Salt Hydrate Adsorption Material-Based Thermochemical Energy Storage for Space Heating Application: A Review

Abstract

Recent years have seen increasing attention to TCES technology owing to its potentially high energy density and suitability for long-duration storage with negligible loss, and it benefits the deployment of future net-zero energy systems. This paper provides a review of salt hydrate adsorption material-based TCES for space heating applications at ~150 °C. The incorporation of salt hydrates into a porous matrix to form composite materials provides the best avenue to overcome some challenges such as mass transport limitation and lower thermal conductivity. Therefore, a systematic classification of the host matrix is given, and the most promising host matrix, MIL-101(Cr)(MOFs), which is especially suitable for loading hygroscopic salt, is screened from the perspective of hydrothermal stability, mechanical strength, and water uptake. Higher salt content clogs pores and, conversely, reduces adsorption performance; thus, a balance between salt content and adsorption/desorption performance should be sought. MgCl₂/rGOA is obtained with the highest salt loading of 97.3 wt.%, and the optimal adsorption capacity and energy density of 1.6 g·g⁻¹ and 2225.71 kJ·kg⁻¹, respectively. In general, larger pores approximately 8–10 nm inside the matrix are more favorable for salt dispersion. However, for some salts (MgSO₄-based composites), a host matrix with smaller pores (2–3 nm) is beneficial for faster reaction kinetics. Water molecule migration behavior, and the phase transition path on the surface or interior of the composite particles, should be identified in the future. Moreover, it is essential to construct a micromechanical experimental model of the interface.

1. Introduction

Recent years have seen a significant increase in global efforts to address climate change challenges. The energy sector accounts for approximately three-quarters of greenhouse gas emissions, particularly carbon dioxide due to the burning of fossil fuels, which is the main cause of climate change. As the largest consumer of energy, the building sector in China accounts for approximately one-third of the total global final energy consumption and nearly 15% of the end-user sector’s direct CO₂ emissions [1], in which space heating has a significant contribution. The use of renewable energy sources, such as solar or wind power, is one of the major approaches to combat greenhouse gas emissions. The increase in renewable demand was particularly faster than that of all other fuels in 2021. Overall, China remains the leader, accounting for 43% of global renewable capacity growth, followed by Europe, the United States, and India (Figure 1a [2], data from Global Energy Review 2021). These four markets alone provide almost 80% of renewable capacity expansion worldwide. Regarding renewable annual net capacity additions by technology in 2015–2021 (Figure 1b [3], data from Renewables 2022.), the global renewable capacity is expected to increase by almost 2400 GW (almost 75%) between 2022 and 2027 as per the IEA (International Energy Agency) main-case forecast, equal to the entire installed power capacity of the People’s Republic of China [3]. However, renewable sources are intermittent and fluctuating, leading to mismatch between energy supply and demand. Energy storage provides an effective avenue to address the mismatch challenge and has thus been regarded as an enabling technology for the future net-zero energy transition [4].

Energy storage technologies are abundant, for example, pumped hydrostorage (currently dominant, accounting for ~95% of global installations), compressed air energy storage with a share of ~1%, thermal energy storage (taking ~2% share of global installations), and electrochemical energy storage (mainly lithium-ion batteries, taking ~2%) [5]. These numbers are mere approximations, and the literature often ignores specific data on thermal energy storage. However, a recent report by the Renewables 2021 Global Status Report (GSR) shows that thermal energy accounts for over 50% of global final energy consumption, highlighting the importance of thermal energy [6]. The preliminary results of a recent UK Royal Society study show that the future zero-carbon energy system requires short-, medium-, and long-term energy storage [7]. For the UK alone, a future renewable energy-dominant energy system requires ~100 to 120 GW/100–200 GWh for short-term storage, ~100 to 130 GW/2–6 TWh for medium-term storage, and 70–80 GW/35–40 TWh for long-term storage [8]. A significant amount of work has attempted to address short-term energy storage, e.g., lithium-ion and sodium-ion batteries, supercapacitors, and flywheels [9,10,11,12], whereas medium- and long-term energy storage types require less effort. For long-term energy storage, high energy density, low energy losses during storage, and low costs are among the top key performance indicators (KPIs) for technology development. Thermal energy storage (TES) has the potential to meet these requirements. For example, solar heat in most countries exceeds the total heat demand in the summer, whereas the reverse occurs in the winter when the heat demand exceeds the solar energy supply. TES can be further subdivided into three subcategories: sensible heat storage (SHS), latent heat storage (LHS), and thermochemical energy storage (TCES). Figure 2 shows the working principle of the salt hydrate adsorption-based TCES system [13], which is cyclically operated as illustrated by the liquid/vapor equilibrium line and the solid/gas equilibrium line in the Clausius-Clapeyron diagram. The two modes of operation are charging (desorption) and discharging (adsorption) for heat energy storage (for cold energy storage, the process is reversed). This work focuses on heat energy storage. In the charging mode, the salt hydrate is heated by a heat source, releasing water vapor to yield a salt hydrate with fewer or zero water molecules, and then heat energy is converted into chemical energy and stored in salt. The stored heat is released when water vapor reacts with the salt. Figure 3 shows the characteristics of these TES technologies, of which the thermochemical-based TES (TCES) could meet the needs of seasonal storage owing to the proper adoption of KPIs (e.g., high energy density and zero loss during storage). However, it has the disadvantage of complex technology and high one-time investment in practical application. SHS systems have good security and excellent heat transfer performance, but the energy density is low and the loss is large. LHS systems, which take advantage of latent heat of phase change, have been attracting notable attention to achieve medium energy storage density compared with the other two heat storage approaches in view of system convenience and operating stability. The latest research includes a novel energy storage tank filled with phase change materials with graded metal foams, which notably helps the melting and solidification processes [14], multi-tube shell, and tube units by increasing the number of tubes, which reduces both the charging and discharging times [15], and a hydronic system equipped with evacuated tubes integrated within a hot water tank to improve in the thermal performance of energy storage [16].

TCES-based research and development has increased, as evidenced by the number of publications over the past decades. In this study, “thermochemical energy storage” was used as the keyword to search global publications in the “Web of Science” database between 2000 and 2022. The retrieved results are shown in Figure 4a–c for geographical distribution, publications and citations, and research areas on TCES technology. In the past 22 years, thermochemical energy storage technology has presented a global development trend and covered dozens of research areas. China, the United States, and Germany have conducted extensive research in this area, and the number of published and cited articles on TCES technology has been increasing every year globally. Only two papers in 2000 lacked citations. However, the number of publications increased to 440 in 2022 at an astonishing rate, and the number of citations surpasses 17,000. In addition, engineering and energy fuels, accounting for approximately half of the total publications, are expected to become the leading research area in the field of TCES, which is attributed to the development of TCES technology to deploy future net-zero energy systems.

According to the nature of the binding force between adsorbate and adsorbent surface molecules, it can be divided into physical adsorption and chemical adsorption. Physical adsorption is caused by the intermolecular attraction between adsorbent and adsorbent; the binding force (van der Waals forces) is weak, the energy storage density is relatively small, and it is easy to desorb, such as the adsorption of activated carbon to gas. Chemical adsorption is caused by the chemical bonding between adsorbate and adsorbent, just like a chemical reaction, and the energy storage density is usually large, up to 1500 kJ/kg, five to ten times that of physical adsorption.

TCES can be further subdivided into adsorption-based and reversible reaction-based technologies (Figure 2). Adsorption-based TCES mainly includes two types: liquid-gas and solid-gas types. As shown in Figure 5, pure adsorption materials have higher energy density, especially for salt hydrates, which are environmentally friendly, low cost, and safe to use and are more suitable for low-temperature TCES systems than ammonium salts.

However, pure salt hydrates are prone to agglomerating and even melting if overhydrated, limiting the transport of water vapor and decreasing the cyclic stability [27,28,29,30,31,32,33]. Therefore, recent efforts to address this challenge include the use of composite structures for loading salt hydrates into the porous matrix, which forms the main part of this review. For the composites, the energy density is significantly lower than that of the pure material mainly because of the addition of the matrix, preparation method, salt content, and other factors. The principle of salt hydrate-based thermochemical energy storage, involving the adsorption (discharging) and desorption (charging) processes of salt (solid-state), is a multiscale problem, from the molecular to the system scale. An overview of the scales involved in the problem and the approaches to modeling are summarized in Figure 6. The studies on the molecular scale of composite particles are limited. Future studies can focus on the migration rules and phase transition paths of water molecules to obtain the microscopic mechanism of adsorption and desorption as guidance for controlling the process in macroscopic reactions. The microstructure of the porous matrix also has a key influence on the performance improvement of composite materials, which will be discussed in this paper. A balance between energy storage performance and the increase in the reaction rate in the reactor cannot always be satisfied. Considering the reaction conditions of the material, the finned structure is the preferred choice to solve this problem.

Previous studies have summarized advances in devices and systems based on salt hydrates and presented current challenges, but only a few studies have systematically classified and summarized the characteristics of host matrices supporting salt hydrates from the perspective of enhancing the energy storage performance of composites in TCES systems. Furthermore, the relationship between salt loading and the improvement of adsorption properties needs to be further analyzed. In line with the need to solve the abovementioned problems, this work focuses on candidate salt hydrates and summarizes their thermophysical properties and reaction kinetics. In addition, various host matrices were classified according to structure type, and the potential host matrix for incorporation with the salts was screened. Then, an in-depth discussion on the correlations between salt content and the adsorption capacity and energy density was given, further explaining how the microstructure and thermal conductivity of the host matrix affected the reaction kinetics and thermal behavior of the composites. Finally, at the device scale, the dependence of operating conditions and structural design on performance optimization was proposed. This article is structured in the following manner (Figure 7). The main content about salt hydrates adsorption material-based thermochemical energy storage for space heating applications is discussed in Section 2. Section 2.1 provides details of the thermophysical and reaction kinetics based on pure materials. Section 2.2 summarizes and screens promising host matrices. The contributing factors for the adsorption and desorption performance of the composites are analyzed in Section 2.3. The prospects for the development of the area are discussed in Section 3. Finally, conclusions are given in Section 4.

