Study on a Quaternary Working Pair of CaCl2-LiNO3-KNO3/H2O for an Absorption Refrigeration Cycle

When compared with LiBr/H2O, an absorption refrigeration cycle using CaCl2/H2O as the working pair needs a lower driving heat source temperature, that is, CaCl2/H2O has a better refrigeration characteristic. However, the crystallization temperature of CaCl2/H2O solution is too high and its absorption ability is not high enough to achieve an evaporation temperature of 5 °C or lower. CaCl2-LiNO3-KNO3(15.5:5:1)/H2O was proposed and its crystallization temperature, saturated vapor pressure, density, viscosity, specific heat capacity, specific entropy, and specific enthalpy were measured to retain the refrigeration characteristic of CaCl2/H2O and solve its problems. Under the same conditions, the generation temperature for an absorption refrigeration cycle with CaCl2-LiNO3-KNO3(15.5:5:1)/H2O was 7.0 °C lower than that with LiBr/H2O. Moreover, the cycle’s COP and exergy efficiency with CaCl2-LiNO3-KNO3(15.5:5:1)/H2O were approximately 0.04 and 0.06 higher than those with LiBr/H2O, respectively. The corrosion rates of carbon steel and copper for the proposed working pair were 14.31 μm∙y−1 and 2.04 μm∙y−1 at 80 °C and pH 9.7, respectively, which were low enough for engineering applications.


Introduction
Absorption refrigeration systems can effectively utilize not only industrial waste heat [1][2][3][4], but also low-grade renewable energy, including solar energy and geothermal energy for refrigeration [5][6][7]. As a traditional working pair, LiBr/H 2 O has been widely used for refrigeration [8][9][10][11][12]. However, studies on new working pairs are still ongoing, because the required temperature of driving heat source for a refrigeration cycle using LiBr/H 2 O even reaches 88.0 • C [13][14][15], which is too high to use for some low-grade heat sources. Lin et al. [16] studied a double-stage air-cooled NH 3 /H 2 O absorption refrigeration system and found that it could effectively lower the temperature of driving heat source for utilizing solar energy. Malinina et al. [17] analyzed the influences of temperature and humidity on a solar energy refrigeration system with LiBr/H 2 O and calculated the minimum heat-collecting temperatures that are based on solar energy in some cities. Mortazavi et al. [18] designed an absorption refrigeration system with a falling-film generator, which could use lower temperature waste heat or solar energy. Bourouis et al. [19] analyzed the performance of LiBr + LiNO 3 + LiCl + LiI + H 2 O in a vertical tube and found that the crystallization temperature was 35 • C lower than LiBr solution. Sun et al. [20] studied LiBr-LiNO 3 (mole ratio: 4:1)/H 2 O and found the alternative working pair had higher COP and less corrisivity than LiBr/H 2 O. Chen et al. [21] studied the performance of an absorption refrigeration system using [emim]Cu 2 Cl 5 /NH 3 as working pair with the UNIFAC model, and results showed that the [emim]Cu 2 Cl 5 /NH 3 system possessed several advantages, including non-crystallization and non-corrosion. Bellos et al. [22] compared the exergy efficiency between LiCl/H 2 O and LiBr/H 2 O, results showed that LiCl/H 2 O performed better at different ambient temperature levels. Wang et al. [23] measured the properties of different ammonia/ionic liquid working pairs. Luo et al. [24][25][26][27][28] studied various lithium nitrate-ionic liquid/water working pairs. They both found that the working pairs with ionic liquid had excellent characteristics for heating, whereas they were not suitable for refrigeration because of insufficient absorption ability. Li et al. [29][30][31][32] measured the thermophysical properties of several CaCl 2 -based working pairs, and found that the CaCl 2 -based working pairs had an excellent refrigeration characteristic. However, their strong corrosivity limited the practical applications.
In this work, to find a new working pair with excellent refrigeration characteristic for absorption refrigeration, various inorganic salts, including NaCl, KCl, LiCl, KNO 3 , and LiNO 3 , were added in CaCl 2 /H 2 O, and their crystallization temperature and saturated vapor pressure were measured. Furthermore, some other thermophysical properties and corrisivity of the proposed working pair were measured and the performance of an absorption refrigeration cycle with the proposed working pair was analyzed. Table 1 shows the purities of the reagents used in this work. Table 2 lists the detailed compositions of carbon steel and copper samples used in the corrosion experiments.

