. 2002, 124, 4974. (3) Yamamuro, O.; Minamimoto, Y.; Inamura, Y.; Hayashi, S.; Hamaguchi, H. Chem. Phys. Lett. 2006, 423, 371. (4) Suarez, P. A. Z.; Einloft, S.; Dullius, J. E. L.; Souza, R. F.; Dupont, J. J. Chem. Phys. 1998, 95, 1626. (5) Huang, J.-F.; Chen, P.-Y.; Sun, I.-W.; Wang, S. P. Inorg. Chim. Acta 2001, 320, 7. (6) Mele, A.; Tran, C. D.; Lacerda, S. H. D. P. Angew. Chem., Int. Ed. 2003, 42, 4364. (7) Dupont, J.; Suarez, P. A. Z.; Souza, R. F.; Burrow, R. A.; Kintzinger, J.-P. Chem. Eur. J. 2000, 2377. (8) Avent, A. G.; Chaloner, P. A.; Day, M. P.; Seddon, K. R.; Welton, T. J. Chem. Soc., Dalton Trans. 1994, 3405. (9) Abdul-Sada, A. K.; Elaiwi, A. E.; Greenway, A. M.; Seddon, K. R. Eur. Mass Spectrom. 1997, 3, 245. (10) Noda, A.; Hayamiz, K.; Watanabe, M. J. Phys. Chem. B 2001, 105, 4603. (11) Tokuda, H.; Hayamizu, K.; Ishii, K.; Susan, M. A. B. H.; Watanabe, M. J. Phys. Chem. B 2004, 108, 16593. (12) Hamaguchi, H.; Saha, S.; Ozawa, R.; Hayashi, S. ACS Symp. Ser. 2005, 901, 68. (13) Hamaguchi, H.; Ozawa, R. AdV. Chem. Phys. 2005, 131, 85. (14) Katanayagi, H.; Hayashi, S.; Hamaguchi, H.; Nishikawa, K. Chem. Phys. Lett. 2004, 392, 460. (15) Saha, S.; Hayashi, S.; Kobayashi, A.; Hamaguchi, H. Chem. Lett. 2003, 32, 740. (16) Hardacre, C.; Holbrey, J. D.; McMath, S. E. J.; Bowron, D. T.; Soper, A. K. J. Chem. Phys. 2003, 118, 273. (17) Matsumoto, K.; Hagiwara, R.; Ito, Y.; Kohara, S.; Suzuya, K. Nucl. Instrum. Methods Phys. Res. Sect. B 2003, 199, 29. (18) Takamuku, T.; Noguchi, Y.; Yoshioka, E.; Kawaguchi, T.; Matsugami, M.; Otomo, T. J. Mol. Liq. 2007, 131-132, 131. (19) Wu, Y. G.; Tabata, M.; Takamuku, T.; Yamaguchi, A.; Kawaguchi, T.; Chung, N. H. Fluid Phase Equilib. 2001, 192, 1. (20) Kitade, S.; Ichikawa, A.; Imura, N.; Takahashi, Y.; Noda, I. J. Rheol. 1997, 41, 1039. (21) Kagaku-Binran. Kiso-Hen, Kaitei 3 Han (Data Handbook for Chemistry), 3rd ed.; The Chemical Society of Japan, Maruzen: Tokyo, 1984; Vol. 2, pp 38-60.
JP800633X