2. Salt Hydrate Composites for Thermochemical Energy Storage

This section reviews salt hydrates with potential for space heating applications. It is organized in the following manner: Section 2.1 discusses the thermophysical and reaction kinetics properties of pure salt hydrates; Section 2.2 summarizes the characteristics of the host matrix; and Section 2.3 examines the improvement of the adsorption/desorption performance of the composites.

2.1. Thermophysical and Reaction Kinetics Properties of Pure Salt Hydrates

The selection of a salt hydrate-based thermochemical energy storage technology involves a few aspects, with materials being one of the most important research areas. For space heating applications, the main requirements include environmental friendliness, good reaction kinetic and thermodynamic properties, and others, such as low cost and commercial availability (Table 1 [34,35,36,37,38,39]). Among them, the thermophysical properties and reaction kinetics affecting the energy storage performance should be given priority to build a reliable adsorption system. Clearly, no ideal material currently meets all of these criteria. Therefore, the comprehensive properties of materials must be investigated when selecting appropriate materials.

Table 1. Selection criteria for salt hydrates adsorption-based TCES for space heating applications [34,35,36,37,38,39].

Environmentally Friendly	Thermophysical Properties	Reaction Kinetics	Others
Non-toxic Non-corrosive Non-flammable	Moderate reaction temperature Suitable melting point Large reaction enthalpy Good thermal conductivity High theoretical energy density	Low activity energy Suitable pre-exponential factor High diffusion coefficient of vapor Fast reaction rate	Low cost commercially available Products easy to separate High thermal stability High mechanical strength Small volume change

Table 1. Selection criteria for salt hydrates adsorption-based TCES for space heating applications [34,35,36,37,38,39].

Environmentally Friendly	Thermophysical Properties	Reaction Kinetics	Others
Non-toxic Non-corrosive Non-flammable	Moderate reaction temperature Suitable melting point Large reaction enthalpy Good thermal conductivity High theoretical energy density	Low activity energy Suitable pre-exponential factor High diffusion coefficient of vapor Fast reaction rate	Low cost commercially available Products easy to separate High thermal stability High mechanical strength Small volume change

2.1.1. Thermophysical Properties

In the process of adsorption and desorption, temperature and pressure interact and influence each other, especially at low temperatures, and water vapor pressure is unfavorable for heat and mass transfer processes. Therefore, the appropriate reaction equilibrium conditions must be determined. The solid-gas equilibrium curves and thermodynamic constraints for the adsorption/desorption processes of the pure salt hydrate materials are summarized in Figure 8 [40,41,42,43,44]. Most pure materials, such as SrCl₂·6H₂O, K₂CO₃·1.5H₂O, Na₂S·9H₂O, MgSO₄·7H₂O, MgCl₂·6H₂O, LiCl·H₂O, CaCl₂·6H₂O, and SrBr₂·6H₂O, can achieve complete adsorption saturation in the range of 25 to 45 °C and 15–25 mbar, and their thermophysical properties, including melt pointing, thermal conductivity, and reaction enthalpy, are mainly discussed in this section.

After several absorption and desorption cycles of pure salt particles, due to excessive water absorption and multiple contraction and expansion processes, the particles adhere to each other and accumulate into blocks; that is, agglomeration occurs, thus limiting gas transfer and causing reversibility issues and low temperature lift. This limits gas transport and causes reversibility problems and low temperature elevation. Once the charging temperature is higher than the melting temperature, the salt layer will melt [45]. The melting of salt hydrates, causing low permeability and a high pressure drop, is a noticeable problem for practical applications, and even prevents the reaction from proceeding. The molten salt layer will be broken if the reaction conditions change, such as increasing the vapor pressure or the temperature, and then it will recover and continue the desorption reaction. As the sample size and reaction paths change, different reaction behaviors occur, causing a more complex multistep reaction. Therefore, the melting temperature must be maintained above the charging temperature [45]. For example, the charging temperature at which SrCl₂·6H₂O decomposes into SrCl₂·2H₂O is 50 °C; however, SrCl₂·6H₂O completes the aforementioned transformation process and forms a saturated salt solution at 61.34 °C [46]. The melting temperatures of LiCl·H₂O and K₂CO₃·1.5H₂O are 99 °C [47] and greater than 80 °C [48], respectively, before which the desorption reaction has already occurred. The melting temperature and thermal conductivity of some salt hydrates are summarized in Figure 9 [48,49,50,51,52,53,54,55,56,57,58,59,60]. Recently, some researchers conducted the desorption process along with melting for MgCl₂·6H₂O, CaCl₂·4H₂O, MgSO₄·7H₂O, and SrBr₂·6H₂O, as shown in Figure 10a–j. Some results can be obtained as follows: MgCl₂·6H₂O can totally change to MgCl₂·4H₂O by reducing the heating rate to below 0.2 K/min without melting, which is similar to MgSO₄·7H₂O and SrBr₂·6H₂O. However, this transition cannot occur in time because of the high heating rates and mass transfer limitations in the practical TCES system [49]; At approximately 30 to 50 °C, the first step occurred, and 1 mol water molecule was lost. Then, desorption occurs from MgSO₄·6H₂O to MgSO₄·xH₂O (x < 0.1) at 50 to 260 °C [51]. MgCl₂·4H₂O undergoes a phase transition as the heating rate increases to 5 K/min at a temperature of approximately 190 °C. For CaCl₂·4H₂O, the first endothermic peak appears at 31 °C, and this process is accompanied by a melting process [61], which makes their use in a thermochemical reactor highly challenging, especially at temperatures higher than 40 °C [62,63,64]. In the range of 41 to 105 °C, 2.2 mol water molecules were lost, and the crystallization of CaCl₂·4H₂O occurred [65,66]. When the heating rate is higher than 8 K/min, melting can be observed simultaneously with the desorption process from SrBr₂·6H₂O to SrBr₂·H₂O [67,68].

The thermal conductivity of salt hydrates related to heat transfer strongly affects the adsorption and desorption rates when considering practical applications. Some results [59] highlighted that the thermal conductivity of the hydrate is higher than that of anhydrous, given the larger porosity for anhydrous thermal conductivity, such as 0.63 W/(m·K) and 0.44 W/(m·K) for SrBr₂·6H₂O and SrBr₂, respectively. As shown in Figure 9, MgSO₄·7H₂O, MgCl₂·6H₂O, and CaCl₂·6H₂O also have low thermal conductivities ranging from 0.4–0.7 W/(m·K) according to the measurements and experiments. Candidate materials with higher reaction enthalpies are preferred in TCES systems. The reaction enthalpies of each step for salt hydrates widely used in space heating are summarized in

Table 2. Reaction enthalpy of each step for the main salt hydrates.

