Liquid Structure and the Ion-Ion Interactions of Ethylammonium Nitrate Ionic Liquid Studied by Large Angle X-Ray Scattering and Molecular Dynamics Simulations
Yasuhiro UMEBAYASHI, Wan-Lin CHUNG, Takushi MITSUGI, Shuhei FUKUDA, Munetaka TAKEUCHI, Kenta FUJII, Toshiyuki TAKAMUKU, Ryo KANZAKI and Shin-ichi ISHIGURO
RTILs are composed of only ions and show their melting point below 100 °C. Typical RTILs are the salts of relatively large organic cations with a flexible alkyl chain such as 1-alkyl-3-methylimidazolium and N-alkylpyridinium combined with various anions. RTILs are one of the promised materials in the wide field of chemistry, i.e., not only organic/inorganic chemistry as Green Solvents , but also separation and/or extraction  and electrochemistry. In particular, electrochemical devices utilizing RTILs are strongly expected as those of high-energy density with high safety . For instance, lithium ion secondary batteries , high-performance electric double layer capacitors , dye-sensitised solar cells , field-effect transistors , electrochemical actuators  and fuel cells  using RTILs have been developed. However, for further development of such noble devices or new RTILs, the low viscosity of RTILs is one of the problems to overcome.
Though numerous numbers of RTILs application have been proposed, fundamental physico-chemical investigation on RTILs to yield their microscopic pictures at a molecular level is still scarce at the present stage. If one attempts to reveal the origin of the low viscous property of RTILs at a molecular level, neutron and X-ray scattering techniques are powerful, as is shown for various liquids including conventional non-aqueous solvents . However, there are some practical problems to study liquid structure of RTILs by means of such a scattering experiment. One of the most troublesome problems is their large and flexible molecular structure, which makes experimental scattering data quite complicated. Therefore, scattering experiments with the aid of molecular simulations become important to reveal complicated liquid structure of RTILs such as Empirical Potential Structure Refinement (EPSR) and Reverse Monte Carlo (RMC) techniques. Recently, we also reported liquid structures of some RTILs by means of X-ray scattering experiments with the aid of molecular dynamics simulations [55 - 57].
One of the most well known RTILs is 1-ethyl-3-methylimidazolium tetrafluoroborate [EMI+][BF4–], which is discovered as current RTILs by Wilkes et al. for the first time . Therefore, recent physico-chemical studies on RTILs are mainly focused on those consisted of 1-alkyl-3-methylimidazolium. On the other hand, owing to some practical reasons and scientific interests, new kinds of cations have been proposed, i.e., the protonated cations is used to yield protic RTILs, which are synthesized by simple neutralization reactions of the adequate acids and bases [3, 12]. However, it has been known since 1814 that the melting point of a simple ethylammonium nitrate [C2H5NH3+][NO3–] (hereafter, EAN is used as an abbreviation) is below ambient temperature . Recently, Angell et al. have shown its availability as a proton conductor for fuel cells .
From the viewpoint of physico-chemical properties of EAN as liquid and/or solvent, its conductivity [15 - 18, 21], viscosity [18, 19], density [18, 20, 21] and phase transitions [18, 21] have been reported. In addition, the acid-base property was discussed in Ref. 21. We demonstrated a direct pH measurement in EAN by using an ion selective field effect transistor (IS-FET) electrode . Moreover, dynamic properties such as a dielectric spectroscopy  and solvation dynamics of EAN ionic liquid [24, 25] have also been reported. For further understanding of both static and dynamic properties of EAN as liquid and/or solvent, it is indispensable to elucidate the liquid structure of EAN at a molecular level.
Very recently, an important structural feature of EAN was reported by Atkin and Warr . They measured small angle neutron scattering (SANS) of EAN and PAN (propylammmonium nitrate) to elucidate nano-scale segregation of the RTILs. In their SANS patterns, significant peaks of 0.66 A–1 for EAN and 0.54 A–1 for PAN were observed, which implies that nano-scale segregation occurs in this kind of ionic liquids, like 1-alkyl-3-methylimidazolium based ones [27 - 31]. Nano- segregation of RTILs is one of the hot topics in the RTILs chemistry, because ion transport and/or solvation properties of RTILs may depend on their nano-segregated liquid structure. On the other hand, structural information of EAN ionic liquid and its analogues are still scarce at the present stage, particularly the liquid structure and the closest ion-ion interactions at a molecular or an atomistic level. To the best of our knowledge, there is no published work on the EAN and its analogous ionic liquids from the structural viewpoint, except Ref. 26.
