Predicting the hygroscopicity of imidazolium-based ILs varying in anion by hydrogen-bonding basicity and acidity

Abstract

Most of ILs are hygroscopic. Hygroscopicity of ILs has a significant influence on the structure and properties of ILs. There should be correlation between hygroscopicity and the other properties of ILs. In this study, eight imidazolium-based ILs varying in anion were used to investigate the relationship between their hygroscopicity parameters (experimental water sorption capacity within 3 h W_3h, saturated water sorption capacity W_∞ derived from the modified two-step model, degree of difficulty to reach sorption equilibrium 1/k, and water sorption rate kW_∞ derived from the modified two step model) and their property parameters (electric field ΔE, polarity E^N_T, hydrogen-bonding acidity α, hydrogen-bonding basicity β, dipolarity-polarizability π*, and the difference between hydrogen-bonding basicity and acidity β − α). We found that β − α had a very strong positive correlation coefficient ρ ≈ 0.99 with the hygroscopicity parameters. However, the respective species, i.e., β and α are less correlated with hygroscopicity. The E^N_T and ΔE are much less correlated with the hygroscopicity of ILs than those of β − α, β and α, while π* has almost no correlation with hygroscopicity of ILs.

---

1. Introduction

Ionic liquids (ILs) are salts which are in the liquid state near room temperature.¹ They have high thermal stability and negligible vapor pressure in normal operating conditions.^2–5 They could be tuned by combinations of anion and cation or functionalization for specific task, thus various ILs with amazing physical and chemical properties could be obtained. Therefore, ILs are widely used for biomass dissolution,^6,7 gas capture,^8–11 organic solvent separation,¹² microemulsion formation,^13,14 material synthesis,^15,16 and so on.

The presence of water has a significant influence on the structure and properties of ILs. After introducing the water molecule, the hydrogen bond between the anion and cation of ILs was weakened.^17–19 Properties, such as viscosity, could be enormously reduced after mixing with water.^20,21 Also, contamination of water might even make ILs (e.g., [BMIM][BF₄], [BMIM][BF₆]) hydrolyzed into poisonous products, such as HF.²² The presence of 1 wt% water could enhance the CO₂ capacity from 1 : 2 to 1 : 1 mol. CO₂/IL.²³ The presence of about 1 wt% water also leads cellulose from being soluble to insoluble in ILs.²⁴

Most of the ILs are hygroscopic, even the so called hydrophobic ILs (e.g., [BMIM][Tf₂N]) could absorb some extent of water.^25–28 Namely, ILs can not avoid contacting with water in normal conditions because humid air is ubiquitous. Unexpectedly, highly hygroscopic IL [BMIM][Ac] absorbed about 16% g g⁻¹ water/IL within 3 h at temperature of 23 °C and relative humidity (RH) of 52%.²⁷ Protic ILs are more hygroscopic. For example, in similar conditions (T = 28.9 °C, RH = 56.6%), diethyl-ammonium formate [DEA][Fo] could absorb about 25% g g⁻¹ water/IL within 24 h.²⁵

The above discussions indicate that the hygroscopicity of ILs is important. Therefore, it is preferable to develop a predictive method for hygroscopicity of ILs. Here, we develop a method to predict the hygroscopicity of ILs from the solvatochromic parameters (i.e., polarity E^N_T, hydrogen-bonding acidity α, hydrogen-bonding basicity β, and dipolarity-polarizability π*). The cue to choose solvatochromic parameters is their good correlations with biomass dissolution,^6,29,30 interaction mechanism,^28,31 acetylene solubility,³² CO₂/CH₄ and H₂S/CH₄ separation,³³ and so on. This research indicated that hydrogen-bonding basicity β is linearly positive correlated with acetylene solubility, CO₂/CH₄ and H₂S/CH₄ selectivity and biomass dissolution. Moreover, Sixta et al.²⁹ studied the role of solvent parameters in the regeneration of cellulose from ionic liquid solutions and found that the net basicity (β − α) was the best parameter to explain the ability of mixtures to dissolve cellulose, rather than β alone. Except solvatochromic parameters, the effect of the electric field (ΔE) on hygroscopicity has also been investigated because the electric field has significant influence on the structure of ILs.^34,35