Reaction	${\mathsf{\Delta }}_{\mathit{r}}\mathit{H}$	Reference
${\mathrm{Na}}_2\mathrm{S}\cdot 9{\mathrm{H}}_2\mathrm{O}\to {\mathrm{Na}}_2\mathrm{S}\cdot 5{\mathrm{H}}_2\mathrm{O}+4{\mathrm{H}}_2\mathrm{O}$	215 ± 20 (kJ/mol)	[58]
${\mathrm{Na}}_2\mathrm{S}\cdot 5{\mathrm{H}}_2\mathrm{O}\to {\mathrm{Na}}_2\mathrm{S}\cdot 2{\mathrm{H}}_2\mathrm{O}+3{\mathrm{H}}_2\mathrm{O}$	176 ± 12 (kJ/mol)
${\mathrm{Na}}_2\mathrm{S}\cdot 2{\mathrm{H}}_2\mathrm{O}\to {\mathrm{Na}}_2\mathrm{S}\cdot \frac{1}{2}{\mathrm{H}}_2\mathrm{O}+1\frac{1}{2}{\mathrm{H}}_2\mathrm{O}$	94 ± 8 (kJ/mol)
${\mathrm{Na}}_2\mathrm{S}\cdot \frac{1}{2}{\mathrm{H}}_2\mathrm{O}+4\frac{1}{2}{\mathrm{H}}_2\mathrm{O}\to {\mathrm{Na}}_2\mathrm{S}\cdot 5{\mathrm{H}}_2\mathrm{O}$	−308 ± 20 (kJ/mol)
${\mathrm{Na}}_2\mathrm{S}\cdot 5{\mathrm{H}}_2\mathrm{O}\to {\mathrm{Na}}_2\mathrm{S}+5{\mathrm{H}}_2\mathrm{O}$	285 (kJ/mol)	[69]
${\mathrm{CaCl}}_2·6{\mathrm{H}}_2\mathrm{O}\to {\mathrm{CaCl}}_2+6{\mathrm{H}}_2\mathrm{O}$	368 (kJ/mol)	[65]
${\mathrm{CaCl}}_2·2.6{\mathrm{H}}_2\mathrm{O}\to {\mathrm{CaCl}}_2+2.6{\mathrm{H}}_2\mathrm{O}$	167 (kJ/mol)
${\mathrm{CaCl}}_2·6{\mathrm{H}}_2\mathrm{O}\to {\mathrm{CaCl}}_2·4{\mathrm{H}}_2\mathrm{O}+2{\mathrm{H}}_2\mathrm{O}$	110 (kJ/mol)	[70]
${\mathrm{CaCl}}_2·2{\mathrm{H}}_2\mathrm{O}\to {\mathrm{CaCl}}_2·2{\mathrm{H}}_2\mathrm{O}+2{\mathrm{H}}_2\mathrm{O}$	118 (kJ/mol)
${\mathrm{CaCl}}_2·2{\mathrm{H}}_2\mathrm{O}\to {\mathrm{CaCl}}_2·{\mathrm{H}}_2\mathrm{O}+{\mathrm{H}}_2\mathrm{O}$	49 (kJ/mol)
${\mathrm{CaCl}}_2·{\mathrm{H}}_2\mathrm{O}\to {\mathrm{CaCl}}_2+{\mathrm{H}}_2\mathrm{O}$	70 (kJ/mol)
${\mathrm{MgSO}}_4·{7\mathrm{H}}_2\mathrm{O}\to {\mathrm{MgSO}}_4·{6\mathrm{H}}_2\mathrm{O}+{\mathrm{H}}_2\mathrm{O}$	50.2 (kJ/mol)	[48,71]
${\mathrm{MgSO}}_4·{7\mathrm{H}}_2\mathrm{O}\to {\mathrm{MgSO}}_4·{0.1\mathrm{H}}_2\mathrm{O}+5.9{\mathrm{H}}_2\mathrm{O}$	319.9 (kJ/mol)
${\mathrm{MgSO}}_4·{0.1\mathrm{H}}_2\mathrm{O}\to {\mathrm{MgSO}}_4+{0.1\mathrm{H}}_2\mathrm{O}$	15.1 (kJ/mol)
${\mathrm{MgSO}}_4·{6.75\mathrm{H}}_2\mathrm{O}\to {\mathrm{MgSO}}_4·{6\mathrm{H}}_2\mathrm{O}+{0.75\mathrm{H}}_2\mathrm{O}$	42.09 (kJ/mol)	[72]
${\mathrm{MgSO}}_4·{6\mathrm{H}}_2\mathrm{O}\to {\mathrm{MgSO}}_4·{\mathrm{H}}_2\mathrm{O}+{5\mathrm{H}}_2\mathrm{O}$	243.81 (kJ/mol)
${\mathrm{MgSO}}_4·{7\mathrm{H}}_2\mathrm{O}\to {\mathrm{MgSO}}_4·{6\mathrm{H}}_2\mathrm{O}+{\mathrm{H}}_2\mathrm{O}$	99.1~−229.0 (kJ·kg−1)	[51]
${\mathrm{MgSO}}_4·{6\mathrm{H}}_2\mathrm{O}\to {\mathrm{MgSO}}_4·{\mathrm{xH}}_2\mathrm{O}+{(6-\mathrm{x})\mathrm{H}}_2\mathrm{O}$	1305.9~−1638.3 (kJ·kg−1)
${\mathrm{MgSO}}_4·{7\mathrm{H}}_2\mathrm{O}\left(\mathrm{s}\right)\to {\mathrm{MgSO}}_4·{7\mathrm{H}}_2\mathrm{O}\left(\mathrm{l}\right)$	69.4 (kJ·kg−1)
${\mathrm{MgCl}}_2·{6\mathrm{H}}_2\mathrm{O}\to {\mathrm{MgCl}}_2·{2\mathrm{H}}_2\mathrm{O}+{4\mathrm{H}}_2\mathrm{O}$	273 (kJ/mol)	[63]
${\mathrm{MgCl}}_2·{2\mathrm{H}}_2\mathrm{O}+{4\mathrm{H}}_2\mathrm{O}\to {\mathrm{MgCl}}_2·{6\mathrm{H}}_2\mathrm{O}$	315 (kJ/mol)	[63]
${\mathrm{MgCl}}_2·{4.7\mathrm{H}}_2\mathrm{O}+{1.3\mathrm{H}}_2\mathrm{O}\to {\mathrm{MgCl}}_2·{6\mathrm{H}}_2\mathrm{O}$	72 (kJ/mol)	[48]
${\mathrm{MgCl}}_2·{6\mathrm{H}}_2\mathrm{O}\to {\mathrm{MgCl}}_2·{4\mathrm{H}}_2\mathrm{O}+{2\mathrm{H}}_2\mathrm{O}$	110 (kJ/mol)	[69,71]
${\mathrm{MgCl}}_2·{4\mathrm{H}}_2\mathrm{O}\to {\mathrm{MgCl}}_2·{2\mathrm{H}}_2\mathrm{O}+{2\mathrm{H}}_2\mathrm{O}$	145 (kJ/mol)
${\mathrm{MgCl}}_2·{6\mathrm{H}}_2\mathrm{O}\to {\mathrm{MgCl}}_2·{4\mathrm{H}}_2\mathrm{O}+{2\mathrm{H}}_2\mathrm{O}$	102.61 (kJ/mol)	[72]
${\mathrm{MgCl}}_2·{4\mathrm{H}}_2\mathrm{O}\to {\mathrm{MgCl}}_2·{2\mathrm{H}}_2\mathrm{O}+{2\mathrm{H}}_2\mathrm{O}$	117.41 (kJ/mol)
${\mathrm{SrBr}}_2·{6\mathrm{H}}_2\mathrm{O}\to {\mathrm{SrBr}}_2·{\mathrm{H}}_2\mathrm{O}+{5\mathrm{H}}_2\mathrm{O}$	56.5 (kJ/mol)	[73]
${\mathrm{SrBr}}_2·{\mathrm{H}}_2\mathrm{O}\to {\mathrm{SrBr}}_2+{\mathrm{H}}_2\mathrm{O}$	72.8 (kJ/mol)	[74]
${\mathrm{SrBr}}_2·{6\mathrm{H}}_2\mathrm{O}\to {\mathrm{SrBr}}_2·{\mathrm{H}}_2\mathrm{O}+{5\mathrm{H}}_2\mathrm{O}$	791 (kJ·kg−1)	[51]
${\mathrm{SrBr}}_2·{6\mathrm{H}}_2\mathrm{O}\left(\mathrm{s}\right)\to {\mathrm{SrBr}}_2·{6\mathrm{H}}_2\mathrm{O}\left(\mathrm{l}\right)$	145.7 (kJ·kg−1)

.

2.1.2. Reaction Kinetics Properties

Adsorption/desorption reaction kinetics and mechanism studies have been key aspects of the research.

Table 3. Kinetic models for solid–gas reactions [51,75].

Reaction Model	Code	$\mathit{f}\left(\mathit{\alpha }\right)$	$\mathit{F}\left(\mathit{\alpha }\right)$ *
One-dimensional diffusion	$D_1$	$0.5{\alpha }^{-1}$	${\alpha }^2$
Two-dimensional diffusion	$D_2$	$[-\mathit{ln}(1-\alpha ){]}^{-1}$	$\alpha +(1-\alpha )\mathit{ln}(1-\alpha )$
Three-dimensional diffusion (sphere)	$D_3$	$1.5(1-\alpha {)}^{2/3}[1-(1-\alpha {)}^{1/3}{]}^{-1}$	$[1-(1-\alpha {)}^{1/3}{]}^2$
Three-dimensional diffusion(cylinder)	$D_4$	$1.5\left[\right(1-\alpha {{)}^{-}}^{1/3}-1{]}^{-1}$	$(1-2\alpha /3)-(1-\alpha {)}^{2/3}$
Two-dimensional phase boundary reaction	$R_2$	$2(1-\alpha {)}^{1/2}$	$1-(1-\alpha {)}^{1/2}$
Two-dimensional phase boundary reaction	$R_3$	$3(1-\alpha {)}^{2/3}$	$1-(1-\alpha {)}^{1/3}$
Nucleation and nuclei growth	$A_1$	$1-\alpha$	$-\mathit{ln}(1-\alpha )$
Nucleation and nuclei growth	$A_2$	$2(1-\alpha )[-\mathit{ln}(1-\alpha ){]}^{1/2}$	$[-\mathit{ln}(1-\alpha ){]}^{1/2}$
Nucleation and nuclei growth	$A_3$	$3(1-\alpha )[-\mathit{ln}(1-\alpha ){]}^{2/3}$	$[-\mathit{ln}(1-\alpha ){]}^{1/3}$
Exponential nucleation	$P_1$	$1$	$\alpha$
Exponential nucleation	$P_2$	$2{\alpha }^{1/2}$	${\alpha }^{1/2}$
Exponential nucleation	$P_3$	$3{\alpha }^{2/3}$	${\alpha }^{1/3}$
Exponential nucleation	$P_4$	$4{\alpha }^{3/4}$	${\alpha }^{1/4}$

* $F\left(\alpha \right)={\int }_0^{\alpha }\frac{d\alpha }{f\left(\alpha \right)}$.

shows a list of models for the analysis of solid-state reactions [51,75]. A comparison of the aforementioned models helps to identify kinetic parameters, such as activation energy and frequency factor corresponding to the specified intermediate state in the reaction path resulting from the influence of different conditions on the reaction behavior. The most frequently used kinetic equations of solid-gas desorption are summarized in

Table 4. Reaction kinetic equation of solid–gas reactions.

	Reaction Equation	Reference
Arrhenius equation	$\frac{d\alpha }{dt}=A{e}^{-\frac{E_a}{RT}f\left(\alpha \right)}$	[51]
Kissinger equation	$\frac{d\left(\ln \frac{\beta }{T_P^2}\right)}{d\left(\frac{1}{T_P}\right)}=-\frac{E_a}{R}$	[76]
	$\log \frac{\beta }{T^2}=\log \frac{AR}{E_a}$	[77]
Doyle equation	$\mathrm{l}\mathrm{n}F\left(\alpha \right)=\ln \frac{A{E}_a}{\beta R}-5.3305-1.053\frac{E_a}{RT}$	[78]
Coats–Redfern equation	$\mathrm{l}\mathrm{n}\frac{F\left(\alpha \right)}{T^2}=\mathrm{l}\mathrm{n}\frac{AR}{\beta E_a}\left(1-\frac{2RT}{E_a}\right)-\frac{E_a}{RT}$
Ozawa equation	$\mathrm{l}\mathrm{o}\mathrm{g}\beta =\log \frac{A{E}_a}{R}-2.315-0.4567\left(\frac{E_a}{R{T}_{max}}\right)$	[79]
Flynn–Wall–Ozawa equation	$\mathrm{l}\mathrm{n}\beta =\mathrm{l}\mathrm{n}\frac{A{E}_a}{Rf\left(\alpha \right)}-5.331-1.052\frac{E_a}{RT}$	[80]
Kissinger–Akahira–Sunose equation	$\mathrm{l}\mathrm{n}\frac{\beta }{T^2}=\mathrm{l}\mathrm{n}\frac{AR}{E_{a}f\left(\alpha \right)}-\frac{E_a}{RT}$	[81]
Starink equation	$\mathrm{l}\mathrm{n}\frac{\beta }{T^{1.92}}=\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{s}\mathrm{t}.-1.0008·\frac{E_a}{RT}$	[82]

where $\beta$ is the heating rate in K/min, $T$ is the temperature in K, $A$ is the pre-exponential factor in K/min, $E_a$ is the activation energy in J/mol, $R$ is the gas constant, $\alpha$ is the degree of conversion, and $f\left(\alpha \right)$ is the reaction model.

