**Ionic Liquids Facilitate the Development of Absorption Refrigeration**

[38] Subramaniam K, Das A, Heinrich G. Development of Conducting Polychloroprene Rubber Using Imidazolium Based Ionic Liquid Modified Multi-Walled Carbon Nanotubes. Composites Science and Technology 2011; 71(11): 1441-9. http:// www.sciencedirect.com/science/article/pii/S0266353811001928 (accessed 2 June 2011).

[39] Yoshizawa M, Ogihara W, Ohno H. Novel Polymer Electrolytes Prepared by Copoly‐ merization of Ionic Liquid Monomers. Polymers Advanced Technologies 2002;13(8): 589-94. http://onlinelibrary.wiley.com/doi/10.1002/pat.261/pdf (accessed 15 August

[40] Pernak J, Walkiewicz F, Maciejewska M, Zaborski M. Ionic Liquids as Vulcanization Accelerators. Industrial Engineering & Chemical Research 2010; 49(10): 5012-17.

[41] Maciejewska M, Walkiewicz F, Zaborski M. Novel Ionic Liquids as Accelerators for the Sulfur Vulcanization of Butadiene–Styrene Elastomer Composites. Industrial En‐ gineering & Chemical Research, 2013, 52(25): 8410–15. http://pubs.acs.org/doi/full/

[42] Flory PJ, Rehner J. Statistical Mechanics of Cross-Linked Polymer Networks. II. Swel‐ ling. Journal of Chemical Physics 1943;11(11): 521-6. http://scitation.aip.org/

[43] Masek A, Zaborski M, Kosmalska A. Derivatives of Flavonoides as Anti-Ageing Sub‐ stances in Elastomers. Comptes Rendus Chimie 2011;14(5): 483-8. http://www.scien‐ cedirect.com/science/article/pii/S1631074811000026 (accessed 15 February 2011).

content/aip/journal/jcp/11/11/10.1063/1.1723792 (accessed 22 December 2004).

http://pubs.acs.org/doi/full/10.1021/ie100151n (accessed 23 April 2012).

10.1021/ie303167z (accessed 17 May 2013).

2012).

622 Ionic Liquids - Current State of the Art

Shiqiang Liang, Wei Chen, Yongxian Guo and Dawei Tang

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/58982

#### **1. Introduction**

The rapid progress in the development of ionic liquids has generated enthusiasm for their application in many traditional fields and renewed interest in absorption refrigeration. New absorption refrigeration working pairs containing ionic liquids have gained widespread attention in the past decade. In a chapter entitled "The Latent Application of Ionic Liquids in Absorption Refrigeration" [1] that we have published 3 years ago with InTech in the book entitled "Applications of Ionic Liquids in Science and Technology" achieved impressive readership results and has so far been accessed more than 4000 times. Over the past 3 years, progress in this field has been outstanding, and a few commercially competitive new working pairs were discovered. In this chapter, we describe the latest progress in the development of a few mentionable new working pairs containing ionic liquids for absorption refrigeration and a type of completely new conceptual absorption refrigeration working pair that was proposed by us and is expected to lead to a major breakthrough in the development of absorption refrigeration.

### **2. Recent progress in absorption refrigeration working pairs containing ionic liquids**

In the past 3 years, enthusiasm for studies on absorption refrigeration working pairs containing ionic liquids seems to have waned. The once preferred ionic liquid working pairs, such as Freon-IL, CO2-IL, and NH3-IL, do not receive attention from researchers any longer. However, some impressive progress is still being made for working pairs composed of a refrigerant and [RR'Im]DMP (1-R,3-R'-imidazolium dimethylphosphate).

© 2015 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and eproduction in any medium, provided the original work is properly cited.

#### **2.1. [mmIm]DMP-CH3OH**

Zhao Jie et al. [2] measured the saturation vapor pressure of [mmIm]DMP-CH3OH at *T* = 303.15–363.15 K and over a low methanol mole fraction range for *x* including 0.529, 0.558, 0.582 and 0.605. Zhao Jin et al. [3] measured the saturation vapor pressure of [mmIm]DMP-CH3OH at *T* = 280–370 K and over the high methanol mole fraction range for *x* including 0.8222, 0.9123, 0.9418 and 0.9652. These experimental results were confirmed by Chen et al. [4] using the UNIFAC model and the Wilson model to predict the vapor pressure and the excess enthalpy, respectively. Figs. 1 and 2 show the predicted vapor pressure and excess enthalpy, respectively, at *T* = 280–380 K and *x* = 0–1.

**Figure 1.** The predicted vapor pressure of [mmIm]DMP-CH3OH.

**Figure 2.** The predicted excess enthalpy of [mmIm]DMP-CH3OH.

= 280–370 K and over the high methanol mole fraction range for *x* including 0.8222, 0.9123, 0.9418 and 0.9652. These experimental results were confirmed by Chen et al. [4] using the UNIFAC model and the Wilson model to predict the vapor pressure and the excess enthalpy, respectively. Figs. 1 and 2 show the predicted vapor pressure and excess enthalpy, respectively, at

*T* = 280–380 K and *x* = 0–1.

The thermodynamic performances of single effect [mmIm]DMP-CH3OH absorption refrigeration **Figure 3.** Effects of operating temperatures on the *f* of the [mmIm]DMP-CH3OH system.

Fig. 3. Effects of operating temperatures on the *f* of the [mmIm]DMP-CH3OH system.

Fig. 2. The predicted excess enthalpy of [mmIm]DMP-CH3OH.

**2.1. [mmIm]DMP-CH3OH**

624 Ionic Liquids - Current State of the Art

at *T* = 280–380 K and *x* = 0–1.

**Figure 1.** The predicted vapor pressure of [mmIm]DMP-CH3OH.

**Figure 2.** The predicted excess enthalpy of [mmIm]DMP-CH3OH.

Zhao Jie et al. [2] measured the saturation vapor pressure of [mmIm]DMP-CH3OH at *T* = 303.15–363.15 K and over a low methanol mole fraction range for *x* including 0.529, 0.558, 0.582 and 0.605. Zhao Jin et al. [3] measured the saturation vapor pressure of [mmIm]DMP-CH3OH at *T* = 280–370 K and over the high methanol mole fraction range for *x* including 0.8222, 0.9123, 0.9418 and 0.9652. These experimental results were confirmed by Chen et al. [4] using the UNIFAC model and the Wilson model to predict the vapor pressure and the excess enthalpy, respectively. Figs. 1 and 2 show the predicted vapor pressure and excess enthalpy, respectively,

> have been simulated and analyzed. Fig. 3 shows the effects of the operating temperatures [condensing temperature (*T*C), evaporating temperature (*T*E), generating temperature (*T*G), and absorption temperature (*T*A)] on the circulation ratio, *f*, of the single system. From Fig. 3, the *f* of [mmIm]DMP-CH3OH is higher than that of LiBr/H2O but still acceptable for operation, and the coefficient of performance (COP) of the [mmIm]DMP-CH3OH absorption refrigeration will remain good, if the heat transfer areas of the regenerator are designed appropriately. Fig. 4 shows the effects of the operating temperatures on the COP of the single system. From The thermodynamic performances of single effect [mmIm]DMP-CH3OH absorption refriger‐ ation have been simulated and analyzed. Fig. 3 shows the effects of the operating temperatures [condensing temperature (*T*C), evaporating temperature (*T*E), generating temperature (*T*G), and absorption temperature (*T*A)] on the circulation ratio, *f*, of the single system. From Fig. 3, the *f* of [mmIm]DMP-CH3OH is higher than that of LiBr/H2O but still acceptable for operation, and the coefficient of performance (COP) of the [mmIm]DMP-CH3OH absorption refrigeration will remain good, if the heat transfer areas of the regenerator are designed appropriately.

> 2 Fig. 4, the COP of the [mmIm]DMP-CH3OH absorption refrigeration is lower than that of LiBr/H2O absorption refrigeration under the same temperature conditions, but higher than that of Fig. 4 shows the effects of the operating temperatures on the COP of the single system. From Fig. 4, the COP of the [mmIm]DMP-CH3OH absorption refrigeration is lower than that of LiBr/ H2O absorption refrigeration under the same temperature conditions, but higher than that of H2O/NH3 absorption refrigeration under most temperature conditions. When the heat source temperature is greater than 400 K, [mmIm]DMP-CH3OH absorption is still possible with a high COP close to that of LiBr/H2O absorption refrigeration. In general, [mmim]DMP/methanol has excellent potential for application as the working pair in absorption refrigeration. H2O/NH3 absorption refrigeration under most temperature conditions. When the heat source temperature is greater than 400 K, [mmIm]DMP-CH3OH absorption is still possible with a high COP close to that of LiBr/H2O absorption refrigeration. In general, [mmim]DMP/methanol has excellent potential for application as the working pair in absorption refrigeration.

Dong et al. [5] studied the thermophysical properties of the [dmIm]DMP-H2O system. The vapor

0.10 to 0.90 were measured and correlated using a non-random two-liquid (NRTL) model. The

range of 303.15–353.15 K were determined by a BT2.15 Calvet microcalorimeter. Fig. 6 presents

in the range of 0.10 to 0.90 and a temperature

, in the range of

pressures of the [dmIm]DMP-H2O system at mass fractions of ionic liquids,

experimental data and the model predictions are presented in Fig. 5.

the experimental data and the corresponding correlation results.

Fig. 4. Effects of operating temperatures on the COP of the [mmIm]DMP-CH3OH system. **Figure 4.** Effects of operating temperatures on the COP of the [mmIm]DMP-CH3OH system.

**2.2 [dmIm]DMP-H2O**

Fig. 5. Vapor pressures of [dmIm]DMP-H2O.

The heat capacities of [dmIm]DMP-H2O at a

3

#### **2.2. [dmIm]DMP-H2O**

Dong et al. [5] studied the thermophysical properties of the [dmIm]DMP-H2O system. The vapor pressures of the [dmIm]DMP-H2O system at mass fractions of ionic liquids, *ω*, in the range of 0.10 to 0.90 were measured and correlated using a non-random two-liquid (NRTL) model. The experimental data and the model predictions are presented in Fig. 5.

**Figure 5.** Vapor pressures of [dmIm]DMP-H2O.

The heat capacities of [dmIm]DMP-H2O at a *ω* in the range of 0.10 to 0.90 and a temperature range of 303.15–353.15 K were determined by a BT2.15 Calvet microcalorimeter. Fig. 6 presents the experimental data and the corresponding correlation results.

**Figure 6.** Heat capacities of [dmIm]DMP-H2O.

The performance characteristics of [dmIm]DMP-H2O and LiBr-H2O single effect absorption refrigeration at *t*<sup>E</sup> = 10 °C, *t*<sup>C</sup> = 40 °C, *t*<sup>A</sup> = 30 °C, and *t*<sup>G</sup> = 80 °C were calculated and are listed in Table 1. It can be seen that the COP of the [dmIm]DMP-H2O system is slightly lower than but close to that of the traditional working pair LiBr-H2O.