In this paper, we carried out large angle X-ray scattering (LAXS) experiments for EAN to reveal its liquid structure and the ion-ion interactions. The inter-molecular X-ray interference function for EAN was successfully evaluated, and thus, the inter-molecular X-ray pair correlation function was obtained. To yield further insight into the closest ion-ion interaction in EAN, molecular dynamics simulations based on the newly developed force fields were performed. The experimental inter-molecular X-ray pair correlation function was reasonably interpreted with the aid of the simulations. The closest ion-ion interactions in EAN will be discussed on the basis of partial atom-atom pair correlation functions derived from simulations at an atomistic level.
2. 1 Materials
The sample liquid EAN was synthesized by neutralizing an ethylamine solution with a concentrated nitric acid solution at -50 °C to avoid coloration. The solvent water was evaporated from the solution with a rotary evaporator to reduce its water content below a few %. Then, the solution was dried in a vacuum desiccator with phosphorus oxide at room temperature for a few weeks. The water content for the finally obtained sample ionic liquid was checked to be negligible (below 100 ppm) by Karl-Fisher titration. The liquid density was measured by using a vibration tube density-meter (Kyoto Electronics DA-310).
2. 2 LAXS measurement
LAXS measurements for the ionic liquid were carried out by using a q - q X-ray diffractometer (RIGAKU Ultima+) equipped with a NaI scintillation counter with a LiF(200) monochrometer. The Mo-Ka tube was used for an incident radiation (l = 0.7107 A) and operated at 50 kV and 40 mA. A sample was sealed in a Teflon® sample vessel with a Kapton® film of 25 mm as an X-ray window. The sample vessel without a sample was also measured for a background correction. Scattered X-ray intensities Iobs(2q) were recorded over the 2q range 1.6°-145°, corresponding to the scattering vector s (= 4psinq/l) of 0.25-16.86 A–1. A fixed count method was employed and Iobs(2q) was accumulated up to 100,000 counts at each point. The measurements were repeated to achieve the statistical reproducibility to be less than 1.0 %.
The Iobs(2q) were corrected for a background, an absorption, a polarization by the conventional methods , and normalized by a high angle and Krough-Moe-Norman methods [33, 34]. Incoherent scatterings multiplied by a Breit-Dirac recoil factor were subtracted to yield observed scattering intensity Icoh(s). With the incoherent scattering, Compton scattering factors reported by Cromer et al. [35, 36] were used and the monochrometer function defined as a leak from the monochrometer, was evaluated by Wakita's manner  in advance and that fitted by the polynomials was used. The total interference function
was defined as follows;
where ni and fi(s) denote the number and the atomic scattering factor  of atom i, respectively. Total correlation function
and total radial distribution function as the form
per the stoichiometric volume containing a pair of ions were obtained by
where r0 is the number density, and B is the damping factor, which is used for evaluation of the inter-molecular radial distribution function (B = 0.023 A2).
In the short range
, a peak at 0.9 A was observed, while the ghost peaks appeared ca 0.6 A due to the unfavorable undulation of a long period in
, which indicates that C-H atom pairs in this system were practically observable. Thus, the undulation was removed from Iobs(2q) by comparing the observed
with the estimated i(s) based on the crystal structures. The procedure for normalization and removal of the undulation was carried out iteratively to achieve the normalization factor difference between the above two means to be less than 0.1 %. For the LAXS data treatment, the program KURVLR  was used.