Apparatus and Methods
To analyze the performance of a working pair, its properties, such as crystallization temperature, saturated vapor pressure, density, viscosity, specific heat capacity, dissolution enthalpy, and corrosion rate, need to be measured.
The crystallization temperature was measured by a dynamic method in a precision thermostat (HX-3010, Bilang, Shanghai). The prepared solution was put in the thermostat at a slightly higher initial temperature. The crystallization temperature was measured by reducing the temperature by 1 • C every 12 hours until crystallization appeared in the solution.
The saturated vapor pressure was measured by a static method. The solution was poured into an autoclave that was assembled with a precision digital absolute pressure gauge (AX-110, Aoxin, Xi'an) and a Pt-100 thermocouple. The autoclave was placed in a precision oil bath (DKU-30, Jinghong, Shanghai) after vacuuming. The data of pressure gauge and thermocouple were obtained, respectively, after stabilization.
The density and viscosity were measured in a precision viscometer oil bath (SYP1003-H, Zhongxi, Beijing). Density measurement was carried out by a capillary pycnometer with a capillary diameter of approximately 1 mm. Ubbelohde capillary viscometers with different fine capillaries was used to carry out the viscosity measurement.
The specific heat capacity and dissolution enthalpy were measured by a micro reaction calorimeter (µRC, THT Co., UK). The measurement of specific heat capacity was conducted by making a 1 • C "step-change" in the measurement temperature. The dissolution enthalpy was measured by an isothermal method, with a solid addition accessory.
The corrosion rate of carbon steel and copper in the solution were measured by a weight loss method. The sample was immerged in the solution for 200 hours. The corrosion rate was calculated according to the mass change of the sample.
References [24][25][26][27][28] give the detailed procedures. All the above experiments were carried out at 101.3 kPa and 25 • C. The properties of water and LiBr/H 2 O were measured and compared with literature values to validate the above methods. In addition, three parallel experiments were carried out for each measurement to verify the reproducibility. Table 3 lists the accuracy of the instruments.

CaCl 2 /H 2 O
To find a working pair with excellent refrigeration characteristic, the saturated vapor pressures of CaCl 2 /H 2 O were measured and are shown in Figure 1a. Figure 1b presents the comparison of the refrigeration characteristic between LiBr/H 2 O and CaCl 2 /H 2 O, it shows that, for an identical pressure of 6.290 kPa, which is a typical pressure in the condenser and generator, CaCl 2 /H 2 O had a lower generation temperature than LiBr/H 2 O, meaning that the refrigeration characteristic of CaCl 2 /H 2 O was better than LiBr/H 2 O. Figure 2 shows the absorption temperature of CaCl 2 /H 2 O under an absorption pressure of 0.872 kPa, which corresponds to the typical evaporation temperature of 5 • C. The crystallization temperature of CaCl 2 /H 2 O is also plotted in Figure 2 to illustrate the limitation from the crystallization of absorbent. The absorption temperature increased with increasing the concentration, and meet the crystallization temperature at 33.0 • C, which was the maximum absorption temperature under the given conditions. Generally, the absorption temperature in absorber for a refrigeration cycle is 37.0 • C, so the binary working pair of CaCl 2 /H 2 O could not be applied for the refrigeration cycle, because of its high crystallization temperature and insufficient absorption ability.
To improve the absorption ability and reduce the crystallization temperature of CaCl 2 /H 2 O, some salts, including NaCl, KCl, LiCl, KNO 3 , and LiNO 3 , were combined with CaCl 2 /H 2 O, and their saturated vapor pressures and crystallization temperatures were measured.  Figure 2 shows the absorption temperature of CaCl2/H2O under an absorption pressure of 0.872 kPa, which corresponds to the typical evaporation temperature of 5 °C. The crystallization temperature of CaCl2/H2O is also plotted in Figure 2 to illustrate the limitation from the crystallization of absorbent. The absorption temperature increased with increasing the concentration, and meet the crystallization temperature at 33.0 °C, which was the maximum absorption temperature under the given conditions. Generally, the absorption temperature in absorber for a refrigeration cycle is 37.0 °C, so the binary working pair of CaCl2/H2O could not be applied for the refrigeration cycle, because of its high crystallization temperature and insufficient absorption ability. To improve the absorption ability and reduce the crystallization temperature of CaCl2/H2O, some salts, including NaCl, KCl, LiCl, KNO3, and LiNO3, were combined with CaCl2/H2O, and their saturated vapor pressures and crystallization temperatures were measured.