**Table 1.** Comparison of performance characteristics between the [dmIm]DMP-H2O and LiBr-H2O systems

**Figure 7.** Effects of *t*G on the COPs of the [dmIm]DMP-H2O and LiBr-H2O systems.

Fig. 7 shows the effects of changes in the *t*G on the COP for the [dmIm]DMP-H2O and LiBr-H2O systems with *t*E = 10 °C, *t*C = 40 °C, and *t*A = 30 °C. It can be seen that as the *t*G is increased, the COPs of the [dmIm]DMP-H2O and LiBr-H2O systems stabilize after a sharp rise, and there is an optimum *t*G, at which the COP reaches the highest value. When reaching the stable stage, the COP of the [dmIm]DMP-H2O system is very close to that of the LiBr-H2O system. Moreover, the operating temperature range has been extended and operational safety has been achieved for the [dmIm]DMP-H2O working pair, because it has no limitation of crystallization. These findings indicate that [dmIm]DMP-H2O has the potential to be novel working pair for absorption refrigeration.

#### **2.3. [emIm]DMP-H2O**

**2.2. [dmIm]DMP-H2O**

626 Ionic Liquids - Current State of the Art

**Figure 5.** Vapor pressures of [dmIm]DMP-H2O.

**Figure 6.** Heat capacities of [dmIm]DMP-H2O.

Dong et al. [5] studied the thermophysical properties of the [dmIm]DMP-H2O system. The vapor pressures of the [dmIm]DMP-H2O system at mass fractions of ionic liquids, *ω*, in the range of 0.10 to 0.90 were measured and correlated using a non-random two-liquid (NRTL)

The heat capacities of [dmIm]DMP-H2O at a *ω* in the range of 0.10 to 0.90 and a temperature range of 303.15–353.15 K were determined by a BT2.15 Calvet microcalorimeter. Fig. 6 presents

the experimental data and the corresponding correlation results.

model. The experimental data and the model predictions are presented in Fig. 5.

Ren et al. [6] measured the vapor pressure of the [emIm]DMP-H2O binary system at different IL mole fractions, *x*, ranging from 0.1 to 0.5, and the experimental data were fitted using the NRTL model (Fig. 8).

**Figure 8.** Vapor pressure of [emIm]DMP-H2O with different mole fractions of IL and at different temperatures.

**Figure 9.** Specific heat capacities of [emIm]DMP-H2O with different mole fractions of IL and at different temperatures.

The specific heat capacities of the [emIm]DMP-H2O binary system were also measured at *T* = 298.15–323.15 K and with different IL mole fractions, *x*, ranging from 0.18 to 1. Fig. 9 presents the experimental data and the model predictions.

Based on the above thermophysical properties, Zhang et al. [7] investigated the thermody‐ namic performance of an absorption chiller employing the [emIm]DMP-H2O working pair. Fig. 10 shows the effect of *t*<sup>E</sup> on the COP of the system a*t t*G = 80 °C, *t*C = 40 °C, and *t*A = 35 °C. The results indicated that, at the same *t*C and *t*A, the COP of the [emIm]DMP-H2O system is less than that of an aqueous solution of LiBr-H2O but still greater than 0.7, whereas the *t*G is less than that of the LiBr-H2O system. Thus, [emIm]DMP-H2O has the potential to be a new working pair for use in an absorption chiller driven by low-grade waste heat or hot water generated by a common solar collector.

**Figure 10.** Effects of *t*E on the COP of the [emIm]DMP-H2O system.

#### **2.4. Summary**

**Figure 8.** Vapor pressure of [emIm]DMP-H2O with different mole fractions of IL and at different temperatures.

**Figure 9.** Specific heat capacities of [emIm]DMP-H2O with different mole fractions of IL and at different temperatures.

The specific heat capacities of the [emIm]DMP-H2O binary system were also measured at *T* = 298.15–323.15 K and with different IL mole fractions, *x*, ranging from 0.18 to 1. Fig. 9 presents

Based on the above thermophysical properties, Zhang et al. [7] investigated the thermody‐ namic performance of an absorption chiller employing the [emIm]DMP-H2O working pair. Fig. 10 shows the effect of *t*<sup>E</sup> on the COP of the system a*t t*G = 80 °C, *t*C = 40 °C, and *t*A = 35 °C. The results indicated that, at the same *t*C and *t*A, the COP of the [emIm]DMP-H2O system is less than that of an aqueous solution of LiBr-H2O but still greater than 0.7, whereas the *t*G is less than that of the LiBr-H2O system. Thus, [emIm]DMP-H2O has the potential to be a new working pair for use in an absorption chiller driven by low-grade waste heat or hot water

the experimental data and the model predictions.

628 Ionic Liquids - Current State of the Art

generated by a common solar collector.

All three of the working pairs described above possess good theoretical cycle characteristics that are better than those of H2O-NH3, but still slightly lower than those of LiBr-H2O. Due to the advantages of the negligible vapor pressure of the absorbent, no corrosion, and no crystallization, these three working pairs can be applied in a wider range of operating conditions than H2O-NH3 or LiBr-H2O. Therefore, it is expected that these three working pairs have enormous potential in industrial applications and strong possibilities for commercial development.

### **3. A new conceptual chemical absorption refrigeration working pair consisting of ammonia and a metal chloride-containing ionic liquid**

#### **3.1. The proposal**

Adsorption refrigeration is a type of environmentally friendly refrigeration that has been studied for many years. The most commonly used working pairs in adsorption refrigeration systems are ammonia–activated carbon, methanol–activated carbon, water–zeolite, ammonia– calcium chloride, and methanol–calcium chloride. The first three are physical adsorption working pairs, and the last two are chemical adsorption working pairs. The following review begins with the NH3-CaCl2 system.

Calcium chloride reacts with ammonia to form coordination compounds:

$$\begin{aligned} \text{CaCl}\_2 \cdot 8\text{NH}\_3 + \Delta H\_1 &\leftrightarrow \text{CaCl}\_2 \cdot 4\text{NH}\_3 + 4\text{NH}\_3 &\text{at } T\_{\text{el}}\\ \text{CaCl}\_2 \cdot 4\text{NH}\_3 + \Delta H\_2 &\leftrightarrow \text{CaCl}\_2 \cdot 2\text{NH}\_3 + 2\text{NH}\_3 &\text{at } T\_{\text{el}}\\ \text{CaCl}\_2 \cdot 2\text{NH}\_3 + \Delta H\_3 &\leftrightarrow \text{CaCl}\_2 + 2\text{NH}\_3 &\text{at } T\_{\text{el}} \end{aligned}$$

where ∆*H*1~∆*H*3 are the enthalpies of the reaction, and *T*e1~*T*e3 are the equilibrium temperatures. Benefiting from the reaction, the most impressive advantage of the NH3-CaCl2 system lies in its higher adsorption capacity compared to the others, while the main disadvantages are the low performances of heat and mass transfer and the phenomena of swelling and agglomeration in the process of adsorption [8]. Much effort has been spent attempting to overcome these defects. For example, Kai Wang et al. [9] proposed a new type of compound adsorbent composed of CaCl2 and an expanded graphite adsorbent, which could mitigate the deteriora‐ tion of the adsorption capacity that occurs in the long-term adsorption/desorption process. Using the compound adsorbent, Liwei Wang et al. [10] designed a multi-effect heat pipe-type adsorption refrigeration system, and a COP for their system of 0.39 was reported at a low *t*E of -20 °C. Obviously, these improvements have little effect, and many other similar efforts [11] proved futile. The essence of all these failures can be attributed to the fact that the adsorbent is a solid. Except for CaCl2, typical metal chlorides used as an ammonia adsorbent include SrCl2, LiCl2, and ZnCl2, among others. If only the solid metal chlorides could be dissolved in ionic liquids, there would be no problem that could not be solved in the absorption or adsorption systems. Fortunately, a few ionic liquids containing metal chlorides have been synthesized, including [bmim]Zn2Cl5 [12], that offer high hydrothermal stability and negligible vapor pressure to perfectly meet the absorbent criteria for absorption refrigeration. Compared with a solid adsorbent or other ionic liquids, the advantage of the ionic liquid [bmim]Zn2Cl5 is self-evident. The chemical reaction between NH3 and Zn2+ will largely enhance the solubility of NH3 in the absorbent and reduce the pressure of vapor phase as well as in the NH3-CaCl2 system, and no defects in heat and mass transfer, swelling, or agglomeration are a problem. Some other metal cations such as Ni2+ [13], Cu2+ [14], and Fe3+ [15] were also found to dissolve in ionic liquids, and thus, a family of new conceptual chemical absorption refrigeration working pairs consisting of ammonia and metal chloride-containing ionic liquids seems ready to be developed.

#### **3.2. VLE behavior of the binary system of NH3-[bmim]Zn2Cl5**

In order to reveal the promising latent application of NH3-[bmim]Zn2Cl5 as a working pair in absorption refrigeration, the vapor pressure data of the binary system of [bmim]Zn2Cl5/NH3 are urgently needed. In our previous work [16], VLE data for the binary system of NH3- [bmim]Zn2Cl5 were measured and fitted using the modified UNIFAC (Dortmund) model.


*3.2.1. Experimental data [16]*

Ionic Liquids Facilitate the Development of Absorption Refrigeration http://dx.doi.org/10.5772/58982 631


**Table 2.** The *P*-*T*-*x* data of binary solutions [bmim]Zn2Cl5 (1) + NH3 (2)

where ∆*H*1~∆*H*3 are the enthalpies of the reaction, and *T*e1~*T*e3 are the equilibrium temperatures. Benefiting from the reaction, the most impressive advantage of the NH3-CaCl2 system lies in its higher adsorption capacity compared to the others, while the main disadvantages are the low performances of heat and mass transfer and the phenomena of swelling and agglomeration in the process of adsorption [8]. Much effort has been spent attempting to overcome these defects. For example, Kai Wang et al. [9] proposed a new type of compound adsorbent composed of CaCl2 and an expanded graphite adsorbent, which could mitigate the deteriora‐ tion of the adsorption capacity that occurs in the long-term adsorption/desorption process. Using the compound adsorbent, Liwei Wang et al. [10] designed a multi-effect heat pipe-type adsorption refrigeration system, and a COP for their system of 0.39 was reported at a low *t*E of -20 °C. Obviously, these improvements have little effect, and many other similar efforts [11] proved futile. The essence of all these failures can be attributed to the fact that the adsorbent is a solid. Except for CaCl2, typical metal chlorides used as an ammonia adsorbent include SrCl2, LiCl2, and ZnCl2, among others. If only the solid metal chlorides could be dissolved in ionic liquids, there would be no problem that could not be solved in the absorption or adsorption systems. Fortunately, a few ionic liquids containing metal chlorides have been synthesized, including [bmim]Zn2Cl5 [12], that offer high hydrothermal stability and negligible vapor pressure to perfectly meet the absorbent criteria for absorption refrigeration. Compared with a solid adsorbent or other ionic liquids, the advantage of the ionic liquid [bmim]Zn2Cl5 is self-evident. The chemical reaction between NH3 and Zn2+ will largely enhance the solubility of NH3 in the absorbent and reduce the pressure of vapor phase as well as in the NH3-CaCl2 system, and no defects in heat and mass transfer, swelling, or agglomeration are a problem. Some other metal cations such as Ni2+ [13], Cu2+ [14], and Fe3+ [15] were also found to dissolve in ionic liquids, and thus, a family of new conceptual chemical absorption refrigeration working pairs consisting of ammonia and metal chloride-containing ionic liquids seems ready

to be developed.

