Review on Absorption Refrigeration Technology and Its Potential in Energy-Saving and Carbon Emission Reduction in Natural Gas and Hydrogen Liquefaction

Abstract

With the requirement of energy decarbonization, natural gas (NG) and hydrogen (H₂) become increasingly important in the world’s energy landscape. The liquefaction of NG and H₂ significantly increases energy density, facilitating large-scale storage and long-distance transport. However, conventional liquefaction processes mainly adopt electricity-driven compression refrigeration technology, which generally results in high energy consumption and carbon dioxide emissions. Absorption refrigeration technology (ART) presents a promising avenue for enhancing energy efficiency and reducing emissions in both NG and H₂ liquefaction processes. Its ability to utilize industrial waste heat and renewable thermal energy sources over a large temperature range makes it particularly attractive for sustainable energy practices. This review comprehensively analyzes the progress of ART in terms of working pairs, cycle configurations, and heat and mass transfer in main components. To operate under different driven heat sources and refrigeration temperatures, working pairs exhibit a diversified development trend. The environment-friendly and high-efficiency working pairs, in which ionic liquids and deep eutectic solvents are new absorbents, exhibit promising development potential. Through the coupling of heat and mass transfer within the cycle or the addition of sub-components, cycle configurations with higher energy efficiency and a wider range of operational conditions are greatly focused. Additives, ultrasonic oscillations, and mechanical treatment of heat exchanger surfaces efficiently enhance heat and mass transfer in the absorbers and generators of ART. Notably, nanoparticle additives and ultrasonic oscillations demonstrate a synergistic enhancement effect, which could significantly improve the energy efficiency of ART. For the conventional NG and H₂ liquefaction processes, the energy-saving and carbon emission reduction potential of ART is analyzed from the perspectives of specific power consumption (SPC) and carbon dioxide emissions (CEs). The results show that ART integrated into the liquefaction processes could reduce the SPC and CE by 10~38% and 10~36% for NG liquefaction processes, and 2~24% and 5~24% for H₂ liquefaction processes. ART, which can achieve lower precooling temperatures and higher energy efficiency, shows more attractive perspectives in low carbon emissions of NG and H₂ liquefaction.

1. Introduction

As global energy demand grows and carbon emission and environmental pollution become increasingly severe, the pursuit and utilization of clean energy have become a global consensus. Among many sustainable energy solutions, natural gas (NG) and hydrogen (H₂) are becoming increasingly critical in the world’s energy landscape. NG is widely considered the vital primary energy to enable the transition from fossil fuels to the renewables era because of its abundant reserves and much lower environmental pollution than oil and coal, which is estimated to reach 25.1% of total energy consumption by 2040 [1]. As an emission-free and sustainable secondary energy source, H₂ is an important energy carrier for a carbon-neutral energy landscape in the future [2]. The utilization of NG and H₂ is not only conducive to reducing the harsh impacts of traditional fuel energy on the environment; meanwhile, they are abundant and widely available, thus effectively alleviating global energy scarcity [3,4].

NG and H₂ production plants are generally far from the application site, thus requiring large-scale, long-distance transport and storage. Liquefaction of NG and H₂ provides an efficient solution for long-distance energy distribution. It significantly reduces the gas volume and increases the energy density, improving transport and storage efficiency [5]. However, conventional NG and H₂ liquefaction processes adopt electricity-driven compression refrigeration technology (CRT). The consumption of large amounts of electricity causes high energy consumption and carbon dioxide emissions in NG and H₂ liquefaction processes, and these factors hinder their utilization and development [6].

On the other hand, H₂ and NG liquefaction plants tend to be far from the power grid. The electrical load relies on gas turbines driven by fuel combustion, which generate a large amount of low-grade waste heat during power generation, usually discharged directly into the environment, resulting in low energy utilization efficiency [7]. Meanwhile, a large amount of low-grade renewable energy sources in the vicinity of the plant are not utilized, such as solar and geothermal energy, resulting in a waste of renewable energy. Considering the urgent demand for global carbon emission reduction, the recovery and utilization of these low-grade energy sources are critical.

Among the various low-grade heat utilization technologies, absorption refrigeration technology (ART) can be driven by different heat sources to generate the cooling effect. It was introduced to address some severe issues, such as the energy crisis, raised fuel expenses, and environmental problems associated with the traditional CRT. Thus, ART is particularly suitable for scenarios with a large amount of low-grade waste heat and a demand for refrigeration. Compared with CRT, ART generally adopts environment-friendly refrigerants, with fewer moving objects and stable operation; it is the most widely used thermal energy-driven refrigeration technology and has received increasingly extensive attention and research under the trend of global energy decarbonization transition [8].

Numerous studies have shown that the integration of ART into NG and H₂ liquefaction processes is an effective way to reduce the energy consumption and carbon dioxide emissions of liquefaction processes [7,9,10]. During periods of energy scarcity or high demand, conventional CRT can place significant strain on electrical grids and consume substantial amounts of electricity. ART, by utilizing heat sources such as natural gas, steam, or waste heat, provides an alternative means of generating cooling without relying solely on electricity. This can help alleviate the burden on energy resources during times of peak demand or limited electrical supply. In addition, CRT generally uses high-GWP (global warming potential) refrigerants and suffers from refrigerant leakage when operating conditions deviate from the rated conditions. ART with environment-friendly refrigerants reduces carbon dioxide emissions caused by the leakage of high-GWP refrigerants by decreasing the cooling loads provided by the compression refrigeration cycles.

However, compared with CRT, the COP of ART is lower, and there are some application defects of the conventional working pairs, such as the severe crystallization and corrosion concerns of lithium bromide solution, and the ammonia solution must be equipped with additional distillation equipment [8]. In recent years, many researchers have studied the performance optimization of ART from different perspectives. For instance, the application defects of conventional working pairs could be effectively alleviated by constructing a multi-component diversified salt absorber; in addition, some researchers have developed new types of working pairs with ionic liquids and deep eutectic solvents as new absorbers and conducted performance assessments [11,12]. Applying heat recovery technology could increase the cycle’s COP and extend the operational conditions. Thus, the generator–absorber heat exchange (GAX) cycle and branch GAX cycle configurations were developed [13]. Based on the single-effect cycle, multi-effect or multi-stage absorption refrigeration cycles have been established by changing the heat and mass coupling approach, thus improving the refrigeration performance at different driven heat source temperatures [14]. The heat and mass transfer processes of the absorption refrigeration cycle have a significant impact on the cycle’s performance, and some researchers studied the enhancement technologies of the heat and mass transfer in the main components [15].

In light of the above analyses, this paper focuses on the process of ART and its potential in energy-saving and carbon emission reduction of NG and H₂ liquefaction. Firstly, the progress on performance optimization of ART is comprehensively reviewed in terms of working pairs, cycle configurations, and heat and mass transfer in main components. Then, for the conventional NG and H₂ liquefaction processes, the energy-saving and carbon emission reduction potential of ART are analyzed from the perspectives of specific power consumption (SPC) and carbon emission (CE). By analyzing the development of ART and its energy-saving and carbon emission reduction potential in NG and H₂ liquefaction, this paper provides a feasible pathway for the energy-saving and low-carbon development of NG and H₂ liquefaction processes, thus facilitating the low-carbon transition of the global energy structure.

2. Progress in Absorption Refrigeration Technology

The absorption refrigeration system is a kind of refrigeration technology powered by low-grade heat sources, including industrial waste heat and renewable thermal energy, such as solar, geothermal, and biomass energy. It is distinguished by its low energy consumption, minimal moving parts, and stable operational performance. It is an effective low-carbon or zero-carbon energy utilization technology for large-scale industrial waste heat recovery or small-scale distributed refrigeration [16,17]. Given the escalating energy demand and the growing concern over carbon emission, the application value of absorption refrigeration technology has become increasingly prominent. Existing research into ART includes but is not limited to solar energy and medium and low temperatures. In the face of new applications, new materials, and new absorption working pairs, novel absorption cycles can be proposed with greater efficiency and wider ranges of heat-source-driven temperatures and solution concentrations.

The genesis of absorption refrigeration technology dates back to 1777 when Edward Nairne first observed that concentrated sulfuric acid could absorb water vapor, thereby cooling the residual water [18]. In 1859, Ferdinand Carre patented a direct-fired ammonia–water absorption refrigeration system, which utilized a coal-fired furnace as the primary heat source [19]. Notably, in the 1940s, Carrier developed the first lithium bromide absorption refrigeration system in the United States, significantly enhancing efficiency and safety compared to earlier units. Distinct from conventional vapor compression refrigeration, absorption refrigeration relies on an external heat source to activate the solution cycle, thereby generating the cooling load (Figure 1 illustrates the basic cycle principle). The system primarily comprises four heat transfer components (generator, absorber, condenser, and evaporator), along with solution pumps and throttling elements. Firstly, the generator heats the circulation solution, causing partial refrigerant evaporation and increasing the absorbent concentration in the concentrated solution. Subsequently, the refrigerant vapor is condensed into a liquid form in the condenser and then depressurized via the throttling element, allowing it to evaporate and absorb heat in the evaporator, thus producing the cooling effect. Finally, the low-pressure refrigerant vapor is absorbed by the concentrated solution to form the dilute solution in the absorber, which is then pumped back to the generator, completing the absorption refrigeration cycle.

Compared to alternative heat-driven refrigeration technologies such as vapor injection, thermoelectric, and thermoacoustic refrigeration, absorption refrigeration technology boasts superior efficiency and a broader application range, establishing it as the predominant heat-driven refrigeration technology. The mature commercial applications of the absorption refrigeration cycle include lithium bromide solution and ammonia/water refrigeration units, notable for their flexible load regulation, high COP, low operational noise, and simple structure. Nonetheless, ART has some application drawbacks. For instance, the lithium bromide/water solution system faces challenges such as metal corrosion, crystallization issues, and the thick liquid film on the surface of the lithium bromide solution, which hinders heat and mass transfer in the absorber and generator, necessitating high airtightness due to low evaporating pressure [21]. The ammonia/water system, employing an ammonia/water working pair, grapples with the toxicity of ammonia and explosiveness. Additionally, the minimal boiling point difference between ammonia and water requires supplementary distillation equipment, alongside high condensing pressures. Furthermore, the inherently larger cooling load of absorption systems, compared to electrically driven systems, complicates the miniaturization of units.

In recent years, researchers worldwide have conducted numerous performance optimization studies to enhance energy utilization efficiency, extend the range of operating conditions, and alleviate the application deficiencies of typical absorption refrigeration units. These optimization efforts mainly focused on three categories: optimizing solution working pairs, improving circulation configurations, and enhancing heat and mass transfer in main components.

2.1. Working Pairs

The selection and performance optimization of working pairs are related to the development direction of the whole absorption refrigeration technology. Currently, working pairs for absorption refrigeration are chosen based on the specific needs of different applications, following rigorous theoretical analyses. An ideal working pair should exhibit several key characteristics [22]: (1) the refrigerant possesses the high latent heat of vaporization and operates under suitable pressure conditions; (2) the absorbent exhibits low viscosity, and the absorbent’s boiling point significantly exceeds that of the refrigerant under working pressure; (3) both the refrigerant and absorbent maintain chemical stability and are non-combustible, non-explosive, environmentally friendly, cost-effective, and easy to produce; (4) the refrigerant demonstrates high solubility within the absorbent, along with a low heat of mixing and specific heat capacity. Based on these criteria, working pairs are classified into five categories according to their refrigerants: water-based, ammonia-based, alcohol-based, Freon-based, and other working pairs.

2.1.1. Water-Based Working Pairs

Water, with its high latent heat of vaporization, excellent chemical and thermal stability, widespread availability, and low cost, is an exceptional refrigerant, particularly effective in forming working pairs with various salts. Water/lithium bromide is the most prevalent solution, offering advantages such as high energy efficiency, environmental friendliness, and low operational pressure. However, several drawbacks are associated with this working pair [22]: (1) the lithium bromide solution is highly corrosive to metals, which compromises the longevity of the unit, and the presence of non-condensable gases generated from the corrosion process affects the stable operation of the unit; (2) it is challenging to generate a cooling temperature below 0 °C; (3) the lithium bromide solution presents a crystallization risk at low operating temperature; (4) the low heat and mass transfer coefficient of the lithium bromide solution necessitates a large heat exchange area, which complicates the miniaturization of units.

One strategy for mitigating the corrosive issue of lithium bromide solution on metal materials involves incorporating corrosion inhibitors. These inhibitors function by forming a passivation film on the metal surface or altering the electrochemical properties of the metal surface. Widely used corrosion inhibitors comprise compounds such as LiOH, LiCrO₄, Li₂WO₄, and Li₂MoO₄ [23]. In applications, the combination of multiple corrosion inhibitors is often employed to enhance the performance of corrosion protection through the synergistic interaction of various mechanisms.

The construction of a multi-salt absorber system alleviates corrosion and crystallization problems and extends the temperature range of the driven heat source. Li et al. [24] examined the performance of working pairs, pairing water with CaCl₂, LiBr, and LiNO₃, in a single-effect solar absorption refrigeration system. The result indicated effectiveness in mitigating crystallization and corrosion issues, with a COP of 0.805 at the heat source temperature of 80.3 °C. Furthermore, Cheng et al. [25] explored the performance of a water/KCOOH working pair. The investigation revealed that the intermolecular forces between water and KCOOH, predominantly hydrogen bonding, are weaker than those in water/lithium bromide. This weaker bonding increases the saturated vapor pressure in the dilute solution of the generator, lowers the required temperature of the heat source, and lessens corrosion and crystallization issues. However, it reduces absorption efficiency in the absorber, leading to an overall refrigeration performance approximately 10% lower than that of the water/lithium bromide.

To enhance the comprehensive performance of traditional absorption systems, ionic liquids are being explored as an innovative, environmentally benign solvent. They have several benefits over traditional absorbents, such as reduced crystallization risk, minimal corrosiveness, virtually zero vapor pressure, and strong hydrophilicity. The exploration of ionic liquids as potential absorbents in ART has attracted widespread attention [26]. Gorakshnath et al. [27] examined the performance of the [emim][EtSO₄]/H₂O working pair in a single-effect absorption refrigeration cycle. The result indicated that the circulation ratio significantly rose with the generator temperature increase before stabilizing. The maximum COP of the [emim][EtSO₄]/H₂O cycle reached 0.66, which is higher than that of the ammonia/H₂O cycle (0.646) and lower than that of the lithium bromide solution cycle (0.833). Notably, [emim][EtSO₄]/H₂O did not encounter the corrosion and crystallization issues inherent to the lithium bromide solution. Hugo et al. [28] assessed the [emim][BF₄]/H₂O and [emim][EtSO₄]/H₂O working pairs’ performance in terms of heat loads, exergy efficiency, and energy efficiency. Additionally, ionic liquids, which are viable for forming working pairs with water, include [dmim][BF₄], [dmim][MPh], [dmim][EPh], [emim][DEP], and so on [29]. However, research in this domain is still in its beginning stage, constrained by limited data on the thermal properties of ionic liquid-based refrigerant working pairs, alongside the challenges posed by complex preparation and elevated costs.

2.1.2. Ammonia-Based Working Pairs

NH₃/H₂O is the most widely applied ammonia-based working pair. As chlorofluorocarbons (CFCs) and hydrochlorofluorocarbons (HCFCs) are increasingly restricted, ammonia has once again become the focus of research as a natural refrigerant. The key advantages of NH₃/H₂O include high miscibility, high latent heat of evaporation, and the capacity to achieve a cooling temperature below 0 °C. However, the application of ammonia is limited due to its toxicity and flammability, alongside the small boiling point difference between NH₃ and H₂O necessitating distillation equipment and its strong corrosivity towards non-ferrous metals [22].

To eliminate dependence on distillation equipment in NH₃/H₂O, researchers have explored the potential of substituting water with LiNO₃ and NaSCN as absorbents [30]. Experimental findings indicate that the COPs of NH₃/LiNO₃ and NH₃/NaSCN, both in single- or double-effect absorption refrigeration units, are comparable and consistently exceed that of NH₃/H₂O. This advantage becomes even more significant at higher operating temperatures. Zhou et al. [31] utilized NH₃/NaSCN as the working pair, achieving the COP range of 0.351–0.651 under driven temperatures between 150 °C and 350 °C. The NH₃/LiNO₃ working pair has been applied in various absorption refrigeration units for waste heat utilization.

Certain ionic liquids exhibit notable ammonia absorption and desorption capacities, eliminating the dependence on distillation equipment due to their extremely low vapor pressure. Wang et al. [32] analyzed nine ionic liquid–ammonia working pairs. Four of these working pairs were found to be theoretically usable in an absorption refrigeration cycle, boasting the high COP of 0.79 for NH₃/[mmim][DMP], 0.74 for NH₃/[emim][TfN], 0.73 for NH₃/[emim][SCN], and 0.70 for NH₃/[bmim][BF₄], all of which outperformed the NH₃/H₂O system. Gong et al. [33] undertook a comparative study on the thermodynamic performance of four ammonia/ionic liquid working pairs versus the ammonia/water in a semi-effective absorption refrigeration cycle. The results indicated that the COPs of all four ammonia/ionic liquid working pairs were higher than that of NH₃/H₂O, with NH₃/[emim][BF4] achieving the highest COP. However, it was noted that the circulation ratio of ammonia/ionic liquid was higher than that of ammonia/water under both high- or low-pressure conditions.