2. 3 MD simulations
Effective pair potentials were used, i.e., Lennard-Jones (LJ) and Coulomb terms were taken into account for the inter-molecular interactions. The force field parameterization for the ethylammonium was in line with that recently proposed by Lopes and Padua et al. [40 - 42], whose force fields for a series of 1-alkyl-3methylimidazolium bis(trifluoromethanesulfonyl) amide ionic liquids can satisfactorily reproduce vaporization enthalpies of the ionic liquids . However, in their previous publication, the force field parameters for the ternary and the quaternary ammonium cations were given, whilst those for the primary and the secondary ammonium cation were not. We employed LJ parameters for ethylammonium and nitrate ions taken from OPLS-AA . Molecular structures of ethylammmonium nitrate and the atom types are shown in Figure 1. Here, it should be noted that LJ parameters, both e and s for the primary ammonium proton were proposed to be zero in OPLS-AA, while Lopes and Padua et al. recommended e = 0.030 kcal mol–1 and s = 2.50 A for the ternary ammonium proton, which are the same values as those for the alkane CH proton in OPLS-AA. We used the latter values in this study. Atomic point charges for an ethylammonium ion were evaluated based on those determined by a ChelpG method  implemented in a Gaussian03 program package  at the density and the geometry from ab initio calculations at the MP2/cc-pvTZ (-f)//HF/6-31G(d) levels of theory. With the nitrate ion, the atomic point charges were taken from Ref. 47. All LJ parameters and atomic point charges used here are listed in Table 1. With regard to the intra-molecular interactions, bond stretching, bending and torsion energies were taken into consideration and the parameters for an ethylammonium ion were based on the OPLS-AA/Amber force fields [44, 48], and those for a nitrate ion were taken from Ref. 33. All simulations reported here were based on the way of OPLS-AA. In our simulations, Gear's predictor-corrector algorithm [49, 50] was employed for the integrals of the equation of motion. The system temperature and pressure were controlled by the Nose's [51, 52] and the Parrinello-Rahman's [53, 54] methods. The long range interactions were treated by Ewald's method with the typical cutoff of 11 A.
Figure 1. Molecular structures and the atom types for ethylammonium nitrate.
Table 1. Lennard-Jones parameters and partial atomic point charges for ethylammonium and nitrate ions.
a in A. b in kcal mol-1.
|atom type||sa||eb||q||atom type||sa||eb||q|
As the first step of the simulations, NTP ensembles at 2,000 K and 10,000 atm were employed and the 256 ion pairs of a low density were set into the cubic box under periodic condition, and were simulated until the equilibration was achieved. After a few hundreds of ps calculations, the systems reached nearly equilibration under this condition. Then, all of the intra-molecular freedoms were frozen and the system ensemble was changed to NTV, where the system volumes were fixed to those of ensemble averages achieved in the former NTP ensembles. The second step equilibrations time typically needed was 500 ps. Thus obtained trajectories were checked whether the component ions were fully mixed, as well as T, P, U, r and g(r). Thus obtained configurations were employed as the initial one for the final simulations. Finally, NTP ensembles of 256 pairs of the flexible ions in a cubic box under periodic condition at 298 K and 1 atm were carried out. The simulation length was 2 ns with the time step of 0.2 fs. It took 1 ns for the system equilibration, followed by the 1 ns production run, and the data obtained in the last 200 ps were analyzed. All simulations were carried out using Fujitsu Materials Explorer 4.0 on the Fujitsu PRIMEQUEST 580 at the Computing and Communications Center, Kyushu University
The total X-ray interference function
was calculated by equation (4),
in which r0 denotes the ensemble average of the number density, and the total number of atoms in the simulation box N is given by
The total X-ray correlation function
and total X-ray radial distribution function as the form
were obtained from
by the same Fourier transformation procedure as for the
3 Results and discussion
3. 1 LAXS experiment
Figure 2 shows X-ray interference function obtained by LAXS experiment. As can be seen, in the low s range of s < 2 A–1, small and intense peaks of ca. 0.62 and 1.7 A–1 appeared. The former is quite unique, while the latter is characteristic of the liquid of a long range ordering. As mentioned above, according to Atkin and Warr , a peak of 0.66 A–1 was observed in their SANS experiments. They evaluated the size of nano segregation to be 8.7 - 9.7 A, and proposed that a local lamellar structure is most probable. The peak of 0.62 A–1 in our LAXS experiment should correspond to their peak of 0.66 A–1. On the other hand, in the larger s range, the oscillation due to the intra-molecular structure can be clearly observed up to 16 A–1. As expected, the oscillation is rather different from those for the other ionic liquids previously reported [55 - 57], suggesting that the intra- and the inter-molecular structures of EAN can be revealed by LAXS experiment. Fourier transform of the X-ray interference function yields X-ray pair correlation function for EAN ionic liquid (Figure 3). Two major peaks of 1.3 and 2.2 A appeared in
. The former, the intense peak of 1.3 A, can be mainly ascribed to the intra-molecular bonding atom-atom correlations such as N-O of 1.26 A in a nitrate ion and both C-C of 1.53 A and C-N of 1.51 A in an ethylammonium ion. Similarly, the latter can be assigned to the intra-molecular non bonding atom-atom correlations such as O...O (NO3–) of 2.17 A and C...N (C2H5NH3+) of 2.54 A.