630 Ionic Liquids - Current State of the Art

*3.2.1. Experimental data [16]*

**3.2. VLE behavior of the binary system of NH3-[bmim]Zn2Cl5**

In order to reveal the promising latent application of NH3-[bmim]Zn2Cl5 as a working pair in absorption refrigeration, the vapor pressure data of the binary system of [bmim]Zn2Cl5/NH3 are urgently needed. In our previous work [16], VLE data for the binary system of NH3- [bmim]Zn2Cl5 were measured and fitted using the modified UNIFAC (Dortmund) model.

**100x2** *pexp/kPa pcal/kPa* **100x2** *pexp/kPa pcal/kPa* **100x2** *pexp/kPa pcal/kPa*

85.78±0.04 67.4 65.6 91.34±0.05 699.6 704.5 87.34±0.05 1287.2 1188.0 86.84±0.04 77.4 77.0 92.39±0.05 829.7 828.8 89.05±0.05 1516.9 1517.5 87.76±0.05 87.7 88.6 93.67±0.06 1040.8 1029.9 90.58±0.06 1787.3 1787.5

90.42±0.06 134.9 137.2 84.79±0.03 488.6 489.8 83.62±0.02 1081.2 1080.8

88.92±0.05 103.5 106.7 *T* = 403.15 K *T* = 483.15 K

*T* = 323.15 K *T* = 383.15 K 86.04±0.04 1148.2 1148.5

The pressure-temperature-composition (*p*-*T*-*x*) data of the binary solutions [bmim]Zn2Cl5 (1) + NH3 (2) with NH3 mole fractions of *x*<sup>2</sup> = 0.83–0.94 at *T* = 323.15, 343.15, 363.15, 383.15, 403.15, 423.15, 443.15, 463.15, 483.15, 503.15, 523.15, 543.15, and 563.15 K are summarized in Table 2. The uncertainties in the NH3 mole fraction in the binary solution, which can be due to the random and systematic errors in the experimental method and the calculation accuracy of the ammonia equation of state (EOS), are also presented in the table.

#### *3.2.2. The modified UNIFAC (Dortmund) model [16]*

Because of the non-volatilization of the ionic liquid [bmim]Zn2Cl5, the vapor phase of the binary system [bmim]Zn2Cl5 (1) + NH3 (2) consists only of NH3, and the total pressure *p* of the binary solution can be given by [17],

$$p = \chi\_2 \gamma\_2 P\_2^{S^\*} \exp\left(\frac{(V\_2^L - B\_2)(p - P\_2^{S^\*})}{RT}\right) \tag{1}$$

where *x*2 is the mole fraction of NH3 in the binary solution, *γ*2 is the activity coefficient of NH3, *V*<sup>2</sup> <sup>L</sup> is the liquid mole volume of NH3, *B*<sup>2</sup> is the second virial coefficient in the ammonia EOS, and *P*<sup>2</sup> S' is the vapor pressure of pure NH3, When the temperature *T* is below the *T*C, *P*<sup>2</sup> S' is equal to the saturation vapor pressure, *P*<sup>S</sup> , which can be calculated by [18],

$$\ln P\_2^{S^\circ} = \frac{1}{T\_r - 0.101947} \left( -4.2522 T\_r^2 + 7.929448 T\_r^3 + 0.3807783 T\_r^7 - 4.039557 \right) \tag{2}$$

where *Tr* is the ratio of the solution temperature *T* and *T*C, When the temperature *T* is higher than the *T*C, the *P*<sup>2</sup> S' is defined as the pure NH3 pressure at *T* and the critical mole fraction *V*C, which can be calculated by the RK type EOS as follows [19]:

$$P\_2^{S^\*} = \frac{RT}{V\_C - b} - \frac{a(T)}{V\_C(V\_C + b)}\tag{3}$$

$$a(T) = 0.42748 \,\text{a}(T) R^2 T\_{\odot} \,^2 \Big/ P\_{\odot} \tag{4}$$

$$b = 0.08664 \, RT\_{\odot} / P\_{\odot} \tag{5}$$

where the temperature dependent term *α*(*T*) can be written by:

$$\alpha(T) = \sum\_{k=0}^{2} \beta\_k \left( 1/T\_r - T\_r \right)^k \tag{6}$$



**Table 3.** EOS constants and critical parameters for NH3

In the UNIFAC model, the excess Gibbs free energy is composed of two contributing parts, the combinatorial part and the residual part, and the activity coefficient *γ<sup>i</sup>* can be given as follows:

$$\ln \gamma\_i = \ln \gamma\_i^C + \ln \gamma\_i^R \tag{7}$$

where *γ<sup>i</sup>* C is the combinatorial activity coefficient and *γ<sup>i</sup>* R is the residual activity coefficient.

The combinatorial activity coefficient *γ<sup>i</sup>* <sup>C</sup> describes the repulsive interaction attributed to the molecular size and shape, which can be calculated by:

$$\ln \chi\_i^C = \ln \frac{\overline{\rho\_i}}{\chi\_i} + \frac{Z}{2} q\_i \ln \frac{\partial\_i}{\partial \overline{\rho\_i}} + l\_i - \frac{\overline{\rho\_i}}{\chi\_i} \sum\_j x\_j l\_j \tag{8}$$

$$dl\_i = \frac{Z}{2}(r\_i - q\_i) - (r\_i - 1) \tag{9}$$

$$\theta\_i = q\_i \mathbf{x}\_i / \sum\_j \mathbf{q}\_j \mathbf{x}\_j \tag{10}$$

$$\mathbf{p}\mathbf{p} = \mathbf{r}\_i \mathbf{x}\_i / \sum\_j \mathbf{r}\_j \mathbf{x}\_j \tag{11}$$

$$\mathbf{q}\_{i} = \sum \mathbf{v}\_{k}^{(i)} \mathbf{Q}\_{k} \tag{12}$$

$$r\_i = \sum \nu\_k^{(i)} R\_k \tag{13}$$

where *Z* is normally set to 10, *vk* (*i*) is the number of group *k* in component *i*, and *Rk* and *Qk* are the volume and surface parameters of the group *k*, respectively. The values of *Rk* and *Qk* for the group used in our experiment are listed in Table 4.


**Table 4.** Volume parameters *Rk* and surface parameters *Qk*

*3.2.2. The modified UNIFAC (Dortmund) model [16]*

binary solution can be given by [17],

632 Ionic Liquids - Current State of the Art

is equal to the saturation vapor pressure, *P*<sup>S</sup>

0.101947

which can be calculated by the RK type EOS as follows [19]:

'

2

where the temperature dependent term *α*(*T*) can be written by:

a

**Table 3.** EOS constants and critical parameters for NH3

*S*

NH3, *V*<sup>2</sup>

EOS, and *P*<sup>2</sup>

2

than the *T*C, the *P*<sup>2</sup>

*S*

*r*

*T*

Because of the non-volatilization of the ionic liquid [bmim]Zn2Cl5, the vapor phase of the binary system [bmim]Zn2Cl5 (1) + NH3 (2) consists only of NH3, and the total pressure *p* of the

22 2

where *x*2 is the mole fraction of NH3 in the binary solution, *γ*2 is the activity coefficient of

( ) ' 23 7

where *Tr* is the ratio of the solution temperature *T* and *T*C, When the temperature *T* is higher

*C CC RT a T <sup>P</sup>*

2 2 ( ) 0.42748 ( ) *C C a T* = a

*V b VV b*

( ) <sup>2</sup>

The EOS constants for NH3 *βk* and the critical parameters *T*C, *V*C, and *P*C are given in Table 3.

**β<sup>0</sup> β<sup>1</sup> β<sup>2</sup> TC / K PC / kPa VC / m3**

1.00027 0.45689 -0.05772 406.15 11424 0.00427

0 () 1 *<sup>k</sup> k rr*

 b *T TT* =

*k*

<sup>1</sup> ln 4.2522 7.929445 0.3807783 4.039557

*P TT T*

<sup>L</sup> is the liquid mole volume of NH3, *B*<sup>2</sup> is the second virial coefficient in the ammonia

S' is the vapor pressure of pure NH3, When the temperature *T* is below the *T*C, *P*<sup>2</sup>

*rr r*

( ) ( )

= -+ + - - (2)

S' is defined as the pure NH3 pressure at *T* and the critical mole fraction *V*C,

æ ö - - <sup>=</sup> ç ÷

( )( ) exp *L S*

'

*<sup>S</sup> V B pP pxP RT*

222

g

'

, which can be calculated by [18],

= - - + (3)

0.08664 *C C b RT P* = (5)

= - å (6)

*T RT P* (4)

è ø (1)

S'

**mol-1**

The residual activity coefficient *γ<sup>i</sup>* R accounts for the intermolecular forces resulting from the corresponding group interaction, which is described as the summation of the group activity coefficient *Γ* for group *k* of component *i*,

$$\ln \gamma\_i^R = \sum\_k^N \nu\_k^{(i)} (\ln \Gamma\_k - \ln \Gamma\_k^{(i)}) \tag{14}$$

where *Γk* and *Γ<sup>k</sup>* (*i*) are the activity coefficients for group *k* in binary solution and in the compo‐ nent *i*, respectively, and can be described as:

$$\ln \Gamma\_k = \mathcal{Q}\_i \left[ 1 - \ln(\sum\_{m=1}^N \theta\_m \phi\_{mk}) - \sum\_{m=1}^N (\frac{\theta\_m \phi\_{km}}{N}) \right] \tag{15}$$

$$\Theta\_m = \mathcal{Q}\_m X\_m \ne \sum\_{n=1}^N \mathcal{Q}\_n X\_n \tag{16}$$

$$X\_m = \sum\_{j=1}^{M} \mathbf{v}\_m^{(j)} \mathbf{x}\_j / \sum\_{j=1}^{M} \sum\_{n=1}^{N} \mathbf{v}\_n^{(j)} \mathbf{x}\_j \tag{17}$$

Eqs. 15–17 can also be used to calculate lnΓ*<sup>k</sup>* (*i*) , except that the group composition variable *Xm* is now the group fraction of group *k* in component *i*. For the modified UNIFAC (Dortmund) model, the group interaction parameters between groups *n* and *m*, *φn,m*, is described as:

$$\phi\_{n,m} = \exp(-\frac{a\_{nm}T^2 + b\_{nm}T + c\_{nm}}{T})\tag{18}$$

where *anm* (K-1), *bnm*, and *cnm* (K) are the adjustable interaction parameters for correlating the experimental vapor pressure data. The corresponding correlation results are listed in Table 5.


**Table 5.** Adjustable interaction parameters for UNIFAC model

( ) ( ) ln (ln ln ) *<sup>N</sup> Ri i i kk k*

ln [1 ln( ) ( )] *N N*

q f

 *Q X QX* =

= ==

(*i*)

*k k m mk N m m*

=- - å å

1 1

1 / *N m mm nn n*

( ) ( ) 1 11 / *M MN j j m mj n j j j n Xx x* n

is now the group fraction of group *k* in component *i*. For the modified UNIFAC (Dortmund) model, the group interaction parameters between groups *n* and *m*, *φn,m*, is described as: 2 , exp( ) *nm nm nm*

*aT bT c T*

where *anm* (K-1), *bnm*, and *cnm* (K) are the adjustable interaction parameters for correlating the experimental vapor pressure data. The corresponding correlation results are listed in Table 5.