It is well known that the linear solvation energy relationship (LSER) is one of the most commonly used approaches to quantitatively study solvent-dependent properties, and was introduced by Kamlet and Taft.^36,37 This method correlates the solvent property and solvatochromic parameters α, β, and π* by multi-parameter equation (XYZ) = (XYZ₀) + a × α + b × β + s × π*. This LSER approach has been successfully used to explain and predict the reaction rate and selectivity of reaction processes in which ionic liquids have served as solvents.^38–42 According to the relationship between hygroscopicity parameters and property parameters of ionic liquids, this LSER correlation should be applied. However, herein we select the correlation between β − α and the hygroscopicity parameters instead of the LSER correlation.

Eight imidazolium-based salts with the same cation [BMIM] were investigated (Table 1). On one hand, imidazolium-based ILs are widely studied and the data are available. On the other hand, the influence of the anion on the physical properties (e.g., hygroscopicity²⁷) and potential industrial application (e.g., CO₂ capture,⁴³ biomass dissolution⁶) is more significant than that of cation type and alkyl chain length. Therefore, eight imidazolium-based ILs varying in anion were selected.

Table 1 Names and chemical structures of 8 imidazolium-based ILs varying in anion which were investigated

No.	ILs	Abbreviation	Chemical structure
1	1-Butyl-3-methyl-imidazolium hexafluorophosphate	[BMIM][PF6]
2	1-Butyl-3-methyl-imidazolium bis(trifluoromethylsulfonyl)imide	[BMIM][Tf2N]
3	1-Butyl-3-methyl-imidazolium tetrafluoroborate	[BMIM][BF4]
4	1-Butyl-3-methyl-imidazolium nitrate	[BMIM][NO3]
5	1-Butyl-3-methyl-imidazolium acetate	[BMIM][Ac]
6	1-Butyl-3-methyl-imidazolium trifluoromethanesulfonate	[BMIM][TFO]
7	1-Butyl-3-methyl-imidazolium trifluoroacetate	[BMIM][TFA]
8	1-Butyl-3-methyl-imidazolium chloride	[BMIM][Cl]

---

2. Data processing

The names and chemical structures of the investigated ILs are listed in Table 1. The hygroscopicity parameters (Table 2) are experimental water sorption capacity within 3 h (W_3h), saturated water sorption capacity derived from the modified two-step model (W_∞), the degree of difficulty to reach sorption equilibrium (1/k), and water sorption rate (10³kW_∞).²⁷ The six property parameters (Table 3) are polarity (E^N_T), hydrogen-bonding acidity (α), hydrogen-bonding basicity (β), dipolarity-polarizability (π*), the difference between hydrogen-bonding basicity and acidity (β − α), and electric field (ΔE).^34,44 The effect of property parameters on hygroscopicity parameters is shown in Fig. 1–4.

Table 2 Hygroscopicity of 8 selected imidazolium-based ILs varying in anion which were investigated^a

No.	ILs	W3h %	W∞ %	1/k min^−1	10^3kW∞ % min^−1
a All hygroscopicity data are conducted at T = 23 °C and RH = 52%, cited from our previous report.^27
1	[BMIM][PF6]	0.43	0.45	75.8	5.9
2	[BMIM][Tf2N]	1.03	1.03	38.5	26.8
3	[BMIM][BF4]	2.98	3.32	80.0	41.5
4	[BMIM][NO3]	7.64	9.61	114.9	83.6
5	[BMIM][Ac]	15.63	24.24	172.4	140.6
6	[BMIM][TFO]	4.28	4.69	84.0	55.8
7	[BMIM][TFA]	9.05	9.88	85.5	115.6
8	[BMIM][Cl]	14.08	20.00	147.1	136.0

---

Table 3 Solvatochromic parameters and electric field of 8 selected imidazolium-based ILs varying in anion^a