Figure 2. X-ray interference function obtained by the LAXS experiment (plots). The intra-molecular X-ray interference function estimated based on the molecular geometries found in crystals is also shown as a solid line.
Figure 3. Total X-ray pair correlation function obtained by LAXS experiment (plots). The intra- and the inter-molecular X-ray pair correlation functions are also shown as thin and thick solid lines, respectively.
To reveal the inter-molecular interactions in the ionic liquid, the intra-molecular atom-atom correlations for ethylammonium and nitrate ions were evaluated based on the molecular geometries found in the crystals including them. The total X-ray interference function
can be written as
represent the intra- and the inter-molecular interference functions, respectively. If the atom-atom distances and their mean square displacements for the intra-molecular atom-atom correlations in the system are known adequately,
can be estimated as follows;
where nij, rij and bij denote number, distance and temperature factor (mean square displacement) of the i-j atom pair, respectively. Thus
, can be evaluated by subtracting
according to the above simple relation.
As shown by a solid line in Figure 2,
was estimated based on the molecular geometries in crystals containing ethylammoniumu and nitrate ions. The intra- and the inter-molecular X-ray pair correlation function
can be obtained by Fourier transform of
, respectively, as shown as thin and thick solid lines in Figure 3. As seen in Figure 2 and Figure 3,
is in good accordance with
in the range of s > 5 A–1, and there is no significant peak in
of r < 2.5 A, indicating that the intra-molecular atom-atom correlations in the system can be satisfactorily estimated and
can be reasonably evaluated. Peaks of 3.0, 3.4, and 4.7 A were found in the
, which may be mainly ascribable to the atom-atom interactions among the closest ions charged oppositely. Here, we can at least assign the peak of 3.0 A to the NH (C2H5NH3+)...O (NO3–) hydrogen bonding safely, because similar atom-atom distance of the O(H2O)...O(NO3–) correlation in the hydrated nitrate ion are reported to be 2.88 - 2.95 A in the literatures. However, it seems to be difficult to ascribe the other peaks in the
to a specific atom-atom correlation. The experimentally evaluated
will be discussed in detail with the aid of molecular dynamics simulations in the following section.
With regard to the long range ordering of EAN ionic liquid, it is convenient to calculate the inter-molecular X-ray radial distribution functions as the form of
, as shown in Figure 4. As can be seen, the broad peaks of ca. 4.7, 8 and 12 A can be found in
together with the peak of 3.4 A, suggesting that long range ordering evidently occurs in EAN ionic liquid. The broad Peak of 4.7 A, may arise from the closest cation-anion interaction. In fact, it is in agreement with the estimation of 5.3 A based on the molar volume by Atkin and Warr .
Figure 4. Inter-molecular X-ray radial distribution functions as the form of D(r) - 4pr2r0 obtained by LAXS experiment (plots). That derived from simulations is also shown as a solid line.
3. 2 MD simulations
It is worth comparing the simulated density with the experimental one. From our simulations, the value of 1.208 (5) g cm–3 was obtained for the system density. As mentioned above, some values of density of EAN were found in the literatures [8, 10 - 12]. The value of 1.211 g cm–3 at 298 K found in Ref. 11 may be the most plausible to be compared with the theoretical value. Thus, our simulations reproduced the system density within 0.2% accuracy, indicating that our simulation can reasonably reproduce the liquid structure and the ion-ion interaction in EAN ionic liquid.
X-ray interference function derived from simulations
was depicted as a solid line in Figure 5, accompanied by the observed one as plots. As can be seen, in the whole s range examined, the
is reasonably in accordance with the
, suggesting that the liquid structure of EAN ionic liquid can be adequately reproduced by simulations. It should be noted that three peaks of 0.70, 1.7 and 2.5 A–1 in the low s range well reproduced the experimental ones, which were observed at 0.62, 1.7 and 2.6 A–1, respectively, though the amplitude of the oscillations seems to be larger relative to the experiment. On the other hand, in the high s range, the accordance of the
is not so good, which may arise from the disagreement of the employed equilibrium values for the intra-molecular structural parameters of the respective ion. However, we mention here again that our main interest focuses on the liquid structure, i.e., the long range ordering and the ion-ion interactions in the ionic liquid. Thus, the inter-molecular part of the X-ray interference functions was compared. Figure 6 shows the inter-molecular X-ray interference function derived from simulations
as a solid line, accompanied by the experimental one as plots. As can be seen in Figure 6,
can be well reproduced by the
, though the intensity of the peak of 1.7 A–1 was slightly larger than the experimental one.