**n m anm/K-1 bnm cnm/K** NH3 [bmim]+ 0.00617 -1.9768 -378.1920 NH3 Zn2+ -0.02754 2.7884 1119.6973 NH3 Cl- 0.03904 -39.2835 11459.2768 [bmim]+ NH3 -0.00915 -2.7566 2059.6872 [bmim]+ Zn2+ -0.01720 -12.6476 5508.8668 [bmim]+ Cl- 22.62142 -82.6723 -5307.7598 Zn2+ NH3 -0.00911 -3.0401 2070.4982 Zn2+ [bmim]+ -0.02991 -6.8487 12385.2753 Zn2+ Cl- 16.45937 66.7508 -2708.8868 Cl- NH3 -0.00281 -1.6807 -548.2453 Cl- [bmim]+ 0.09208 313.3495 -86.4292 Cl- Zn2+ 0.19296 59.8376 -5434.4813

 n

= =

<sup>=</sup> å *Γ Γ* - (14)

<sup>=</sup> å (16)

<sup>=</sup> å åå (17)

+ + = - (18)

, except that the group composition variable *Xm*

å (15)

are the activity coefficients for group *k* in binary solution and in the compo‐

1

=

q f

*n*

*m km*

q f

*n nm*

*k*

 n

g

*Γ Q*

q

*n m*

f

**Table 5.** Adjustable interaction parameters for UNIFAC model

where *Γk* and *Γ<sup>k</sup>*

(*i*)

634 Ionic Liquids - Current State of the Art

nent *i*, respectively, and can be described as:

Eqs. 15–17 can also be used to calculate lnΓ*<sup>k</sup>*

**Figure 11.** *P*-*T*-*x* phase diagrams (Lines: calculated with the UNIFAC model. Symbols: experimental data) [16].

The *P*-*T*-*x* phase diagrams with symbols for experimental data and lines for the UNIFAC model calculations are shown in Fig. 11. It can be seen from the figure that with an increase in the NH3 mole fraction, the vapor pressure also increases, and the rising trend becomes increasingly more obvious. With an increase in the binary solution temperature, the vapor pressure increases rapidly. When the temperature is below the *T*C, the rate of increase becomes more rapid, but when the temperature is higher than the *T*C of NH3, the rate of increase tends to slow and the vapor pressure even declines slightly.

**Figure 12.** Comparison of experimental data and the UNIFAC model calculations [16].

A comparison of the experimental data and the UNIFAC model calculations is shown in Fig. 12. All deviations are below 5.0% and are mainly produced by the uncertainties in the volumes of the high pressure vessel (0.7%) and liquid phase of binary system (1.1%), the weights of [bmim]Zn2Cl5 (0.01%) and NH3 (0.05%), the temperature distribution in the high pressure vessel (0.7%), EOS calculation accuracy (1.2%), and the fitting uncertainty (1.6%). Based on the above uncertainties, the total uncertainty of the measurements is estimated to be within 4.3%.

#### *3.2.3. Comparison with normal ionic liquids and ZnCl2*

Fig. 13 compares the vapor pressures of [bmim]Zn2Cl5/NH3 solutions with *x*2 = 0.9507 and 0.9231 to those of ZnCl2 6NH3 [23] at *T* = 330–420 K. When *x*<sup>2</sup> = 0.9507, the NH3 mass fraction of the binary solution is equal to the NH3 mass fraction in ZnCl2 6NH3, and when *x*2 = 0.9231, the mole ratio of NH3 and Zn2+ is 6, which is equal to that in ZnCl2 6NH3. From Fig. 13, the vapor pressures of NH3 in [bmim]Zn2Cl5 with *x*<sup>2</sup> = 0.9507 and 0.9231 are 2–3 times higher than that of ZnCl2 6NH3, which can be attributed to the effects of the ionic liquid [bmim]Cl on the complexation reaction production of NH3 and Zn2+ ions.

**Figure 13.** Comparison of vapor pressures of NH3-[bmim]Zn2Cl5 solution and ZnCl2 6NH3 [16].

Fig. 14 compares the vapor pressure of the NH3-[bmim]Zn2Cl5 solution and ammonia solutions containing ionic liquids [19] [emim][Ac], [emim][SCN], [emim][EtOSO3], and [DMEA][Ac] at *T* = 348 K and *x*2 = 0–1. From Fig. 14, the vapor pressure of NH3 in [bmim]Zn2Cl5 is one order of magnitude smaller than that in normal ionic liquids at *x*2 = 0–0.95, which means that the complexion reaction of NH3 and Zn2+ ions can largely reduce the vapor pressure of NH3 in the ionic liquid and largely enhance the solubility of NH3 in the ionic liquid.

Based on the results in Figs. 13 and 14, the absorption characteristics of [bmim]Zn2Cl5 are much better than those of normal ionic liquids but slightly lower than those of ZnCl2. Additionally,

**Figure 14.** Comparison of vapor pressures of [bmim]Zn2Cl5/NH3 solution and ammonia solutions containing the ionic liquids [emim][Ac], [emim][SCN], [emim][EtOSO3] and [DMEA][Ac] [16].

the liquid form of [bmim]Zn2Cl5 offers a major advantage over ZnCl2, which will completely resolve the abovementioned limitations to improve the cycle performance for ZnCl2/NH3 adsorption refrigeration. Therefore, working pairs of NH3-[bmim]Zn2Cl5 have good latent application potential in absorption refrigerator and heat pump operation.

#### **3.3. Heat capacities and excess enthalpies of the NH3-[bmim]Zn2Cl5 system**

In order to investigate the cycle characteristics of [bmim]Zn2Cl5/NH3 absorption refrigeration, data for the heat capacities of [bmim]Zn2Cl5 and excess enthalpies of [bmim]Zn2Cl5/NH3 are urgently needed. In our previous work, the heat capacities of [bmim]Zn2Cl5 for *T* = 210.15– 383.15 K were obtained by differential scanning calorimetry (DSC), and the excess enthalpies of [bmim]Zn2Cl5/NH3 at various ammonia mole fractions for *T* = 288.15–333.15 K were measured experimentally. The data for excess enthalpies were fit by a five-parameter NRTL model. Based on the heat capacity of [bmim]Zn2Cl5 and the excess enthalpy of [bmim]Zn2Cl5/ NH3, the enthalpies of [bmim]Zn2Cl5/NH3 solution at *x*1 = 0–1 for *T* = 273.15–343.15 K were calculated.

#### *3.3.1. Heat capacities of [bmim]Zn2Cl5*

A comparison of the experimental data and the UNIFAC model calculations is shown in Fig. 12. All deviations are below 5.0% and are mainly produced by the uncertainties in the volumes of the high pressure vessel (0.7%) and liquid phase of binary system (1.1%), the weights of [bmim]Zn2Cl5 (0.01%) and NH3 (0.05%), the temperature distribution in the high pressure vessel (0.7%), EOS calculation accuracy (1.2%), and the fitting uncertainty (1.6%). Based on the above uncertainties, the total uncertainty of the measurements is estimated to be within 4.3%.

Fig. 13 compares the vapor pressures of [bmim]Zn2Cl5/NH3 solutions with *x*2 = 0.9507 and 0.9231 to those of ZnCl2 6NH3 [23] at *T* = 330–420 K. When *x*<sup>2</sup> = 0.9507, the NH3 mass fraction of the binary solution is equal to the NH3 mass fraction in ZnCl2 6NH3, and when *x*2 = 0.9231, the mole ratio of NH3 and Zn2+ is 6, which is equal to that in ZnCl2 6NH3. From Fig. 13, the vapor pressures of NH3 in [bmim]Zn2Cl5 with *x*<sup>2</sup> = 0.9507 and 0.9231 are 2–3 times higher than that of ZnCl2 6NH3, which can be attributed to the effects of the ionic liquid [bmim]Cl on the

*3.2.3. Comparison with normal ionic liquids and ZnCl2*

636 Ionic Liquids - Current State of the Art

complexation reaction production of NH3 and Zn2+ ions.

**Figure 13.** Comparison of vapor pressures of NH3-[bmim]Zn2Cl5 solution and ZnCl2 6NH3 [16].

ionic liquid and largely enhance the solubility of NH3 in the ionic liquid.

Fig. 14 compares the vapor pressure of the NH3-[bmim]Zn2Cl5 solution and ammonia solutions containing ionic liquids [19] [emim][Ac], [emim][SCN], [emim][EtOSO3], and [DMEA][Ac] at *T* = 348 K and *x*2 = 0–1. From Fig. 14, the vapor pressure of NH3 in [bmim]Zn2Cl5 is one order of magnitude smaller than that in normal ionic liquids at *x*2 = 0–0.95, which means that the complexion reaction of NH3 and Zn2+ ions can largely reduce the vapor pressure of NH3 in the

Based on the results in Figs. 13 and 14, the absorption characteristics of [bmim]Zn2Cl5 are much better than those of normal ionic liquids but slightly lower than those of ZnCl2. Additionally, Fig. 15 shows the result of TG scanning for [bmim]Zn2Cl5, which was determined using a TGA/ SDT instrument. Onset of a 2.5% weight loss in a nitrogen atmosphere occurs at 676.15 K, and approximately 40% of the mass is gone by 774.15 K, which is a typical volatilization tempera‐ ture for imidazolium salts. Continued heating of [bmim]Zn2Cl5 eventually results in a constant weight near 1043.15 K, with a residual weight of 36.4%. These results indicate that [bmim]Zn2Cl5 possesses high thermal stability at *T* < 673.15 K.

**Figure 15.** Thermogravimetric (TG) scan results for [bmim]Zn2Cl5.

**Figure 16.** DSC scanning results for [bmim]Zn2Cl5: (a) variations in heat flow, *qm*, and specific heat capacity, *cp*, along with the temperature of [bmim]Zn2Cl5; (b) *cp*-*T* diagram of [bmim]Zn2Cl5 for *T* = 243.15–383.15 K.

Fig. 16(a) shows the variation in the specific heat flow, *qm*, and the specific heat capacity, *cp*, along with temperature variations for [bmim]Zn2Cl5. The specific heat flow increases with an increase in temperature. The rate of increase of the specific heat flow rises at *T* < 243.15 K, and the variation in the rate of increase of the specific heat flow is minor for *T* > 243.15 K. The specific heat capacity increases with an increase in the [bmim]Zn2Cl5 temperature at *T* < 243.15 K, and the rate of increase also increases. The specific heat capacity presents an increasing trend after a decline for *T* > 243.15 K. These results indicate that the melting temperature of [bmim]Zn2Cl5 is near 243.15 K. By careful observation, it is found that variations in the specific heat capacity with *T* > 251.15 K can be well fitted by the following quadratic equation:

$$\mathcal{L}\_p = 2.39327 \text{--} 0.00691T + 0.000011767 \, T^2 \tag{19}$$

Fig. 16(b) shows the *cp*-*T* diagram for *T* = 243.15–383.15 K. The symbols represent the experi‐ mental data, and the lines represent the calculated conic curve. It can be seen that variation in the specific heat capacity with temperature is accurately described by the quadratic curve.

#### *3.3.2. Experimental data for excess enthalpy of NH3-[bmim]Zn2Cl5*

Temperature-component-molar excess enthalpy (*T*-*x-H*E) data for the binary systems [bmim]Zn2Cl5 (2) + NH3 (1) at ammonia mole fractions of *x*<sup>1</sup> = 0.60–0.95 and *T* = 288.15, 303.15, 318.15, and 333.15 K are summarized in Table 6. Uncertainties in the temperature, ammonia mole fraction, and molar excess enthalpy are also presented in the table. The uncertainties are due to random errors as well as systematic errors for the experimental apparatus and the calculation accuracy of the UNIFAC model for the VLE of NH3/[bmim]Zn2Cl5. With an increase in the ammonia mole fraction, *x*1, the molar excess enthalpy presents an increasing trend after an initial decline.