No.	ILs	PolarityE^NT	Hydrogen-bonding acidity α	Hydrogen-bonding basicity β	β − α	Dipolarity-polarizability π*	Electric field ΔE/MV cm^−1
a The remaining data are cited from the ref. 34.b “–” indicates that the corresponding data is not available.c Supercooled liquid.
1	[BMIM][PF6]	0.667	0.630	0.196	−0.434	0.991	−4.8
2	[BMIM][Tf2N]	0.64	0.647	0.306	−0.341	0.903	−6.6
3	[BMIM][BF4]	0.669	0.643	0.407	−0.236	0.991	−4.2
4	[BMIM][NO3]	0.634	0.517	0.644	0.127	1.06	−2.2
5	[BMIM][Ac]	0.566	0.437	1.122	0.685	0.974	−1.7
6	[BMIM][TFO]^44	0.667 (ref. 44 and 49)	0.630 (ref. 44 and 50)	0.46 (ref. 44 and 50)	−0.17	1 (ref. 42 and 44)	—^b
7	[BMIM][TFA]^44,51	0.623	0.560	—^b	—^b	0.56	—^b
8	[BMIM][Cl]^c	0.614	0.470	0.87	0.4	1.1	0

---

Fig. 1 Effect of the solvatochromic parameters and electric field on the water sorption capacity W_3h. The value of R² is used to represent the efficiency of a linear regression. (a) N/A means the value of R² is negative, which is statistically meaningless. N/A indicates not available.

Fig. 2 Effect of the solvatochromic parameters and electric field on the water sorption capacity W_∞. The value of R² is used to represent the efficiency of a linear regression. (a) N/A means the value of R² is negative, which is statistically meaningless. N/A indicates not available.

Fig. 3 Effect of the solvatochromic parameters and electric field on the equilibrium difficulty 1/k. The value of R² is used to represent the efficiency of a linear regression. (a) N/A means the value of R² is negative, which is statistically meaningless. N/A indicates not available.

Fig. 4 Effect of the solvatochromic parameters and electric field on the water sorption rate 10³kW_∞. The value of R² and R′² are used to represent the efficiency of a linear regression and an exponential regression, respectively. (a) N/A means the value of R² is negative, which is statistically meaningless. N/A indicates not available.

Two parameters are used to analyze the correlations between hygroscopicity and property parameters. The first one is the adjusted R square R² of the fitted linear equation. The maximum value of R² is 1. Another parameter is the correlation coefficient ρ. The sign + and – of ρ means positive correlation and negative correlation, respectively. The closer of ρ to 1 or −1 means the better correlation. The tendency of R² and ρ is same, so we mainly use R² as the indicator of the linear relationship. If a change of direction is considered, ρ is also discussed. The overall presentation of R² and ρ is presented in Fig. 5 and 6, respectively. All data processing was conducted using Origin 8.0.

Fig. 5 Overall presentation of R² for the fitted linear equation between the hygroscopicity parameters (W_3h, W_∞, 1/k, and 10³kW_∞) and other property parameters (E^N_T, α, β, β − α, π*, and ΔE). W_3h, W_∞, 1/k, and 10³kW_∞ represent the water sorption capacity within 3 h, steady-state water sorption capacity, the difficulty to reach water sorption equilibrium, sorption rate of ILs, respectively. E^N_T, α, β, β − α, π*, and ΔE represent polarity, hydrogen-bonding acidity, hydrogen-bonding basicity, the difference between hydrogen-bonding basicity and acidity, dipolarity-polarizability, and electric field of ILs, respectively. Note that the correlation between hygroscopicity parameters and π* is statistically meaningless, because the value of R² is negative as shown in the figure.

Fig. 6 Overall presentation of the correlation coefficient ρ for the linear relationship between hygroscopicity parameters (W_3h, W_∞, 1/k, and 10³kW_∞) and other property parameters (E^N_T, α, β, β − α, π*, and ΔE).