. The desorption process along with the melting process can be described by the Arrhenius equation [51]. The activation energy is the energy required for a molecule to change from a normal state to an active state that is prone to chemical reactions, the amount of which can reflect the difficulty of chemical reactions, and it can be calculated by the Kissing equation [76,77]. If the heating rate is constant, then the Doyle [78] equation, Coats-Redfern equation [78] and Ozawa equation [79] can be used to perform the kinetic analysis based on the TG curve. The isoconversional integral Flynn-Wall-Ozawa (FWO) [80], Kissinger-Akahira-Sunose (KAS) [81], and Starink [82] methods were used to calculate the activation energy and pre-exponential factor at different conversion rates.

Mamania et al. [83] performed isothermal desorption experiments of MgCl₂·6H₂O at four different temperatures of 70 °C, 80 °C, 90 °C, and 100 °C with TG and DSC, and they obtained the kinetic parameters of MgCl₂·6H₂O between the initiation of the desorption reaction and the time when 2.34 H₂O were lost. In particular, the activation energy equaled 103.74 kJ/mol, and the frequency factor equaled 2.088 × 1013 s⁻¹ by using the kinetics models R2, R3, A2, and A3 (Table 4), which is relatively close to previous research results via the Freeman-Carroll method [77]. However, these data differ from the data obtained by Huang et al. [75], who used the kinetics models R2, R3, D4, and A1 and found a much lower activation energy of 66.8 kJ/mol and frequency factor of 3.6·109 s⁻¹ for the first desorption stage by using the Doyle, Coats-Redfern, and Malek methods. MgSO₄·11H₂O was discovered as early as 1837 [84], and Peterson and Wang et al. [85] proved the existence of such structures containing only 11 water molecules by crystal X-ray diffraction. According to previous studies [51], the melting process of MgSO₄·7H₂O required 88.4 kJ/mol, increasing the activation energy of the whole reaction and reducing the reaction rate, mainly due to the limitations of heat and mass transfer. Okhrimenko et al. [86] investigated the desorption mechanism under various water vapor pressures and temperatures based on the nonstoichiometric model of hydrates, and the results obtained for MgSO₄ are in good agreement with the localized water molecule model. Fortes et al. [87] identified that the temperature dependence of the lattice parameters over the intervening range was fitted with a modified Einstein oscillator model, which was used to obtain the coefficients of the thermal expansion tensor. In addition, the crystal structure and thermal expansion tensor of MgSO₄·11H₂O were obtained, as shown in Figure 11a–d. This work provided the basis for computer modeling of the structure and interpretation of high-pressure studies of this substance.

The crystalline structures of various forms of MgSO₄ and slow kinetics increase the complexity of the application. Therefore, the molecular-level investigation of the factors affecting the extent of the reaction of the materials is crucial for understanding the mechanism. Many findings focusing on the characterization of hydrogen bond networks indicated that hydrogen bonds in MgSO₄ have a key effect on the microstructure [88,89]. Iype et al. [90] identified that these hydrogen bonds led to distortions and limited the coordination of water molecules, forming lower-energy isomers. In addition, the hexahydrated structure exhibited an intramolecular proton-transfer reaction, as shown in Figure 11e,f. The findings suggest that the strong hydrogen bond interactions potentially dissociated water molecules during adsorption. Similar to MgSO4, the crystal structures of CaCl₂ [91] and CaCl₂·2H₂O [92], as well as MgCl₂·2H₂O [93] and MgCl₂·2H₂O [94], have a strong anisotropic morphology, as shown in Figure 12. A new ReaxFF force field was developed to describe crystal surface energies, reaction enthalpies, thermal conductivity, desorption kinetics, and radial distribution functions [95]. Some results show that the initial desorption is 1.9–2.5 times lower in the z-direction. Although cracks and pores significantly increased the rate of reaction by creating pathways through outer layers of the evaporation of inner water molecules, superficial layers prevented dehydrating of core layers similar to the results [96], unlike MgCl₂, which ignored this effect [94]. Recently, two new transferable force fields for MgCl₂- and CaCl₂-based steady-state non-equilibrium molecular dynamics (SS-NEMD) were performed to investigate the diffusion coefficient of water through MgCl₂·nH₂O (n = 1 to 6) in the range of 10⁻¹¹ to 10⁻⁹ m²/s, which is similar to 9.13 × 10⁻¹¹ to 4.4710⁻⁹ m²/s [93] and higher than 0.37 × 10⁻¹⁰ m²/s obtained by ReaxFF-MD (molecular dynamic) simulations [96]. In addition, a lower thermal conductivity of CaCl₂·2H₂O with 0.3 W/(m·K), and 0.3–0.9 W/(m·K) for MgCl₂·nH₂O [93] was obtained. Another investigation [94] used ReaxFF in combination with density functional theory (DFT) calculations to study the desorption and hydrolysis kinetics of MgCl₂·H₂O and MgCl₂·2H₂O. Some reaction kinetic parameters included a pre-exponential factor of 1.55 × 10⁻⁵ for MgCl₂·2H₂O and 7.72 × 10⁻¹¹ for MgCl₂·H₂O and an activation energy of 32.22 kJ/mol for MgCl₂·2H₂O and 69.28 kJ/mol for MgCl₂·H₂O.

Figure 11. Crystal structure for MgSO₄·11H₂O: (a–d) [87] and for MgSO₄·6H₂O; (e) along with bond lengths for MgSO₄·6H₂O (f) [90]: Reprinted with permission from Ref. [87]. Copyright © Springer- Chemical Society Verlag 2008 All rights reserved; with permission from Ref. [90]. Copyright © 2012 American Chemical Society All rights reserved.

Figure 11. Crystal structure for MgSO₄·11H₂O: (a–d) [87] and for MgSO₄·6H₂O; (e) along with bond lengths for MgSO₄·6H₂O (f) [90]: Reprinted with permission from Ref. [87]. Copyright © Springer- Chemical Society Verlag 2008 All rights reserved; with permission from Ref. [90]. Copyright © 2012 American Chemical Society All rights reserved.

Figure 12. Crystal structure for (a) CaCl₂ [91]; (b) CaCl₂·2H₂O [92]; (c) MgCl₂·6H₂O [93] and (d) MgCl₂·2H₂O [94]: Reprinted with permission from Ref. [91] Copyright © 1963 American Mineralogist Crystal Structure Database All rights reserved; with permission from Ref. [92] Copyright © 1977 Wiley All rights reserved; with permission from Ref. [93] Copyright © 2019 Elsevier Ltd. All rights reserved; with permission from Ref. [94] Copyright © 2016 the Owner SocietiesAll rights reserved.

Figure 12. Crystal structure for (a) CaCl₂ [91]; (b) CaCl₂·2H₂O [92]; (c) MgCl₂·6H₂O [93] and (d) MgCl₂·2H₂O [94]: Reprinted with permission from Ref. [91] Copyright © 1963 American Mineralogist Crystal Structure Database All rights reserved; with permission from Ref. [92] Copyright © 1977 Wiley All rights reserved; with permission from Ref. [93] Copyright © 2019 Elsevier Ltd. All rights reserved; with permission from Ref. [94] Copyright © 2016 the Owner SocietiesAll rights reserved.

Molecular-level investigations, such as hydrogen bond networks affecting the coordination of water molecules during adsorption, provided the basis for water transport in the microstructure. In addition, models, including the ReaxFF force field and molecular dynamic simulations, calculated the crystalline structures of the materials and the diffusion coefficient of water; hence, their usefulness as tools for gaining insights into the adsorption and desorption reactions.

In summary, technical challenges exist in the use of pure salt hydrates, including mass transport limitation of water vapor within salt hydrate grains, decomposition of salt hydrate resulting in microstructural change, performance degradation due to melting and agglomeration, and lower thermal conductivity. The incorporation of salt hydrates into a porous supporting matrix to form composite thermochemical materials provides an avenue to overcome some of these challenges, which has been a topic of great interest in recent years. This aspect will be covered in Section 2.2 and Section 2.3.