Figure 5. X-ray interference function derived from the simulations shown as a solid line. The experimental one, which is the same data shown in Figure 3, is also plotted for comparison.
Figure 6. Inter-molecular X-ray interference functions extracted by the LAXS experiment (plots) and derived from the simulations (solid line).
The inter-molecular X-ray pair correlation function from simulations
is shown as a solid line in Figure 7 with the experimental one. The experimentally observed peaks of 3.0, 3.4, and 4.7 A may be ascribable to the broad peaks of ca. 3.4 and 4.6 A from the simulations, indicating that our simulation at least qualitatively agrees with the closest ion-ion interactions in EAN ionic liquid experimentally revealed. In addition, the inter-molecular radial distribution function as the form of
calculated from simulations is represented in Figure 4. As expected by the comparison of
, the long range ordering of the EAN already mentioned in the LAXS experiment was appropriately reproduced by our simulations, i.e., the positions for the peak of 3.4 A and the broad ones of 4.7, 8 and 12 A in
agree well with those found in
, though the peak intensity is slightly larger.
Figure 7. Inter-molecular X-ray pair correlation function extracted by the LAXS experiment (plots) and derived from the simulations (solid line).
In order to yield further insight into the closest ion-ion interaction in EAN ionic liquid, it is important to assign the peaks found in the
more clearly. Therefore, we attempted to calculate the inter-molecular partial atom-atom correlation functions gX-X(r) from our simulations. Figure 8(a), (b) and (c) shows gX-X(r) of the corresponding cation-anion, the cation-cation and the anion-anion correlations, respectively. As shown in Figure 8, it seems to be difficult to assign the observed peaks of 3.0, 3.4 and 4.7 A to a specific atom-atom correlation except the peak of 3.0 A. It is clearly ascribable to the N-O correlation between cation and anion, in other words, the NH...O hydrogen bonding. However, the peaks of 3.4 A can be safely ascribed to the CE-O (methyl carbon and oxygen in nitrate) and the C1-O (methylene carbon and oxygen in nitrate) correlations in the cation-anion interaction. It should be noted that the peak of 4.7 A in
significantly includes the cation-cation and the anion-anion interactions, as clearly shown in Figure 8 (b) and (c).
Figure 8. Partial atom-atom pair correlation functions for the cation-anion (a), the cation-cation (b) and the anion-anion (c), respectively.
It is worth arguing the respective gX-X(r) though qualitatively. The shortest correlation is found for the N-O correlation as expected, whose peak position of the first peak is 3.16 A, followed by the C1-O and the CE-O correlations, i.e., rather intense peaks of 3.3 A for the former and 3.5 A for the latter, respectively, are found as the first peak. This indicates that nitrate ions significantly interact with not only the ammonium moiety but also the terminal methyl and the methylene groups of an ethylammonium ion. Spatial distribution function (SDF) is suitable to look at this situation. Figure 9 displays the iso-probability surface of the most neighboring oxygen atom in the nitrate anion around the ammonium nitrogen in the cation. SDF clearly shows that the oxygen atom in the nitrate anion interacts at the CH proton of the terminal methyl and methylene groups in the ethylammonium cation. As shown in Table 1, the positive charge in an ethylammonium ion is concentrated around the ammonium and the neighboring methylene groups. Therefore, van der Waals force mainly operates in the CE...O interaction between cation and anion. It should be noted that it is generally considered that Coulombic interaction in ionic liquids are strong.
Figure 9. SDF of the most neighboring oxygen in the nitrate around the nitrogen in the cation. The iso-probability surface (red cloud) corresponds to two times larger probability than bulk one. Grey, sky blue and blue atoms represent C, H and N, respectively.