Measurement of Crystallization Temperature TC
The TC of CaCl2-NaCl/H2O, CaCl2-KCl/H2O, CaCl2-LiCl/H2O, CaCl2-KNO3/H2O, CaCl2-LiNO3/H2O, and CaCl2-LiNO3-KNO3/H2O were measured. Figure 3 gives the comparison of     Figure 2 shows the absorption temperature of CaCl2/H2O under an absorption pressure of 0.872 kPa, which corresponds to the typical evaporation temperature of 5 °C. The crystallization temperature of CaCl2/H2O is also plotted in Figure 2 to illustrate the limitation from the crystallization of absorbent. The absorption temperature increased with increasing the concentration, and meet the crystallization temperature at 33.0 °C, which was the maximum absorption temperature under the given conditions. Generally, the absorption temperature in absorber for a refrigeration cycle is 37.0 °C, so the binary working pair of CaCl2/H2O could not be applied for the refrigeration cycle, because of its high crystallization temperature and insufficient absorption ability. To improve the absorption ability and reduce the crystallization temperature of CaCl2/H2O, some salts, including NaCl, KCl, LiCl, KNO3, and LiNO3, were combined with CaCl2/H2O, and their saturated vapor pressures and crystallization temperatures were measured.

Measurement of Crystallization Temperature T C
The  Figure 3 gives the comparison of T C between these CaCl 2 -based working pairs and CaCl 2 /H 2 O. Here, the concentration is the solution's total solute mass concentration.   TC between these CaCl2-based working pairs and CaCl2/H2O. Here, the concentration is the solution's total solute mass concentration.  Figure 3a shows CaCl2-NaCl/H2O with adding 5.0 g NaCl to CaCl2/H2O, in which CaCl2 were from 42.9 g to 61.3 g and H2O was 95.0 g. The crystallization temperatures of CaCl2-NaCl/H2O were higher than those of CaCl2/H2O under the same concentrations. Figure 3b shows CaCl2-KCl/H2O with adding 5.0 g KCl to CaCl2/H2O, in which CaCl2 were from 78.6 g to 117.4 g and H2O was 95.0 g. The crystallization temperatures of CaCl2-KCl/H2O were approximately 20.0 °C lower than those of CaCl2/H2O under the same concentrations.  Corresponding to the concentrations ranging from 55.0 wt.% to 62.0 wt.%, which is a practical concentration range for an absorption refrigeration cycle, the crystallization temperatures of CaCl 2 -LiNO 3 /H 2 O were from -10.0 • C to 7.0 • C, which are sufficiently low to solve the absorbent crystallization problem in summer. However, the addition amount of 35.0 g LiNO 3 was relatively large, and it is a disadvantage from the aspect of cost due to LiNO 3 being much more expensive than CaCl 2 .
To depress the cost increase, a part of LiNO 3 was replaced with KNO 3 for CaCl 2 -LiNO 3 /H 2 O. Figure 3f shows CaCl 2 -LiNO 3 -KNO 3 /H 2 O with adding 25.0 g LiNO 3 and 5.0 g KNO 3 to CaCl 2 /H 2 O, in which CaCl 2 were from 66.7 g to 117.4 g and H 2 O was 70.0 g. A reduction of crystallization temperature up to 30.0 • C was achieved from 58.0 wt.% to 65.0 wt.%, which indicated that the crystallization problem would not occur in this concentration range.   The measured p was fitted by Equation (1) [33][34][35].
where A i , B i , and C i are the regression parameters. Equation (2) obtains the average absolute relative deviation (AARD) between the measured values and the fitted values.
where N is total number of data, P exp is the measured or obtained value, and P fit is the fitted value. The regression parameters and AARD were obtained and are shown in Tables 6 and 7, respectively.     The p of this working pair was measured in order to analyze the cycle performance with   The p of this working pair was measured in order to analyze the cycle performance with CaCl2-LiNO3-KNO3(15.5:5:1)/H2O, as shown in Table 8.  The p of this working pair was measured in order to analyze the cycle performance with CaCl 2 -LiNO 3 -KNO 3 (15.5:5:1)/H 2 O, as shown in Table 8. The measured p was fitted by Equation (3) and the AARD was obtained to be 1.82% by Equation (2).
(3) Figure 6 shows the measured p and fitted value. The measured p agreed well with the fitted value, which indicated that the p of CaCl 2 -LiNO 3 -KNO 3 (15.5:5:1)/H 2 O could be obtained with the given corresponding concentration and temperature.
(3) Figure 6 shows the measured p and fitted value. The measured p agreed well with the fitted value, which indicated that the p of CaCl2-LiNO3-KNO3(15.5:5:1)/H2O could be obtained with the given corresponding concentration and temperature.