3. Results and discussion

3.1. Correlations between water the sorption capacity and property parameters of ILs

Water sorption capacity is directly related to hygroscopicity. It is one of the most studied parameters. Greater water sorption capacity is deemed as stronger affinity of ILs with water, hence higher hygroscopicity.⁴⁵

The water amount (mass ratio as g g⁻¹ H₂O/IL) absorbed by ILs within 3 h is denoted as W_3h.²⁷ The correlations between W_3h and solvatochromic parameters (i.e., polarity E^N_T, hydrogen-bonding acidity α, hydrogen-bonding basicity β, and dipolarity-polarizability π*) and electric field (ΔE) were conducted.

Results show that W_3h has almost no linear correlation with π* (R² is negative), a moderately linear correlation with ΔE (R² = 0.7535), E^N_T (R² = 0.8143), and α (R² = 0.8839), a good linear correlation with β (R² = 0.9605), while an almost perfect correlation with β − α (R² = 0.9875) (Fig. 1 and Table 4). The order of correlation coefficient ρ and R² is same as: π* < ΔE < E^N_T < α < β < β − α (Fig. 1 and Table 4). E^N_T and α have a negative correlation with W_3h, while π*, β and β − α have positive correlation. W_3h = 6.40 + 14.38(β − α), with R² = 0.9875 and ρ = 0.9948.

Table 4 Fitted linear equation and correlation coefficient between hygroscopicity and solvatochromic parameters and between hygroscopicity and the electric field, for the 8 selected imidazolium-based ILs varying in anion^a,b,c

	W3h	W∞	1/k	10^3kW∞
a The columns and rows of the table are indicated by x and y, respectively.b ρ and R^2 indicate the correlation coefficient between x and y and adjusted R square of the fitted linear equation of x and y, respectively. Contents in bold indicate the good correlation.c The table cell with and without bold text is positively and negatively correlated, respectively. The purpose for the bold text is to guide the eye.d N/A means the value of R^2 is negative, which is statistically meaningless. N/A represents not available.
E^NT	ρ = −0.9170	ρ = −0.9380	ρ = −0.8065	ρ = −0.8765
	y = 95.81 − 140.61x	y = 148.30 − 220.04x	y = 683.94 − 923.75x	y = 832.13 − 1196.13x
	R^2 = 0.8143	R^2 = 0.8598	R^2 = 0.5921	R^2 = 0.7297
α	ρ = −0.9490	ρ = −0.9528	ρ = −0.9178	ρ = −0.8968
	y = 38.31 − 56.26x	y = 57.40 − 86.40x	y = 326.71 − 406.41x	y = 339.90 − 473.11x
	R^2 = 0.8839	R^2 = 0.8924	R^2 = 0.8160	R^2 = 0.7716
β	ρ = 0.9834	ρ = 0.9841	ρ = 0.9419	ρ = 0.9776
	y = −3.95 + 18.45x	y = −7.25 + 28.56x	y = 26.29 + 132.33x	y = −19.42 + 156.73x
	R^2 = 0.9605	R^2 = 0.9622	R^2 = 0.8645	R^2 = 0.9469
β − α	ρ = 0.9948	ρ = 0.9935	ρ = 0.9539	ρ = 0.9869
	y = 6.40 + 14.38x	y = 8.77 + 22.21x	y = 100.52 + 103.26x	y = 68.50 + 121.90x
	R^2 = 0.9875	R^2 = 0.9843	R^2 = 0.8920	R^2 = 0.9687
π*	ρ = 0.02957	ρ = 0.12147	ρ = 0.32108	ρ = −0.1256
	y = 5.92 + 1.02x	y = 3.06 + 6.43x	y = 21.20 + 83.02x	y = 112.33 − 38.67x
	R^2 N/A^d	R^2 N/A^d	R^2 N/A^d	R^2 N/A^d
ΔE	ρ = 0.8960	ρ = 0.8649	ρ = 0.9159	ρ = 0.8932
	y = 15.04 + 2.49x	y = 21.75 + 3.68x	y = 166.55 + 19.01x	y = 141.66 + 21.31x
	R^2 = 0.7535	R^2 = 0.6852	R^2 = 0.7986	R^2 = 0.7472