Researchers have introduced deep eutectic solvents (DESs) as novel absorbents for ammonia-based refrigerants to overcome the challenges associated with the complex and costly preparation of ionic liquids. Characterized by a eutectic mixture of hydrogen bond recipients (HBAs) and hydrogen bond donors (HBDs), a DES features properties similar to those of ionic liquids, such as high thermal stability and low vapor pressure. The most commonly used DESs are those formed from amides, polyols, and sugars (as HBDs) combined with organic salts (as HBAs), favored for their ease of preparation and compatibility with water [34]. Haghbakhsh et al. [35] examined the performance of three DES absorbents comprising urea, ethylene glycol, and glycerol mixed with choline chloride. The results revealed that the refrigeration efficiency of NH₃/DES slightly surpassed that of the NH₃/H₂O system. However, the notable drawback was the significantly increased power consumption for the pump, which was more than triple that required by the NH₃/H₂O system. Presently, the application of DESs as absorbents in absorption refrigeration is still relatively limited, and further investigation is needed.

2.1.3. Alcohol-Based Working Pairs

Alcohol-based working pairs, characterized by their high latent heat of evaporation, can generate a cooling temperature below 0 °C and exhibit excellent adaptability to low-grade driven heat sources. Research on alcohol-based refrigerants mainly focused on methanol (CH₃OH), trifluoroethanol (TFE), and hexafluoroisopropanol (HFIP), with salts, organic solvents, and ionic liquids as absorbents.

For methanol, Iyokli et al. [26] analyzed and compared the performance of CH₃OH/LiBr, CH₃OH/ZnBr₂, CH₃OH/LiI, CH₃OH/ZnBr₂+LiBr, CH₃OH/ZnCl₂+LiBr, CH₃OH/ZnBr₂+LiBr+LiI, and CH₃OH/ZnBr₂+LiI working pairs in double-stage and double-effect cycles. The results revealed that CH₃OH/LiBr exhibited the highest COP in both system configurations; CH₃OH/ZnBr₂+LiI and CH₃OH/ZnCl₂+LiBr had the broader temperature adaptation range. The study also noted that the high viscosity of these alcohol-based working pairs adversely decreases the heat and mass transfer rate. In the study of ionic liquids as absorbents, the focus has primarily been on [mmim]DMP [36]. Chen et al. [37] conducted a detailed investigation of CH₃OH/[mmim]DMP, employing experimental data of saturated vapor pressure to build a revised UNIFAC and Wilson model. Their work culminated in predicting gas–liquid equilibrium phase diagrams of the CH₃OH/[mmim]DMP working pair. Concerning refrigeration performance, the COP of the CH₃OH/[mmim]DMP system is 10% lower than that of the water/lithium bromide system but 20% higher than that of the ammonia/water system.

Trifluoroethanol (TFE), a colorless and transparent liquid, is miscible with water and many organic solvents. In absorption refrigeration, TFE is commonly paired with organic solvents like tetraethylene glycol dimethyl ether (TEGDME), and N-Methyl-2-pyrrolidone (NMP) to form working pairs. Notably, TFE/TEGDME stands out for its significant boiling point difference between refrigerant and absorbent, offering a broad working temperature range, excellent thermal stability, and non-corrosiveness to metals, thereby becoming the most extensively studied TFE working pair [38]. Long et al. [39] studied a TFE/TEGDME diffusion–absorption refrigeration system utilizing helium as the equilibrium gas and analyzed its performance. Luo et al. [40] examined the impact of generation, absorption, evaporation, and condensation temperatures on the performance of TFE/NMP, TFE/TEGDME, and NH₃/H₂O systems, focusing on COP and circulation ratio. The TFE/NMP system exhibited the highest COP, followed by TFE/TEGDME, with NH₃/H₂O showing the lowest. However, the circulation ratio showed the opposite trend. The minor boiling point difference between TFE and NMP necessitates distillation equipment.

HFIP, an ethanol-based refrigerant with a high melting point, exhibits excellent miscibility with water and various organic solvents, alongside robust thermal stability. The investigation of HFIP-based working pairs mainly focuses on HFIP/DTG, HFIP/DMPU, and HFIP/DMETEG [41]. Yi et al. [42] assessed and compared the performance of HFIP/DMETEG and TFE/DMETEG in an absorption refrigeration system through group contribution method simulations. The result indicates that the COP of the HFIP/DMETEG surpasses that of the TFE/DMETEG under the generation temperature of 130~150 °C.

2.1.4. Freon-Based Working Pairs

Freon, a term for halogenated hydrocarbon refrigerants, typically forms working pairs with organic solvents, offering exceptional intermiscibility and a broad temperature working range. Historically, CFCs and HCFCs, such as R11, R21, and R22, were extensively utilized in refrigeration systems due to their favorable thermophysical properties, chemical stability, and high refrigeration efficiency [22]. However, subsequent research uncovered their significant environmental impact, notably their implications for ozone layer depletion (CFCs and some HCFCs) and global warming (HCFCs and HFCs). The adoption of the “Paris Agreement (2016)” and the “Kigali Amendment to the Montreal Protocol (2016)” introduced stringent criteria for Freon refrigerants, demanding zero ozone depletion potential (ODP) and low global warming potential (GWP). Consequently, research on Freon refrigerants has shifted towards exploring low-GWP hydrofluorocarbons (HFCs) and hydrofluoroolefins (HFOs), with absorbents predominantly comprising organic solvents, ionic liquids, and deep eutectic solvents.

Research on hydrofluorocarbon (HFC) refrigerants for absorption refrigeration systems has predominantly focused on R134a, R32, R161, and R152a. These refrigerants emerged as alternatives to R22 and have garnered considerable interest due to their superior physicochemical properties and refrigeration performance. Wang et al. [43] simulated and analyzed the performance of an absorption refrigeration system employing R134a/DMF as working pairs. The study focused on the reflux ratio, generating temperature, and solution composition. The COP reached the maximum when the R134a molar component percentage was 56%, and the generating and absorbing temperatures were 131 °C and 28°C, respectively. Zhang et al. [44] selected DMETEG as the absorbent and evaluated R32, R152a, and R161 as refrigerants. They predicted solubility, enthalpy, and entropy using the modified NRTL activity coefficient model. The results in both single-effect and compressor-assisted absorption systems revealed that R32/DMETEG and R152a/DMETEG were particularly effective, with the latter achieving the highest COP of 0.5 at generating temperatures between 80 °C and 135 °C. In the study of ionic liquids as absorbents, Sujatha et al. [45] assessed the [hmim][Tf₂N] pair with R32, R152, R125, and R1234ze using Aspen Plus software. The investigation indicated R32 and R152a as the most compatible refrigerants for [hmim][Tf₂N], with R32/[hmim][Tf2N] achieving the highest COP of 0.51.

With the “Kigali Amendment to the Montreal Protocol” coming into effect in 2019, several high-GWP hydrofluorocarbons (HFCs), such as R134a, R245fa, R32, and R152a, have become regulated refrigerants, which means their usage will be progressively restricted. Driven by related policies and regulations, low-GWP HFOs and related mixed refrigerants have received extensive attention. Carbon–carbon double bonds in HFOs allow them to react with atmospheric hydroxyl groups, resulting in an exceptionally short atmospheric lifetime and minimal GWP, aligning with environmental sustainability goals. However, some of the HFOs exhibit flammability and mild toxicity [46]. Salhi et al. [47] explored the performance of an R1234yf pair with DMETEG and NMP in a single-effect absorption refrigeration cycle. The results, focusing on the impact of generating, absorbing, and evaporating temperatures on the COP and circulation ratio, revealed that the COPs of the R1234yf/DMETEG system were 0.053~0.409, and the COPs of R1234yf/NMP were 0.025~0.333. Moreover, they discovered that incorporating a compressor to augment heat and mass transfer in the absorption process could further enhance the COP.

In recent years, ionic liquids have garnered considerable interest as absorbents for hydrofluoroolefins (HFOs). Sun et al. [48] utilized R1234yf as the refrigerant and six ionic liquids—[hmim][OTf], [emim][BF4], [hmim][PF6], [hmim][BF4], [hmim][Tf2N], and [omim][BF4]—as absorbents to compose working pairs. They assessed the performance of these working pairs in both the single-effect absorption refrigeration cycle and the compressor-assisted absorption cycle. The results indicated that systems with compressor assistance significantly outperformed single-effect cycles due to the limited absorptive capacity of ionic liquids, with [hmim][Tf₂N] achieving the highest COP in both systems, and the highest COP was 0.35. Wu et al. [49] investigated four working pairs, each pairing R1234ze with [emim][BF4], [hmim][BF4], [omim][BF4], and [hmim][Tf2N]. The results demonstrated that R1234ze/[hmim][Tf₂N] displayed the best performance in both the single-effect absorption refrigeration cycle and the compressor-assisted absorption cycle, with the highest COP of 0.25 and 0.43, respectively. Overall, the COP of HFOs/ionic liquids is generally lower than that of NH₃/H₂O and H₂O/LiBr, primarily due to inadequate absorption. While augmenting compressor assistance or increasing the reflux ratio can mitigate this issue, it also escalates the power consumption and investment costs. Therefore, improving the interaction between ionic liquids and HFOs and enhancing their solubility are effective strategies for optimizing performance.

Researchers have explored the potential of DESs as absorbents for HFOs. Abedin et al. [50] performed molecular dynamics simulation on three working pairs, utilizing R1336mzzE, R1234zeE, and R245fa as refrigerants, with a DES mixture of ethylene glycol/acetyl propionic acid and choline chloride in 2:1 molar ratio serving as the absorbent. The results indicate that R245fa exhibited superior refrigeration efficiency compared to R1336mzzE and R1234zeE in a single-effect absorption refrigeration cycle driven by 80 °C thermal energy. Although research on DESs as absorbers for HFOs is limited, the field holds promising development potential, considering its inherent advantages in terms of low vapor pressure, affordability, ease of fabrication, and designable composition [51].

2.1.5. Other Working Pairs

The exploration of working pairs for absorption refrigeration extends beyond the four categories above. Motivated by environmental considerations, researchers have started investigating the viability of other natural refrigerants, mainly focused on hydrocarbons and CO₂.

Hydrocarbons are easily accessible and environmentally benign. They have garnered significant attention in the replacement of Freon refrigerants, with R290 and R600a being particularly prominent in absorption refrigeration research. The absorbents studied with these refrigerants include organic solvents, absorbent oils, and ionic liquids. Jia et al. [52] investigated the performance of R600a/squalane in single-effect and compressor-assisted absorption refrigeration systems. Their research focused on the COP, circulation ratio, and exergy efficiency. The results indicated that increased evaporating temperature notably enhances the COP. However, exergy efficiency exists at a maximum value within the evaporating temperature range of −13~27 °C. For the single-effect absorption refrigeration system, the maximum COP was 0.45 at the generating temperature of 87 °C, while the COP of the compressor-assisted absorption system reached 0.675. Yang et al. [53] conducted experimental tests to assess the compatibility of R290 with various absorbent oils, such as mineral oil, POE oil, PAG oil, and AB oil, providing analytical data for using R290 in absorption refrigeration cycles. Fukuta et al. [54] theoretically analyzed and experimentally tested the performance of R290/mineral oil working pairs in an absorption–compression combined cycle, and the results show that the system could operate stably. Moreover, Jia et al. [55] evaluated the performance of the ionic liquid [P6,6,6,14][Cl] paired with R290 and R600a refrigerants in a single-effect cycle and compared their performance against HFOs/[P6,6,6,14][Cl] working pairs.

CO₂ stands out as a refrigerant with zero ODP and low GWP, and its advantages include non-toxicity, non-flammability, non-corrosiveness to equipment, low-pressure ratio, and high refrigeration capacity per unit of volume. Ionic liquids are predominantly utilized as the absorbent for CO₂. Martin et al. [56] explored CO₂/[bmim][Tf₂N] in an absorption refrigeration cycle, calculating the phase equilibrium characteristics of the solution and conducting a comprehensive analysis of each stage within the CO₂/[bmim][Tf₂N] refrigeration cycle system. The result revealed that under the studied conditions, the COP and circulation ratio were 0.21 and 24, respectively. Further, He et al. [57] developed and tested a dual-temperature transcritical CO₂/[emim][Tf₂N] absorption refrigeration system, assessing the impact of generation, absorption, and evaporation temperatures on system performance.

According to the refrigerants, this section reports the progress of five kinds of working pairs: water-based, ammonia-based, alcohol-based, Freon-based, and other types of working pairs.

Table 1. Summary of the characteristics of various working fluid pairs.