2.2. Physics and Chemical Characterization of Host Matrices

Salt hydrate composites can be obtained by impregnating salt into a structural material covering micropores (pore size < 2 nm), mesopores (2 nm < pore size < 50 nm), and macropores (pore size > 50 nm). These porous materials with larger pore volume, specific surface area, and pore size not only provide support for the salt hydrates, allowing them to be evenly distributed while increasing the contact surface area and mass transfer channels for water vapor transport, which is beneficial for the adsorption capacity. The basic characteristics of common and potential host matrices are shown in Table 5 and Table 6, respectively. Carbon-based materials (e.g., expanded natural graphite, carbon nanotube, activated carbon fiber felt, and foam), alumina-silicates (e.g., zeolites, activated alumina), silicates (e.g., silica gel, silica aerogels, Wakkanai siliceous shale, MCM-41, and SBA-15), and natural mineral clays (e.g., bentonite, goethite, kaolinite, and attapulgite (palygorskite), expanded perlite, sepiolite, diatomite, (expanded) vermiculite, and hydroxyapatite, among others) have been applied widely to support pure salt hydrates in TCES systems. Some potential matrices, including metal-aluminophosphates (e.g., AlPOs, SAPOs, VAPOs, and TAPOs, among others), metal–organic frameworks (MOFs, e.g., MILs, UiOs, PCNs, DUTs, BUTs, NUs, and ZIFs), and microporous/macroporous silicon foam, have been gradually developed in recent years and may be regarded as promising porous materials in the future. The key indicators to be considered in the application of composite materials as matrices consist mainly of higher thermal conductivity, hydrothermal stability, mechanical strength, hydrophilicity, adsorption capacity, and cost-effectiveness. Candidate materials are prioritized from front to back, as shown in Figure 13. The dashed arrows indicate that the performance of candidates’ host matrices decreased in sequence according to the key indicators. A more in-depth discussion of the above-mentioned matrices is given below.

Nanoparticles (Cu, Al₂O₃) with high thermal conductivity are beneficial to improve the thermal performance of TES systems in the following aspects: stable output temperature, uniform heat transfer, and high exergy efficiency [97]. The active carbon fiber felts (ACF FELT) fabricated by viscose-based fibers, as well as their excellent mechanical properties and low thermal expansion, accelerate intraparticle adsorption kinetics. In addition, a large number of fibers create discontinuous texture grooves, and wedge-shaped cracks are highly suitable for salt to adhere (Figure 14a) [98]. Active carbon foam (ACF) is a widely applied material in the TCES process with strong surface features, as well as a nearly ordered homogeneous spherical microstructure with pores and interconnected threads, as shown in Figure 14b, to confirm the higher porosity and specific surface area [99]. Expanded graphite is hydrophobic [100] with a higher thermal conductivity of 200 W/(m·K) [101], and its ordered lamellar structure is shown in Figure 14c [102]. The highest thermal conductivity belongs to carbon nanotubes (CNTs) with 3000–4000 W/(m·K) [103] among carbon-based matrices. CNTs are categorized into two groups, single-walled carbon nanotubes (SWCNTs) and multi-walled carbon nanotubes (MWCNTs), and their sophisticated graphite layers with a one-dimensional tubular nanostructure are shown in Figure 14d [104]. Similar to CNTs, graphene-based materials have substantial thermal conductivity up to 5300 W/(m·K), extraordinarily high specific surface area of 2620 m²/g, high chemical stability, and low volume change [100]. Thus, they exhibit wide applications such as catalysis, and energy storage, and can store 77% energy from sunlight. In addition, hybrid materials such as N-doped and some metal oxides can be introduced to improve the significance of these materials [100]. In addition, unlike Expanded graphite, the superhydrophobic surface of CNTs can be changed using the matrix hygroscopic group of –OH to prepare hydrophilic composite materials. For example, CaCl₂-MWCNT composites modified using the PVP water-based adhesive improved the water-retaining ability and adsorption rate, and an adsorption capacity up to 0.75 g·g⁻¹ was determined at a low relative humidity of 50% [105].

Alumina-silicate adsorbents, such as activated alumina (AA) and zeolites, have been proven to be excellent host matrices owing to their higher porous structure and bulk density, and their microstructure is shown in Figure 14e,f [106,107]. The structural stability of AA is rarely fragmented in adsorption/desorption cycles [108], whereas for zeolites, the weak mechanical strength can lead to performance degradation and sorbent losses [106] and a higher desorption temperature of 130 to 180 °C, which is unfavorable to the reaction process. In addition, more advantages exist in zeolites and zeotype matrices, such as uniform pore size distribution [109], strong hydrophilicity [110], higher water uptake at high humidity [111], energy density of 107–185 kWh/m³ [112], and better thermal and hydrothermal stability. Ion-exchange characteristics in zeolites allow for tuning their adsorption capacity. A 42% increase in water uptake was achieved by stirring zeolite with magnesium nitrate solution at P/P0 = 0.2. However, no obvious increase in the water uptake capacity and characteristic adsorption energy of Mg/Na-Y zeolites was observed when the ion exchange degree exceeded 64.1% [109].

For natural mineral materials (bentonite, goethite, kaolinite, attapulgite (palygorskite), expanded perlite, sepiolite, diatomite, vermiculite (expanded), and hydroxyapatite, among others), their cost-effectiveness, ease of access, and (chain) layered structure (Figure 14g for attapulgite [113] and Figure 14h for expanded vermiculite [114]) provide many adsorption sites. In particular, the diatomite presented a flat round shape with a large number of pores inside the particles (Figure 14i) and excellent hydrophilicity [115], promoting the mass transfer of water vapor and significantly increasing the water uptake capacity [115,116,117,118]. Hydroxyapatite (HAP), as a member of the calcium phosphate family, has a higher heat transfer performance (a thermal conductivity of 0.15–0.2 K) [119]), and biocompatibility with surrounding materials. Nevertheless, insufficient thermal and hydrothermal stability for these mineral materials may be an obstacle to practical application.

Silica gel has outstanding properties, such as a high surface area, controllable pore size, and uniform channel inside the particle [120], but suffers from relatively low hydrophilicity, especially for conditions at 35 °C and 1.2 kPa during adsorption [121]; thus, a high relative pressure is needed to maximize its capacity. For a limited pressure range, such as P/P0 < 0.3 in the closed system, a larger porous structure is necessary to enhance the water uptake capacity. The TCES system has a lower energy density of 50 kWh/m³ (a theoretical value of 200–300 kWh/m³) under conditions less than 100 °C [122]. Silica aerogels as a novel adsorption matrix have attracted great interest owing to their high porosity (80% to 99.8% [123]) and controllable pore size to facilitate mass transfer and water uptake behavior [124]. Nevertheless, extremely low thermal conductivity with 0.014–0.02 W/(m·K) [125] and low density of 80 kg/m³ [123] fail to achieve a considerable effect unless they are specially regulated or incorporated into high thermal conductivity materials. The microstructures of silica gel and silica aerogels are shown in Figure 14j [126] and Figure 14k [127], respectively. Other ordered mesoporous silica materials, such as MCM-41 and SBA-15, have more satisfactory characteristics with better hydrothermal stability and controlled porous structures that are easily accessible by water molecules moving back and forth from the atmosphere to internal pores [128,129]. Slightly different from MCM-41, SBA-15 exhibited a rod structure with a two-dimensional hexagonal channel structure (Figure 14l [130] and Figure 14m [128], respectively). Studies have identified that pore size [131,132] and Si/Al ratios [133] mainly affect the hydrophobic-hydrophilic properties and cycling performance of MCM-41 and SBA-15. MCM-41-(Al₂)x and SBA-15-(Al₂)x composites exhibited higher adsorption capacities (0.17 g·g⁻¹ and 0.09, respectively) than pure host materials (0.04 g·g⁻¹ for MCM-41 and 0.02 g·g⁻¹ for SBA-15, respectively) at lower relative pressure values of 0.3, and the highest energy density of 612 kJ·kg⁻¹ for the MCM-41-(Al₂)7 composite was obtained [134]. When P/P0 < 0.3, more water (0.47 g·g⁻¹) was adsorbed by SBA-15/CaCl₂. In addition, adsorption along with the melting process occurred rapidly as long as the salt particles were loaded into the smaller pores (8 nm) [131].

Table 5. Basic characteristics of common porous matrices.