On the other hand, our simulation revealed that relatively weak van der Waals interaction between the oppositely charged ions (probably including polarization effect) is also significant. It should be remarked that, in the cation-cation interactions shown in Figure 8(b), the CE-CE correlation shows rather significant attractive interaction, i.e., their first peak appears at 4.0 A with the peak height of ca. 1.8. This evidently suggests that the van der Waals attractive force operates among the methyl groups of ethlammonium ions. A typical snapshot in our simulation is shown in Figure 10, where the polar parts of the ionic liquid, -NH3 and NO3, are colored red, and the other relatively non-polar part, -CH2CH3, is shown by white, respectively. From the snapshot, the polar- and the non-polar parts may form the respective networked liquid structure. Thus, our simulation supports relatively short alkyl chain (ethyl and propyl group) aggregation in the ionic liquids proposed by Atkin and Warr.
Figure 10. Typical snapshot of liquid structure of EAN. See text on the atom coloring.
As a final remark, the small bump of 0.7 A–1 found in the
may be ascribable to the observed peak of 0.62 A–1. However, the peak is not so clear, because of relatively weak methyl group aggregation in EAN ionic liquid. Thus, further investigation for this kind of ionic liquid with a longer alkyl chain such as propyl and butyl groups is needed, and is now in progress.
This work has been financially supported by Grant-in-Aids for Scientific Research Nos. 18850017, 19003963, 19350033, 19550022, 19750062 and 20350037 and by a Grant-in-Aid for the Global COE Program, "Science for Future Molecular Systems" from the Ministry of Education, Culture, Sports, Science and Technology.
[ 1] a) R. D. Rogers, K. R. Seddon, S. Volkov, Ed., Green Industrial Applications of Ionic Liquids, Kluwer Academic Publishers, Dordrecht (2000).
b) R. D. Rogers, K. R. Seddon, Ed., Ionic Liquids Industrial Applications to Green Chemistry, ACS Symposium Series 818, American Chemical Society, Washington, DC (2002).
c) R. D. Rogers, K. R. Seddon, Ionic liquids as Green SolVents, Progress and Properties, ACS Symposium Series 856, American Chemical Society, Washington, DC (2003).
d) D. J. Adams, P. D. Dyson, S. J. Tavener, Chemistry in Alternative Reaction Media, John & Wiley Sons ltd., West Sussex (2004).
[ 2] a) M. Koel, Cri. Rev. Anal. Chem., 35, 177-192 (2005).
b) G. A. Baker, S. N. Baker, S. Pandey, F. V. Bright, Analyst, 130, 800-808 (2005).
c) J. L. Anderson, D. W. Armstrong, G.-T. Wei, Anal. Chem., 78, 2892-2902 (2006).
d) M. L. Dietz, Sep. Sci. Tech., 41, 2047-2063 (2006).
e) F. Kubota, M. Goto, Sol. Extra. Res. Dev. Jpn., 13, 23-36 (2006).
f) T. Kakiuchi, Anal. Chem., 79, 6442-6449 (2007).
[ 3] H. Ohno Ed., Electrochemical Aspects of Ionic Liquids, John & Wiley Sons, Inc., Hoboken, New Jersey (2005).
[ 4] a) H. Sakaebe and H. Matsumoto, Electrochem. Commun., 5, 594 (2003).
b) H. Nakagawa, S. Izuchi, K. Kuwana, T. Nukuda and Y. Aihara, J. Electrochem. Soc., 150, A695 (2003).
c) S. Seki, Y. Kobayashi, H. Miyashiro, Y. Ohno, A. Usami, Y. Mita, N. Kihira, M. Watanabe and N. Terada, J. Phys. Chem. B, 110, 10228 (2006).
d) S. Seki, Y. Kobayashi, H. Miyashiro, Y. Ohno, A. Usami, Y. Mita, M. Watanabe and N. Terada, Chem. Commun., 2006, 544.
[ 5] a) A. B. McEwen, H. L. Ngo, K. LeComte and J. L. Goldman, J. Electrochem. Soc., 146, 1687 (1999).
b) M. Ue, M. Takeda, A. Toriumi, A. Kominato, R. Hagiwara and Y. Ito, J. Electrochem. Soc., 150, A499 (2003).