Working Pair	Method	Cycle Configuration	Operating Temps. (°C)	COP	Remarks	Refs.
[Water-based]
H2O/CaCl2+LiBr+LiNO3	Experiment	single-effect	Tg = 80.3 Te = 7.0 Ta = 37.0 Tc = 37.0	0.81	Compared to H2O/LiBr, the COP of the new working pair is 0.04 higher, and Tg is 7.7 °C lower.	[24]
H2O/KCOOH	Experiment	single-effect	71 < Tg < 81 Te = 15 28 < Ta < 34 28 < Tc < 34	0.54~0.65	Compared to H2O/LiBr, H2O/KCOOH can operate in lower Tg, while the COP is approximately 10% lower; less corrosion and crystallization issues.	[25]
H2O/[emim][EtSO4]	Simulation	single-effect	65 < Tg < 100 5 < Te < 15 30 < Ta < 40 30 < Tc < 40	0.32~0.68	The COP of H2O/[emim][EtSO4] is higher than NH3/H2O, but lower than H2O/LiBr under the same operational conditions; less corrosion and crystallization issues.	[27]
H2O/[emim][EtSO4]	Simulation	single-effect	70 < Tg < 150 6 < Te < 12 35 < Ta < 45 35 < Tc < 45	0.14~0.75	At lower Ta and Tc and higher Te, the refrigeration performance of H2O/[EMIM][EtSO4] is comparable to H2O/LiBr; less corrosion and crystallization issues.	[28]
H2O/[dmim][Cl] H2O/[dmim][DMP]	Simulation	single-effect	70 < Tg < 130 5 < Te < 15 30 < Ta < 40 30 < Tc < 40	0.68~0.84	Compared to H2O/LiBr, H2O/[dmim][Cl] has higher COP under Tg and Te at around 100 and 15 °C.	[29]
H2O/[dmim][Cl]	Simulation	double-effect	100 < Tg < 160 5 < Te < 15 30 < Ta < 40 30 < Tc < 40	0.85~1.35	When the Tg is below 120 °C, the COPs of H2O/[dmim][Cl] in single- and double-effect are comparable.	[29]
[Ammonia-based]
NH3/LiNO3	Simulation	single-effect	65 < Tg < 100 −10 < Te < 10 20 < Ta < 40 20 < Tc < 40	0.57~0.7	The COP of NH3/LiNO3 is higher than NH3/H2O with no requirement for a rectifier; the circulation ratio and viscosity of NH3/LiNO3 are slightly higher than NH3/H2O.	[30]
NH3/NaSCN	Simulation	single-effect	60 < Tg < 100 −10 < Te < 10 20 < Ta < 40 20 < Tc < 40	0.55~0.72	The COP of NH3/NaSCN is higher than NH3/H2O with no requirement for a rectifier; NH3/NaSCN is not suitable for refrigeration below −10 °C due to crystallization.	[30]
NH3/NaSCN	Experiment	single-effect	100 < Tg < 120 −24 < Te < 5 36 < Ta < 42 30 < Tc < 33	0.351-0.653	The NH3-NaSCN is potentially applied in wide ranges including both freezing and air-conditioning applications.	[31]
NH3/[emim][BF4] NH3/[dema][Ac] NH3/[dmim][DMP] NH3/[emim][SCN]	Simulation	double-stage	65 < Tg < 90 5 < Te < 10 30 < Ta < 40 40 < Tc < 45	[emim][BF4]: 0.32~0.40 [dema][Ac]: 0.12~0.39 [dmim][DMP]: 0.13~0.36 [emim][SCN]: 0.10~0.36	The COPs of four NH3/IL are higher than the COP of NH3/H2O; the circulation ratios of NH3/ILs are all higher than NH3/H2O.	[33]
NH3/ethaline (DES) NH3/reline (DES) NH3/glyceline (DES)	Simulation	single-effect	70 < Tg < 120 Te = 2 20 < Ta < 40 Tc = 45	ethaline: 0.53~0.67 reline: 0.51~0.66 glyceline: 0.52~0.65	Compared to NH3/H2O, NH3/DES has higher COP and ECOP, while the circulation ratio is significantly higher (about 3 times); NH3/ethaline exhibits the best refrigeration performance.	[35]
[Alcohol-based]
CH3OH/LiBr CH3OH/ZnCl2+LiBr CH3OH/ZnBr2 CH3OH/LiI CH3OH/ZnBr2+LiI	Simulation	double-effect	120 < Tg < 170 Te = −10 Ta = 30 Tc = 30	LiBr: 1.15~1.18 ZnCl2+LiBr: 0.98~1.03 ZnBr2: 0.77~0.79 LiI: 0.83~0.84 ZnBr2+LiI: 1.06~1.11	The viscosity of these alcohol-based working pairs is higher than NH3/H2O and H2O/LiBr.	[58]
CH3OH/ZnCl2+LiBr CH3OH/LiI CH3OH/ZnBr2+LiI CH3OH/ZnCl+ZnBr2	Simulation	double-stage	65 < Tg < 165 Te = −10 Ta = 30 Tc = 30	ZnCl2+LiBr: 0.31~0.33 LiI: 0.24~0.27 ZnBr2+LiI: 0.29~0.34 ZnCl+ZnBr2: 0.26~0.27	CH3OH/LiBr exhibited the highest COP; CH3OH/ZnBr2+LiI and CH3OH/ZnCl2+LiBr have the wider temperature operational range.	[58]
CH3OH/[mmim]DMP	Simulation	single-effect	92 < Tg < 112 2 < Te < 12 22 < Ta < 37 37 < Tc < 57	0.4~0.8	The COPs of CH3OH/[mmim]DMP are higher than NH3/H2O by 20%, while lower than H2O/LiBr by 10%, under typical operation conditions.	[37]
TFE/TEGDME	Simulation	diffusion absorption refrigeration	100 < Tg < 190 −17 < Te < 12 40 < Tc < 50	0.02~0.62	At low cooling temperatures, the TFE/TEGDME cycle is a good alternative to NH3/H2O.	[39]
TFE/TEGDMETFE/NMP	Simulation	single-effect	70 < Tg < 120 −15 < Te < 20 20 < Ta < 40 25 < Tc < 45	TEGDME: 0.45~0.82 NMP: 0.35~0.92	The COPs of TFE/NMP and TFE/TEGDME are higher than NH3/H2O, while the circulation ratios showed the opposite trend.	[40]
[Freon-based]
R134a/DMF	Simulation	single-effect	70 < Tg < 100 13 < Te < 21 25 < Ta < 35 25 < Tc < 35	0.39~0.53	-	[59]
R32/DMETEG R152a/DMETEG R161/DMETEG	Simulation	single-effect	80 < Tg < 180 Te = 5 Ta = 35 Tc = 40	R32: 0.07~0.43 R152a: 0.23~0.51 R161: 0.08~0.55	R161/DMETEG yields the highest COP when the Tg is above 135 °C, followed by R161/DMETEG when below 135 °C.	[44]
R32/DMETEG R152a/DMETEG R161/DMETEG	Simulation	compressor-assisted absorption	60 < Tg < 180 Te = 5 Ta = 35 Tc = 40	R32: 0.17~0.56 R152a: 0.34~0.62 R161: 0.33~0.69	A compressor can effectively enhance the absorptivity of working pairs, thus improving refrigeration performances.	[44]
R32/[hmim][TF2N] R152a/[hmim][TF2N] R125/[hmim][TF2N] R1234zeE/[hmim][TF2N] R1234yf/[hmim][TF2N]	Simulation	single-effect	Tg = 100 0 < Te < 25 30 < Ta < 40 40 < Tc < 50	R32: 0.05~0.62 R152a: 0.25~0.60 R125: 0.09~0.34 R1234zeE: 0.07~0.41 R1234yf: 0.02~0.35	The refrigeration performance of [hmim][TF2N]/HFCs (R32 and R152a) is better than that of [hmim][TF2N]/HFOs (R1234yf and R1234zeE).	[45]
R1234yf/DMETEG R1234yf/NMP	Simulation	single-effect	60 < Tg < 95 −5 < Te < 10 20 < Ta < 30 20 < Tc < 30	DMETEG: 0.053~0.409 NMP: 0.025~0.333	Incorporation of a compressor can improve the COP of the cycle and extend the operation range.	[47]
R1234yf/[emim][BF4] R1234yf/[hmim][TF2N] R1234yf/[hmim][BF4] R1234yf/[omim][BF4] R1234yf/[hmim][PF6] R1234yf/[hmim][TfO]	Simulation	single-effect	62 < Tg < 95 −11 < Te < 15 20 < Ta < 40 20 < Tc < 40	[emim][BF4]: 0.01~0.07 [hmim][TF2N]: 0.01~0.35 [hmim][BF4]: 0.01~0.24 [omim][BF4]: 0.01~0.30 [hmim][PF6]: 0.01~0.16 [hmim][TfO]: 0.01~0.22	R1234yf/[hmim][Tf2N] shows the best performance.	[48]
R1234yf/[emim][BF4] R1234yf/[hmim][TF2N] R1234yf/[hmim][BF4] R1234yf/[omim][BF4] R1234yf/[hmim][PF6] R1234yf/[hmim][TfO]	Simulation	compressor-assisted absorption	45 < Tg < 95 −20 < Te < 15 20 < Ta < 40 20 < Tc < 40	[emim][BF4]: 0.01~0.11 [hmim][TF2N]: 0.01~0.4 [hmim][BF4]: 0.01~0.28 [omim][BF4]: 0.02~0.31 [hmim][PF6]: 0.01~0.27 [hmim][TfO]: 0.01~0.30	Compared to the single-effect cycle, the compression-assisted cycle effectively improves the cooling performance, reduces the circulation ratio, and extends the operational conditions.	[48]
[Other types]
R600a/squalane DME/squalane	Simulation	single-effect	60 < Tg < 120 −8 < Te < 20 Ta = 30 Tc = 30	R600a: 0.02~0.95 DME: 0.02~0.81	Compared to DME/squalane, R600a/squalane performs better.	[52]
R600a/squalane	Simulation	compressor-assisted absorption	40 < Tg < 100 −12 < Te < 14 Ta = 30 Tc = 30	0.25~0.82	-	[52]
R290/[P6,6,6,14][Cl] R600a/[P6,6,6,14][Cl] DME/[P6,6,6,14][Cl]	Simulation	single-effect	65 < Tg < 110 30 < Ta < 50	R290: 0.08~0.35 R600a: 0.05~0.31 DME: 0.04~0.54	DME/[P6,6,6,14][Cl] exhibits the highest COP and ECOP at Tg above 72 °C and Ta below 42 °C.	[55]
CO2/[bmim][Tf2N]	Simulation	single-effect	100 < Tg < 185 −3 < Te < 7	0.13~0.21	Due to the limitation of absorbability, the circulation ratio of CO2/IL is much higher than NH3/H2O, and the COP is slightly lower.	[56]
CO2/[emim][Tf2N]	Simulation	novel configuration	60 < Tg_l < 90 100 < Tg_h < 120 −5 < Te < 11 19 < Ta < 27 19 < Tc < 27	0.25~0.70	The COP of a CO2/[emim][Tf2N] single-effect absorption cycle was improved by 50% through configuration optimization.	[57]

summarizes the performance characteristics of the different types of working pairs. The utilization of multi-component salt absorbents alleviates the crystallization and corrosion issues of H₂O/LiBr and extends the range of driven heat source temperatures. To avoid the necessity of additional distillation equipment for NH₃/H₂O, NH₃/LiNO₃ and NH₃/NaSCN have been extensively investigated. Since LiNO₃ and NaSCN are salts with a large difference in boiling temperature from NH₃, no distillation equipment is required, and the specific heat capacities of both NH₃/LiNO₃ and NH₃/NaSCN are smaller than that of NH₃/H₂O, which is conducive to reducing the heat exchanger area required. Compared with NH₃/H₂O, NH₃/LiNO₃ and NH₃/NaSCN have higher COPs under most operating conditions. However, both of them have higher viscosities than NH₃/H₂O, which increases the flow and mass transfer resistance, and there is a potential risk of crystallization when the mass fractions of LiNO₃ and NaSCN are high. Alcohol-based and Freon-based working pairs are suitable for low-temperature driven heat sources and could generate cooling temperatures below 0 °C. With the growing global demand for environmentally friendly refrigerants, natural refrigerants such as R290, R600, and CO₂ are gaining significant attention, providing a new research direction for absorption working pairs. The exploitation of new types of working pairs has been the research focus of ART in recent years, in which the working pairs with ionic liquids and deep eutectic solvents as new absorbers have shown promising developmental potential. Ionic liquids could be well miscible with water, ammonia, alcohols, halogenated hydrocarbons, or hydrocarbon refrigerants to form good working pairs. Notably, imidazolium-based ionic liquids paired with polar refrigerants, such as NH₃/[emim][Tf2N] and H₂O/[emim][DMP], exhibit excellent refrigeration performance, achieving a COP comparable to that of H₂O/LiBr, without corrosion and crystallization. However, ionic liquids pose challenges due to their high cost, complex synthesis, and high viscosity. Alternatively, deep eutectic solvents, which could also form working pairs with different types of refrigerants, are easy to synthesize and have lower costs than ionic liquids. However, current research on deep eutectic solvents is scant and predominantly theoretical; further experimental study is necessary to evaluate their feasibility. In addition, the common shortcoming of ionic liquids and DESs is that they generally have poor absorptivity for refrigerant vapor, and the circulation ratio is higher than that of H₂O/LiBr and NH₃/H₂O. To achieve a high COP, a compressor-assisted absorption refrigeration cycle needs to be adopted for the working pairs of ionic liquids and DESs.

2.2. Cycle Configurations

This section analyzes seven distinct configurations of the absorption refrigeration cycle: single-effect, multi-effect, multi-stage, auto-cascade, GAX, ejector-assisted, and compressor-assisted absorption cycles. This section briefly outlines the working principle for each cycle configuration and reports the latest advancements in optimizing its operating conditions and thermodynamic performance.

2.2.1. Single-Effect Absorption Cycle

The single-effect absorption refrigeration cycle, characterized by its simplicity in structure, holds an important position in the refrigeration domain. Its operational principle is illustrated in Figure 1. Numerous researchers have conducted comprehensive evaluations and optimization studies on this cycle from different aspects, including theoretical analysis, computational simulation, and experimental verification. These studies primarily focus on optimizing the performance of working pairs, analyzing thermodynamic operational conditions, and refining operational strategies [60]. The optimization of working pair performance is discussed in Section 2.1.

Soheil et al. [61] utilized EES software to assess the steady-state performance of a small solar absorption refrigeration system, examining the impact of solution concentration, generation temperature, and condensation temperature on the COP and identifying the optimal thermodynamic operating conditions. Muhammad et al. [62] devised the correlation equation to predict the COP of a single-effect cycle for 27 working pairs derived from the physical properties of the working pairs and cycle operating parameters. The literature validation results showed that the deviation of the COP prediction was less than 30%. Liu et al. [63] proposed a dual-loop energy-saving control strategy to address the limitation of typical single-closed-loop control strategies, which meet refrigeration capacity demands but fail to operate with high efficiency. Established on the particle swarm optimization algorithm, the novel control strategy aimed to enhance overall system energy efficiency by dynamically adjusting operating parameters in response to system conditions and load demands.

2.2.2. Multi-Effect Absorption Cycle

To enhance the energy efficiency of the single-effect cycle under high temperatures of the driven heat source, researchers have introduced the multi-effect absorption refrigeration cycle. This process utilizes external heat input for one time, through internal heat coupling within the cycle, enabling the generation of multiple streams of refrigerant vapors, thereby optimizing the high-temperature heat source utilization efficiency. The double-effect absorption cycle is the most widely used multi-effect absorption cycle. It is categorized into parallel and series double-effect absorption cycles based on the flowing approach of the concentrated solution from the absorber to two separate generators, with schematics illustrated in Figure 2.

Muhammad et al. [52,64] developed a thermodynamic and economic optimization model for a double-effect absorption refrigeration system based on a genetic algorithm. With the optimization objective of minimizing the total annual cost, the heat transfer area and thermodynamic operational conditions of series and parallel units were optimized. The results showed that the total annual cost of the series and parallel units could be reduced by 33% and 26%, respectively. Arun et al. [65] compared the double-effect series and parallel cycles, discovering the efficiency of the parallel cycle surpasses the series. However, the refrigeration performance of the parallel cycle is significantly influenced by the low-temperature solution heat exchanger, and it faces challenges in controlling the solution distribution ratio, proving more complex than series configurations. Azhar et al. [66] adopted exergy efficiency as the objective function to optimize the operational parameters of the primary and intermediate generators, condenser, and solution distribution ratio in a lithium bromide double-effect parallel absorption cycle. They compared the exergy efficiencies between double-effect parallel and series cycles. The result indicated that the maximum exergy efficiency of the parallel cycle was 3~6% higher than that of the series cycle.

The triple-effect and quadruple-effect absorption cycles were proposed to further enhance the energy efficiency under the higher driven heat source temperature. Gomri [67] conducted a comprehensive analysis comparing the energy and exergy efficiencies of single-, double-, and triple-effect cycles, analyzing the impact of generation and refrigeration temperatures on cycle performances.

Currently, research on the triple-effect cycle is primarily based on simulation, whereas quadruple- or higher-effect cycles remain in the conceptual stage basically, and further experimental studies are needed to verify their practical performance in the future [68,69].

2.2.3. Multi-Stage Absorption Cycle

The COP of the single-effect cycle significantly diminishes as heat source temperature decreases. Multi-stage absorption cycles, which lower the temperature requirements of the heat source, are proposed to address this issue. The multi-stage absorption cycles output the cooling load for one time while inputting driven thermal energy for two or more times through heat and mass coupling within the cycle. The multi-stage absorption cycle represents a more sophisticated and efficient version of traditional absorption refrigeration systems, offering enhanced COP, flexibility, and energy utilization capabilities through the integration of multiple absorption and desorption stages. This advanced configuration is particularly well suited for applications where high efficiency and adaptability to varying heat sources are essential considerations. The two-stage absorption cycle is the most widely studied and practically applied, with its schematic illustrated in Figure 3.

Maryami et al. [68] compared the two-stage, single-effect, and double-effect lithium bromide absorption units, focusing on their energy and exergy efficiencies in various operating conditions. Arora A. et al. [70] theoretically analyzed the impact of generation temperature, solution heat exchanger efficiency, and generator pressure on the performance of a two-stage cycle. They determined that the COP at optimal operating conditions ranged from 0.415 to 0.438, while exergy efficiency varied between 6.96% and 13.74%. Ziegler et al. [71] explored different configurations of multi-stage absorption cycles and carried out a performance comparison with other types of absorption cycles in thermodynamic analysis [72].

The driven heat source required for multi-stage cycles is of low grade and generally freely available, such as solar energy or industrial waste heat. Hence, the “electrical COP” is also an important performance indicator. Du [73] and Aprile et al. [74] conducted experimental research on an air-cooled two-stage cycle using the NH₃/H₂O working pair. The results revealed that the COP varied between 0.25 and 0.5, with electrical COP ranging from 5.1 to 10. Furthermore, two-stage absorption cycles can employ both the single solution and dual solutions for the high- and low-pressure stages, such as lithium chloride and lithium bromide solutions, respectively [75].

2.2.4. Auto-Cascade Absorption Cycle

The auto-cascade absorption refrigeration cycle, employing multiple non-azeotropic refrigerants, utilizes the principle of multi-stage condensation with mixed refrigerants to achieve low-temperature refrigeration. Implementing the auto-cascade technique in a compression refrigeration cycle could significantly decrease the number of compressors, enhance the heat coupling capacity of circulation, and lower the evaporating temperature. The auto-cascade absorption cycle is particularly useful for achieving very low temperatures that may not be feasible with single-effect refrigeration systems. It allows for efficient cooling at temperatures significantly below those achievable with conventional refrigeration cycles. Components of refrigerants with lower boiling points are progressively separated and condensed through multi-stage sub-condensation, and eventually evaporated to provide the low-temperature cooling load. The absorption cycle replaces the compressor in the compression refrigeration cycle with the solution cycle, and the schematic is shown in Figure 4.

He et al. [76] conducted theoretical and experimental investigations with an auto-cascade absorption refrigeration cycle, successfully achieving refrigeration at −47.2 °C with a generating temperature of 163 °C. Through further optimizations of the working pairs and operating conditions, He et al. [77] used the R23+R32+R134a/DMF auto-cascade cycle to generate an evaporation temperature of −52.9 °C under the generation temperature of 122.5 °C. He et al. [78] developed a novel auto-cascade absorption cycle adopting two absorbers, reaching cooling temperatures down to −58 °C. In 2020, He et al. [79] formulated equivalent cycles for reversible and irreversible auto-cascade absorption refrigeration cycles and assessed the thermodynamic perfection index.

2.2.5. GAX Absorption Cycle

The generator–absorber heat exchange (GAX) cycle represents a continuously variable-effect absorption refrigeration cycle that utilizes the characteristics of the temperature overlap between the generator and the absorber, thus effectively reducing the heat input through continuous internal heat recovery. The operational principle is depicted in Figure 5.

In the 1980s, Rojey et al. [80] originally proposed the concept of the GAX cycle. Subsequently, Sivalingam et al. [81] conducted experimental evaluations of the GAX cycle using the NH₃/H₂O working pair, achieving the maximum COP of 0.63, representing an improvement of 10~30% over the traditional single-effect cycle. Wang [82] conducted a theoretical study of a GAX cycle with a NH₃-H₂O-LiBr solution as a working pair. The results showed that the advantage of the GAX absorption cycle improved with an increase in the heat source temperature. Under the design parameters of the evaporating temperature of 5 °C, generating temperature of 140 °C, condensation temperature of 39 °C, and lithium bromide mass fraction of 10%, the COP of the GAX cycle significantly surpasses that of the single-effect cycle by 18%. However, the GAX cycle does not offer the notable performance enhancement effect at generation temperatures below 113 °C. To further augment the efficiency of the GAX cycle, Erickson et al. [83] devised a branched GAX configuration, which increases the solution flow rate to enhance the generation–absorption heat exchange by adding a solution pump between the hot-steam side of the absorber and the cold-steam side of the generator.