**Table 6.** Mole excess enthalpy of binary systems [bmim]Zn2Cl5 (1) + NH3 (2).

#### *3.3.3. NRTL model*

**Figure 15.** Thermogravimetric (TG) scan results for [bmim]Zn2Cl5.

638 Ionic Liquids - Current State of the Art

**Figure 16.** DSC scanning results for [bmim]Zn2Cl5: (a) variations in heat flow, *qm*, and specific heat capacity, *cp*, along

Fig. 16(a) shows the variation in the specific heat flow, *qm*, and the specific heat capacity, *cp*, along with temperature variations for [bmim]Zn2Cl5. The specific heat flow increases with an increase in temperature. The rate of increase of the specific heat flow rises at *T* < 243.15 K, and the variation in the rate of increase of the specific heat flow is minor for *T* > 243.15 K. The specific heat capacity increases with an increase in the [bmim]Zn2Cl5 temperature at *T* < 243.15 K, and the rate of increase also increases. The specific heat capacity presents an increasing trend after a decline for *T* > 243.15 K. These results indicate that the melting temperature of [bmim]Zn2Cl5 is near 243.15 K. By careful observation, it is found that variations in the specific

heat capacity with *T* > 251.15 K can be well fitted by the following quadratic equation:

<sup>2</sup> 2.39327-0.00691 0.000011767 *<sup>p</sup> c TT* = + (19)

with the temperature of [bmim]Zn2Cl5; (b) *cp*-*T* diagram of [bmim]Zn2Cl5 for *T* = 243.15–383.15 K.

Based on the local composition representation of the excess Gibbs energy, *G*E, Renon and Prausnitz [24] proposed the NRTL model. The *G*E for the NRTL model can be described by:

$$\frac{G^{\ominus}}{RT} = \mathbf{x}\_1 \mathbf{x}\_2 \left( \frac{\mathbf{r}\_{21} G\_{21}}{\mathbf{x}\_1 + \mathbf{x}\_2 G\_{21}} + \frac{\mathbf{r}\_{12} G\_{12}}{\mathbf{x}\_2 + \mathbf{x}\_1 G\_{12}} \right) \tag{20}$$

$$
\pi\_{12} = \frac{\mathbf{g}\_{12} - \mathbf{g}\_{22}}{RT}, \pi\_{21} = \frac{\mathbf{g}\_{21} - \mathbf{g}\_{11}}{RT} \tag{21}
$$

$$G\_{12} = \exp\left(-a\tau\_{12}\right), G\_{21} = \exp\left(-a\tau\_{21}\right) \tag{22}$$

where *g*ij and *g*jj are the interaction energy between ij and jj component pairs, respectively, and *α* is the non-random parameter. The relationship between *G*E and the activity coefficient is given by:

$$\log G^{\ominus} = RT \sum\_{l} \ln \chi\_{l} + \sum\_{k \ast l} \left(\frac{\partial G^{\ominus}}{\partial \mathbf{x}\_{l}}\right)\_{T, P, x\_{/\partial k}} \tag{23}$$

Therefore, the activity coefficients of components 1 and 2 in a binary mixture can be written as:

$$\ln \mathcal{V}\_1 = \mathbf{x}\_2^2 \left[ \tau\_{21} \left( \frac{G\_{21}}{\mathbf{x}\_1 + \mathbf{x}\_2 G\_{21}} \right)^2 + \left( \frac{\tau\_{12} G\_{12}}{\left( \mathbf{x}\_2 + \mathbf{x}\_1 G\_{12} \right)^2} \right) \right] \tag{24}$$

$$\ln \gamma\_2 = \mathbf{x}\_1^2 \left[ \tau\_{12} \left( \frac{G\_{12}}{\mathbf{x}\_2 + \mathbf{x}\_1 G\_{12}} \right)^2 + \left( \frac{\tau\_{21} G\_{21}}{\left( \mathbf{x}\_1 + \mathbf{x}\_2 G\_{21} \right)^2} \right) \right] \tag{25}$$

The definition of the activity coefficient for ammonia, *γ*1, is presented in our previous work [16]. For the NRTL model, the interaction energy between the ij and jj component pairs are defined as:

$$\mathbf{g}\_{12} - \mathbf{g}\_{22} = A\_{\text{l}} + B\_{\text{l}}T \tag{26}$$

$$\mathbf{g}\_{21} - \mathbf{g}\_{11} = A\_2 + B\_2 T \tag{27}$$

The Gibbs-Helmholtz equation for excess enthalpy is:

$$-\frac{H^{\ominus}}{T^{2}} = \left[\frac{\partial \left(G^{\ominus}/T\right)}{\partial T}\right]\_{p,\chi} \tag{28}$$

For the five-parameter NRTL model, the excess enthalpy can be calculated by:

$$\mathbf{H}^{\ominus} = -\mathbf{x}\_{\mathbf{l}}\mathbf{x}\_{2}\left\{ \frac{A\_{2}G\_{21}\left[\mathbf{x}\_{1}\tau\_{21}\left(\alpha\tau\_{21}-1\right) - \mathbf{x}\_{2}G\_{21}\right]}{\left(\mathbf{x}\_{1}+\mathbf{x}\_{2}G\_{21}\right)^{2}} + \frac{A\_{\rm l}G\_{12}\left[\mathbf{x}\_{2}\tau\_{12}\left(\alpha\tau\_{12}-1\right) - \mathbf{x}\_{\rm l}G\_{12}\right]}{\left(\mathbf{x}\_{2}+\mathbf{x}\_{\rm l}G\_{12}\right)^{2}} \right\} \tag{29}$$

The correlation results are shown in Table 7.


**Table 7.** Binary parameters and non-random parameters for NRTL model

<sup>12</sup> () () 12 21 <sup>21</sup> *G G* =- =- exp , exp at

ln

*<sup>G</sup> G RT <sup>x</sup>* g

*i*

*G G <sup>x</sup> x xG x xG*

*G G <sup>x</sup> x xG x xG*

( ) <sup>E</sup> <sup>E</sup>

E 2 21 1 21 21 2 21 1 12 2 12 12 1 12 1 2 2 2

 <sup>ë</sup> û ë <sup>û</sup> <sup>=</sup> - + í ý ï ï + + î þ

1 2 21 2 1 12

*AG x* 1 1 *x G AG x x G*

*x xG x xG* ì ü ï ï é -- ù é -- ù

ê ú ¶ ë û

*H G T T T* é ù ¶ - = ê ú

*P x*,

2

( ) ( )

t at

The correlation results are shown in Table 7.

For the five-parameter NRTL model, the excess enthalpy can be calculated by:

E

 t

> t

ln

ln

g

The Gibbs-Helmholtz equation for excess enthalpy is:

g

given by:

640 Ionic Liquids - Current State of the Art

defined as:

*H xx*

where *g*ij and *g*jj are the interaction energy between ij and jj component pairs, respectively, and *α* is the non-random parameter. The relationship between *G*E and the activity coefficient is

*i ki i TPx*

Therefore, the activity coefficients of components 1 and 2 in a binary mixture can be written as: 2 2 21 12 12 1 2 21 2

é ù æ öæ ö = + ê ú ç ÷ç ÷ è øè ø + + ë û

1 2 21 2 1 12

2 1 12 1 2 21

The definition of the activity coefficient for ammonia, *γ*1, is presented in our previous work [16]. For the NRTL model, the interaction energy between the ij and jj component pairs are

2 2 12 21 21 2 1 12 2

é ù æ öæ ö = + ê ú ç ÷ç ÷ è øè ø + + ë û

E

<sup>¹</sup> <sup>¹</sup> æ ö ¶

, ,

( )

( )

t

t

*j ik*

= + ç ÷ ¶ è ø å å (23)

12 22 1 1 *g g A BT* - =+ (26)

21 11 2 2 *g g A BT* - =+ (27)

( ) ( )

t at

at

(22)

(24)

(25)

(28)

(29)

**Figure 17.** *T*-*x-H*E diagram of [bmim]Zn2Cl5/NH3 for *T* = 288.15–333.15 K.

Fig. 17 shows the *T*-*x-H*E diagram for the binary system of [bmim]Zn2Cl5/NH3 at *T* = 288.15– 333.15 K. The symbols represent the experimental data, and the lines represent the calculations from the NRTL model. From Fig. 17, with an increase in the NH3 mole fraction, the excess enthalpy shows an increasing trend after a decline. There are minimum excess enthalpies for each temperature: -6555.7, -6707.1, -6846.3, and -6974.7) J/mol appear at *x*<sup>1</sup> = 0.772, 0.774, 0.776, and 0.777 for *T* = 288.15, 303.15, 318.15, and 333.15 K, respectively.

**Figure 18.** Absolute deviations and relative deviations with the NRTL model.

Fig. 18 shows the absolute deviations and relative deviations between the experimental data and the values predicted by the NRTL model for excess enthalpy data. The results indicate that all deviations for excess enthalpy data are less than 3.9%. The measurement deviations are mainly produced by the uncertainties in volumes of the high pressure vessels (0.5%), the little tank (0.5%) and the liquid phase of binary system (0.2%); the weights of [bmim]Zn2Cl5 (0.01%), NH3 (0.05%) and water (0.01%); temperature distributions in the water bath (1.4%) and the bath container (1.2%); and the UNIFAC calculation accuracies (0.9%). Based on the above uncertainties, the total uncertainty of measurement is estimated to be less than 4.8 %.

#### *3.3.4. Enthalpy of [bmim]Zn2Cl5/NH3 solution*

The enthalpy of a solution of [bmim]Zn2Cl5/NH3 at *T* and a given NH3 mass fraction *ω1* can be calculated by:

$$h = (1 - \alpha\_1)h\_1 + \alpha\_1 h\_2 + h^E \tag{30}$$

The enthalpy of [bmim]Zn2Cl5, *h*1, can be calculated by:

$$h\_t = \int\_{T\_0}^{T} c\_p \, dT \tag{31}$$

where *cp* is the specific heat capacity of [bmim]Zn2Cl5, which can be calculated using Eq. (1), and *T*0 is defined as 273.15 K. The enthalpy of NH3 can be calculated by [19]:

$$h\_2 = \sum\_{l=0}^{6} a\_l \left( T - 273.15 \right)^l \tag{32}$$

**Figure 19.** Calculations for enthalpies of [bmim]Zn2Cl5/NH3 solution at *x*1 = 0–1 for *T* = 273.15–343.15 K.

Fig. 19 shows the calculated enthalpies for [bmim]Zn2Cl5/NH3 solution at *ω*1 = 0–1 and *T* = 273.15–343.15 K. Based on the VLE properties and the enthalpies of [bmim]Zn2Cl5/NH3 solutions, the thermodynamic performances of the [bmim]Zn2Cl5/NH3 absorption system can be investigated.

#### **3.4. Thermodynamic analysis of an absorption system using NH3-[bmim]Zn2Cl5 as the working pair [25]**

In our previous work, the modified UNIFAC model was used to describe the VLE properties of [bmim]Zn2Cl5/NH3 [16] and the NRTL model was used to predict the excess enthalpic properties of [bmim]Zn2Cl5/NH3. Based on a single-effect absorption refrigeration model, the thermodynamic performance of the [bmim]Zn2Cl5/NH3 absorption system was simulated and compared with that of the NaSCN/NH3 adsorption system. The coefficients of performance for cooling (COP) and heating (COP\* ) and circulation ratios under the condition of a subzero evaporating temperature were calculated and analyzed.