Measurement of Density ρ
ρ of CaCl2-LiNO3-KNO3(15.5:5:1) /H2O was measured by a capillary pycnometer method. Table  9 lists the results.  The measured ρ of CaCl2-LiNO3-KNO3(15.5:5:1)/H2O was fitted by Equation (4) and AARD was obtained to be 0.22% by Equation (2).  Figure 8 shows the measured ρ and the fitted value. The measured ρ was highly consistent with the fitted value. The density linearly decreased with the temperature increasing, and it increased with the concentration increasing.  Table 9 lists the results.  Table 10 shows the results. The measured η of CaCl2-LiNO3-KNO3(15.5:5:1)/H2O was fitted by Equation (5) and the AARD was obtained to be 0.82% by Equation (2).  Figure 9 shows the measured η and the fitted value. The measured η agreed well with the fitted value. η exponentially decreased with the temperature increasing, whereas it increased with the concentration increasing.   Table 10 shows the results. The measured η of CaCl 2 -LiNO 3 -KNO 3 (15.5:5:1)/H 2 O was fitted by Equation (5) and the AARD was obtained to be 0.82% by Equation (2).

Measurement of Specific Heat Capacity Cp
Cp of CaCl2-LiNO3-KNO3(15.5:5:1)/H2O was measured with a micro reaction calorimeter. Table  11 lists the results. The measured Cp was fitted by Equation (6) and AARD was obtained to be 0.21% by Equation (2) Figure 10 shows the measured Cp and the fitted value. Cp linearly increased with increasing the temperature.  Table 11 lists the results. The measured C p was fitted by Equation (6) and AARD was obtained to be 0.21% by Equation (2).

Cp of CaCl2, LiNO3, KNO3 and H2O
The Cp of solid KNO3 was measured and is shown in Table 12, Cp of CaCl2, LiNO3, and H2O are given in Reference Literature [29].   Table 13.  [25,36,37]. Table 14 lists the obtained results.  The C p of solid KNO 3 was measured and is shown in Table 12, C p of CaCl 2 , LiNO 3 , and H 2 O are given in Reference Literature [29].
The obtained h was fitted by Equation (7) and AARD was obtained to be 0.07% by Equation (2). Figure 11 shows the obtained h and the fitted value. h linearly increased with increasing the temperature, and the slope of line slightly increased with reducing the concentration.  The obtained h was fitted by Equation (7) and AARD was obtained to be 0.07% by Equation (2) Figure 11 shows the obtained h and the fitted value. h linearly increased with increasing the temperature, and the slope of line slightly increased with reducing the concentration.

Calculation of Specific Entropy s
s of a solution can be also obtained from the measured Cp and ΔHmix [38]. s of CaCl2-LiNO3-KNO3(15.5:5:1) /H2O was obtained and is shown in Table 15.

Calculation of Specific Entropy s
s of a solution can be also obtained from the measured C p and ∆H mix [38]. s of CaCl 2 -LiNO 3 -KNO 3 (15.5:5:1) /H 2 O was obtained and is shown in Table 15.  Figure 12 shows the obtained s and the fitted value. s increased with the temperature increasing and decreased with the concentration increasing when the temperature was above 28 • C, whereas it changed little with the concentration when the temperature was below 28 • C.
Entropy 2019, 21, x 16 of 23 Figure 12 shows the obtained s and the fitted value. s increased with the temperature increasing and decreased with the concentration increasing when the temperature was above 28 °C, whereas it changed little with the concentration when the temperature was below 28 °C.