2.2.6. Ejector-Assisted Absorption Refrigeration Cycle

An ejector can effectively utilize fluid pressure energy when a significant pressure differential exists within the system. The integration of ejection and absorption represents an effective method for improving the performance of the absorption cycle, in which the ejector recuperates the pressure potential energy of high-pressure gas [84]. The typical configuration of the ejector-assisted absorption refrigeration cycle is illustrated in Figure 6.

To improve the single-effect cycle performance and reduce the generating temperature of the double-effect cycle, Hong et al. [85] integrated an ejection into the single-effect cycle, which improved the performance of the cycle by more than 20% compared to the conventional single-effect cycle. Meanwhile, the generating temperature varied between the single- and the double-effect cycle in variable effect. Farshi et al. [86] conducted a comparative thermal-economic analysis of the double-effect ejector-assisted absorption cycle. The result showed that the cycle with the ejector makes full use of the vapor exergy from the generator and the pressure potential energy of the solution, resulting in superior thermal economy. Shi et al. [87] proposed the application of an ejector to a two-stage absorption refrigeration cycle, where the refrigerant vapor from the high-pressure generator directly ejected the refrigerant vapor from the low-pressure generator to the condensation. The results revealed that the COP of the two-stage ejector-assisted absorption cycle with the H₂O/LiBr working pair increased by more than 20%. Ghariri et al. [88] adopted flash tanks to optimize the basic ejector-assisted absorption cycle, which increased the entrainment rate of the ejector. The first flash tank is positioned between the condenser and evaporator, and the second between the absorber and generator lowers the operational temperature of the generator, reducing the thermal energy required. This modification increases the COP and exergy efficiency by 56% and 22%, respectively.

2.2.7. Compressor-Assisted Absorption Refrigeration Cycle

The compressor is pivotal in the mechanical compression refrigeration cycle. The compressor-assisted absorption cycle has been developed to decrease the refrigerant’s evaporation pressure in the evaporator and lower the refrigeration temperature. The compressor generally be arranged after the evaporator or generator because of the high dryness requirements of the fluid; thus, processes are categorized into low-pressure or high-pressure processes [89]. The schematics are shown in Figure 7.

Hu et al. [90] investigated the impact of intermediate pressure on the efficiency of low- and high-pressure compressor-assisted absorption cycles powered by renewable ocean energy, using an ammonia/water working pair. The result showed that the low-pressure cycle exhibited higher energy and exergy efficiencies. Sun et al. [48] explored using R1234yf/ionic liquid as working pairs. They analyzed the impact of generation, absorption, evaporation, condensation temperatures, and compression ratio on the performance of both single-effect and compressor-assisted absorption cycles. The result indicates that the compressor-assisted absorption cycle significantly enhances absorption efficiency, improves refrigeration performance, reduces circulation ratio, and widens the operational temperature range for generation, absorption, and evaporation. Kumar et al. [91] demonstrated that positioning the compressor between the absorber and evaporator within the GAX cycle increased the COP by 30%. Kim et al. [92] conducted a comparative study on the different arranges of compressors in a triple-effect lithium bromide solution cycle. The results reveal that the cycle with the compressor lowers the generation temperature and mitigates the high-temperature corrosion of the lithium bromide solution. Further, investigations by some researchers [93] on applying compressors in the two-stage cycle, where the compressor links the high-pressure-stage absorber and the low-pressure-stage generator, have also been reported.

Integrating the absorption and compression refrigeration cycles to form an absorption–compression cascade cycle constitutes a prevalent method for achieving lower refrigeration temperatures. The high-temperature refrigeration side of the cycle adopts the absorption cycle, and the low-temperature refrigeration side adopts the compression cycle. The absorption cycle provides the cooling load for the condenser of the compression cycle; relevant studies can be found in the literature [94,95,96].

A variety of absorption refrigeration cycle configurations have been developed to extend the operating temperature range and improve energy utilization efficiency. Seven configurations of absorption refrigeration are analyzed, including the basic single-effect absorption cycle, the multi-effect absorption cycle for high-temperature driving heat sources, the multi-stage absorption cycle used in low-temperature driving heat sources, the auto-cascade absorption cycle for very low refrigeration temperatures, and the ejector-assisted absorption cycle and compressor-assisted absorption cycle, which incorporate sub-components to improve system energy utilization efficiency and extend the operating conditions.

Table 2. Summary of the characteristics of different types of absorption cycles.

Cycle Configuration	Method	Working Pair	Operating Temps. (°C)	COP	Remarks	Refs.
Single-effect	Experiment	H2O/LiBr	85 < Tg < 92 Te = 6 36 < Ta < 45 30 < Tc < 43	0.05~0.77	Optimization of operating conditions can significantly improve cycle performance.	[61]
	Experiment	H2O/LiBr	Tg = 85 Te = 12 Ta = 30 Tc = 30	0.4175	The novel double-closed-loop control strategy improves energy efficiency by 19.3% under typical conditions.	[63]
Double-effect (series)	Simulation	H2O/LiBr	115 < Tg < 190 4 < Te < 10 33 < Ta < 39 33 < Tc < 39	0.74~1.32	The COP of the double-effect cycle is about twice that of the single-effect, while the required Tg is higher.	[67]
Double-effect (parallel and series)	Simulation	H2O/LiBr	2 < Te < 10 25 < Ta < 40 30 < Tc < 45	parallel: 1.32~1.50 series: 1.08~1.43	The COPs of parallel cycles are higher than series cycles in most conditions, while the distribution ratio of parallel cycles needs to be regulated.	[65]
Triple-effect (series)	Simulation	H2O/LiBr	135 < Tg < 225 4 < Te < 10 33 < Ta < 39 33 < Tc < 39	1.07~1.63	The triple-effect cycle can work efficiently at high Tg.	[67]
Two-stage	Simulation	H2O/LiBr	50 < Tg < 85 4 < Te < 10 33 < Ta < 39 33 < Tc < 39	0.18~0.44	Compared to the single-effect cycle, the COP of a double-stage cycle is higher under a low Tg (around 65 °C and below).	[68]
	Experiment	NH3/H2O	69 < Tg < 88 −6 < Te < 8 21 < Ta < 29 32 < Tc < 36	0.13~0.29	The performance of the air-cooled two-stage NH3/H2O cycle driven by solar energy has been tested in a prototype.	[73]
	Simulation	H2O/LiCl and H2O/LiBr	60 < Tg < 70 1 < Te < 9 30 < Ta < 40 30 < Tc < 40	0.28~0.40	The cycle with dual solutions shows a higher COP than the traditional two-stage cycle, with a maximum COP improvement of 35%.	[75]
Auto-cascade	Experiment	R23+R134a/DMF	Tg = 157 Te = −47.2 Ta = 28 Tc = 18	0.011	The cooling temperature of ARAC is much lower than the traditional absorption refrigeration cycle.	[76]
	Experiment	R23+R32+R134a/DMF	Tg = 122 Te = −52.9 Ta = 28 Tc = 17	0.011	-	[77]
	Simulation	R23+R134a/DMF	Tg = 160 −57.5 < Te < −56.5 Ta = 30 Tc = 30	0.399~0.415	The COP of a novel ACAR with a double absorber is 20% higher than that of an ACAR with a single absorber.	[78]
GAX	Experiment	NH3/H2O	25 < Tg < 50 −15 < Te < 20 0 < Ta < 80 Tc = 24	0.22~0.63	The maximum COP of the GAX cycle is higher (0.1–0.3) than that of the single-effect cycle.	[81]
	Simulation	NH3/H2O	110 < Tg < 180 Te = 5 Ta = 30 Tc = 39	0.45~0.65	The performance advantage of the GAX cycle is even more significant at high Tg.	[82]
Ejector-assisted absorption refrigeration	Simulation	H2O/LiBr	120 < Tg < 155 2 < Te < 14 Ta = 40 Tc = 40	0.72~1.5	The pressure potential energy in a single-effect cycle has been utilized by the ejector; thus, the COP improves by 0~80%.	[85]
	Simulation	H2O/LiBr	70 < Tg < 90 1 < Te < 15 Ta = 40 Tc = 40	0.40~0.79	The pressure potential energy in a two-stage cycle has been utilized by the ejector; thus, the COP improves by 0~78%.	[87]
	Simulation	NH3/H2O	70 < Tg < 125 −10 < Te < 10 30 < Ta < 40 30 < Tc < 40	0.04~1.86	The addition of two flash tanks enhances the performance of the ejector-assisted absorption cycle.	[88]
Compressor-assisted absorption refrigeration	Simulation	R1234yf/[emim][BF4] R1234yf/[hmim][TF2N] R1234yf/[hmim][BF4] R1234yf/[omim][BF4] R1234yf/[hmim][PF6] R1234yf/[hmim][TfO]	62 < Tg < 95 −11 < Te < 15 20 < Ta < 40 20 < Tc < 40	[emim][BF4]: 0.01~0.12 [hmim][Tf2N]: 0.01~0.40 [hmim][BF4]: 0.02~0.30 [omim][BF4]: 0.01~0.32 [hmim][PF6]: 0.01~0.28 [hmim][TfO]: 0.01~0.31	The utilization of a compressor significantly improves refrigeration performance, reduces the circulation ratio, and widens the temperature range of operational conditions.	[48]
	Simulation	H2O/LiBr	45 < Tg < 75 4 < Te < 10 32 < Ta < 38 32 < Tc < 38	0.38~0.44	Compared to the conventional two-stage cycle, the Tg of a compressor-assisted two-stage cycle is lower by 7~10 °C with higher COP.	[93]
	Simulation	H2O/LiBr	Tg = 162.8 Te = 5 Ta = 35 Tc = 40	1.70~1.74	Four types of compressor-assisted triple-effect cycles were investigated; the Tg of a compressor-assisted cycle is lower by 40 °C, approximately.	[92]

summarizes the operating conditions and performance characteristics of the different cycle configurations.

2.3. Heat and Mass Transfer in Main Components

The distinguishing characteristic of an absorption refrigeration system is the pressurization process, which is achieved through the refrigerant absorbed and desorbed from the circulating solution. The heat and mass transfer processes of the absorber and generator significantly influence the COP of the system. This section reviews the progress of typical absorbers and generators and heat and mass transfer enhancement technologies in terms of additives, ultrasonic oscillations, and structural optimization.

2.3.1. Heat and Mass Transfer in Absorber

In the absorber, the concentrated solution absorbs the refrigerant vapor from the evaporator to ensure the stable operation of the system. Enhancing the heat and mass transfer facilitates the miniaturization of the unit design and reduces the flow rate of the cooling source. Based on the different mechanisms of the absorption process, absorbers are generally classified into four types: falling-film, bubble, spray, and membrane media absorbers.

The falling-film absorber, renowned for its high-efficiency heat and mass transfer coefficients and large gas–liquid contact area, has garnered widespread research and application [97]. The addition of additives, such as nanoparticles and alcohol additives, to the base fluid could significantly enhance its thermal conductivity, thereby increasing the heat and mass transfer rate of the absorption process. Gao et al. [98] developed a heat and mass transfer coupling model based on the finite-element approach for the H₂O/LiBr falling-film absorption. The results show that adding CuO nanoparticles into the lithium bromide solution augments the average mass transfer rate in the falling-film absorber by 28~75%. Lee et al. [99] experimentally tested the enhancement effect of nanoparticles (CNTs and Al₂O₃) on the ammonia absorption process. They demonstrated that a 0.02% volume fraction of CNT increased the absorption and heat transfer rates by 17% and 16%, respectively. Similarly, a 0.02% volume fraction of Al₂O₃ increased the absorption and heat transfer rates by 29% and 18%, respectively. Kang et al. [100] experimentally analyzed the impact of carbon nanotubes (CNTs) and Fe nanoparticles on the heat and mass transfer in the H₂O/LiBr falling-film absorption process. The results showed that the enhancement of the mass transfer rate was more significant than that of the heat transfer rate, and the effect of CNTs was better, with 0.1 wt.% CNTs increasing the mass transfer rate by 2.48 times. Zhang et al. [101] analyzed the impact of three mechanical treatments of tube surfaces and two surface activators (2-ethyl-1-hexanol and 1-octanol) in the falling-film absorption process. They discovered that each kind of tube’s heat and mass transfer coefficients decreased progressively as the solution moved down the tube bundle, and the influence of surfactants on heat and mass transfer coefficients was more significant than that of mechanical surface treatments. The results showed that 2-ethyl-1-hexanol significantly improved the heat and mass transfer rates by 400% and 350%, respectively. Jiang [102] conducted a comprehensive study about the impact of TiO₂ nanoparticles on the absorption process, generation process, and overall system performance through theoretical analysis and experimental investigation. It was determined that the system’s optimal performance occurred at a TiO₂ concentration of 0.5 wt.%, where the COP increased by 25%. Subsequently, the researchers introduced 0.02 wt.% SDBS as a surfactant of the TiO₂ nanoparticles to enhance the nanoparticles’ dispersion stability in the base solution. This modification resulted in nanofluids with higher thermal conductivity and reduced surface tension, further increasing the system’s COP. Zhao et al. [103] designed a novel falling-film absorber that consisted of microchannel flat tubes and rhombic mesh sieves. They compared coverage and liquid film thickness between the novel and conventional absorber. They explored the liquid film’s thickness, fluctuation, and internal mixing based on the three-dimensional VOF model. The reinforced structure of the rhombic mesh sieve improved surface coverage and promoted liquid film fluctuation and mixing. The ultrasonic oscillation technique could effectively increase the heat and mass transfer rate through cavitation effects and mechanical oscillations. He et al. [104] simulated NH₃/H₂O falling-film absorption characteristics, which were enhanced by ultrasonic atomization. The results showed that the optimal position for the ultrasonic nebulizer is 0.45m from the tube’s top, resulting in a 15.1% increase in absorption rate and elevating the NH₃ mass fraction of the outlet solution by 1.2%. Installing an ultrasonic atomizer on the absorber significantly amplified the liquid–vapor contact area. Zhou et al. [105] theoretically analyzed the effect of the atomizer installation method, atomization ratio and droplet parameters on heat and mass transfer in a NH₃/H₂O falling-film absorber. The results show that the ultrasonic atomizer improves the absorption rate and system energy efficiency under a large range of operating conditions.

The bubble absorber, with the advantages of simple structure and minimal vibration interference, has received wide attention in developing miniaturized air-cooled absorption refrigeration units in recent years [106]. The dynamics of bubble morphology during absorption substantially influence the heat and mass transfer characteristics. Li [107] focused on the H₂O/LiBr solution bubble absorption, constructing an experimental setup to explore the flow and heat transfer characteristics. Through experimental visualization, Li discussed the changes in bubble flow morphology and size, assessing the effect of various factors on the heat transfer performance of bubble absorption. A convective heat transfer model of bubble absorption was established based on the experimental data. The heat transfer characteristics of the bubble absorption process and the flow and temperature fields in the vertical tube bubble absorber were numerically analyzed. Kong et al. [108] theoretically and experimentally analyzed the heat and mass transfer of R134a/DMF in a bubble absorber, focusing on the flow pattern changes during the absorption process, including stirred, elastoplastic, vesicular, and single-phase flows. Wang et al. [109] proposed the application of ultrasonic oscillation and TiO₂ nanoparticles in the bubble absorption process to address ammonia absorption issues caused by LiBr in the NH₃-H₂O-LiBr ternary solution. Experimental results demonstrated significant enhancements in the ammonia absorption process by this hybrid strategy. When 68 kHz ultrasonic treatment and 0.1 wt.% nanoparticles were employed, the absorption rate increased by 26%. Kim et al. [110] experimentally tested the heat and mass transfer enhancement effect of nanoparticles (Cu, CuO, Al₂O₃) on the NH₃/H₂O bubble absorption process. The results revealed that Cu nanoparticles showed the most significant enhancement effect, which increases the absorption rates by 3.21 times with 0.1 wt.%.

The spray absorber, which ejects the concentrated solution of the generator into the absorber chamber, facilitates the absorption of refrigerant vapor to form a dilute solution. This process disperses the dilute solution within the chamber in droplet form, enhancing mass transfer by reducing solution resistance and accelerating the refrigerant vapor absorption rate. Palacios et al. [111] examined the vapor absorption rate in an adiabatic spray absorber, finding that the decomposition rate of the conical liquid film significantly impacts absorption efficiency. Notably, liquid film decomposition reached 60% merely 0.004 s post ejection, with the average mass transport coefficient of the conical liquid film reaching 0.002 m/s. Wang et al. [112] proposed the incorporation of phase change microcapsules (PCMs) into the circulation solution to augment the effective heat capacity, thereby diminishing solution recirculation and atomization power consumption. Their analysis, utilizing an adiabatic spray absorption model for lithium bromide solution, indicated that PCMs contribute to lower atomization power consumption at equivalent cooling capacities, but absorption times were extended simultaneously. Jiang et al. [113] introduced a novel absorber design that integrated spray and falling-film absorption techniques, featuring two refrigerant vapor inlets below the nozzles and the falling-film tubes. This innovative configuration improved refrigerant vapor’s uniform distribution within the absorber, thus enhancing absorption efficiency.