[ 6] a) P. Wang, S. M. Zakeeruddin, P. Comte, I. Exnar and M. Gratzel, J. Am. Chem. Soc., 125, 1166 (2003).
b) P. Wang, S. M. Zakeeruddin, J. E. Moser, M. K. Nazeeruddin, T. Sekiguchi and M. Gratzel, Nature Mater., 2, 402 (2003).
c) N. Yamanaka, R. Kawano, W. Kubo, N. Masaki, T. Kitamura, Y. Wada, M. Watanabe and S. Yanagida, J. Phys. Chem. B, 111, 4763 (2007).
[ 7] a) J. Lee, M. J. Panzer, Y. He, T. P. Lodge and C. D. Frisbie, J. Am. Chem. Soc., 129, 4532 (2007).
b) S. Ono, S. Seki, R. Hirahara, Y. Tominari and J. Takeya, Appl. Phys. Lett., 92, 103313 (2008).
[ 8] W. Lu, A. G. Fadeev, B. Qi, E. Smela, B. R. Mattes, J. Ding, G. M. Spinks, J. Mazurkiewicz, D. Zhou, G. G. Wallace, D. R. MacFarlane, S. A. Forsyth and M. Forsyth, Science, 397, 983 (2002).
[ 9] a) M. Yoshizawa, W. Xu and C. A Angell, J. Am. Chem. Soc., 125, 15411 (2003).
b) M. A. B. H. Susan, A. Noda, S. Mitsushima and M. Watanabe, Chem. Commun., 2003, 938.
 a) H. Ohtaki, Mon. Chem., 132, 1237-1268 (2001).
b) H. Ohtaki and S. Ishiguro, Chemistry of Nonaqueous Solutions Current Progress, ed. by G. Mamantov and A. I. Popov, Ed., VCH Publishers, New York (1994), p. 179.
c) H. Ohtaki and H. Yamatera Ed., Structure and Dynamics of Solutions, Elsevier, Amsterdam (1992).
d) H. Ohtaki, Pure Appl. Chem., 59, 1143 (1987).
e) Y. Marcus, Introduction to Liquid State Chemistry, John Wiley & Sons, Ltd., New York (1977).
f) R. F. Kruh, Chem. Rev., 62, 319 (1962).
g) K. Furulawa, Rep. Pro. Phys., 25, 395 (1962).
 J. S. Wilkes, M. J. Zaworotko, J. Chem. Soc., Chem. Commun., 1992, 965-964.
 T. L. Greaves and C. J. Drummond, Chem. Rev., 108, 206-237 (2008).
 P. Walden, Bull. Acad. Sci. St. Petersburg, 1914, 405-422.
 W. Xu and C. A. Angell, Science, 302, 422-425 (2003).
 C. Reichardt, Green Chem., 7, 339 (2005).
 J.-P. Belieres, D. Gervasiob and C. A. Angell, Chem. Commun., 2006, 4799-4801.
 T. L. Greaves, A. Weerawardena, C. Fong, I. Krodkiewska, C. J. Drummond, J. Phys. Chem. B, 110, 22479 (2006).
 D. F. Evans, A. Yamauchi, A.; R. Roman, E. Z. Casassa, J. Colloid Interface Sci., 88, 89 (1982).
 M. Yamada, I. Honma, Polymer, 45, 8349 (2004).
 I. B. Malham, P. Letellier, A. Mayaffre, M. Turmine, J. Chem. Thermodynamics, 39, 1132-1143 (2007).
 J.-P. Belieres and C. A. Angell, J. Phys. Chem. B, 111, 4926-4937 (2007).
 R. Kanzaki, K. Uchida, S. Hara, Y. Umebayashi, S. Ishiguro, and S. Nomura, Chem. Lett., 36, 684-685 (2007).
 H. Weinga1rtner, A. Knocks, W. Schrader and U. Kaatze, J. Phys. Chem. A, 105, 8646-8650 (2001).
 M. Halder, L. S. Headley, P. Mukherjee, X. Song, and J. W. Petrich, J. Phys. Chem. A, 110, 8623-8626 (2006).
 D. L. Rosen, Appl. Spectrosc, 60, 1453-1460 (2006).
 R. Atkin and G. G. Warr, J. Phys. Chem. B, 112, 4164-4166 (2008).
 J. N. C. Lopes, A. A. H. Padua, J. Phys. Chem. B, 110, 3330 (2006).
 S. M. Urahata, M. C. C. Ribeiro, J. Chem. Phys., 120, 1855 (2004).
 Y. Wang, G. A. Voth, J. Am. Chem. Soc., 127, 12192 (2005).
 Y. Wang, G. A. Voth, J. Phys. Chem. B, 110, 18601 (2006).
 A. Triolo, O. Russina, H. J. Bleif, E. DiCola, J. Phys. Chem. B, 111, 4641 (2007).
 G. Licheri, G. Paschina, G. Piccaluga, G. Pinna, M. Magini Ed., X-ray Diffraction of Ions in Aqueous Solutions: Hydration and Complex Formation, CRC press inc.CRC press inc., New York (1988).