Class	Porous Matrix	Physical Properties	Basic Characteristics
Mesoporous Nanoparticles (Cu, Al2O3)	Cu nanoparticles	Thermal conductivity: 400 W/(m·K) [97] Density: 8954 kg/m3 [97]	Uniform heat transfer [97] Stable output temperature [97] High exergy efficiency [97]
Macroporous Carbon-based	Expanded natural graphite	Pore volume: 3.95 cm3/g [135] Specific surface area: 15.57 m2/g [135] Average pore diameter: 998.84 nm [135] Thermal conductivity: 200 W/(m·K) [101]	High thermal conductivity [101] High porosity [101] Hydrophobic surfaces [101]
Mesoporous Carbon-based	Carbon nanotube (SWCNTs, MWCNTs)	Specific surface area: 495 m2/g [103] Thermal conductivity: 3000–4000 W/(m·K) [103] Average pore diameter: 2–3 nm [103] Bulk density: 1600–2600 kg/m3 [136]	Superhydrophobic surfaces [137] High toughness [137] Ultra-high specific strength [137] Excellent mechanical properties [137]
	Activated carbon foam	Pore volume: 0.7–1.3 cm3/g [138] Specific surface area: 2032.6732 m2/g [139] Thermal conductivity: 0.23–0.3 W/(m·K) [138] Average pore diameter: 3.5–5 nm [140]	Large surface area [139] Large pores volume [139] Excellent mechanical properties [139]
	Activated carbon fiber felt	Pore volume: 0.67 cm3/g [141] Specific surface area: 2175.0 m2/g [142] Thermal conductivity: 0.2–0.35 W/(m·K) [143] Average pore diameter: 2.6 nm [141]
Mesoporous Alumina-silicates (Zeolites 13X, 3A, 4A, Activated alumina, etc.)	Classical zeolites Zeolite 13X (3A, 4A, Na-X/Y, etc.)	Pore volume: 0.28 cm3/g [106] Specific surface area: 336 m2/g [106] Thermal conductivity: 0.15 W/(m·K) [144] Average pore diameter: 2.08 nm [145] Adsorption capacity: 0.202 g·g−1 [110]	Uniform pore size distribution [109] Interconnected channel [109] Ion-exchange capacity [109] Low mechanical strength [106]
	Activated alumina	Pore volume: 0.39 cm3/g [108] Specific surface area: 271 m2/g [108] Thermal conductivity: 0.28 W/(m·K) [108] Average pore diameter: 5.53 nm [108] Bulk density: 908 kg/m3 [108]	High mechanical strength [108] High bulk density [108]
Mesoporous/Macroporous Natural mineral materials (bentonite, goethite, kaolinite, attapulgite (palygorskite), expanded perlite, sepiolite, diatomite, (expanded) vermiculite, hydroxyapatite, etc.)	Expanded vermiculite	Pore volume: 2.8 cm3/g [146] Specific surface area: 18 m2/g [146] Thermal conductivity: 0.065 W/(m·K) [147] Average pore diameter: 600 nm [147] Bulk density: 127.74 kg/m3 [60] Adsorption capacity: 0.03 g·g−1 [60]	Cost-effective [148] Excellent hydrophilic [117] Weak hydrothermal stability [60]
	Expanded perlite	Pore volume: 3.7 cm3/g [116] Specific surface area: 20.29 m2/g [116] Average pore diameter: 720 nm [116] Adsorption capacity: 0.17 g·g−1 [116]
	Attapulgite	Specific surface area: 98 m2/g [148] Average pore diameter: 64 nm [148]
	Diatomite	Pore volume: 0.0438 cm3/g [117] Specific surface area: 22 m2/g [117] Average pore diameter: 6.7042 nm [117] Bulk density: 1900–2400 kg/m3 [149]
	Speiolite	Pore volume: 0.19 cm3/g [118] Specific surface area: 58.68 m2/g [118] Average pore diameter: 17.68 nm [118]
	Hydroxyapatite	Pore volume: 0.664 cm3/g [150] Specific surface area: 113.2 m2/g [150] Thermal conductivity: 0.15–0.2 W/(m·K) [119] Adsorption capacity: 0.039 g·g−1 [119]	Superior compatibility [150] Higher thermal conductivity [119]
Mesoporous Silicates (silica gel, silica aerogels, Wakkanai siliceous shale, MCM-41, SBA-15, etc.)	Silica gel	Pore volume: 0.28 cm3/g [120] Specific surface area: 716 m2/g [120] Thermal conductivity: 0.174 W/(m·K) [120] Average pore diameter: 2.2 nm [120] Adsorption capacity: 0.45 g·g−1 [151]	Low mechanical strength [120] High moisture adsorption capability [120] High porosity [120]
	Silica aerogels	Pore volume: 0.97 cm3/g [123] Specific surface area: 509 m2/g [123] Thermal conductivity: 0.014–0.02 W/(m·K) [125] Average pore diameter: 13 nm [152] Bulk density: 80 kg/m3 [123] Adsorption capacity: 1.15–1.35 g·g−1 [124]
	Ordered mesoporous silica MCM-41/48	Pore volume: 0.988 cm3/g [128] Specific surface area: 1043.085 m2/g [128] Average pore diameter: 3.62 nm [128] Bulk density: 147.365 kg/m3 [128] Adsorption capacity: 0.04 g·g−1 [134]	High porosity [128] Controllable pore size [128] Uniform channel [128] Excellent mechanical strength [128]
	Ordered mesoporous silica SBA-15 (MAS, etc.)	Pore volume: 1.307 cm3/g [128] Specific surface area: 514.68 m2/g [128] Average pore diameter: 9.69 nm [128] Bulk density: 101.423 kg/m3 [128] Adsorption capacity: 0.02 g·g−1 [134]

Table 5. Basic characteristics of common porous matrices.

Class	Porous Matrix	Physical Properties	Basic Characteristics
Mesoporous Nanoparticles (Cu, Al2O3)	Cu nanoparticles	Thermal conductivity: 400 W/(m·K) [97] Density: 8954 kg/m3 [97]	Uniform heat transfer [97] Stable output temperature [97] High exergy efficiency [97]
Macroporous Carbon-based	Expanded natural graphite	Pore volume: 3.95 cm3/g [135] Specific surface area: 15.57 m2/g [135] Average pore diameter: 998.84 nm [135] Thermal conductivity: 200 W/(m·K) [101]	High thermal conductivity [101] High porosity [101] Hydrophobic surfaces [101]
Mesoporous Carbon-based	Carbon nanotube (SWCNTs, MWCNTs)	Specific surface area: 495 m2/g [103] Thermal conductivity: 3000–4000 W/(m·K) [103] Average pore diameter: 2–3 nm [103] Bulk density: 1600–2600 kg/m3 [136]	Superhydrophobic surfaces [137] High toughness [137] Ultra-high specific strength [137] Excellent mechanical properties [137]
	Activated carbon foam	Pore volume: 0.7–1.3 cm3/g [138] Specific surface area: 2032.6732 m2/g [139] Thermal conductivity: 0.23–0.3 W/(m·K) [138] Average pore diameter: 3.5–5 nm [140]	Large surface area [139] Large pores volume [139] Excellent mechanical properties [139]
	Activated carbon fiber felt	Pore volume: 0.67 cm3/g [141] Specific surface area: 2175.0 m2/g [142] Thermal conductivity: 0.2–0.35 W/(m·K) [143] Average pore diameter: 2.6 nm [141]
Mesoporous Alumina-silicates (Zeolites 13X, 3A, 4A, Activated alumina, etc.)	Classical zeolites Zeolite 13X (3A, 4A, Na-X/Y, etc.)	Pore volume: 0.28 cm3/g [106] Specific surface area: 336 m2/g [106] Thermal conductivity: 0.15 W/(m·K) [144] Average pore diameter: 2.08 nm [145] Adsorption capacity: 0.202 g·g−1 [110]	Uniform pore size distribution [109] Interconnected channel [109] Ion-exchange capacity [109] Low mechanical strength [106]
	Activated alumina	Pore volume: 0.39 cm3/g [108] Specific surface area: 271 m2/g [108] Thermal conductivity: 0.28 W/(m·K) [108] Average pore diameter: 5.53 nm [108] Bulk density: 908 kg/m3 [108]	High mechanical strength [108] High bulk density [108]
Mesoporous/Macroporous Natural mineral materials (bentonite, goethite, kaolinite, attapulgite (palygorskite), expanded perlite, sepiolite, diatomite, (expanded) vermiculite, hydroxyapatite, etc.)	Expanded vermiculite	Pore volume: 2.8 cm3/g [146] Specific surface area: 18 m2/g [146] Thermal conductivity: 0.065 W/(m·K) [147] Average pore diameter: 600 nm [147] Bulk density: 127.74 kg/m3 [60] Adsorption capacity: 0.03 g·g−1 [60]	Cost-effective [148] Excellent hydrophilic [117] Weak hydrothermal stability [60]
	Expanded perlite	Pore volume: 3.7 cm3/g [116] Specific surface area: 20.29 m2/g [116] Average pore diameter: 720 nm [116] Adsorption capacity: 0.17 g·g−1 [116]
	Attapulgite	Specific surface area: 98 m2/g [148] Average pore diameter: 64 nm [148]
	Diatomite	Pore volume: 0.0438 cm3/g [117] Specific surface area: 22 m2/g [117] Average pore diameter: 6.7042 nm [117] Bulk density: 1900–2400 kg/m3 [149]
	Speiolite	Pore volume: 0.19 cm3/g [118] Specific surface area: 58.68 m2/g [118] Average pore diameter: 17.68 nm [118]
	Hydroxyapatite	Pore volume: 0.664 cm3/g [150] Specific surface area: 113.2 m2/g [150] Thermal conductivity: 0.15–0.2 W/(m·K) [119] Adsorption capacity: 0.039 g·g−1 [119]	Superior compatibility [150] Higher thermal conductivity [119]
Mesoporous Silicates (silica gel, silica aerogels, Wakkanai siliceous shale, MCM-41, SBA-15, etc.)	Silica gel	Pore volume: 0.28 cm3/g [120] Specific surface area: 716 m2/g [120] Thermal conductivity: 0.174 W/(m·K) [120] Average pore diameter: 2.2 nm [120] Adsorption capacity: 0.45 g·g−1 [151]	Low mechanical strength [120] High moisture adsorption capability [120] High porosity [120]
	Silica aerogels	Pore volume: 0.97 cm3/g [123] Specific surface area: 509 m2/g [123] Thermal conductivity: 0.014–0.02 W/(m·K) [125] Average pore diameter: 13 nm [152] Bulk density: 80 kg/m3 [123] Adsorption capacity: 1.15–1.35 g·g−1 [124]
	Ordered mesoporous silica MCM-41/48	Pore volume: 0.988 cm3/g [128] Specific surface area: 1043.085 m2/g [128] Average pore diameter: 3.62 nm [128] Bulk density: 147.365 kg/m3 [128] Adsorption capacity: 0.04 g·g−1 [134]	High porosity [128] Controllable pore size [128] Uniform channel [128] Excellent mechanical strength [128]
	Ordered mesoporous silica SBA-15 (MAS, etc.)	Pore volume: 1.307 cm3/g [128] Specific surface area: 514.68 m2/g [128] Average pore diameter: 9.69 nm [128] Bulk density: 101.423 kg/m3 [128] Adsorption capacity: 0.02 g·g−1 [134]

A series of metal aluminophosphates (AlPOs, SAPOs, VAPOs, and TAPOs, among others) presented a highly stable structure and regenerability [153,154,155], especially for AlPOs and SAPOs with high mechanical strength, and the open structures enhanced the surface areas and resulted in a good balance between hydrophobicity and hydrophilicity [156,157]. The microstructures of AlPOs, SAPOs, and VAPOs are shown in Figure 14n–p [155,158,159], respectively. In previous studies on SAPOs and AlPOs [160,161,162,163], the water adsorption mechanism was revealed, especially for structural phase-dependent water diffusivity at various concentrations and Si/Al ratios, by means of molecular dynamics (MD), density functional theory (DFT) computations [162], nuclear magnetic resonance (NMR) [160], and X-ray diffraction (XRD) [160,163]. AlPOs and SAPOs can load larger amounts of water than silica gel, showing a superior ability to bind to water [164]. Both of them process a higher energy storage capacity of 200–240 Wh/kg at a moderate desorption temperature of 75 to 140 °C [165]. SAPO-5 exhibited the best adsorption capacity at a lower pressure of 0.38 owing to the introduction of extrahydrohydrophilic sites (SiO₄⁴⁻) in the SAPO framework, a hierarchical porous structure, and a larger mesoporous volume compared with AlPO4-5 [155]. Therefore, SAPOs are still a potential adsorbent for water in combination with salt hydrates; although, extant research on this material is limited.