Application for an Absorption Refrigeration Cycle
3.9.1. Absorption Refrigeration Cycle using CaCl2-LiNO3-KNO3(15.5:5:1)/H2O Figure 13a shows the schematic of an absorption refrigeration cycle. Figure 13b is the P-T diagram of the cycle, and the points that are marked in the two diagrams are one-to-one correspondence. The working conditions are given, as follows: the evaporation temperature was 5.0 °C; the absorption temperature and condensation temperature were 37.0 °C; and, the evaporation and condensation pressures were 0.872 kPa and 6.290 kPa, respectively. The concentration of dilute solution for CaCl2-LiNO3-KNO3(15.5:5:1)/H2O was figured out to be 60.5 wt.% by Equation (3), and the strong solution was 63.5 wt.%, with a concentration difference of 3.0 wt.%, thus, the generation temperature of the cycle was determined to be 74.0 °C by Equation (3). The same method was applied to calculate the generation temperature while using LiBr/H2O and other CaCl2-based  Figure 13a shows the schematic of an absorption refrigeration cycle. Figure 13b is the P-T diagram of the cycle, and the points that are marked in the two diagrams are one-to-one correspondence. Figure 12 shows the obtained s and the fitted value. s increased with the temperature increasing and decreased with the concentration increasing when the temperature was above 28 °C, whereas it changed little with the concentration when the temperature was below 28 °C.

Application for an Absorption Refrigeration Cycle
3.9.1. Absorption Refrigeration Cycle using CaCl2-LiNO3-KNO3(15.5:5:1)/H2O Figure 13a shows the schematic of an absorption refrigeration cycle. Figure 13b is the P-T diagram of the cycle, and the points that are marked in the two diagrams are one-to-one correspondence. The working conditions are given, as follows: the evaporation temperature was 5.0 °C; the absorption temperature and condensation temperature were 37.0 °C; and, the evaporation and condensation pressures were 0.872 kPa and 6.290 kPa, respectively. The concentration of dilute solution for CaCl2-LiNO3-KNO3(15.5:5:1)/H2O was figured out to be 60.5 wt.% by Equation (3), and the strong solution was 63.5 wt.%, with a concentration difference of 3.0 wt.%, thus, the generation temperature of the cycle was determined to be 74.0 °C by Equation (3). The same method was applied to calculate the generation temperature while using LiBr/H2O and other CaCl2-based  Table 16 lists the results. As seen in Table 16, the generation temperature was reduced by 7.0 • C through the use of CaCl 2 -LiNO 3 -KNO 3 (15.5:5:1)/H 2 O instead of LiBr/H 2 O. The generation temperature differences between the four CaCl 2 -based working pairs were relatively small.

Analysis of COP and Exergy Efficiency
To analyze the performance of a refrigeration cycle with CaCl 2 -LiNO 3 -KNO 3 (15.5:5:1)/H 2 O, the state parameters of typical points in Figure 13 were obtained by Equations (3), (7) and (8). Table 17 lists the results. The coefficient of performance (COP) for the absorption refrigeration cycle can be defined as: where α represents circulating ratio. COP was obtained to be 0.801 when using CaCl 2 -LiNO 3 -KNO 3 (15.5:5:1)/H 2 O as the working pair. The COPs for other working pairs were obtained with the same method, and the results are listed in Table 18.
where T 0 represents the environment temperature that was taken as 25 • C in this paper. The exergy destructions for each part of the absorption refrigeration cycle were obtained as follows [39]. Evaporator: Condenser: Absorber: Generator: Heat exchanger:  The η E of the absorption refrigeration cycle with CaCl 2 -LiNO 3 -KNO 3 (15.5:5:1)/H 2 O and LiBr/H 2 O were obtained to be 0.327 and 0.272, respectively. When compared with COP, the difference in exergy efficiency between the two working pairs was more distinct, which further showed the advantage of CaCl 2 -LiNO 3 -KNO 3 (15.5:5:1)/H 2 O as an alternative working pair. Figures 14 and 15 show the changes of generation temperature and efficiencies (COP and η E ) for CaCl 2 -LiNO 3 -KNO 3 (15.5:5:1)/H 2 O, with the evaporation temperature varying from 5 • C to 15 • C. As shown in Figure 14a, the generation temperature decreased almost linearly with increasing the evaporation temperature. As shown in Figure 14b, the COP of the absorption refrigeration cycle increased with the evaporation temperature increasing, whereas the exergy efficiency decreased with the evaporation temperature increasing. difference in exergy efficiency between the two working pairs was more distinct, which further showed the advantage of CaCl2-LiNO3-KNO3(15.5:5:1)/H2O as an alternative working pair. Figure 14 and Figure 15 show the changes of generation temperature and efficiencies (COP and ηE) for CaCl2-LiNO3-KNO3(15.5:5:1)/H2O, with the evaporation temperature varying from 5 °C to 15 °C. As shown in Figure 14a, the generation temperature decreased almost linearly with increasing the evaporation temperature. As shown in Figure 14b, the COP of the absorption refrigeration cycle increased with the evaporation temperature increasing, whereas the exergy efficiency decreased with the evaporation temperature increasing.  Figure 15 shows the variations of COP and ηE with the solution heat exchanger efficiency. COP and ηE increased almost linearly with the heat exchanger efficiency increasing, and the increasing slope of COP was greater than that of ηE.