Membrane media, known for their simplicity, high separation coefficients, large interfacial areas, and efficient heat and mass transfer performance, have led researchers to propose the use of hydrophobic porous membranes (HPMs) in an absorber, which could reduce the absorber’s size, weight, and cost [97]. The HPM medium is positioned between the solution and vapor, allowing only vapor to pass and preventing solution penetration into the vapor side. The vapor pressure differential between the two sides of the membrane drives mass transfer [114]. Isfahani et al. [115] utilized the HPM in 80% porosity and 1 μm pore size as a separation layer between lithium bromide solution and water vapor, exploring the impact of solution membrane thickness and flow rates on water vapor absorption. The result indicated a 35% increase in absorption rate when the membrane thickness decreased from 160 to 100 μm, attributed to the reduced interfacial concentration between the solution and vapor, thus lowering mass transfer resistance. Zhai et al. [116] conducted experiments on the microchannel membrane absorber in different operational conditions to modify the existing correlation formulas, establishing new calculations for the dimensionless number (Nusselt and friction factors). Lv [117] focused on a hydrophobic membrane plate and frame absorber made of microporous polytetrafluoroethylene. A new numerical model was developed by integrating vapor transport across the membrane with the absorption process. They studied the heat and mass transfer mechanisms and the influence of geometrical parameters on performance. Some researchers have applied membrane media to ammonia absorption units and carried out performance studies [118].

2.3.2. Heat and Mass Transfer in Generator

A generator absorbs the thermal energy from a driven heat source and desorbs the refrigerant vapor from the circulating solution. Immersed generators are simple in structure and developed for maturity basically. The falling-film generator and the membrane generator are more efficient and are still in the developmental stage; thus, this section focuses on the progress of these two types of generators.

The falling-film generator is distinguished by its high heat and mass transfer coefficients on the solution side, minimal filling volume of solution, and negligible temperature difference between the solution and the metal wall. In a falling-film generator, the heat and mass transfer process occurs in the solution film; thus, the refrigerant vapor diffuses directly out of the solution film. The solution membrane boils with increased heat flux and superheating temperature of the wall. The key factors influencing the performance of falling-film generators include the mass flow rate of solution, generating temperature, spraying density, and geometrical parameters [97]. Romo et al. [119] utilized falling-film technology in a generator and distiller to meet the high-temperature demands of the driven heat source in an ammonia absorption refrigeration cycle, assessing the impact of driven heat source temperature, solution concentration, solution flow rate, and distillation temperature on the heat and mass transfer efficiency. Hu et al. [120] designed and experimentally tested a novel plate-type falling-film generator, which consists of a directly connected falling-film overflow liquid distributor, to improve the solution distribution uniformity and the stability of the falling-film flow. They also proposed a separable plate falling-film heat exchanger coupling device that could be used for the coupling of the generator and the condenser, or the absorber and the evaporator. Experimental results indicated heat transfer coefficients ranging from 0.345 to 0.660 kW/(m²·K) and mass transfer coefficients between 0.000027 and 0.000078 m/s for H₂O/LiBr [121]. Wirtz M et al. [122] designed a combined generator–distiller unit incorporating a falling-film plate heat exchanger for a single-effect ammonia absorption system, analyzing generator efficiency through numerical simulations of the outlet vapor ammonia mass fraction, mass flow rate of the hot-side fluid, and solution inlet temperature. Lee et al. [123] tested seven different heated tube bundle surface configurations in a falling-film generator, finding that surface-modified heating tube bundles significantly improved shell-side heat and mass transfer (around 60%) compared to conventional bare tube bundles. Enhancements in the heat transfer characteristics of a falling-film generator were also observed with increased driven heat source temperature and inlet Reynolds number of the recirculating solution, along with reduced generation pressure. Jiang [102] conducted a comprehensive study on the impact of TiO₂ nanoparticles on the generation process and overall system performance through theoretical and experimental investigation. The results show that the system’s optimal performance occurred at the TiO₂ concentration of 0.5 wt.%, where the COP increased by 25%. Subsequently, the researcher added 0.02 wt.% SDBS as the surfactant to enhance the dispersion stability of TiO₂ in the base solution, leading to a further increase in the COP. Li [124] analyzed the effects of ZnFe₂O₄, SiC, and TiN nanoparticles on the generation rate of an ammonia falling-film generator from theoretical and experimental perspectives. The results showed that ZnFe₂O₄ had the most significant enhancement effect on heat and mass transfer. The combination of the surfactant SDBS further improved the mass transfer performance by increasing the dispersion of ZnFe₂O₄ in solution. The result showed that the addition of 0.1 wt.% ZnFe₂O₄ and 0.05 wt.% SDBS increased the generation rates by 60%. Ultrasonic oscillation enhances the heat and mass transfer rate of the solution boiling process through the cavitation effect. Tang et al. [125] experimentally investigated the effect of ultrasonic waves on the heat and mass transfer in the generator of a lithium bromide refrigeration unit. The results showed that ultrasonic waves increase the generation rate by 20~60% when the driving heat source temperature was 65~80 °C, and the strengthening effect is more significant with a lower heat source temperature. Zhu et al. [126] experimentally investigated the effect of ultrasonic waves at a frequency of 28 kHz on a lithium bromide solution generator, with variables including the height of the solution level in the generator. The results showed that the cooling capacity and COP of the unit increased by 19.6% and 13.8%, respectively, when the liquid level was 8~10 cm above the center line of the ultrasonic transducer. Hou et al. [127] analyzed the effects of solution temperature, ultrasonic frequency, and solution concentration on ultrasonic enhancement efficiency based on mathematical models. Hou et al. [128] further investigated the impact of multiple ultrasonic synergistic enhancements on the lithium bromide solution generation process. The experimental results showed that multiple ultrasonic vibrators had a more significant cavitation effect on the solution in the generator compared to a single ultrasonic vibrator. The generation rate was increased by 10.26% for dual ultrasonic vibrators and 5.69% for single ultrasonic vibrators at a frequency of 25 kHz and total ultrasonic power of 60 W. Zheng et al. [129] analyzed the impact of generator surface structure optimization (smooth, screwed and finned) and ultrasonic vibration (21 and 45 kHz) on the heat transfer of a lithium bromide solution generator through a combination of theoretical and experimental methods. The results showed that the generator with finned tubes and ultrasonic vibration of 21 kHz increases the heat transfer coefficient by 17.85%.

The microchannel membrane generator could separate the refrigerant vapor lower than the solution’s boiling temperature, thus extending the temperature range of driven heat sources. The pressure differential resulting from the temperature difference between the two sides of the membrane drives the refrigerant vapor mass transfer. At the liquid–gas interface within the membrane pores, the refrigerant evaporates at the thermal boundary, traversing the membrane pores in a gaseous state before condensing on the cooler side. He et al. [130] conducted an experiment and simulation to assess the heat and mass transfer properties of a generator that features porous polytetrafluoroethylene hollow fiber membranes. The result revealed that an elevated solution flow rate enhances the desorption rate of water vapor, with desorption experiments using a lithium bromide solution achieving a concentration differential of 0.4%. Vega et al. [131] proposed an air-cooled lithium bromide solution membrane generator/condenser employing porous polytetrafluoroethylene membranes of 45 μm pore size. They conducted experiments to evaluate the influence of solution flow rate, heat source temperature, and solution solubility on desorption rates and the solution pressure drop. The result showed that the desorption rates ranged from 0.003 to 0.014 kg/(m²·s), and the solution pressure drop ranged from 70 to 105 Pa/m. Venegas et al. [132] experimentally investigated the desorption performance of microporous polytetrafluoroethylene membranes with lithium bromide solution as the recirculating solution, using hot water at 62~66 °C as the driven heat source. The experimental results showed that increasing the inlet temperature of hot water increased the desorption rate, which was due to the higher water temperature increasing the water pressure in the solution, thus improving the pressure potential energy of desorption; the effective desorption area accounted for 85~89% of the contact area, and the actual desorption rate was 0.0016~0.0042 kg/m².

The enhancement of heat and mass transfer in the absorber and generator is crucial to the performance improvement of ART. It is conducive to improving energy efficiency and promoting the miniaturization of ART. This section analyzes different types of absorbers (falling-film, bubble, spray, and membrane-based absorbers) and generators (falling-film and membrane-based generators) regarding their working mechanisms and heat and mass transfer enhancement approaches. The main enhancement techniques can be classified into additives, ultrasonic oscillation, and surface structure treatment.

Table 3. Summary of enhancement technologies on heat and mass transfer.

Enhancement Strategy	Method	Enhanced Process	Working Pair	Remarks	Refs.
additive	Simulation	falling-film absorption	H2O/LiBr	The CuO nanoparticles augment the average mass transfer rate of the absorber by 28~75%.	[98]
	Experiment	falling-film absorption	H2O/LiBr	The mass transfer enhancement was 2.48 times with the 0.1 wt.% CNT and 1.9 times with the 0.1 wt.% Fe nanoparticles.	[100]
	Simulation	absorption process	H2O/LiBr	The addition of Al2O3 nanoparticles increases the COP by 1–35% under Tg of 87~120 °C; the maximum COP occurred with 0.2% volume fraction of Al2O3.	[133]
	Experiment	falling-film absorption	H2O/LiBr	2-ethyl-1-hexanol improved the heat and mass transfer rates by 400% and 350%; 1-octanol improved the heat and mass transfer rates by 350% and 155%.	[101]
	Experiment	bubble absorption	NH3/H2O	The addition of nanoparticles (Cu, CuO, Al2O3) enhances the bubble absorption performance; Cu with 0.1 wt.% enhances the absorption rates up to 3.21 times.	[110]
	Experiment	absorption process	NH3/H2O	CNT with 0.02% volume fraction increases the absorption rate and heat transfer rate by 17% and 16%; Al2O3 with 0.02% volume fraction increases the absorption rate and heat transfer rate by 29% and 18%.	[99]
	Experiment	-	NH3/LiNO3	CNT nanoparticles with 0.01 wt.% enhance thermal conductivity by 7.5%.	[134]
	Simulation and experiment	generation process	NH3/H2O	The addition of TiO2 (0.5 wt.%, nanoparticle) and SDBS (0.02 wt.%, surfactant) increases the COP of the system by 28%.	[113]
	Simulation and experiment	falling-film generation	NH3/H2O	The addition of ZnFe2O4 (0.1 wt.%, nanoparticle) and SDBS (0.05 wt.%, surfactant) increases the generation rates by 60%.	[124]
ultrasonic oscillation	Simulation	falling-film absorption	H2O/LiBr	Ultrasonic atomization increases the absorption rate by 15.1% and elevates the ammonia mass fraction of a strong solution by 1.2%.	[104]
	Experiment	bubble absorption	NH3/H2O+LiBr	Ultrasonic oscillations and nanoparticles have a synergistic optimization effect; under operational conditions (ultrasonic frequency of 68 kHz, TiO2 of 0.1 wt.%), the absorption rate increased by 26%.	[109]
	Experiment	generation process	H2O/LiBr	Ultrasonic waves increase the generation rate by 20~60% under a Tg of 65~80 °C; the strengthening effect is more significant as the Tg decreases.	[125]
	Experiment	generation process	H2O/LiBr	Ultrasonic waves increase the cooling capacity and COP by 19.6% and 13.8%.	[126]
	Simulation	generation process	H2O/LiBr	Ultrasonic waves increase the cooling capacity and COP by 33.2% and 31.3%.	[127]
	Experiment	generation process	H2O/LiBr	The generation rate was increased by 10.26% for dual ultrasonic vibrators and 5.69% for single ultrasonic vibrators.	[128]
	Simulation	falling-film absorption	NH3/H2O	Ultrasonic atomizers can significantly improve the absorption rate and system energy efficiency under a wide range of operating conditions.	[105]
surface treatment	Experiment	falling-film absorption	H2O/LiBr	Three tube surfaces (plain tube, floral finned tube, and floral tube) are tested; the average heat and mass transfer coefficients of the floral finned tube are the largest.	[101]
	Simulation	falling-film absorption	H2O/LiBr	Incorporating rhombic mesh into the falling-film absorber improves surface coverage and promotes liquid film fluctuation and mixing.	[103]
	Experiment	falling-film generation	H2O/LiBr	The heat transfer coefficients of seven surface-treated tubes are tested; the heat transfer coefficient of the NF surface-modified tube increases by 60%.	[123]
	Simulation and experiment	generation process	H2O/LiBr	A generator with finned tubes and ultrasonic vibration at 21 kHz increases the heat transfer coefficient by 17.85%.	[129]

summarizes the different heat and mass transfer enhancement techniques. Additives include nanoparticles, surfactants, and alcohol-based additives. Nanoparticles, in particular, have attracted much attention due to their wide applicability and high enhancement efficiency. The key parameters affecting their properties are the mass fraction and dispersion uniformity of nanoparticles in solution. Ultrasonic oscillation enhances the heat and mass transfer rate by perturbing the solution through mechanical oscillation and cavitation. Meanwhile, when interacting with nanoparticles, it can improve the dispersion homogeneity of nanoparticles, thus generating a synergistic enhancement effect. Nanoparticles and ultrasonic oscillations, as an efficient synergistic approach for enhancing heat and mass transfer, show promising development potential. Surface structure treatment could increase the heat transfer area, reduce the thickness of the boundary layer, and improve the microscopic flow by changing the shape and structure of the heat exchanger surface. Common surface treatments include adding corrugations and fins to the tube surface.

3. Energy-Saving and Carbon Emission Reduction Potential of Absorption Refrigeration Technology in Natural Gas Liquefaction

In 1914, Godfrey Lowell Cabot received the first patent for natural gas (NG) liquefaction, storage, and transport [135]. The first large-scale NG liquefaction plant was built in Algeria in 1964 [136]. Since then, numerous researchers have conducted studies on developing and optimizing novel NG liquefaction. Over more than half of a century, a variety of effective NG liquefaction processes have been developed, which can be mainly categorized into expansion refrigeration liquefaction process and mixed refrigerant liquefaction process according to the refrigeration units (expander and Joule–Thompson valve) and refrigerant components [137].

Absorption refrigeration technology (ART) driven by low-grade heat sources, such as industrial waste heat, solar energy, and geothermal energy, could provide refrigeration temperatures of −60 °C and above. Its integration into the precooling stage of NG liquefaction processes significantly lowers energy consumption and carbon dioxide emissions. This section analyzes the energy-saving and carbon emission reduction potential of ART in various NG liquefaction processes. There are relatively few studies on ART integrated into the expansion refrigeration liquefaction process, so this section mainly focuses on the mixed refrigerant liquefaction process.

3.1. Carbon Emission Calculation Method

Studies about the potential of ART application in the NG and H₂ liquefaction processes mainly focused on the perspective of energy consumption reduction. The data about energy utilization efficiency, such as specific power consumption, energy efficiency, and exergy efficiency, are generally available from existing literature. However, few studies have analyzed and published data about CO₂ emission in the liquefaction processes integrated with ART. Thus, this paper establishes a methodology for analyzing the CO₂ emission reduction potential of ART in various liquefaction processes [6]. The CO₂ emission of the liquefaction process consists of two parts: the CO₂ emission from refrigerant leakage and the electricity production process. Refrigerant leakage generally occurs when the compressor operating conditions deviate from the rated operating conditions, and it can be calculated by Equations (1) and (2) [6].

$$GWPI={\displaystyle \sum \left(F_{ref,i}\cdot GWP{I}_{ref,i}\right)}$$

(1)

$$C{E}_{\mathrm{LR}}=\frac{m_{\mathrm{MR}}\times LR\times GWPI}{m_{\mathrm{LNG}}}$$

(2)

where GWPI is the global warming potential index of refrigerant; F_ref,i is the mass fraction of refrigerant (i) used in liquefaction process, %; GWPI_ref,i is the global warming potential index of refrigerant (i); CE_LR is the carbon dioxide emission of mixed refrigerant leakage, kg_CO2/kg_LNG or kg_CO2/kg_LH2; m_MR is the mass flow rate of mixed refrigerant, kg/h; LR is the leakage rate of mixed refrigerant in the compressor, which is set to 0.5% [6]; m_LNG is the production rate of LNG or LH2, kg/h.

The electricity demand consumed by the liquefaction process is assumed to be fully supplied by a power plant, with the energy consumed from coal, and the CO₂ emission of the electricity production process can be calculated by Equations (3) and (4) [6].

$$m_{\mathrm{coal}}=\frac{SPC\cdot m_{\mathrm{LNG}}}{{\eta }_{\mathrm{plant}}\times LN{V}_{\mathrm{coal}}}$$

(3)

$$C{E}_{\mathrm{ele}}=\frac{m_{\mathrm{coal}}\cdot \alpha }{m_{\mathrm{LNG}}}$$

(4)

where m_coal is the consumption rate of coal in plant, kg/h; SPC is the specific power consumption, kWh/kg; η_plant is the efficiency of the power plant, which is set to 30% [6]; LHV_coal is the lower heating value of coal, equal to 8.06 kWh/kg_coal; CE_ele is the carbon dioxide emission from the electricity production process, kg_CO2/kg_LNG or kg_CO2/kg_LH2; α is the mass of CO₂ produced per unit mass of coal in power plant, equal to 2.42 kg_CO2/kg_coal.