Compared with other matrices, metal–organic frameworks (MOFs, aluminum fumarate, MILs, CPOs, UiOs, PCNs, DUTs, BUTs, NUs, and ZIFs, among others) are highly hydrophilic materials composed of inorganic metal ions or metal clusters that form crystalline inorganic-organic structures, creating abundant porous structures. Some MOFs, such as MIL-101 [166], aluminum fumarate [167], UiO-67 [168], and PCN-33 [169], have microstructures that are presented in Figure 14q–t. In particular, a series of MILs or UiOs comprising three- and four-valent metal cations (Cr³⁺, Al³⁺, Fe³⁺, Zr⁴⁺, and Ti⁴⁺) and aromatic polycarboxylate linkers (Figure 15) were considered for water adsorption applications [170,171]. Additionally, the most promising materials regarding thermal and chemical stability are HKUST-1 [172], MIL-101(Cr) [173], UiO-66/67 [173], MIL-53 [173,174,175], and MIL-160 [176] (but with lower thermal conductivities of 0.065–0.117 W/(m·K) [176]) as highly hydrophilic; even higher water uptake capacity was demonstrated for MIL-101(Cr), aluminum fumarate, and CPO-27(Ni), as shown in Figure 16 [170], while UiO-67, MIL-101(Al), and ZIF-8 are rather hydrophobic [173]. Especially for MILs, there is a huge energy storage capacity of 350–680 Wh/kg at lower desorption temperatures of 70–80 °C [177,178]. Moreover, MOF-related materials have been investigated [179,180], and it is clear that major challenges still exist in the adsorption process for some MOF materials such as poor hydrothermal stability and weak affinity to water leading to lower temperature lift and then limiting large-scale application.

Figure 14. SEM images of activated carbon fiber felt (a) [98], activated carbon foam (b) [99], expanded natural graphite (c) [102], carbon nanotubes (d) [104], activated alumina (e) [107], zeolite 13X (f) [106], attapulgite (g) [113], expanded vermiculite (h) [114], diatomite (i) [115], silica gel (j) [126], silica aero-gels (k) [127], MCM-41 (l) [130], SBA-15 (m) [128], AlPO4-5 (n) [155], SAPO-34 (o) [159], VAPO-5 (p) [158], MIL-101 (q) [166], Aluminum fumarate (r) [167], UiO-67 (s) [168], and PCN-33 (t) [169]: Reprinted with permission from Ref. [98] Copyright © 2016 Elsevier Ltd. All rights reserved; with permission from Ref. [99] Copyright © 2022 Elsevier Ltd. All rights reserved. All rights reserved; with permission from Ref. [102] Copyright © 2019 Elsevier Ltd. All rights reserved; with permission from Ref. [104] Copyright © 2005 Elsevier Ltd. All rights reserved; with permission from Ref. [107] Copyright © 2016 Elsevier Ltd. All rights reserved; with permission from Ref. [106] Copyright © 2019 Elsevier B.V. All rights reserved; with permission from Ref. [113] Copyright © 2021 Elsevier B.V. All rights reserved; with permission from Ref. [114] Copyright © 2016 Elsevier Ltd. All rights reserved; with permission from Ref. [115] Copyright © 2022 Elsevier Ltd. All rights reserved; with permission from Ref. [126] Copyright © 2021 Elsevier Ltd. All rights reserved.; with permission from Ref. [127] Copyright © 2022 Elsevier B.V. All rights reserved; with permission from Ref. [130] Copyright ©2022 The Author(s) All rights reserved; with permission from Ref. [128] Copyright © 2020 Elsevier Masson SAS. All rights reserved; with permission from Ref. [155] Copyright © 2018 Elsevier B.V. All rights reserved; with permission from Ref. [159] Copyright © 2022 Elsevier Ltd. All rights reserved; with permission from Ref. [158] Copyright © 2000 Elsevier Science B.V. All rights reserved; with permission from Ref. [166] Copyright © 2021 Elsevier Ltd. All rights reserved; with permission from Ref. [167] Copyright © 2022 Elsevier Ltd. All rights reserved; with permission from Ref. [168] Copyright © 2022 Elsevier Ltd. All rights reserved; with permission from Ref. [169] Copyright © 2021 Elsevier B.V. All rights reserved.

Figure 14. SEM images of activated carbon fiber felt (a) [98], activated carbon foam (b) [99], expanded natural graphite (c) [102], carbon nanotubes (d) [104], activated alumina (e) [107], zeolite 13X (f) [106], attapulgite (g) [113], expanded vermiculite (h) [114], diatomite (i) [115], silica gel (j) [126], silica aero-gels (k) [127], MCM-41 (l) [130], SBA-15 (m) [128], AlPO4-5 (n) [155], SAPO-34 (o) [159], VAPO-5 (p) [158], MIL-101 (q) [166], Aluminum fumarate (r) [167], UiO-67 (s) [168], and PCN-33 (t) [169]: Reprinted with permission from Ref. [98] Copyright © 2016 Elsevier Ltd. All rights reserved; with permission from Ref. [99] Copyright © 2022 Elsevier Ltd. All rights reserved. All rights reserved; with permission from Ref. [102] Copyright © 2019 Elsevier Ltd. All rights reserved; with permission from Ref. [104] Copyright © 2005 Elsevier Ltd. All rights reserved; with permission from Ref. [107] Copyright © 2016 Elsevier Ltd. All rights reserved; with permission from Ref. [106] Copyright © 2019 Elsevier B.V. All rights reserved; with permission from Ref. [113] Copyright © 2021 Elsevier B.V. All rights reserved; with permission from Ref. [114] Copyright © 2016 Elsevier Ltd. All rights reserved; with permission from Ref. [115] Copyright © 2022 Elsevier Ltd. All rights reserved; with permission from Ref. [126] Copyright © 2021 Elsevier Ltd. All rights reserved.; with permission from Ref. [127] Copyright © 2022 Elsevier B.V. All rights reserved; with permission from Ref. [130] Copyright ©2022 The Author(s) All rights reserved; with permission from Ref. [128] Copyright © 2020 Elsevier Masson SAS. All rights reserved; with permission from Ref. [155] Copyright © 2018 Elsevier B.V. All rights reserved; with permission from Ref. [159] Copyright © 2022 Elsevier Ltd. All rights reserved; with permission from Ref. [158] Copyright © 2000 Elsevier Science B.V. All rights reserved; with permission from Ref. [166] Copyright © 2021 Elsevier Ltd. All rights reserved; with permission from Ref. [167] Copyright © 2022 Elsevier Ltd. All rights reserved; with permission from Ref. [168] Copyright © 2022 Elsevier Ltd. All rights reserved; with permission from Ref. [169] Copyright © 2021 Elsevier B.V. All rights reserved.

Figure 15. Polyhedral representation of MOFs. The cavity space is indicated by yellow van der Waals spheres (Fe, green; Cr, purple; Zr, blue; Ti, orange; Al, pink; C, gray; N, blue; O, red; and H, white) [170,171]: Reprinted with permission from Ref. [170] Copyright © 2016 Elsevier B.V. All rights reserved; with permission from Ref. [171] Copyright ©2017 The Royal Society of Chemistry All rights reserved.

Figure 15. Polyhedral representation of MOFs. The cavity space is indicated by yellow van der Waals spheres (Fe, green; Cr, purple; Zr, blue; Ti, orange; Al, pink; C, gray; N, blue; O, red; and H, white) [170,171]: Reprinted with permission from Ref. [170] Copyright © 2016 Elsevier B.V. All rights reserved; with permission from Ref. [171] Copyright ©2017 The Royal Society of Chemistry All rights reserved.

Figure 16. Adsorption capacity of various host matrices at specified conditions [60,110,124,134,150,151,170,181,182].

Figure 16. Adsorption capacity of various host matrices at specified conditions [60,110,124,134,150,151,170,181,182].

Microporous or macroporous silicon foams are well interconnected, creating a continuity network and effective vapor permeability to allow for vapor flow into or out of pores for both low and high salt contents during the adsorption and desorption phases [164]. Good elasticity and flexibility of this foam can improve the mechanical stability and cyclicity, especially for salt hydrate volume expansion during the adsorption process [164,183,184].

Table 6. Basic characteristics of potential porous matrices.