---

The saturated water sorption capacity W_∞ is another parameter for characterizing water sorption capacity, which may be derived from the modified two-step model W = W_∞(1 − e^−kt).^25–27 The measurement of W_∞ is inconvenient owing to the very long time to reach equilibrium. Deriving W_∞ from the modified two-step model is relatively easy. The correlations between W_∞ and the parameters (i.e., E^N_T, α, β, π*, β − α and ΔE) were also investigated.

The results show that the extent of the linear relationship between W_∞ and property parameters is ordered as: π* < ΔE < E^N_T < α < β < β − α (Fig. 2 and Table 4). The correlation of β (R² = 0.9627) and α (R² = 0.8924) with W_∞ is less than that of β − α (R² = 0.9843). Specifically, the linear relationship between W_∞ and β − α could be expressed as W_∞ = 8.77 + 22.21(β − α). Instead, π* has no correlation with W_∞; ΔE and E^N_T have a moderate correlation with W_∞. It is the same as that with W_3h. It suggests that designing ILs with greater water sorption capacity could be achieved by selecting IL with higher β − α, and vice versa.

3.2. Correlations between the degree of difficulty to reach sorption equilibrium and the property parameters of ILs

Based on the previous studies,^25–27 the degree of difficulty to reach water sorption equilibrium (1/k) at a fixed temperature and relative humidity can be derived from a modified two-step model. A higher value of 1/k means that it is harder to reach sorption equilibrium. That is to say, if the equilibrium is difficult to reach, the difficulty to reach some percentage (e.g., 90% (ref. 43)) of the equilibrium could also be used. Thus, the prediction of 1/k is of much importance.

Fig. 3 and Table 4 show that the correlations between parameters (E^N_T, α, β, π*, β − α and ΔE) and 1/k are weaker than that for water sorption capacity. Although the most correlated parameter to 1/k is also β − α, the absolute value of R² is only 0.8919, which is much lower than that of W_3h (R² = 0.9875) and W_∞ (R² = 0.9843). It is also corroborated by the value of correlation coefficient ρ, i.e., ρ_{1/k&β − α} (0.9539) < ρ_{W∞&β − α} (0.9935) ≈ ρ_{W3h&β − α} (0.9948) (Table 4).

This indicates that factors affecting 1/k might be very complicated except for solvatochromic parameters and electric field. For example, the size of ion, the charge on the ion surface and the free volume are also important.^46,47 For example, [BMIM][Tf₂N] (β = 0.306) shows a negative deviation in the correlation between 1/k and β, which might be caused by the relatively big size of the anion Tf₂N. A bigger size of anion is not favorable for ILs for interacting with water, hence easy to reach equilibrium. Also, the positive deviation for [BMIM][PF₆] (β = 0.196) in the correlation between 1/k and β might be caused by the relatively large free volume of the anion PF₆. A larger free volume of ILs might be helpful to load more water, hence more difficult to reach equilibrium.

3.3. Correlations between water sorption rate and property parameters of ILs

The water sorption rate of ILs is related to how fast ILs would absorb water from humid air.^25–27 It could be indicated by 10³kW_∞ based on previous reports.^25–27,45

Similarly, 10³kW_∞ is correlated with the parameters (E^N_T, α, β, π*, β − α and ΔE). Fig. 4 and Table 4 show that 10³kW_∞ is negatively correlated to E^N_T, α, and π*, but positively correlated to ΔE, β, and β − α. The linear fitted efficiency for β − α (R² = 0.9687) is better than that of β (R² = 0.9469) and α (R² = 7716). It suggests that both β and α contribute to 10³kW_∞. The value of 10³kW_∞ could thus be directly predicted by β − α using the equation 10³kW_∞ = 68.50 + 121.90(β − -α) (Table 4). It means that designing ILs with a higher water sorption rate could also be obtained by tethering groups with a higher value of β − α, and vice versa.