The total CO₂ emission per unit mass of LNG or LH2 produced in the liquefaction process is calculated by Equation (5) [6].

$$C{E}_{\mathrm{tot}}=C{E}_{\mathrm{LR}}+C{E}_{\mathrm{ele}}$$

(5)

where CE_tot is the total carbon dioxide emission produced in the liquefaction process, kg_CO2/kg_LNG or kg_CO2/kg_LH2.

Based on the above equations, the CO₂ emission before and after the integration of ART into the NG and H₂ liquefaction processes can be determined. The CO₂ emission reduction data presented in this paper are calculated using this method.

3.2. Mixed Refrigerant Liquefaction Process

In the mixed refrigerant liquefaction process, the refrigerant is a mixture of nitrogen and hydrocarbons such as methane, ethane, ethylene, and propane, with the component percentages designed based on the composition and pressure of the feedstock NG [137]. The mixed refrigerant liquefaction process involves sequentially compressing and cooling the refrigerants at varying temperatures to facilitate the condensation and separation of the heavy and light refrigerant components. Subsequently, the different refrigerants flow through throttle valves into cryogenic heat exchangers, progressively cooling the feed gas in multiple stages. Through multiple stages of cooling and condensation, the NG is ultimately throttled to produce LNG.

Depending on the type of mixed refrigerants and the stages of refrigeration cycles, the mixed refrigerant liquefaction processes are mainly categorized into four types: the single mixed refrigerant (SMR) process, the propane precooling mixed refrigerant (C3MR) process, the dual mixed refrigerant (DMR) process, and the cascade process. This section reviews the principles of processes and analyzes the energy-saving and carbon emission reduction potential of ART in these processes.

3.2.1. Single Mixed Refrigerant Process

The SMR process was designed by Air Products and Chemicals Inc. (APCI), and it is mainly composed of low-temperature multi-flow heat exchangers, throttle valves, compressors, and coolers. The flow of the SMR process in the basic configuration is shown in Figure 8. NG enters the LNG heat exchanger at a high pressure and ambient temperature, and it is cooled and condensed by the mixed refrigerant (MR). The NG leaves the heat exchanger as a super-cooled liquid and is then flashed through a throttle valve to a pressure slightly above atmospheric to obtain LNG. In the MR circuit, MR reaches the high-pressure state through a multiple-stage compression process, with each compressor followed by an inter-stage cooler. After compression, the high-pressure MR flows through a throttle valve to decrease temperature and pressure, absorbs heat from NG, vaporizes completely, and then flows back to the compressor to begin the process anew. Black & Veatch Pritchard in Manila, Philippine, and Korea Gas Corporation in Seongnam, Korea, have optimized the basic configuration by proposing the poly-refrigerant integrated cycle operation (PRICO) process and the Korea single mixed refrigerant (KSMR) process, respectively [137].

To improve the energy utilization efficiency of the SMR process, researchers integrated the absorption refrigeration cycle (ARC) into the SMR as a precooling cycle for NG and MR liquefaction cycles. Ghorbani et al. [138] developed a hybrid system combining an SMR process, an ammonia/water ARC, and a multi-effect desalination (MED) process and conducted energy and exergy analyses. The study evaluates two alternative driven heat sources for the ARC: one powered by an NG power plant and the other by the dish collector of a solar steam power plant. The results show that the hybrid system operating in the second option has higher energy and exergy efficiencies, which are 87.31% and 91.12%, respectively, and increases LNG production by 4.7%. After coupling with the ARC, the specific power consumption (SPC) was reduced by 31.9% to 0.19kWh/kg. The economic analysis showed that the cost of the LNG product for Option 1 and Option 2 was 0.2580 and 0.1784 USD/kg, respectively. Meanwhile, the carbon emission calculations showed that carbon emission (CE) was reduced by 19.5%, from 0.434 to 0.349 kg_CO2/kg_LNG.

Ghorbani et al. [139] proposed an innovative process comprising an ARC precooling SMR process, an NG oxygen-enriched combustion, and a MED process. The ARC is driven by exhaust heat from oxygen-enriched combustion. The results show that the integrated process produces 593.3 t/h of LNG and 74.58 t/h of fresh water. The ARC with the precooling temperature of −26.55 °C reduces the SPC and CE by 35.84% and 22.09%, respectively. Exploring solar energy as a driven energy, Ghorbani et al. [140] assessed a novel hybrid system incorporating an organic Rankine cycle (ORC) to reduce power requirements and improve solar energy efficiency. The ammonia/water ARC in this process is powered by solar energy through parabolic trough collectors. The exergy efficiency of the novel hybrid system is 88.97%, and it produces 14.5 t/h of LNG, 1.693 t/h of desalinated water, and 2.611 t/h of liquid carbon dioxide. Economic analysis indicates the payback period is 6 years, with LNG production costs of 0.242 USD/kg.

Mehrpooya et al. [6] introduced a novel NG liquefaction process that integrated a NH₃/H₂O diffusion absorption refrigeration (DAR) cycle into the SMR process. The driven thermal energy of the DAR system was supplied by a solar parabolic trough collector (PTC), with an additional heater to compensate for thermal energy in periods of insufficient sunlight. The DAR cycle employed hydrogen as an inert gas to promote refrigerant evaporation. The study investigated the impact of hydrogen-to-ammonia ratio, ammonia solution concentration, and DAR operating pressure on system performance. The results showed that the COP of the DAR system was 0.157, which provided a precooling temperature of −29.32 °C. After the combination of DAR, the energy and exergy efficiencies of the hybrid system were 90% and 38%, respectively, and the SPC of LNG was 0.225 kWh/kg. Compared with the vapor compression propane precooling liquefaction process, the application of the DAR cycle reduced the electric power consumption by 19.36% and the CO₂ emission by 19.4%.

To achieve a lower precooling temperature than the single-effect ARC, researchers propose an absorption–compression cascade cycle consisting of an ammonia/water ARC with a CO₂ vapor compression refrigeration cycle. To eliminate the power demand of the cascade cycle, Mehrpooya et al. [141] further incorporated an ORC into the cascade cycle. The steam waste heat of 350 °C drove the ORC and ARC sequentially to achieve the energy gradient utilization. The results showed that the absorption–compression cascade cycle integrated with ORC could operate independently (without an external power supply) and provide a precooling load of −54.62 °C for NG liquefaction. The SPC of the novel integrated process was 0.189 kWh/kg, with energy and exergy efficiencies of 32.5% and 91.68%, respectively. The liquefaction process combined with the ARC reduced CE by 35.86%. Economic analysis indicates the operating cost is 0.1959 USD/kg. Ghorbani et al. [142] further coupled solar collectors to the integrated process and achieved LNG, natural gas liquid (NGL), and electricity tri-generation. The integrated system comprised an SMR process, an absorption–compression cascade refrigeration cycle, an ORC, and a solar parabolic trough collector. Solar energy drove the ARC and ORC to generate the precooling and electrical loads, respectively. In the paper, a multi-flow heat exchanger network was optimized using the pinch method to improve the overall thermodynamic efficiency. The results showed that the integrated system could produce NGL at 54.12 kg/s, LNG at 66.52 kg/s, and a net electrical power of 278.5 MW. The SPC of LNG was 0.3771 kWh/kg, and energy and exergy efficiencies were 78.38% and 84.47%, respectively. Ebrahimi et al. [143] considered that biomass, as a clean and renewable energy source, released a large amount of thermal energy in the gasification process that could be used to drive the ARC and the ORC in the above system. Therefore, the combination of biomass energy into the NG liquefaction process was proposed. The result showed that employing the absorption–compression precooling cycle driven by biomass reduced the power consumption by 4.195%, with an SPC of 0.7673 kWh/kg and an energy efficiency of 54.29%.

3.2.2. Propane Precooling Mixed Refrigerant Process

To improve the energy utilization efficiency of the SMR process, Air Products And Chemicals Inc. (APCI) in Pennsylvania, USA, developed the propane precooling mixed refrigerant liquefaction (C3MR) process. Based on the SMR process, the C3MR process introduces a propane precooling vapor-compression cycle to precool the NG and partially liquefy the mixed refrigerant (MR) flow stream. The liquefaction and sub-cooling of NG are accomplished in a mixed refrigerant cycle, as shown in Figure 9.

Ansarinasab et al. [144] proposed a novel process by replacing the propane precooling cycle with a single-effect ammonia/water ARC. The simulation revealed that the ARC, driven by gas turbine exhaust heat, could achieve a precooling temperature of −30 °C with a COP of 0.49. Compared to the C3MR process, the novel process resulted in a 20.38% reduction in power consumption and the SPC of LNG of 0.21 kWh/kg. Carbon emission calculations showed that the combination of an ARC reduced the CE of the C3MR process by 22.9%. Kalinowski et al. [145] compared single- and double-stage ammonia/water ARC as alternatives to the propane precooling cycle in a C3MR process. The result showed that the single-stage ARC provides the precooling temperature of 18 °C in COP of 0.57 and −42 °C in COP of 0.17, and the overall COP was 0.41. The two-stage cycle’s overall COP was 0.47 under the same operating condition, which was higher than that of the single-stage cycle. However, the two-stage ARC was insufficient for the second stage precooling (−42 °C) when the first stage precooling (18 °C) was ideal. Thus, the single-stage ARC presented a preferable performance alternative. Theoretical analyses demonstrated that by harnessing 5.2 MW waste heat discharged by a 9 MW gas turbine power plant, the single-stage ARC substituted the propane precooling cycle, resulting in a 1.9 MW reduction in electricity consumption. Ghorbani et al. [146] integrated a single-stage ammonia/water ARC into a co-production system consisting of NGL, LNG, and a nitrogen recovery unit (NRU). The results showed that the ARC applied to the initial C3MR process reduced electricity consumption by 18.4% and carbon emission by 29.2%, and the SPC was reduced from 0.359 to 0.257 kWh/kg. Considering that the precooling process was driven by low-grade thermal energy instead of electricity, the overall energy efficiency was reduced by 5.6%. The energy and exergy efficiencies of the novel process were 87.0% and 58.1%, respectively.

The ARC using a lithium bromide solution working pair has a higher COP than that of the ammonia/water ARC, which has been proposed to cool the condenser of the propane precooling cycle and the inter-stage cooler of the mixed refrigerant liquefaction cycle, considering that its refrigeration temperature is above 0 °C.

Rodgers et al. [147] found that when seawater was used as cooling water for the C3MR cycle, the energy efficiency of the propane cycle decreased significantly with increasing seawater temperature. Thus, they proposed two enhancement schemes of the propane cycle based on the H₂O/LiBr ARC, including propane sub-cooling after the condensation or precooling the cooling water to reduce the condensing pressure of the propane cycle. Three absorption refrigeration configurations were assessed, including single-effect, double-effect, and single-effect–double-effect cascade refrigeration cycles. The results showed that using a single-effect–double-effect cascade refrigeration cycle to sub-cool propane exhibited the best performance. When the seawater temperature increased to 35 °C, this scheme improved the cooling capacity and COP of the propane precooling cycle by 23% and 13%, respectively, and it achieved a good energy match with the exhaust thermal energy of the gas turbine. Mortazavi et al. [148] explored the energy-saving and carbon emission reduction potential of a double-effect lithium bromide/water ARC in a C3MR process powered by gas turbine exhaust heat. Examining eight operational scenarios of the ARC, the study found that optimal energy efficiency for the overall LNG process occurred when the ARC supplied precooling temperatures of 9 °C and 22 °C. This scenario provided sufficient cooling loads for the mixed refrigerant cycle’s inter-stage cooler and the propane cycle’s condenser, recovering 97% waste heat of the gas turbine, reducing 21.32% of the electrical power consumption and 11.86% of CE. The SPC and exergy efficiency of the novel process were 0.244 kWh/kg and 89.6%, respectively.

3.2.3. Dual Mixed Refrigerant Process

To overcome the inherent limitation of the large size of the pure propane refrigerant compressor in the C3MR, the dual mixed refrigerant liquefaction (DMR) process was developed; the schematic is illustrated in Figure 10. Similar to the C3MR process, the DMR process features two vapor compression refrigeration cycles: precooling and liquefaction. The difference is that the precooling cycle uses the mixed refrigerant cycle, typically consisting of ethane and propane.

To diminish the energy consumption and production costs of the DMR process, Ghorbani et al. [146] developed an integrated system comprising NGL, LNG, and a nitrogen recovery unit (NRU). The system substituted the vapor compression precooling cycle with an ammonia/water ARC. The result demonstrated that incorporating the ARC into the traditional DMR process reduced power consumption by 12.6%, with the SPC decreasing from 0.351 kWh/kg to 0.257 kWh/kg, and CE reducing from 0.659 kg_CO2/kg_LNG to 0.462 kg_CO2/kg_LNG. Furthermore, this study conducted a thorough economic analysis of the novel system, indicating a 25% reduction in capital costs compared to the DMR process and a 27% improvement in net annual benefits.

Based on the DMR process, Mehrpooya et al. [10] conceptualized and developed an innovative integrated system by combining NG liquefaction with power production. The system comprised five cycles: an MR liquefaction cycle, an ammonia/water ARC, a molten carbonate fuel cell (MCFC), a supercritical CO₂ Brayton cycle, and a gas turbine (GT). Notably, the ARC, powered by the waste heat of MCFC, provided a 28.3 MW cooling load of −33 °C with a COP of 0.45, enhancing the total cooling capacity of the system by 10%. The results indicated that the integrated system produces 6300 kmol/h LNG and 146.55 MW of electricity, achieving an exergy efficiency of 95% and an energy efficiency of 85%. To further elevate the energy utilization efficiency of this integrated system and extract helium from the NG feedstock, Mehrpooya et al. [149] introduced a helium recovery unit into the integrated process. Crude helium and LNG were produced after the feedstock NG passed sequentially through the ammonia/water ARC, helium recovery, and LNG production units. A portion of the methane-rich NG entered the MCFC cycle, generating electricity and supplying the driven thermal energy for ARC. The ARC provided a cooling capacity of 219.7 MW at −30 °C. The novel process’s energy efficiency was 94.91%, the helium extraction rate was 91.42%, and the SPC of LNG was 0.2086 kWh/kg, while the process generated 4793 MW of net electricity.

Waste heat at different temperatures is discharged in the liquefaction process. To improve the energy efficiency of the ARC under different driven heat source temperatures, Lu et al. [7] designed a novel NH₃/H₂O absorption refrigeration prototype by optimizing the generator structure, which could recover different-temperature waste heat through the reboiler and the distillation section. Experimental results demonstrated that the COP of the novel system ranged between 0.29 and 0.35 at the evaporating temperature of −10 °C, with the waste heat recovery rate improving by 150%. The performance of the novel ARC was analyzed in a 3000 m³/d LNG plant to substitute the conventional compression precooling cycle. The results showed that the optimized process reduced power consumption by 30% and CE by 31.18%, with SPC and energy efficiency being 0.391kWh/kg and 39.8%, respectively.

The refrigeration temperatures of the ammonia/water ARC are generally above −42 °C due to the limitation of the standard boiling point of NH₃. To further lower the precooling temperature of the ARC, He et al. [77] evaluated the precooling performance of an auto-cascade absorption refrigeration (ACAR) cycle for the NG liquefaction process. The ACAR cycle employed R23 and R134a as refrigerants and DMF as an absorbent. Theoretical analyses indicate that the ACAR cycle offers broader heat source adaptability and achieves lower refrigeration temperature ranges than the ammonia–water ARC. Experimental tests demonstrated that the ACAR system could attain refrigeration temperatures of −52.9 °C with a COP of 0.011 at a generation temperature of 122.5 °C, and −56.7 °C with a COP of 0.013 at a generation temperature of 155.2 °C.

3.2.4. Cascade Processes

According to the refrigerants used, the cascade NG liquefaction processes are generally divided into two categories: the single-component refrigerant cascade liquefaction process and the mixed fluid cascade liquefaction process.

The single-component refrigerant cascade liquefaction process typically consists of three separate vapor compression refrigeration cycles using pure methane, ethylene, and propane as refrigerants, respectively. The schematic is illustrated in Figure 11. NG feedstock sequentially flows through propane, ethylene, and methane refrigeration cycles for precooling, liquefaction, and sub-cooling at temperatures around −30 °C, −90 °C, and −150 °C, respectively. Then, the LNG product is obtained after throttling and gas–liquid separation. The process has good operational stability due to the low interaction between refrigeration systems. However, its widespread application is hindered by its lengthy process flow, requirement for multiple refrigeration units, high-purity refrigerants, and unsuitability for NG liquefaction with a high nitrogen component.

Mouneer et al. [150] integrated an ARC into a single-component refrigerant cascade liquefaction cycle to substitute the air-cooled condenser in the propane precooling refrigeration cycle with a single-effect H₂O/LiBr ARC. The results showed that the energy efficiency of the novel process was slightly lower than that of the conventional vapor compression cascade cycle due to the utilization of low-grade energy, with the energy efficiency of the nine-stage compressor cascade cycle decreasing from 0.741 to 0.704, and that of the eight-stage compressor cascade decreasing from 0.838 to 0.709. The SPC of the novel process was about 0.1875~0.225 kWh/kg.