#### *3.4.1. System description and simulation*

Fig. 18 shows the absolute deviations and relative deviations between the experimental data and the values predicted by the NRTL model for excess enthalpy data. The results indicate that all deviations for excess enthalpy data are less than 3.9%. The measurement deviations are mainly produced by the uncertainties in volumes of the high pressure vessels (0.5%), the little tank (0.5%) and the liquid phase of binary system (0.2%); the weights of [bmim]Zn2Cl5 (0.01%), NH3 (0.05%) and water (0.01%); temperature distributions in the water bath (1.4%) and the bath container (1.2%); and the UNIFAC calculation accuracies (0.9%). Based on the above uncertainties, the total uncertainty of measurement is estimated to be less than 4.8 %.

The enthalpy of a solution of [bmim]Zn2Cl5/NH3 at *T* and a given NH3 mass fraction *ω1* can be

 w

where *cp* is the specific heat capacity of [bmim]Zn2Cl5, which can be calculated using Eq. (1),

( ) <sup>6</sup>

273.15 *<sup>i</sup>*

1 1 12 *h h hh* =- + + (1 ) w

> 0 1

and *T*0 is defined as 273.15 K. The enthalpy of NH3 can be calculated by [19]:

0

*i i h aT* =

**Figure 19.** Calculations for enthalpies of [bmim]Zn2Cl5/NH3 solution at *x*1 = 0–1 for *T* = 273.15–343.15 K.

2

*T*

E

(30)

*<sup>p</sup> <sup>T</sup> h c dT* <sup>=</sup> ò (31)

= - å (32)

*3.3.4. Enthalpy of [bmim]Zn2Cl5/NH3 solution*

The enthalpy of [bmim]Zn2Cl5, *h*1, can be calculated by:

calculated by:

642 Ionic Liquids - Current State of the Art

Fig. 20 shows a schematic diagram of the single effect absorption system. The main system is composed of the generator (G), the absorber (A), the condenser (C), the evaporator (E), the regenerator (R), the valves (V), and the solution pump (P). In Fig. 20, the status point numbers are given, and the fluids at each point are marked: *i* denotes the [bmim]Zn2Cl5/NH3 solution from the absorber with a high NH3 mass fraction, *ii* denotes the [bmim]Zn2Cl5/NH3 solution from the generator with a low NH3 mass fraction, and *iii* denotes the refrigerant NH3. The symbols *q*E, *q*C, *q*A, and *q*G represent the heat flow of the evaporator, the condenser, the absorber, and the generator, respectively.

**Figure 20.** Schematic diagram of the single effect absorption system [25].

In order to simulate the thermodynamic performance of an absorption system using [bmim]Zn2Cl5/NH3 as a working pair, several assumptions were made as follows [25]:


The mass and energy conservation equations for the evaporator are given by:

$$m\_{\S} = m\_{\S} \tag{33}$$

$$q\_{\rm E} = m\_6 h\_6 - m\_8 h\_8 \tag{34}$$

The mass and energy conservation equations for the condenser are given by:

$$m\_1 = m\_4 \tag{35}$$

$$q\_C = m\_1 h\_1 - m\_4 h\_4 \tag{36}$$

For the absorber, the mass conservation equation of the solution, the mass conservation equation of IL, and the energy conservation equation are given by:

$$m\_{\uparrow} = m\_{6} + m\_{8} \tag{37}$$

$$m\_{\uparrow}(1-\alpha\_{\uparrow}) = m\_{6} + m\_{8}(1-\alpha\_{8})\tag{38}$$

$$q\_A = m\_6 h\_6 + m\_8 h\_8 - m\_7 h\_7 \tag{39}$$

For the generator, the mass conservation equation of the solution, the mass conservation equation of IL, and the energy conservation equation are given by:

$$m\_2 = m\_1 + m\_3 \tag{40}$$

Ionic Liquids Facilitate the Development of Absorption Refrigeration http://dx.doi.org/10.5772/58982 645

$$m\_2(\mathbf{l} - \alpha\_2) = m\_1 + m\_3(\mathbf{l} - \alpha\_3) \tag{41}$$

$$q\_G = m\_1 h\_1 + m\_3 h\_3 - m\_2 h\_2 \tag{42}$$

For the regenerator, the energy conservation equation is given by:

$$m\_2(h\_2 - h\_9) = m\_3(h\_{10} - h\_8) \tag{43}$$

Based on the above assumptions and the conservation equations for mass and energy conser‐ vation, the heat flow values of *q*G, *q*C, *q*E, and *q*A; mass flow values of *m*2 and *m*3; and mass fractions of *ω*2 and *ω*3, can be calculated. The circulation ratio (*f*) is calculated by:

$$f = \frac{m\_3}{m\_1} = \frac{1 - \alpha\_3}{\alpha\_2 - \alpha\_3} \tag{44}$$

The COP for cooling is defined by:

In order to simulate the thermodynamic performance of an absorption system using [bmim]Zn2Cl5/NH3 as a working pair, several assumptions were made as follows [25]:

**2.** The vapor pressure losses are neglected, the pressure of the evaporator is equal to that of the absorber, and the pressure of the condenser is equal to that of the generator;

**3.** The refrigerant flowing out of the condenser is in a saturated liquid state, and the

*m m* 5 6 = (33)

*m m* 1 4 = (35)

C 11 4 4 *q mh mh* = - (36)

*mmm* 7 68 = + (37)

A 66 88 77 *q mh mh mh* =+- (39)

*m mm* <sup>213</sup> = + (40)

(38)

E 66 55 *q mh mh* = - (34)

refrigerant flowing out of the evaporator is in a saturated gas state;

The mass and energy conservation equations for the evaporator are given by:

The mass and energy conservation equations for the condenser are given by:

equation of IL, and the energy conservation equation are given by:

equation of IL, and the energy conservation equation are given by:

For the absorber, the mass conservation equation of the solution, the mass conservation

7 7 68 8 *m mm* (1 ) (1 ) - =+ -

For the generator, the mass conservation equation of the solution, the mass conservation

 w

w

**4.** The heat recovery rate of the regenerator is set to 0.80 [26]; and

**5.** The thermal losses and pumping work are negligible.

**1.** The simulation is conducted under steady state;

644 Ionic Liquids - Current State of the Art

$$COP = \frac{q\_{\oplus}}{q\_{\oplus}} \tag{45}$$

The exergy efficiency (*η*ex) for cooling is given by:

$$\eta\_{\rm ex} = \frac{q\_{\rm E} \left(\frac{T\_0}{T\_{\rm E}} - 1\right)}{q\_{\rm G} \left(1 - \frac{T\_0}{T\_{\rm G}}\right)}\tag{46}$$

The COP\* for heating is defined as:

$$\text{LCOP}^\* = \frac{q\_\text{A} + q\_\text{C}}{q\_\text{G}} = \text{l} + \frac{q\_\text{E}}{q\_\text{G}} \tag{47}$$

The exergy efficiency for cooling (*η\** ex) is given by:

$$\eta\_{\alpha}^{\*} = \frac{q\_{\mathbb{C}} \left(1 - \frac{T\_0}{T\_{\mathbb{C}}}\right) + q\_{\mathbb{A}} \left(1 - \frac{T\_0}{T\_{\mathbb{A}}}\right)}{q\_{\mathbb{G}} \left(1 - \frac{T\_0}{T\_{\mathbb{G}}}\right)}\tag{48}$$

#### *3.4.2. Results and discussion*

Fig. 22 shows variations in the COP and *η*ex of [bmim]Zn2Cl5/NH3 absorption refrigeration with variations in *t*A and *t*<sup>C</sup> at a *t*<sup>G</sup> = 90 °C and *t*<sup>E</sup> = -10 °C. These results show that both the COP and *η*ex decline with increases in *t*A and *t*C. This is because increases in *t*A and *t*C lead to a decrease in the mass fraction of solution from the absorber (*ω*2) and an increase in the mass fraction of solution from the generator (*ω*3). These changes in both *ω*2 and *ω*3 result in a decrease of *q*E. The slopes of both the COP and *η*ex curves are less steep when *t*A and *t*C are lower, and as *t*<sup>A</sup> and *t*C continue to increase, the slopes become increasingly steep. This can be explained by the fact that,with the continuous increases in *t*A and *t*C, the difference between *ω*2 and *ω*3 continues to become smaller. By comparison, the thermal performance of the [bmim]Zn2Cl5/NH3 system is better than that of the NaSCN/NH3 system when *t*A and *t*<sup>C</sup> are low. However, when *t*A and *t*C are high, the thermal performance of the NaSCN/NH3 system is better than that of the [bmim]Zn2Cl5/NH3 system, and the upper operating limit of *t*A and *t*C for NaSCN/NH3 is higher than that for the [bmim]Zn2Cl5/NH3 system. This can also be explained by the properties of NH3 solubility in [bmim]Zn2Cl5 and NaSCN. The higher solubility of NH3 in [bmim]Zn2Cl5 ensures that the [bmim]Zn2Cl5/NH3 system possesses better thermal performance than the NaSCN/NH3 system with operating conditions of low *t*A and *t*C. The stronger combination of NH3 and [bmim]Zn2Cl5 demonstrates that the upper operating limit of *t*A and *t*C for the [bmim]Zn2Cl5/NH3 system are lower than those of the NaSCN/NH3 system.

**Figure 21.** Variations in the COP and *η*ex of [bmim]Zn2Cl5/NH3 absorption refrigeration with changes in *T*G for *t*A = *t*C = 25 °C and *t*E= -10 °C [25].

Fig. 23 shows the effects of *t*<sup>G</sup> on the COP for *t*<sup>G</sup> = 110–230 °C with *t*A = *t*<sup>C</sup> = 35 °C and *t*E= -10 °C, -20 °C, -30 °C, or -40 °C. With an increase in *t*C, the COP presents a trend of first increasing and the decreasing. The reason for this trend is that the increase in *t*G has both positive and negative effects on the COP. The positive and negative effects are the same as indicated by the analysis of the trend in COP shown in Fig. 2. When *t*<sup>G</sup> is lower, the positive effect is predominant, and

*3.4.2. Results and discussion*

646 Ionic Liquids - Current State of the Art

25 °C and *t*E= -10 °C [25].