Both the relationship between 10³kW_∞ and β, between 10³kW_∞ and β − α, could be well regressed by an exponential equation, such as y = y₀(1 − e^−kx), where y represents 10³kW_∞, and x represent β or β − α (Fig. 4). However, the value of R² for the exponential relationship suggested above is much less than the linear relationship. Specifically, the linear relationship for β and β − α is R² = 0.9469, and R² = 0.9687, respectively (Table 4). However, for the exponential relationship, the values of R² decrease to 0.9047, and −1.4153, respectively. Thus, it could be concluded that the linear relationship is more favorable than the exponential relationship for 10³kW_∞.

3.4. Overall view on the correlation result

An overall presentation of R² is displayed in Fig. 5. It shows that β − α has the best correlation with the hygroscopicity parameters (W_3h, W_∞, 1/k, and 10³kW_∞) compared to other property parameters (E^N_T, α, β, π*, and ΔE) of ILs. Particularly, R² for the linear relationship between β − α and W_3h reaches the highest value, i.e., 0.9875, while both β and α have less correlation with the hygroscopicity than β − α.

The correlation coefficient ρ is also investigated for the above relationship. The results show that the tendency of R² and ρ is identical, i.e., β − α > β > α > ΔE ≈ E^N_T ≫ π* (Fig. 5 and 6). But the difference of ρ between the above correlations is not as significant as that of R² (Fig. 5 and 6). However, ρ could directly show a positive or negative correlation. It could be seen from Fig. 6 that E^N_T and α are negatively correlated to all the parameters of hygroscopicity. ΔE, β and β − α are positively correlated to all the parameters of hygroscopicity, while π* is positively correlated to W_3h, W_∞, and 1/k, and negatively correlated to 10³kW_∞.

All the hygroscopicity parameters are more strongly related to β − α (positive correlation) than β (positive correlation) and α (negative correlation) (Fig. 5 and 6). Namely, a higher value of β − α contributes to higher hygroscopicity. In addition, β is more strongly related to hygroscopicity than α (Fig. 5 and 6). This indicates that water mainly interacts with the anion rather than the cation, because β and α are mainly determined by the anion and cation in terms of the imidazolium-based ILs, respectively.⁴⁸ Our previous report also suggested that the anion played a more important role in water sorption.²⁷

4. Conclusion

The correlations between hygroscopicity with solvatochromic parameters (E^N_T, α, β, π*, and β − α) and electric field (ΔE) of ILs were investigated. The parameters of hygroscopicity include water sorption capacity within 3 h (W_3h), saturated water sorption capacity (W_∞), the degree of difficulty to reach water sorption equilibrium (1/k), and water sorption rate (10³kW_∞).

The most important finding is that β − α is strongly positively correlated to all parameters of hygroscopicity, β, α, E^N_T, and ΔE are moderately correlated to the hygroscopicity, and π* is nearly not correlated to these parameters. These findings are very helpful for predicting hygroscopicity of ILs. They also us a hint that for designing hydrophilic or hydrophobic ILs, β − α of the ILs should be paid enough attention. It would also be helpful to synthesize new ILs when water is present intended for a homogenous or heterogeneous chemical reaction, phase separation, drying agent, water proof materials, and so on. Note that only eight imidazolium-based ILs varying in anion were investigated because of the greatest popularity and most abundant data for imidazolium salts. More work still needs to be done for other kinds of ILs.

Acknowledgements

This work was supported by the Plan Project of Science and Technology of Beijing Municipal Education Committee (KM201210016007) and the National Natural Science Foundation of China (21173267).