Measurement of Corrosion Rate RC
Generally, carbon steel is used as the structural material and copper is used as the heat exchange material for absorption heat pump. The RC of carbon steel and copper in 63.5 wt.% solution of CaCl2-LiNO3-KNO3(15.5:5:1)/H2O were measured at 80.0 °C and pH 9.7. Figure 16 gives the difference in exergy efficiency between the two working pairs was more distinct, which further showed the advantage of CaCl2-LiNO3-KNO3(15.5:5:1)/H2O as an alternative working pair. Figure 14 and Figure 15 show the changes of generation temperature and efficiencies (COP and ηE) for CaCl2-LiNO3-KNO3(15.5:5:1)/H2O, with the evaporation temperature varying from 5 °C to 15 °C. As shown in Figure 14a, the generation temperature decreased almost linearly with increasing the evaporation temperature. As shown in Figure 14b, the COP of the absorption refrigeration cycle increased with the evaporation temperature increasing, whereas the exergy efficiency decreased with the evaporation temperature increasing.  Figure 15 shows the variations of COP and ηE with the solution heat exchanger efficiency. COP and ηE increased almost linearly with the heat exchanger efficiency increasing, and the increasing slope of COP was greater than that of ηE.

Measurement of Corrosion Rate RC
Generally, carbon steel is used as the structural material and copper is used as the heat exchange material for absorption heat pump. The RC of carbon steel and copper in 63.5 wt.% solution of CaCl2-LiNO3-KNO3(15.5:5:1)/H2O were measured at 80.0 °C and pH 9.7. Figure 16 gives the  Figure 15 shows the variations of COP and η E with the solution heat exchanger efficiency. COP and η E increased almost linearly with the heat exchanger efficiency increasing, and the increasing slope of COP was greater than that of η E .

Measurement of Corrosion Rate R C
Generally, carbon steel is used as the structural material and copper is used as the heat exchange material for absorption heat pump. The R C of carbon steel and copper in 63.5 wt.% solution of CaCl 2 -LiNO 3 -KNO 3 (15.5:5:1)/H 2 O were measured at 80.0 • C and pH 9.7. Figure 16    Although the corrosivity of CaCl2-LiNO3-KNO3(15.5:5:1)/H2O to carbon steel was stronger than that of LiBr/H2O, it was still acceptable for practical applications. On the other hand, the corrosivity of CaCl2-LiNO3-LiBr(8.72:1:1)/H2O to carbon steel was too strong to be applied, even though it had the lowest generation temperature among the CaCl2-based working pairs. The Rc of copper in 63.5 wt.% solution of CaCl2-LiNO3-KNO3(15.5:5:1)/H2O was 2.04 μm•y −1 , which was smaller than that in 59.4 wt.% solution of LiBr/H2O and it could meet the requirements for engineering applications.

Conclusions
1. When compared with LiBr/H2O, for an identical adsorption temperature at 0.872 kPa, which is a typical pressure of absorber, CaCl2/H2O had a lower absorption temperature at 6.290 kPa, which is a typical pressure of generator, meaning that CaCl2/H2O basically had a better refrigeration characteristic for an absorption refrigeration cycle. However, the absorption ability of CaCl2/H2O was not strong enough for achieving an evaporation temperature of 5 °C or lower, because of its high crystallization temperature. 2. The crystallization temperature was significantly lowered when combining CaCl2/H2O with LiNO3 or LiNO3+KNO3. As a result, the absorption ability of CaCl2-LiNO3/H2O or CaCl2-LiNO3-KNO3/H2O was essentially improved. 3. For an absorption refrigeration cycle using CaCl2-LiNO3-KNO3(15.5:5:1)/H2O as the working pair, the generation temperature that is required for achieving an evaporation temperature of 5 °C was 74.0 °C, which was 7.0 °C lower than that using LiBr/H2O.