Fig. 22 shows variations in the COP and *η*ex of [bmim]Zn2Cl5/NH3 absorption refrigeration with variations in *t*A and *t*<sup>C</sup> at a *t*<sup>G</sup> = 90 °C and *t*<sup>E</sup> = -10 °C. These results show that both the COP and *η*ex decline with increases in *t*A and *t*C. This is because increases in *t*A and *t*C lead to a decrease in the mass fraction of solution from the absorber (*ω*2) and an increase in the mass fraction of solution from the generator (*ω*3). These changes in both *ω*2 and *ω*3 result in a decrease of *q*E. The slopes of both the COP and *η*ex curves are less steep when *t*A and *t*C are lower, and as *t*<sup>A</sup> and *t*C continue to increase, the slopes become increasingly steep. This can be explained by the fact that,with the continuous increases in *t*A and *t*C, the difference between *ω*2 and *ω*3 continues to become smaller. By comparison, the thermal performance of the [bmim]Zn2Cl5/NH3 system is better than that of the NaSCN/NH3 system when *t*A and *t*<sup>C</sup> are low. However, when *t*A and *t*C are high, the thermal performance of the NaSCN/NH3 system is better than that of the [bmim]Zn2Cl5/NH3 system, and the upper operating limit of *t*A and *t*C for NaSCN/NH3 is higher than that for the [bmim]Zn2Cl5/NH3 system. This can also be explained by the properties of NH3 solubility in [bmim]Zn2Cl5 and NaSCN. The higher solubility of NH3 in [bmim]Zn2Cl5 ensures that the [bmim]Zn2Cl5/NH3 system possesses better thermal performance than the NaSCN/NH3 system with operating conditions of low *t*A and *t*C. The stronger combination of NH3 and [bmim]Zn2Cl5 demonstrates that the upper operating limit of *t*A and *t*C for the

[bmim]Zn2Cl5/NH3 system are lower than those of the NaSCN/NH3 system.

**Figure 21.** Variations in the COP and *η*ex of [bmim]Zn2Cl5/NH3 absorption refrigeration with changes in *T*G for *t*A = *t*C =

Fig. 23 shows the effects of *t*<sup>G</sup> on the COP for *t*<sup>G</sup> = 110–230 °C with *t*A = *t*<sup>C</sup> = 35 °C and *t*E= -10 °C, -20 °C, -30 °C, or -40 °C. With an increase in *t*C, the COP presents a trend of first increasing and the decreasing. The reason for this trend is that the increase in *t*G has both positive and negative effects on the COP. The positive and negative effects are the same as indicated by the analysis of the trend in COP shown in Fig. 2. When *t*<sup>G</sup> is lower, the positive effect is predominant, and

**Figure 22.** Variations in the COP and *η*ex of [bmim]Zn2Cl5/NH3 absorption refrigeration with varying *t*A and *t*C for *t*G = 90 °C and *t*E = -10 °C [25].

**Figure 23.** Effects of *T*G on the COP for *t*G = 110–230 °C, *t*A = *t*C = 35 °C and *t*E= -10 °C, -20 °C, -30 °C, or -40 °C [25].

the COP increases with an increase in *t*G. With a further increase in *t*G, the negative effect is gradually enhanced, and the rate at which the COP increases is continually reduced until it finally becomes negative. When *t*G is higher, the negative effect is predominant, and the COP decreases with an increase in *t*G. For *t*E= -10 °C, -20 °C, -30 °C, and -40 °C, the maximum COPs for the [bmim]Zn2Cl5/NH3 system of 0.54, 0.48, 0.42, and 0.35 appear at *t*<sup>G</sup> =133 °C, 161 °C, 188 °C, and 225 °C, respectively. When *t*<sup>E</sup> = -10 °C and -20 °C, the maximum COPs for the NaSCN/ NH3 system occurs at *t*G = 81 °C and 98 °C, respectively [27]. These results indicate that the required temperature of the heat source for the [bmim]Zn2Cl5/NH3 system is higher than that of the NaSCN/NH3 system.

**Figure 24.** Effects of *t*G on *f* for *t*G = 110–230 °C, *t*A = *t*C = 35 °C and *t*E= -10 °C, -20 °C, -30 °C, or -40 °C [25].

Fig. 24 shows effects of *t*<sup>G</sup> on the *f* for *t*<sup>G</sup> = 110–230 °C with *t*A = *t*<sup>C</sup> = 35 °C and *t*E= -10 °C, -20 °C, -30 °C, or -40 °C. With an increase in *t*G, the *f* declines, because the increase in *t*G is conducive to desorption of NH3 in the generator. With an increase in *t*E, the *f* grows, because the increase in *t*<sup>E</sup> will decrease the absorption pressure of the absorber. Thus, the absorption ability of the absorber will be decreased. The *f* is an important parameter for absorption refrigeration. An increase in the *f* will lead to an increase in the amount of energy used to heat the solution from *t*A to *t*G. If the *f* is greater than 10, the COP decreases, even when the efficiency of the regenerator is greater than 0.9. For *t*E = -10 °C, -20 °C, and -30 °C, the circulation ratios are less than 10 when *t*<sup>G</sup> is greater than 115°C, 139 °C, or 187 °C, respectively. Based on the results shown in Figs. 23 and 24, the [bmim]Zn2Cl5/NH3 system can be used when *t*E = -10 to -30 °C.

**Figure 25.** Effects of *t*G on the COP\* for *t*G = 170–300 °C with *t*A = *t*C = 55 °C, 57 °C, 59 °C, or 61 °C and *t*E = -10 °C [25].

Fig. 25 shows the effects of *t*G on the COP\* for *t*<sup>G</sup> = 170–300 °C with *t*A = *t*<sup>C</sup> = 55 °C, 57 °C, 59 °C, or 61 °C and *t*E = -10 °C. The COP\* presents a trend of declining after increasing with an increase in *t*G. For *t*A = *t*C = 55 °C, 57 °C, 59 °C, and 61 °C, the maximum COP\* values of 1.447, 1.422, 1.390, and 1.348 appear at *t*G =266 °C, 276 °C, 283 °C, and 289 °C, respectively. The COP\* decreases with increases in *t*A and *t*C, because the increase in *t*<sup>A</sup> is not conducive to absorption of NH3 in the absorber. In addition, the increase in *t*<sup>C</sup> is not conducive to desorption of NH3 in the generator.

**Figure 26.** Effects of *t*G on *η*\* ex for *t*G = 170–300 °C with *t*A = *t*C = 55 °C, 57 °C, 59 °C, or 61 °C and *t*E= -10 °C [25].

**Figure 24.** Effects of *t*G on *f* for *t*G = 110–230 °C, *t*A = *t*C = 35 °C and *t*E= -10 °C, -20 °C, -30 °C, or -40 °C [25].

and 24, the [bmim]Zn2Cl5/NH3 system can be used when *t*E = -10 to -30 °C.

**Figure 25.** Effects of *t*G on the COP\*

648 Ionic Liquids - Current State of the Art

or 61 °C and *t*E = -10 °C. The COP\*

Fig. 25 shows the effects of *t*G on the COP\*

Fig. 24 shows effects of *t*<sup>G</sup> on the *f* for *t*<sup>G</sup> = 110–230 °C with *t*A = *t*<sup>C</sup> = 35 °C and *t*E= -10 °C, -20 °C, -30 °C, or -40 °C. With an increase in *t*G, the *f* declines, because the increase in *t*G is conducive to desorption of NH3 in the generator. With an increase in *t*E, the *f* grows, because the increase in *t*<sup>E</sup> will decrease the absorption pressure of the absorber. Thus, the absorption ability of the absorber will be decreased. The *f* is an important parameter for absorption refrigeration. An increase in the *f* will lead to an increase in the amount of energy used to heat the solution from *t*A to *t*G. If the *f* is greater than 10, the COP decreases, even when the efficiency of the regenerator is greater than 0.9. For *t*E = -10 °C, -20 °C, and -30 °C, the circulation ratios are less than 10 when *t*<sup>G</sup> is greater than 115°C, 139 °C, or 187 °C, respectively. Based on the results shown in Figs. 23

for *t*G = 170–300 °C with *t*A = *t*C = 55 °C, 57 °C, 59 °C, or 61 °C and *t*E = -10 °C [25].

for *t*<sup>G</sup> = 170–300 °C with *t*A = *t*<sup>C</sup> = 55 °C, 57 °C, 59 °C,

presents a trend of declining after increasing with an increase

Fig. 26 shows the effects of *t*G on *η*\* ex for *t*G = 170–300 °C with *t*A = *t*<sup>C</sup> = 55 °C, 57 °C, 59 °C, or 61 °C and *t*E= -10 °C. With an increase in *t*G, the *η*\* ex presents a trend of decreasing after initially increasing. For *t*A = *t*C = 55 °C, 57 °C, 59 °C, and 61 °C, the maximum values of *η*\* ex of 0.341, 0.368, 0.384, and 0.402 appear at *t*<sup>G</sup> =193 °C, 206 °C, 221 °C, and 246 °C, respectively. It can be seen that the optimal*η*\* ex occurs at a lower *t*G than did the optimal COP\* . This is because the increase in *t*<sup>G</sup> leads to an increase in the exergy proportion in *q*E, which induces a decreasing trend in *η*\* ex but has no effect on the COP\* .

**Figure 27.** Effects of *t*G on the *f* for *t*G = 170–300 °C with *t*A = *t*C = 55 °C, 57 °C, 59 °C, or 61 °C and *t*E= -10 °C [25].

Fig. 27 shows the effects of *t*<sup>G</sup> on the *f* for *t*<sup>G</sup> = 170–300 °C with *t*A = *t*<sup>C</sup> = 55 °C, 57 °C, 59 °C, or 61 °C and *t*E= -10 °C. The variation in the *f* is the same as that for the COP for cooling in Fig. 6. For *t*A = *t*C = 55 °C, 57 °C, 59 °C, and 61 °C, the *f* values are less than 10 when *t*<sup>G</sup> is greater than 220 °C, 244 °C, 263 °C, and 285 °C, respectively. Overall, the results in Figs. 21–27 indicate that the [bmim]Zn2Cl5/NH3 absorption system is suitable for use in heating applications.

**Figure 28.** Effects of *t*G on the COP for *t*G = 75–130 °C with *t*A = 35°C, *t*C = 40 °C, and *t*E = 5 °C.

The theoretical cycle characteristic of the [bmim]Zn2Cl5/NH3 absorption system is also compared with that of the LiBr/H2O system. Fig. 28 shows the effects of *t*<sup>G</sup> on the COP for *t*G = 75–130 °C with *t*<sup>A</sup> = 35°C, *t*<sup>C</sup> = 40 °C, and *t*<sup>E</sup> = 5 °C and a heat recovery rate of the regenerator of 0.75. For both systems, the COP initially exhibits a significant increase as the *t*G increases. As *t*G continues to increase though, the slope of the COP curve for the [bmim]Zn2Cl5/NH3 system becomes less steep, whereas the COP curve for the LiBr/H2O system presents a trend of a slight decrease after the increase. When *t*G < 95°C, the COP of the [bmim]Zn2Cl5/NH3 system is slightly higher than that of the LiBr/H2O system. When 95 < *t*<sup>G</sup> < 115 °C, the COP of the [bmim]Zn2Cl5/NH3 system is slightly less than that of the LiBr/H2O system. When *t*G > 115 °C, the COP of the [bmim]Zn2Cl5/NH3 system is higher than that of the LiBr/H2O system. As *t*<sup>G</sup> continues to increase, the COP curve for the [bmim]Zn2Cl5/NH3 system still maintains the upward trend with a small slope, but the COP curve of the LiBr/H2O system shows a down‐ ward trend with a small slope. Although the COP of the [bmim]Zn2Cl5/NH3 system is less than the COP of the LiBr/H2O system at some specific temperatures, the overall theoretical cycle characteristic of the [bmim]Zn2Cl5/NH3 system is slightly better than that of the LiBr/H2O system, especially at higher *t*G values.