References

1. J. S. Wilkes, Green Chem., 2002, 4, 73–80.
2. Y. Chen, Y. Cao, Y. Shi, Z. Xue and T. Mu, Ind. Eng. Chem. Res., 2012, 51, 7418–7427.
3. T. Mu and B. Han, in Structure and Bonding, Springer-Verlag Berlin and Heidelberg GmbH & Co. KG, 2014, vol. 151, pp. 107–140.
4. M. J. Earle, J. Esperanca, M. A. Gilea, J. N. C. Lopes, L. P. N. Rebelo, J. W. Magee, K. R. Seddon and J. A. Widegren, Nature, 2006, 439, 831–834.
5. C. Maton, N. De Vos and C. V. Stevens, Chem. Soc. Rev., 2013, 42, 5963–5977.
6. Q. Chen, A. Xu, Z. Li, J. Wang and S. Zhang, Green Chem., 2011, 13, 3446–3452.
7. W. N. Liu, Y. C. Hou, W. Z. Wu, S. H. Ren, Y. Jing and B. G. Zhang, Ind. Eng. Chem. Res., 2011, 50, 6952–6956.
8. X. Zhang, H. Dong, Z. Zhao, S. Zhang and Y. Huang, Energy Environ. Sci., 2012, 5, 6668–6681.
9. Y. Zhang, S. Zhang, X. Lu, Q. Zhou, W. Fan and X. P. Zhang, Chem.–Eur. J., 2009, 15, 3003–3011.
10. J. Sun, J. Wang, W. Cheng, J. Zhang, X. Li, S. Zhang and Y. She, Green Chem., 2012, 14, 654–660.
11. S. Ren, Y. Hou, S. Tian, X. Chen and W. Wu, J. Phys. Chem. B, 2013, 117, 2482–2486.
12. D. Xiong, H. Wang, Z. Li and J. Wang, ChemSusChem, 2012, 5, 2255–2261.
13. J. Li, J. Zhang, Y. Zhao, B. Han and G. Yang, Chem. Commun., 2012, 48, 994–996.
14. J. Li, J. Zhang, B. Han, L. Peng and G. Yang, Chem. Commun., 2012, 10562–10564.
15. K. Ding, Z. Miao, B. Hu, G. An, Z. Sun, B. Han and Z. Liu, Langmuir, 2010, 26, 10294–10302.
16. K. Ding, Z. Miao, B. Hu, G. An, Z. Sun, B. Han and Z. Liu, Langmuir, 2009, 26, 5129–5134.
17. Q. Zhang, N. Wang, S. Wang and Z. Yu, J. Phys. Chem. B, 2011, 115, 11127–11136.
18. Q. Zhang, N. Wang and Z. Yu, J. Phys. Chem. B, 2010, 114, 4747–4754.
19. L. Zhang, Z. Xu, Y. Wang and H. Li, J. Phys. Chem. B, 2008, 112, 6411–6419.
20. Z. Wang, L. Fu, H. Xu, Y. Shang, L. Zhang and J. Zhang, J. Chem. Eng. Data, 2012, 57, 1057–1063.
21. W. Li, Z. Zhang, B. Han, S. Hu, Y. Xie and G. Yang, J. Phys. Chem. B, 2007, 111, 6452–6456.
22. M. G. Freire, C. M. S. S. Neves, I. M. Marrucho, J. A. P. Coutinho and A. M. Fernandes, J. Phys. Chem. A, 2010, 114, 3744–3749.
23. S. J. Zhang, J. M. Zhang, K. Dong, Y. Q. Zhang, Y. Q. Shen and X. M. Lv, Chem.–Eur. J., 2006, 12, 4021–4026.
24. R. P. Swatloski, S. K. Spear, J. D. Holbrey and R. D. Rogers, J. Am. Chem. Soc., 2002, 124, 4974–4975.
25. Y. Chen, Y. Cao, X. Lu, C. Zhao, C. Yan and T. Mu, New J. Chem., 2013, 37, 1959–1967.
26. Y. Cao, Y. Chen, L. Lu, Z. Xue and T. Mu, Ind. Eng. Chem. Res., 2013, 52, 2073–2083.
27. Y. Cao, Y. Chen, X. Sun, Z. Zhang and T. Mu, Phys. Chem. Chem. Phys., 2012, 14, 12252–12262.