To improve the temperature matching of the single-component refrigerant cascade process, Statoil and Linde designed the mixed fluid cascade (MFC) process [151], as depicted in Figure 12. Compared to the single-component refrigerant cascade process, it exhibited higher liquefaction efficiency due to the use of mixed refrigerants. The MR commonly consisted of nitrogen and hydrocarbons such as methane, ethane, and propane. The percentage of refrigerant components varied in each refrigeration cycle. The main disadvantage of this process was the need to adjust the composition of the MR when the feed gas composition changed.

Mehrpooya et al. [152] integrated an ammonia/water ARC into the MFC process, replacing the compression precooling cycle. The results showed that the ARC with a COP of 0.48 led to a 30% reduction in the SPC compared to the traditional MFC process, representing a decrease from 0.245 to 0.172 kWh/kg. However, the ARC also reduced the energy efficiency to 90.73%, so the heat rejection of the liquefaction process increased by about 117%. Carbon emission calculations showed that the ARC reduces the CE by 29%, from 0.414 to 0.294 kg_CO2/kg_LNG. To further refine the novel process, Mehrpooya et al. [153] designed an integrated mathematical and thermodynamic optimization method based on Mixed Integer Linear Programming (MILP) and a genetic algorithm (GA). A detailed economic analysis was also conducted. The results showed that the optimization of the liquefaction process using this methodology resulted in a reduction in the SPC, capital cost, and main cost of the product by 38.28%, 31.9%, and 15.31%, respectively. The SPC and exergy efficiency of the optimized liquefaction process were 0.18 kWh/kg and 58.11%, respectively. To simultaneously extract helium and produce LNG from NG feedstock gas, Zaitsev et al. [154] further combined a helium extraction unit with LNG production. The results showed that SPC for MFC and novel process were 0.265 kWh/kg and 0.184 kWh/kg, respectively. The authors performed an exergoeconomic analysis of the novel process, which showed that the heat exchanger (58.19%), distillation column (18.94%), and compressor (14.87%) exhibited the primary rate of exergy destruction. Exergy and energy efficiencies were 87.16% and 88.96%, respectively. In terms of operating cost, although the initial investment cost of the novel process increased by 31%, the production cost decreased by 6.28%, which is economically sound for a long operational period.

Ghorbani et al. [155] replaced the propane precooling cycle of the MFC process with an ammonia/water ARC for the co-production of LNG and NGL. A thermodynamic analysis of the novel process was conducted, which showed that the overall energy efficiency of the novel process was reduced by 12.72% because the absorption precooling cycle was driven by low-grade thermal energy. However, the power consumption of the optimized process declined by 35.66%, reducing the SPC from 0.423 kWh/kg to 0.272 kWh/kg. The exergy efficiency of the optimized process was 48.93%. An economic sensitivity analysis was conducted to analyze the impact of product and utility prices on the annualized cost of the system, which showed that the annualized cost of the modified process was reduced by 4.32% and the annualized net profit increased by 6.2%.

Nitrogen in NG diminishes its calorific value, so separating nitrogen from NG could further increase the energy density of NG. Integrating NGL, LNG, and a nitrogen removal unit (NRU) is an efficient treatment process. Ghorbani et al. [156] incorporated an NRU into the MFC process, substituting the compression precooling cycle with an ammonia/water ARC. The novel process was analyzed in terms of exergy cost, exergy economic factor, exergy, and exergy efficiency. The results indicated that the ammonia/water ARC offered a higher energy economic factor and a lower unit energy cost.

3.3. Expansion Refrigeration Liquefaction Process

The expansion refrigeration liquefaction process utilizes nitrogen or methane as the refrigerant typically, as depicted in Figure 13. The expansion refrigeration cycle consists of a multi-stage compression unit, an inter-stage cooler (air or water), expanders, and heat exchangers. The high-pressure refrigerant gas undergoes isentropic expansion via the expander, significantly reducing the temperature and pressure. Then, the low-temperature refrigerant cools and liquefies NG feedstock. The refrigerant remains in the gaseous state throughout the cycle, and the heat transfer is in sensible heat. Thus, it is also known as the “reverse Brayton cycle”. The liquefaction process exhibits significant adaptability to NG feedstock composition variations, along with superior flexibility and simplicity. However, its drawbacks include the need for a large temperature difference and heat transfer area between NG and refrigerant, leading to lower energy efficiency—typically around 40% lower than that of the mixed refrigeration process.

Zhang et al. [157] integrated ammonia/water single-effect ARC into an expansion refrigeration liquefaction process, capable of producing 5~500 tons of LNG per day, to precool the feedstock gas and the expansion refrigeration cycle. They analyzed four kinds of processes: single nitrogen expansion (SN), single methane expansion (SM), absorption precooling single nitrogen expansion (SNA), and absorption precooling single methane expansion (SMA). The ammonia/water ARC was driven by industrial waste heat from turbine exhaust gas (temperature of 180 °C and pressure of 550 kPa). Thermodynamic and economic analyses revealed that the ARC significantly reduces SPC and production costs by 26~35% and 13~17%, respectively. Among these processes, the SMA process demonstrated the best performance, compared to the other processes, achieving a 28~48% decrease in SPC and a 13–43% reduction in production cost. In terms of CE, compared to the SM process, the SMA process reduces CE by 32.2%.

During transportation, the evaporation of LNG generates boil-off gas (BOG) that necessitates re-liquefaction. Traditional BOG re-liquefaction processes commonly utilize the parallel nitrogen expansion process (PNEP). Yin et al. [158] integrated an ammonia/water single-effect ARC powered by solar energy into the PNEP. The simulation was carried out by Aspen HYSYS software, and the results showed that the SPC of the novel process was 0.7878 kWh/kg, which was 11.0% lower than that of the conventional PNEP of 0.8856 kWh/kg. Furthermore, the exergy and energy efficiencies of the novel process were 36.57% and 26.36%, respectively, marking improvements of 12.6% and 11.8%. The CE was reduced by 9.9%. Kim et al. [159] proposed the application of an ammonia/water ARC in BOG re-liquefaction of LNG carriers. The ARC was powered by the exhaust heat of the carriers’ engines. The result revealed that the ARC with a cooling load of −30 °C reduced the gasification rate of LNG in carriers through waste heat recovery.

Table 4. The performance overview of various NG liquefaction processes integrated with the absorption refrigeration cycle.

Initial Process	Coupled Absorption Refrigeration Cycle				Optimized NG Liquefaction Process			Ref.
	Configuration	Working Pair	Cooling Temp./(°C)	Heat source	SPC /(kWh/kg)	Energy Efficiency/(%)	Exergy Efficiency/(%)
SMR	Single effect	Ammonia/H2O	−26.55	Solar energy	0.19	87.31	91.12	[138]
	Single effect	Ammonia/H2O	−26.55	Industrial waste heat	0.18	---	62.33	[139]
	Single effect	Ammonia/H2O	−26.55	Solar energy	0.18	---	88.97	[140]
	Diffusion–absorption	Ammonia/H2O	−29.32	Solar energy	0.23	90.00	38.00	[6]
	Absorption–compression cascade	Ammonia/H2O+CO2	−54.62	Industrial waste heat	0.19	32.50	91.68	[141]
C3MR	Single effect	Ammonia/H2O	−30	Industrial waste heat	0.21	78.84	---	[144]
	Double effect	H2O/LiBr	9/22	Industrial waste heat	0.24	89.60	---	[148]
DMR	Single effect	Ammonia/H2O	−10	Industrial waste heat	0.39	39.80	---	[7]
	Single effect	Ammonia/H2O	−29.5	---	0.26	87.00	58.10	[146]
MFC	Single effect	Ammonia/H2O	−29.5	Industrial waste heat	0.17	90.73	---	[152]
	Single effect	Ammonia/H2O	−29.5	---	0.18	---	58.11	[153]
	Single effect	Ammonia/H2O	−29.5	---	0.18	87.16	88.96	[154]
	Single effect	Ammonia/H2O	−29.5	Industrial waste heat	0.27	---	48.93	[155]
Methane expansion	Single effect	Ammonia/H2O	−29.6	Industrial waste heat	0.32	---	---	[157]
Nitrogen expansion	Single effect	Ammonia/H2O	−29.6	Industrial waste heat	0.37	---	---	[157]
	Single effect	Ammonia/H2O	−29.6	Solar energy	0.79	23.36	36.57	[158]

summarizes the performance indicators of various NG liquefaction processes integrated with the ARC, including SPC, energy efficiency, and exergy efficiency. It can be seen that the ARC, which is driven by solar energy and industrial waste heat generally, mainly adopts the single-effect NH₃/H₂O cycle as the precooling cycle, which could provide a precooling temperature around −30 °C for the NG feed gas and low-temperature liquefaction cycle. Figure 14 shows that the ARC could reduce the SPC of the initial liquefaction process by 9.91~38.28%, and the different SPC reductions mainly depend on the precooling method of the ARC and the type of the initial liquefaction process.

Table 5. Carbon emission reduction benefits of various NG liquefaction processes integrated with the absorption refrigeration cycle.

Initial Process	Coupled Absorption Refrigeration Cycle			CE of Initial NG Liquefaction Process/(kgCO2/kgLNG)			CE of Optimized NG Liquefaction Process/(kgCO2/kgLNG)			Reduction in CEtot/(%)	Ref.
	Configuration	Working pairs	Cooling Temp./(°C)	CELR	CEele	CEtot	CELR	CEele	CEtot
SMR	Single effect	Ammonia/H2O	−26.55	0.167	0.267	0.434	0.159	0.190	0.349	19.5	[138]
	Single effect	Ammonia/H2O	−26.55	0.165	0.267	0.432	0.158	0.179	0.337	22.1	[139]
	Single effect	Ammonia/H2O	−26.55	0.165	0.267	0.432	0.158	0.179	0.337	22.1	[140]
	Diffusion–absorption	Ammonia/H2O	−29.32	0.228	0.267	0.495	0.184	0.225	0.409	17.34	[6]
	Absorption–compression cascade	Ammonia/H2O+ CO2	−54.62	0.250	0.305	0.555	0.167	0.189	0.356	35.86	[141]
C3MR	Single effect	Ammonia/H2O	−30	0.247	0.253	0.500	0.175	0.210	0.386	22.92	[144]
	Single effect	Ammonia/H2O	−29.5	0.235	0.359	0.595	0.163	0.257	0.421	29.24	[146]
	Double effect	H2O/LiBr	9/22	0.247	0.310	0.557	0.247	0.244	0.491	11.86	[148]
DMR	Single effect	Ammonia/H2O	−10	0.308	0.558	0.865	0.205	0.391	0.596	31.18	[7]
	Single effect	Ammonia/H2O	−29.5	0.308	0.351	0.659	0.205	0.257	0.462	29.83	[146]
MFC	Single effect	Ammonia/H2O	−29.5	0.190	0.224	0.414	0.122	0.172	0.294	29.0	[152]
	Single effect	Ammonia/H2O	−29.5	0.215	0.265	0.480	0.142	0.185	0.327	31.9	[154]
	Single effect	Ammonia/H2O	−29.5	0.253	0.369	0.622	0.162	0.272	0.434	30.3	[155]
Methane Expansion	Single effect	Ammonia/H2O	−29.6	0.799	0.410	1.209	0.559	0.260	0.819	32.2	[157]
Nitrogen Expansion	Single effect	Ammonia/H2O	−29.6	0	0.505	0.505	0	0.367	0.367	27.4	[157]
	Single effect	Ammonia/H2O	−29.6	0	0.875	0.875	0	0.788	0.788	9.9	[158]

summarizes the CE of refrigerant leakage and electricity consumption and the impact of the ARC on the various liquefaction processes’ total CE. The ARC reduces the CE of the initial liquefaction process from two aspects: one is a reduction in the CE from the refrigerant leakage of the compression cycle, and the other is a reduction in the CE from the power production process by reducing the SPC. It can be seen that the ARC could reduce the CE of liquefaction processes by 9.9~35.7%, and the difference in carbon reduction benefits depends on the precooling method of the ARC and the GWP of the refrigerants in the processes. Notably, an ARC with a lower precooling temperature has a more noticeable improvement potential in the initial liquefaction process.

4. Energy-Saving and Carbon Emission Reduction Potential of Absorption Refrigeration Technology in Hydrogen Liquefaction

In 1898, James Dewar achieved H₂ liquefaction for the first time (4 mL/min) by throttling 20 MPa H₂ which was cooled by CO₂ and liquid air. Over more than a century, there have been significant advancements in the efficiency and capacity of H₂ liquefaction processes. Presently, large-scale H₂ liquefaction processes are generally categorized into three types: the Linde–Hampson cycle, generating cooling load by throttling; the Claude cycle, featuring H₂ self-expansion refrigeration; and the Joule–Brayton cycle, which includes independent expansion refrigeration cycles [160]. The liquefaction of H₂ spans a large temperature range, from ambient temperature to 20 K. Absorption refrigeration technology could provide a precooling load and modify the liquefaction process temperature matching problem, thus improving energy utilization efficiency. Conventional hydrogen liquefaction processes often rely on energy-intensive compression and cooling techniques, which can result in significant greenhouse gas emissions. Absorption refrigeration technology, on the other hand, offers a more environmentally friendly alternative by minimizing or eliminating the need for mechanical compression. The specific energy-saving and emission reduction potential of absorption refrigeration technology in hydrogen liquefaction will depend on various factors, including the scale of the operation, the heat source availability, and the overall process design. This chapter analyzes the energy-saving and carbon emission reduction potential of ART in H₂ liquefaction.

4.1. Linde–Hampson Cycle

The Linde–Hampson cycle is a thermodynamic process used for gas liquefaction, named after Carl von Linde and William Hampson. This cycle is particularly important in the context of cryogenics and industrial gas production, such as the liquefaction of air or hydrogen. To achieve the throttling cooling effect, the temperature of the H₂ before throttling must be lower than the maximum conversion temperature (204.6 K). Thus, the liquid nitrogen (LN2) precooling Linde–Hampson cycle is typically used for H₂ liquefaction, with the schematic and T-S diagram shown in Figure 15a,b, respectively. In this process, after isothermal compression at ambient temperature, H₂ is cooled isobarically through the liquid nitrogen precooling and low-temperature reflux H₂, followed by throttling expansion to achieve partial liquefaction of the gas. Recognized as the simplest method for H₂ liquefaction, the Linde–Hampson cycle offers the benefits of simplicity and stable operation. However, its lower energy utilization efficiency generally leads to high SPC and production costs [161].

Ratlamwala et al. [162] proposed an integrated process for H₂ liquefaction and power co-generation, which consisted of a binary power plant driven by geothermal energy, a quadruple-effect ARC, and a LN2 precooling Linde–Hampson cycle. The ARC precooled the H₂ feedstock gas. The impacts of geothermal source temperature, ambient temperature, and ammonia solution concentration on the energy and exergy efficiencies of the process were analyzed through thermodynamic analysis. Further, Ratlamwala et al. [163] coupled solar collectors with the above integrated process, enabling solar energy as the second driven heat source for the triple-effect ARC. The impacts of operational parameters on the performance of the integrated system were analyzed based on the first and second laws of thermodynamics. The results showed that the integrated systems’ energy and exergy efficiencies were 3.7~5.9% and 13~21%, respectively. Yilmaz et al. [164] investigated seven approaches to achieve H₂ production and liquefaction using geothermal energy. The basic liquefaction process was the LN2 precooling Linde–Hampson cycle. The results showed that incorporating the geothermal energy-driven ammonia/water ARC as the precooling cycle could markedly reduce energy consumption and production costs of the liquefaction process by minimizing irreversible losses. The production cost of LH2 for different geothermal energy utilization approaches ranged from 0.979 to 2.615 USD/kg, and that coupled with an ARC ranged from 0.979 to 0.981 USD/kg, which was significantly lower than that of other utilization approaches.

In summary, while the Linde–Hampson cycle is a fundamental and historically significant process in gas liquefaction, its efficiency can be limited. Advances in cryogenic engineering continue to develop more efficient methods and cycles to address these challenges, particularly for the liquefaction of gases with extremely low boiling points.

4.2. Claude Cycle

The Claude cycle is a thermodynamic process used for the liquefaction of gases, particularly those with very low boiling points like hydrogen and helium. It was developed by Georges Claude in the early 20th century and is widely used in large-scale cryogenic applications, with the schematic and T-S diagram shown in Figure 16a,c, respectively. Liquid nitrogen precooling is typically integrated into the Claude cycle to enhance hydrogen liquefaction efficiency, as illustrated in Figure 16b. The reflux gas and liquid nitrogen initially cool the high-pressure H₂. Subsequently, the H₂ is partially expanded by an expander and converges with the reflux gas to augment the cooling capacity of the reflux gas. The remaining gas is further cooled by heat exchange and eventually practically liquefied through throttling. Unlike the Linde–Hampson cycle, which relies solely on throttling refrigeration, the Claude cycle recovers the energy of the high-pressure H₂ through expansion, thus improving energy utilization efficiency.