#### **3.5. Summary**

The vapor pressures of the binary solution of [bmim]Zn2Cl5/NH3 with NH3 mole fraction *x*2 = 0.83–0.94 at *T* = 323.15–563.15 K were measured via a static method with a total uncertainty of measurement below 4.3% [16]. The experimental data were fit using the modified UNIFAC model, and new group interaction parameters between any two of the four tested groups were obtained with a maximum deviation less than 5% [16]. Vapor pressures were compared between ZnCl2 6NH3, ammonia solutions containing normal ionic liquids ([emim][Ac], [emim] [SCN], [emim][EtOSO3] and [DMEA][Ac]), and [bmim]Zn2Cl5/NH3. The results indicate that the absorption characteristics of [bmim]Zn2Cl5 are much better than those of normal ionic liquids but slightly lower than those of ZnCl2. However, the liquid form of [bmim]Zn2Cl5 offers a major advantage over ZnCl2. Therefore, working pairs of [bmim]Zn2Cl5/NH3 have good latent application in absorption refrigerator and heat pump operation.

TG scanning of [bmim]Zn2Cl5 was carried out using a TGA/SDT instrument over the range of *T* = 323.15–1173.15 K. The results indicate that [bmim]Zn2Cl5 possesses high thermal stability for *T* < 637.15 K. Heat capacity data at *T* = 210.15–383.15 K were obtained by using a DSC 910S operated with a rate of temperature increase of 5 Kmin-1 and a nitrogen volume flow of 40 cm3 min-1. Molar excess enthalpy data for [bmim]Zn2Cl5/NH3 at *x*<sup>1</sup> = 0.60–0.95 for *T* = 288.15, 303.15, 318.15, and 333.15 K were measured. The excess enthalpy data were fit using the NRTL model. Measurement uncertainties and the maximum deviation of correlations for the excess enthalpy data were lower than 4.8% and 3.9%, respectively. With an increase in the NH3 mole fraction, excess enthalpy showed a trend of increasing after declining. Minimum excess enthalpies of -6555.7, -6707.1, -6846.3, and -6974.7 J/mol appeared at *x*<sup>1</sup> = 0.772, 0.774, 0.776, and 0.777 for *T* = 288.15, 303.15, 318.15, and 333.15 K, respectively. Based on the heat capacity of [bmim]Zn2Cl5 and the excess enthalpy of [bmim]Zn2Cl5/NH3, the enthalpies of [bmim]Zn2Cl5/ NH3 solutions can be calculated, which makes it feasible to investigate the thermodynamic performances of the [bmim]Zn2Cl5/NH3 absorption system.

Based on the modified UNIFAC model and the NRTL model, the thermodynamic performance of a single effect absorption system using [bmim]Zn2Cl5/NH3 as the working pair was simu‐ lated and compared with those of the NaSCN/NH3 adsorption system [25] and the H2O-LiBr absorption system. The thermal performance of the [bmim]Zn2Cl5/NH3 system is better than that of the NaSCN/NH3 system when the *t*G is high and *t*A and *t*C are low and also better than that of the H2O-LiBr absorption system in some cases. With an increase in *t*G, the COP and COP\* present trends of declining after increasing, and the circulation ratios show a decreasing trend. When *t*E = -30 °C and *t*A = *t*C =35 °C, the maximum COP of the [bmim]Zn2Cl5/NH3 system is still greater than 0.42. When *t*<sup>E</sup> = -10 °C and *t*A = *t*<sup>C</sup> = 60 °C, the maximum COP\* is still greater than 1.40. Under these two operating conditions, the circulation ratios remain acceptable. Although the COP of the [bmim]Zn2Cl5/NH3 system is less than that of the LiBr/H2O system in some specific temperature ranges, the overall theorertical cycle characteristic of the [bmim]Zn2Cl5/NH3 system is slightly better than that of the LiBr/H2O system, especially when tG is high. Overall, these results indicate that the [bmim]Zn2Cl5/NH3 absorption system offers good thermal performance for use in both cooling and heating applications.

#### **4. Conclusions and outlook**

Fig. 27 shows the effects of *t*<sup>G</sup> on the *f* for *t*<sup>G</sup> = 170–300 °C with *t*A = *t*<sup>C</sup> = 55 °C, 57 °C, 59 °C, or 61 °C and *t*E= -10 °C. The variation in the *f* is the same as that for the COP for cooling in Fig. 6. For *t*A = *t*C = 55 °C, 57 °C, 59 °C, and 61 °C, the *f* values are less than 10 when *t*<sup>G</sup> is greater than 220 °C, 244 °C, 263 °C, and 285 °C, respectively. Overall, the results in Figs. 21–27 indicate that

the [bmim]Zn2Cl5/NH3 absorption system is suitable for use in heating applications.

**Figure 28.** Effects of *t*G on the COP for *t*G = 75–130 °C with *t*A = 35°C, *t*C = 40 °C, and *t*E = 5 °C.

system, especially at higher *t*G values.

**3.5. Summary**

650 Ionic Liquids - Current State of the Art

The theoretical cycle characteristic of the [bmim]Zn2Cl5/NH3 absorption system is also compared with that of the LiBr/H2O system. Fig. 28 shows the effects of *t*<sup>G</sup> on the COP for *t*G = 75–130 °C with *t*<sup>A</sup> = 35°C, *t*<sup>C</sup> = 40 °C, and *t*<sup>E</sup> = 5 °C and a heat recovery rate of the regenerator of 0.75. For both systems, the COP initially exhibits a significant increase as the *t*G increases. As *t*G continues to increase though, the slope of the COP curve for the [bmim]Zn2Cl5/NH3 system becomes less steep, whereas the COP curve for the LiBr/H2O system presents a trend of a slight decrease after the increase. When *t*G < 95°C, the COP of the [bmim]Zn2Cl5/NH3 system is slightly higher than that of the LiBr/H2O system. When 95 < *t*<sup>G</sup> < 115 °C, the COP of the [bmim]Zn2Cl5/NH3 system is slightly less than that of the LiBr/H2O system. When *t*G > 115 °C, the COP of the [bmim]Zn2Cl5/NH3 system is higher than that of the LiBr/H2O system. As *t*<sup>G</sup> continues to increase, the COP curve for the [bmim]Zn2Cl5/NH3 system still maintains the upward trend with a small slope, but the COP curve of the LiBr/H2O system shows a down‐ ward trend with a small slope. Although the COP of the [bmim]Zn2Cl5/NH3 system is less than the COP of the LiBr/H2O system at some specific temperatures, the overall theoretical cycle characteristic of the [bmim]Zn2Cl5/NH3 system is slightly better than that of the LiBr/H2O

The vapor pressures of the binary solution of [bmim]Zn2Cl5/NH3 with NH3 mole fraction *x*2 = 0.83–0.94 at *T* = 323.15–563.15 K were measured via a static method with a total uncertainty of measurement below 4.3% [16]. The experimental data were fit using the modified UNIFAC

Ten years have passed since ionic liquids were introduced in the field of absorption refriger‐ ation, and unfortunately, the research progress pertaining to absorption refrigeration working pairs containing ionic liquids has been disappointing to us. Most of the working pairs proposed by researchers from all over the world are gradually fading from view due to a lack of practical applications. Particular attention was paid to ionic liquid working pairs containing [RR'Im]DMP (1-R,3-R'-imidazolium dimethylphosphate). When applied in absorption refrigeration, the cycle characteristics of the three representative working pairs of [dmIm]DMP-H2O, [emIm]DMP-H2O and [mmIm]DMP-CH3OH are better than that of H2O-NH3, but still slightly lower than those of LiBr-H2O. So far, the new conceptual chemical absorption refrigeration working pairs containing an ionic liquid, with [bmim]Zn2Cl5/NH3 as the representative, are the most ideal ionic liquid-type working pairs for absorption refriger‐ ation. The thermodynamic performances of absorption refrigeration using the proposed chemical working pairs are comparable to those achieved with the LiBr-H2O system. Addi‐ tionally, the ranges of operating conditions for the chemical working pairs are wider than those of the conventional working pairs. At present, promoting the industrial application of [bmim]Zn2Cl5/NH3 is our next priority. The discovery of chemical absorption refrigeration working pairs containing an ionic liquid is a milestone in the development of absorption refrigeration technology. It is foreseeable that the application of ionic liquids in absorption refrigeration will achieve a major breakthrough in the development of this technology, with the continued discovery of similar ionic liquid working pairs based on the chemical reaction.

#### **Acknowledgements**

This work was supported by the National Basic Research Program of China (973 Program) under Grant No. 2015CB251503 and the National Natural Science Foundation of China under Grant No. 51276180.

#### **Author details**

Shiqiang Liang1\*, Wei Chen2 , Yongxian Guo1 and Dawei Tang1

\*Address all correspondence to: liangsq@mail.etp.ac.cn

1 Institute of Engineering Thermophysics, Chinese Academy of Sciences, Beijing, P.R. China

2 College of Electromechanical Engineering, Qingdao University of Science and Technology, Qingdao, P.R. China

#### **References**

[1] S Q Liang, W Chen, K Y Cheng et al. The latent application of ionic liquids in absorp‐ tion refrigeration. In: Application of ionic liquid in science and technology. InTe‐ chOpen, Croatia 2011:467-494.

[2] J Zhao, S Q Liang, J Chen, et al. VLE for the High Concentration [MMIm]DMP-Meth‐ anol Solutions. Chemical Engineering, 38 (2010) 52-56

by researchers from all over the world are gradually fading from view due to a lack of practical applications. Particular attention was paid to ionic liquid working pairs containing [RR'Im]DMP (1-R,3-R'-imidazolium dimethylphosphate). When applied in absorption refrigeration, the cycle characteristics of the three representative working pairs of [dmIm]DMP-H2O, [emIm]DMP-H2O and [mmIm]DMP-CH3OH are better than that of H2O-NH3, but still slightly lower than those of LiBr-H2O. So far, the new conceptual chemical absorption refrigeration working pairs containing an ionic liquid, with [bmim]Zn2Cl5/NH3 as the representative, are the most ideal ionic liquid-type working pairs for absorption refriger‐ ation. The thermodynamic performances of absorption refrigeration using the proposed chemical working pairs are comparable to those achieved with the LiBr-H2O system. Addi‐ tionally, the ranges of operating conditions for the chemical working pairs are wider than those of the conventional working pairs. At present, promoting the industrial application of [bmim]Zn2Cl5/NH3 is our next priority. The discovery of chemical absorption refrigeration working pairs containing an ionic liquid is a milestone in the development of absorption refrigeration technology. It is foreseeable that the application of ionic liquids in absorption refrigeration will achieve a major breakthrough in the development of this technology, with the continued discovery of similar ionic liquid working pairs based on the chemical reaction.

This work was supported by the National Basic Research Program of China (973 Program) under Grant No. 2015CB251503 and the National Natural Science Foundation of China under

1 Institute of Engineering Thermophysics, Chinese Academy of Sciences, Beijing, P.R. China

2 College of Electromechanical Engineering, Qingdao University of Science and Technology,

[1] S Q Liang, W Chen, K Y Cheng et al. The latent application of ionic liquids in absorp‐ tion refrigeration. In: Application of ionic liquid in science and technology. InTe‐

and Dawei Tang1

, Yongxian Guo1

\*Address all correspondence to: liangsq@mail.etp.ac.cn

chOpen, Croatia 2011:467-494.

**Acknowledgements**

652 Ionic Liquids - Current State of the Art

Grant No. 51276180.

**Author details**

Qingdao, P.R. China

**References**

Shiqiang Liang1\*, Wei Chen2