28. Y. Cao, X. Sun, Y. Chen and T. Mu, ACS Sustainable Chem. Eng., 2014 DOI:10.1021/sc4003246.
29. L. K. Hauru, M. Hummel, A. W. King, I. Kilpeläinen and H. Sixta, Biomacromolecules, 2012, 13, 2896–2905.
30. T. V. Doherty, M. Mora-Pale, S. E. Foley, R. J. Linhardt and J. S. Dordick, Green Chem., 2010, 12, 1967–1975.
31. Y. Wu, T. Sasaki, K. Kazushi, T. Seo and K. Sakurai, J. Phys. Chem. B, 2008, 112, 7530–7536.
32. J. Palgunadi, S. Y. Hong, J. K. Lee, H. Lee, S. D. Lee, M. Cheong and H. S. Kim, J. Phys. Chem. B, 2011, 115, 1067–1074.
33. P. J. Carvalho and J. A. Coutinho, Energy Environ. Sci., 2011, 4, 4614–4619.
34. S. Zhang, Y. Zhang, X. Ma, L. Lu, Y. He and Y. Deng, J. Phys. Chem. B, 2013, 117, 2764–2772.
35. S. G. Zhang, R. Shi, X. Y. Ma, L. J. Lu, Y. D. He, X. H. Zhang, Y. T. Wang and Y. Q. Deng, Chem.–Eur. J., 2012, 18, 11904–11908.
36. M. J. Kamlet, J. L. M. Abboud, M. H. Abraham and R. Taft, J. Org. Chem., 1983, 48, 2877–2887.
37. M. J. Kamlet, R. M. Doherty, M. H. Abraham, Y. Marcus and R. W. Taft, J. Phys. Chem., 1988, 92, 5244–5255.
38. L. Crowhurst, N. L. Lancaster, J. M. Pérez Arlandis and T. Welton, J. Am. Chem. Soc., 2004, 126, 11549–11555.
39. R. Bini, C. Chiappe, V. L. Mestre, C. S. Pomelli and T. Welton, Org. Biomol. Chem., 2008, 6, 2522–2529.
40. G. Ranieri, J. P. Hallett and T. Welton, Ind. Eng. Chem. Res., 2008, 47, 638–644.
41. T. P. Wells, J. P. Hallett, C. K. Williams and T. Welton, J. Org. Chem., 2008, 73, 5585–5588.
42. M. Ab Rani, A. Brant, L. Crowhurst, A. Dolan, M. Lui, N. Hassan, J. Hallett, P. Hunt, H. Niedermeyer and J. Perez-Arlandis, Phys. Chem. Chem. Phys., 2011, 13, 16831–16840.
43. Y. Chen, J. Han, T. Wang and T. Mu, Energy Fuels, 2011, 25, 5810–5815.
44. P. G. Jessop, D. A. Jessop, D. Fu and L. Phan, Green Chem., 2012, 14, 1245–1259.
45. F. Di Francesco, N. Calisi, M. Creatini, B. Melai, P. Salvo and C. Chiappe, Green Chem., 2011, 13, 1712–1717.
46. M. Klähn, C. Stüber, A. Seduraman and P. Wu, J. Phys. Chem. B, 2010, 114, 2856–2868.
47. L. E. Ficke and J. F. Brennecke, J. Phys. Chem. B, 2010, 114, 10496–10501.
48. S. K. Shukla, N. D. Khupse and A. Kumar, Phys. Chem. Chem. Phys., 2012, 14, 2754–2761.
49. M. J. Muldoon, C. M. Gordon and I. R. Dunkin, J. Chem. Soc., Perkin Trans. 2, 2001, 433–435.
50. T. Welton, L. Crowhurst, P. R. Mawdsley, J. M. Perez-Arlandis and P. A. Salter, Phys. Chem. Chem. Phys., 2003, 5, 2790–2794.
51. H. Tokuda, K. Hayamizu, K. Ishii, M. A. B. H. Susan and M. Watanabe, J. Phys. Chem. B, 2004, 108, 16593–16600.

---