 J. Krough-Moe, Acta Crystallogr., 9, 351 (1956).
 N. Norman, Acta Crystallogr., 10, 370 (1956).
 D. T. Cromer and J. B. Mann, J. Chem. Phys., 47, 1892 (1967).
 D. T. Cromer, J. Chem. Phys., 50, 4857 (1969).
 Y. Tajiri, M. Ichihashi, H. Wakita, X-sen bunsekino shinpo, 17, 207-214 (1986), (in Japanese).
 A. J. C. Wilson, E. Prince, International Tables For Crystallography Vol. C: Mathematical, Physical and Chemical Tables, Int. Uni. Crystallogr., Kluwer Academic Publishers, Dorfrecht (1999).
 G. Johanson, M. Sandstorm, Chem. Scr., 4, 195 (1973).
 J. N. C. Lopes, J. Deschamps and A. A. H. Padua, J. Phys. Chem. B, 108, 2038-2047 (2004).
 J. N. C. Lopes and A. A. H. Padua, J. Phys. Chem. B, 108, 16893-16898 (2004).
 J. N. C. Lopes and A. A. H. Padua, J. Phys. Chem. B, 110, 19586-19592 (2006).
 L. M. N. B. F. Santos, J. N. C. Lopes, J. A. P. Coutinho, J. M. S. S. Esperanca, L. R. Gomes, I. M. Marrucho and L. P. N. Rebelo, J. Am. Chem. Soc., 129, 284-285 (2007).
 W. L. Jorgensen, D. S. Maxwell, J. Tirado-Rives, J. Am. Chem. Soc., 118, 11225 (1996).
 M. Baaden, M. Burgard, and G. Wipff, J. Phys. Chem. B, 105, 11131-11141 (2001).
 C. M. Breneman and K. B. Wiberg, J. Comp. Chem., 11, 361 (1990).
 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople, Gaussian 03, Rev. B.04, Gaussian, Inc., Pittsburgh PA (2003).
 W. D. Cornell, P. Cieplak, C. I. Bayly, I. R. Gould, K. M. J. Merz, D. M. Ferguson, D. C. Spellmeyer, T. Fox, J. W. Caldwell, P. A. Kollman, J. Am. Chem. Soc., 117, 5179 (1995).
 G. W. Gear, Numerical Initial Value Problems in Ordinary Differential Equations, Prentice-Hall, Inc. (1971).
 H. J. C. Berendsen and W. F. Van Gunsteren, Proceeding of the Enrico Fermi Summer School on Molecular dynamics simulation of statistical mechanical system, ed. by G. Ciccotti and G. Hover, Ed., North Holland (1986), p. 43.
 S. Nose, Mol. Phys., 52, 255 (1984).
 S. Nose, J. Chem. Phys., 81, 511 (1984).
 M. Parrinello and A. Rahman, J. Appl. Phys., 52, 7182 (1981).
 M. Parrinello and A. Rahman, Phys. Rev. Lett., 45, 1196 (1980).
 K. Fujii, Y. Soejima, Y. Kyoshoin, S. Fukuda, R. Kanzaki, Y. Umebayashi, T. Yamaguchi, S. Ishiguro, and T. Takamuku, J. Phys. Chem. B, 112, 4329-4336 (2008).
 S. Fukuda, M. Takeuchi, K. Fujii, R. Kanzaki, T. Takamuku, K. Chiba, H. Yamamoto, Y. Umebayashi, S. Ishiguro, J. Mol. Liq., 143, 2-7 (2008).
 K. Fujii, S. Seki, S. Fukuda, T. Takamuku, S. Kohara, Y. Kameda, Y. Umebayashi and S. Ishiguro, J. Mol. Liq., 143, 64-69 (2008).