In 1937, the low-pressure liquefaction cycle, known as the Kapitza cycle, was proposed by Kapitza in the former Soviet Union based on the Claude cycle. Compared with the Claude cycle, the Kapitza cycle uses a turbine expander with high adiabatic efficiency and a reversible heat exchanger with high efficiency and eliminates the low-temperature stage heat exchanger, with the process flow diagrams and the temperature–entropy diagrams shown in Figure 17a,b. The Kapitza cycle operates in a relatively low pressure range, and the enthalpy drops for both the isenthalpic throttling and adiabatic expansion processes are small. Hence, the liquefaction ratio is generally less than 6%. A simple process; low specific energy consumption, metal consumption, and initial investment; and simplicity of operation characterize the Kapitza liquefaction process.

For the simple Claude liquefaction process, Majid et al. [165] designed a novel H₂ liquefaction process by coupling two ammonia/water ARCs to the simple Claude cycle. The ARC precooled the H₂ feed gas and the inter-stage cooler within the Claude cycle. The paper comprehensively analyzed the system components through exergoeconomic and exergoenvironmental analyses. The results showed that the SPC, energy efficiency, and exergy efficiency of the novel process were 12.7 kWh/kg, 9.56%, and 31.6%, respectively. Carbon emission calculations show that combining two absorption precooling cycles reduced CE by 19.1%, from 15.713 to 12.711 kg_CO2/kg_LH2.

For the LN2 precooling Claude process, Mehmet et al. [166] implemented a single-effect ammonia/water ARC driven by geothermal energy to precool the H₂ feed gas, thereby reducing the energy consumption of the low-temperature stage heat exchanger. The results showed that the ARC precools the H₂ feed gas to −26.9 °C with a COP of 0.556 and exergy efficiency of 67.0%. The energy and exergy efficiencies of the integrated process were 16.2% and 67.9%, respectively. The application of ART reduced the SPC and CE of the original H₂ liquefaction process by 25.4% and 20.3%, respectively. The study further analyzed the impacts of H₂ precooling temperature, geothermal temperature, and compressor pressure on the performance of the integrated process. The result showed that higher geothermal temperatures enhanced the liquefaction rate of H₂ and reduced power consumption. To evaluate the effectiveness of utilizing geothermal energy via the ARC, Mehmet et al. [167] designed and integrated five geothermal energy utilization approaches to generate and liquefy H₂. The energy efficiencies in different approaches were analyzed and compared. The impact of geothermal temperature on the thermodynamic performances of the various approaches was investigated. The results showed that utilizing geothermal energy by the ARC was one of the optimal approaches for reducing the SPC and production cost of H₂ liquefaction. Yilmaz et al. [168] incorporated a single-effect ammonia/water ARC driven by geothermal energy into the LN2 precooling Claude process, precooling H₂ feed gas to −30 °C. The COPs of the ARC and Claude liquefaction cycles were 0.556 and 0.2017, respectively. And the energy and exergy efficiencies of the integrated process were 34.6% and 69.4%, respectively. The thermo-economics of the process was optimized through a genetic algorithm, the SPC was reduced from 11.88 to 11.52 kWh/kg, the liquefaction rate increased from 6.028 to 8.711 kg/s, and the exergoeconomic cost decreased from 1.555 to 1.349 USD/kg_LH2. The combination of the ARC reduced the CE of the liquefaction process by 24.3%. To enhance the energy utilization efficiency of geothermal energy in H₂ liquefaction, Yilmaz [169] further integrated a geothermal energy isobutane binary power generation cycle into the above process, which provided the electrical load for the H₂ liquefaction process. Optimization through a genetic algorithm showed that the energy and exergy efficiencies of the integrated process were further improved to 40.8% and 76.1%, respectively, and the unit exergy cost of H₂ liquefaction was reduced to 1.114 USD/kg. The SPC was 10.06 kWh/kg_LH2.

The mixed refrigerant precooling cycle was proposed to modify the temperature matching in the thermostatic precooling process of liquid nitrogen. In 2023, Yan et al. [170] introduced a H₂ and NG integrated liquefaction process that utilized the mixed refrigerant precooling of the Claude cycle. Meanwhile, the integrated process was coupled with a single-effect ammonia/water ARC driven by solar energy. The ARC cooled the inter-stage cooler within the Claude cycle. The integrated process was optimized using particle swarm optimization algorithms. The results showed that the SPC of the integrated process was 5.2201 kWh/kg, marking a 10.67% reduction from the base case. The exergy efficiency was 62.21%, which was 6.63% higher than the base case.

The Claude cycle is a highly efficient process for the large-scale liquefaction of gases with low boiling points. Its use of a turbo expander and counterflow heat exchangers contributes to its effectiveness. However, its complexity and higher costs make it more suitable for industrial-scale applications rather than small-scale operations.

4.3. Joule–Brayton Cycle

The Joule–Brayton cycle is a thermodynamic cycle that is commonly used in gas turbine engines, including jet engines and power plants. It is an idealized cycle that is composed of four main processes: compression, heat addition, expansion, and heat rejection. The Joule–Brayton liquefaction process used in H₂ liquefaction comprises the H₂ flow stream loop and the reverse Brayton expansion refrigeration cycle. The reverse Brayton cycle, which adopts helium as the working fluid, is also known as the “helium expansion refrigeration cycle” and is typically coupled with the LN2 precooling cycle, as shown in Figure 18a. In recent years, the mixed refrigerant (MR) precooling Joule–Brayton cycle constructed by KRASAE-IN et al. [171] has garnered wide attention for its low SPC. The configuration of a standard MR precooling Joule–Brayton cycle (J-B cycle) is illustrated in Figure 18b, where the H₂ feed gas is first cooled to the liquid nitrogen temperature region via the MR precooling cycle before being further cooled and liquefied through a multi-stage reverse Brayton cycle (3~4 stages). The MR precooling cycle utilizes the mixture of He, Ne, N₂, CF₄, and light hydrocarbons, significantly reducing the temperature differential between hot and cold streams compared to the traditional LN2 precooling cycle, thus increasing energy utilization efficiency.

Majid et al. [172] integrated a single-effect ammonia/water ARC into the MR precooling Joule–Brayton process, which provided the cooling load of −25 °C to the inter-stage cooler of the MR cycle and the multi-stage J-B cycle. The results showed that the SPC of LH2 was 6.47 kWh/kg, and the energy and exergy efficiencies were 20.34% and 45.5%, respectively. The impact of operating parameters on the process performance was investigated, including the circulating solution mass flow rate of ARC, the temperature distribution in the heat exchanger, and the adiabatic efficiency of the compressor and expander. In addition, a detailed economic analysis was conducted. Majid et al. [173] combined solar collectors and an ORC into the process to further remove the external power requirement of the above process, where the solar collectors provided the driven energy for the ORC and the ARC. The ORC provided the electrical load for the ARC and the MR precooling J-B cycle. The results showed that the ARC decreased the SPC and CE of the liquefaction process by 8.84% and 5.0%, respectively. The energy and exergy efficiencies of the novel integrated process were 20.02% and 73.57%, respectively. In addition, the impacts of variations in the number of solar collectors on the ARC and the integrated process were analyzed. Using the 954 MW Parand gas-fired power plant in Iran as the energy source, Hamed et al. [174] analyzed the feasibility of an MR precooling J-B process combined with an ARC. The single-effect ammonia/water ARC, driven by the 546 °C exhaust gas discharged from the gas turbine, precooled the H₂ feedstock gas to −30 °C and provided the cooling load for the inter-stage cooler of the MR cycle and the J-B cycle. The simulation of the novel process was conducted based on actual production data, and the trial and error method was used to optimize the operational variables and the composition of the MR. The results showed that the Parand gas-fired power plant could provide sufficient electricity and thermal energy for the 2000 t/d H₂ liquefaction plant. The energy efficiency and SPC were 0.271 and 4.54 kWh/kg, respectively. Zhang et al. [175] developed an integrated process for H₂ liquefaction and electricity generation, which consisted of a solar steam power generation system, an ammonia/water absorption cascade cycle, and an MR precooling J-B cycle. The ARC, driven by the waste heat of the solar power generation system, provided a cooling load of −59.41 °C for the inter-stage cooler of the MR cycle and the J-B cycle, which reduced the CE by 18.50%. The results showed that the SPC, energy efficiency, and exergy efficiency were 5.413 kWh/kg_LH2, 14.33%, and 86.99%, respectively. The process could produce 367.2 tons of LH2 per day and 1278.85 and 1353.78 MWh of net electricity during day and night.

Forms of renewable energy, such as geothermal, solar, and biomass energy, are widely available and environmentally friendly. Converting renewable energy into LH2 is an effective approach to achieve high-density energy storage. ART has the unique advantage of utilizing renewable energy due to its good adaptability to low-grade thermal energy over a large temperature range. Mahdieh et al. [176] proposed a process for the production and liquefaction of H₂ based on renewable energy (solar and geothermal energy), which consisted of a photovoltaic power generation system, an electrolytic hydrogen production system, an ammonia/water single-effect ARC, and an MR precooling J-B H₂ liquefaction system. The photovoltaic power generation system provided the electricity for the H₂ production and liquefaction, while the ARC driven by geothermal energy provided a −30 °C precooling load for the H₂ feed gas. Simulations showed that the liquefaction process incorporating an ARC reduced the SPC from 5.07 to 4.97 kWh/kg_LH2, increasing the exergy efficiency from 44 to 57%. Saman et al. [177] developed an integrated process for H₂ production and liquefaction based on geothermal energy, which consisted of an ORC, a proton membrane electrolyzer (PEM), an ammonia/water ARC, and a H₂ liquefaction cycle. The ORC provided the electricity for the PEM and H₂ liquefaction cycle, and the ARC provided the precooling load for the H₂ feed gas. The energy, exergy, and economic analyses showed that the SPC and energy efficiency of the integrated process were 8.81 kWh/kg and 0.49, respectively, and the liquefaction cost was 1.84 USD/kg.

Table 6. Performance overview of various H₂ liquefaction processes integrated with the absorption refrigeration cycle.

Initial Process	Coupled Absorption Refrigeration Cycle				Optimized H2 Liquefaction Process			Ref.
	Configuration	Working Pair	Cooling Temp./°C	Heat Source	SPC /(kWh/kg)	Energy Efficiency/(%)	Exergy Efficiency/(%)
Linde–Hampson	Triple effect	Ammonia/H2O	−13.45	Solar/geothermal energy	---	5.90	21.00	[163]
Simple Claude	Single effect	Ammonia/H2O	−29.5	Solar energy	12.70	9.56	31.6	[165]
Claude with LN2 precooling	Single effect	Ammonia/H2O	−26.9	Geothermal energy	6.78	20.26	16.20	[165]
	Single effect	Ammonia/H2O	−30	Geothermal energy	11.88	34.6	69.44	[168]
	Single effect	Ammonia/H2O	−30	Geothermal energy	13.80	40.81	76.1	[169]
JT with MR precooling	Single effect	Ammonia/H2O	−25	Solar energy	6.47	20.34	45.50	[172]
	Single effect	Ammonia/H2O	−29.5/0	Solar energy	4.02	20.02	73.57	[173]
	Single effect	Ammonia/H2O	−30	Industrial waste heat	4.54	27.10	---	[174]
	Single-effect cascade	Ammonia/H2O	−59.41	Solar energy	5.41	14.33	86.99	[174]
	Single effect	Ammonia/H2O	−30	Geothermal energy	4.97	---	57.00	[176]
	Single effect	Ammonia/H2O	−26.9	Geothermal energy	8.87	49.00	---	[177]

summarizes the performance indicators of various H₂ liquefaction processes integrated with the ARC, including SPC, energy efficiency, and exergy efficiency. The single-effect ammonia/water cycles driven by solar, geothermal, and industrial waste thermal energy are generally adopted and can provide precooling temperatures of around −59 °C and above. From Figure 19, it can be seen that the ARC cycle could reduce the SPC of the initial liquefaction process by 1.97~24.33%, and the different reductions in SPC mainly depend on the precooling temperature, precooling method, and type of the initial liquefaction process.

Table 7. Carbon emission reduction benefits of various H₂ liquefaction processes integrated with the absorption refrigeration cycle.

Initial Process	Coupled Absorption Refrigeration Cycle			CE of Initial H2 Liquefaction Process/(kgCO2/kgLH2)			CE of Optimized H2 Liquefaction Process/(kgCO2/kgLH2)			Reduction in CEtot/%	Ref.
	Configuration	Working Pair	Cooling Temp./(°C)	CELR	CEele	CEtot	CELR	CEele	CEtot
Simple Claude	Single effect	Ammonia/H2O	−29.5	0	15.713	15.713	0	12.711	12.711	19.1	[165]
Claude with LN2 precooling	Single effect	Ammonia/H2O	−26.9	0	8.509	8.509	0	6.786	6.786	20.3	[166]
	Single effect	Ammonia/H2O	−30	0	15.713	15.713	0	11.890	11.890	24.3	[168]
	Single effect	Ammonia/H2O	−30	0	15.713	15.713	0	13.811	13.811	12.1	[169]
JT with MR precooling	Single effect	Ammonia/H2O	−29.5/0	71.458	4.414	75.871	68.049	4.023	72.072	5.0	[173]
	Single-effect cascade	Ammonia/H2O	−59.41	71.458	6.475	77.933	58.053	5.417	63.470	18.5	[175]

summarizes the CE of refrigerant leakage and electricity consumption, and the total CE for various H₂ liquefaction processes before and after integration of the ARC. It can be seen that the ARC could reduce the CE of liquefaction processes by 5.0~24.3%. For the Claude cycle, there is no refrigerant leakage CE because the GWP of H₂ is zero, and the absorption precooling cycle could reduce the CE by 12.1~24.3% by reducing the SPC of the Claude cycle. For the MR precooling J-B liquefaction process, the total CE is much higher than that for the Claude cycle due to the high-GWP mixed refrigerants. The ARC could reduce the CE of the J-B liquefaction process by 5.0–18.5%.

5. Conclusions

As global energy demand increases and carbon emission and environmental pollution intensify, NG and H₂ become crucial in mitigating the energy crisis and promoting the development of energy decarbonization. Liquefied NG and H₂ are some of the most economical means for large-scale storage and long-distance transport. However, the conventional NG and H₂ liquefaction processes are confronted with high energy consumption, costs, and carbon dioxide emissions. Absorption refrigeration technology, driven by waste heat and renewable thermal energy sources across a wide temperature range, offers a promising solution for enhancing energy efficiency and reducing emissions in both NG and hydrogen liquefaction processes. This review analyzes the progress of ART from three perspectives: solution working pairs, cycle configurations, and heat and mass transfer in main components. For the conventional NG and H₂ liquefaction processes, the energy-saving and carbon emission reduction potential of ART are analyzed from the perspectives of specific power consumption (SPC) and carbon dioxide emissions (CEs). The widespread adoption of absorption refrigeration in NG and hydrogen industries will depend on continued advancements in technology, cost-effectiveness, and integration with existing infrastructure. As these technologies evolve, their potential to contribute significantly to energy-saving and emission reduction goals will continue to grow. The following conclusions are drawn:

(1)

To match different driven heat sources and refrigeration temperatures, working pairs present a diversified development trend, among which environment-friendly and high-efficiency working pairs with ionic liquids and deep eutectic solvents as new absorbers exhibit promising development potential. Ionic liquids and deep eutectic solvents could form good working pairs with various environmentally friendly refrigerants, but their refrigerant vapor absorbability still needs to be enhanced.

(2)

Through the heat and mass transfer coupling within the cycle and the addition of sub-components, cycle configurations with high energy efficiency, which are suitable for different driven heat sources and refrigeration temperatures, are developed.

(3)

Additives, ultrasonic oscillations, and mechanical treatment of heat transfer surfaces are efficient approaches to enhance the ARC’s heat and mass transfer. A synergistic enhancement effect exists between ultrasonic oscillation and nanoparticle additives, as ultrasonic oscillations improve the dispersion homogeneity of nanoparticles in the circulating solution.

(4)

The ARC, driven by industrial waste heat or renewable thermal energy, integrated into conventional NG liquefaction processes, including SMR, C3MR, DMR, cascade, and expansion refrigeration liquefaction processes, by providing precooling temperatures of −13~−54 °C, results in a reduction in SPC and CE by 10~38% and 10~36%, respectively.

(5)

The ARC integrated into conventional H₂ liquefaction processes, including Linde–Hampson, Claude, and Joule–Brayton liquefaction processes, by providing precooling temperatures of 9~−59 °C, results in a reduction in SPC and CE by 2~24% and 5~24%, respectively.

(6)

The ARC, which can achieve lower precooling temperatures and higher energy efficiency, exhibits greater energy-saving and carbon emission reduction potential in NG and H₂ liquefaction. Compared to the lithium bromide cycle, which only provides a cooling temperature above 0 °C for the inter-stage cooler, the auto-cascade absorption, absorption–compression cascade, and absorption-cascade cycles could provide precooling temperatures of −50~−60 °C, which could substitute the compression or expansion precooling cycle, thus exhibiting more significant energy-saving and carbon emission reduction potential.

From the perspective of energy consumption and carbon emissions, ART integrated into the precooling stage of the NG and H₂ liquefaction processes shows significant potential in energy-saving and carbon emission reduction, providing a new pathway for the future development of NG and H₂ liquefaction processes. Future studies are recommended to analyze the economic impacts of ART on the liquefaction processes to comprehensively assess its practical engineering application value.