Class	Porous Matrix	Physical Properties	Basic Characteristics
Mesoporous Metal-aluminophosphates (AlPOs, SAPOs, VAPOs, TAPOs, etc.)	AlPO4-5	Pore volume: 0.1167 cm3/g [155] Specific surface area: 325 m2/g [155] Average pore diameter: 2.6–9.9 nm [155]	highly stable structure [155] Regenerability [155] High surface area [185] Tunable surface features [186]
	SAPO-34	Pore volume: 0.26 cm3/g [185] Specific surface area: 477 m2/g [185] Adsorption capacity:0.31 g·g−1 [181]
	VAPO-5	Pore volume: 0.101cm3/g [158] Specific surface area: 305 m2/g [158] Adsorption capacity: 0.14 g·g−1 [158]
	TAPO-5	Pore volume: 0.111 cm3/g [187] Specific surface area: 305 m2/g [187]
Mesoporous Metal–organic frameworks (MOFs) (MILs, UiOs, PCNs, DUTs, BUTs, NUs, ZIFs, etc.)	HKUST-1	Pore volume: 0.75 cm3/g [173] specific surface area: 1568.5 m2/g [173] Average pore diameter: 5.8 nm [173] Thermal conductivity: 0.27 W/(m·K) [188]	High stability [189] High adsorption capacity [176] Ordered pore structure [176,190] Controllable pore size [190]
	MIL-101(Cr)	Pore volume: 1.6 cm3/g [191] specific surface area: 2327.9 m2/g [191] Adsorption capacity: 1.47 g·g−1 [170]
	MIL-100(Fe)	Pore volume: 0.82 cm3/g [192] specific surface area: 1549 m2/g [192]
	CPO-27(Ni)	Pore volume: 0.5 cm3/g [170] specific surface area: 1225.7 m2/g [170] Average pore diameter: 12.1 nm [170] Crystal density:1200 kg/m3 [170] Adsorption capacity: 0.47 g·g−1 [170]
	Aluminum fumarate	Pore volume: 0.436 cm3/g [170] specific surface area: 1211.5 m2/g [170] Crystal density:1087 kg/m3 [170] Adsorption capacity: 0.53 g·g−1 [170]
	DUT-4	Pore volume: 0.79 cm3/g [192] specific surface area: 1360 m2/g [192]
	UiO-67	Pore volume: 2.02 cm3/g [193] specific surface area: 1790 m2/g [193] Average pore diameter: 1.0–2.8 nm [193]
	ZIF-8	Pore volume: 0.64 cm3/g [192] specific surface area: 1255 m2/g [192]
	PCN-223	Pore volume: 0.87 cm3/g [194] specific surface area: 1884.9 m2/g [194] Average pore diameter: 1.85 nm [194]
Microporous/Macroporous foam	Microcellular (Silicon foam)	Average pore diameter: 1.82 nm [195] Bulk density: 975 kg/m3 [195]	High salt loading capacity [195] High porosity [195] Effective vapor permeability [184] Good elasticity and flexibility [184]
	Macrocellular (Silicon foam)	Average pore diameter: 1.18 mm [183] Bulk density: 969 kg/m3 [183]

Table 6. Basic characteristics of potential porous matrices.

Class	Porous Matrix	Physical Properties	Basic Characteristics
Mesoporous Metal-aluminophosphates (AlPOs, SAPOs, VAPOs, TAPOs, etc.)	AlPO4-5	Pore volume: 0.1167 cm3/g [155] Specific surface area: 325 m2/g [155] Average pore diameter: 2.6–9.9 nm [155]	highly stable structure [155] Regenerability [155] High surface area [185] Tunable surface features [186]
	SAPO-34	Pore volume: 0.26 cm3/g [185] Specific surface area: 477 m2/g [185] Adsorption capacity:0.31 g·g−1 [181]
	VAPO-5	Pore volume: 0.101cm3/g [158] Specific surface area: 305 m2/g [158] Adsorption capacity: 0.14 g·g−1 [158]
	TAPO-5	Pore volume: 0.111 cm3/g [187] Specific surface area: 305 m2/g [187]
Mesoporous Metal–organic frameworks (MOFs) (MILs, UiOs, PCNs, DUTs, BUTs, NUs, ZIFs, etc.)	HKUST-1	Pore volume: 0.75 cm3/g [173] specific surface area: 1568.5 m2/g [173] Average pore diameter: 5.8 nm [173] Thermal conductivity: 0.27 W/(m·K) [188]	High stability [189] High adsorption capacity [176] Ordered pore structure [176,190] Controllable pore size [190]
	MIL-101(Cr)	Pore volume: 1.6 cm3/g [191] specific surface area: 2327.9 m2/g [191] Adsorption capacity: 1.47 g·g−1 [170]
	MIL-100(Fe)	Pore volume: 0.82 cm3/g [192] specific surface area: 1549 m2/g [192]
	CPO-27(Ni)	Pore volume: 0.5 cm3/g [170] specific surface area: 1225.7 m2/g [170] Average pore diameter: 12.1 nm [170] Crystal density:1200 kg/m3 [170] Adsorption capacity: 0.47 g·g−1 [170]
	Aluminum fumarate	Pore volume: 0.436 cm3/g [170] specific surface area: 1211.5 m2/g [170] Crystal density:1087 kg/m3 [170] Adsorption capacity: 0.53 g·g−1 [170]
	DUT-4	Pore volume: 0.79 cm3/g [192] specific surface area: 1360 m2/g [192]
	UiO-67	Pore volume: 2.02 cm3/g [193] specific surface area: 1790 m2/g [193] Average pore diameter: 1.0–2.8 nm [193]
	ZIF-8	Pore volume: 0.64 cm3/g [192] specific surface area: 1255 m2/g [192]
	PCN-223	Pore volume: 0.87 cm3/g [194] specific surface area: 1884.9 m2/g [194] Average pore diameter: 1.85 nm [194]
Microporous/Macroporous foam	Microcellular (Silicon foam)	Average pore diameter: 1.82 nm [195] Bulk density: 975 kg/m3 [195]	High salt loading capacity [195] High porosity [195] Effective vapor permeability [184] Good elasticity and flexibility [184]
	Macrocellular (Silicon foam)	Average pore diameter: 1.18 mm [183] Bulk density: 969 kg/m3 [183]

Figure 16 summarizes the water uptake capacity of host matrices under specified conditions. Compared with other materials, MIL-101(Cr), silica gel, and silica aerogels can adsorb larger amounts of water vapor, especially MIL-101(Cr), which is more suitable for hydrolysis salts, such as CaCl₂, while pure salts with high deliquescence relative humidity (MgSO₄, and MgCl₂) are more suitable for silica gel and silica aerogels to improve adsorption performance. For zeolites, other MOF, and SAPO materials, their reaction conditions and the kinetic properties of salts can be considered comprehensively to form composite materials.

2.3. Improvement of Adsorption/Desorption Performance on the Composites

Recently, composite salt hydrate materials have attracted significant attention and have the potential to address some of the challenges related to the application of salt hydrates. This section reviews the work on composite salt hydrates (Figure 17). The preparation method and salt content of the composites, host matrix types and their microstructures (pore volume/diameter), and structure design of the device and operating conditions, among others, have significant effects on the adsorption/desorption performance of the composites. A suitable preparation method is the first step to improve the performance. A preference for host matrices must be given to composites owing to their larger porosity, further contributing to better mass transfer by enlarging the contact area between water vapor and salt. Different salt contents have a significant effect on the reaction behavior in the same host matrix. In general, a higher salt content of composites contributes to higher water uptake and energy density. However, leakage of salt from the composite structure is undesirable during adsorption for practical applications of composite sorbents. This phenomenon can occur when the volume of the salt exceeds the pore volume of the host matrices during thermal cycling, leading to performance degradation of the sorbents. Therefore, optimizing the salt content is crucial under the given host matrix and experimental conditions. Table 7 gives a summary of the literature data on the salt content, adsorption capacity, and energy density of some composite salt hydrates.

2.3.1. Preparation Method of the Composite Materials

The suitable preparation method mainly depended on the host matrix type and specific functional structure requirements. Wet impregnation is commonly used for preparation of most salt hydrates, but for MOF-based composite materials, the vacuum impregnation and encapsulation method by using spray-drying was more suitable to obtain excellent mechanical properties, hydrothermal stability, and adsorption rate. The foam synthesis method is mainly aimed at macroporous silicone foam matrix loading salt hydrate, which was beneficial for the better dispersion of salt and combining the matrix with salt.

A suitable preparation method was selected according to the host matrix type and functional structure requirements. The preparation methods of salt hydrate composite materials are commonly used, as shown in Figure 18: wet impregnation [110], dry impregnation, vacuum impregnation, encapsulation (spray-drying approach), foam synthesis [183], and straightforward mixing [105]. Different preparation methods have a significant influence on the adsorption and desorption properties of materials; thus, the appropriate preparation method must be selected according to the characteristics of the host matrix. Detailed results are shown in Table 7.

The wet impregnation method is shown in Figure 18a. First, the host matrix was dried in an oven at 200 to 300 °C for approximately 12 h to remove a small amount of water. Second, anhydrous salts were added to deionized water to form a specific salt solution. Subsequently, the host matrix and salt hydrate were mixed evenly with a saturated salt solution with continuous stirring to obtain a wet composite. The composite was then dried at approximately 200–300 °C for approximately 12 h to ensure complete desorption. Finally, the dehydrated composite was shaped under different pressing processes and conditions as needed. For the dry impregnation method, a certain amount of salt solution is dispersed on the matrix rather than directly mixing the salt and host matrix, for example, by dropping by drop. Vacuum impregnation is basically the same as the traditional wet impregnation method but requires the matrix to be vacuumed and then mixed with salt solution. Another encapsulation step was usually produced by the spray-drying method (Figure 18b). Anhydrous salts were added to deionized water to complete the preparation of the saturated salt solution. Second, spraying the saturated salt solution into the frame structure resulted in a core-shell composite material composed of salt (core) and framework (shell), and then the obtained composites were dried in an oven to ensure thorough dehydration. This method is often used for the combination of metal organic frame (MOF) matrices and salt hydrates. A novel method, foam synthesis [184], was proposed. The salt-silicone foam sample preparation process was implemented as follows (Figure 18c): first, a proper polymer with high permeability for water vapor was employed to form a silicone foam as the host matrix for salts; then, the salts were mixed with the siloxane matrix until a homogeneous slurry was formed; finally, the foaming reaction of the composites was performed in an oven under controlled temperature.

A suitable preparation method mainly depends on the host matrix type and specific functional structure requirements. Wet impregnation is commonly used for the preparation of most salt hydrates. However, for MOF-based composite materials, vacuum impregnation, and encapsulation methods via the spray-drying technique are more suitable to obtain excellent mechanical properties, hydrothermal stability, and adsorption rates. The foam synthesis method is mainly aimed at a macroporous silicone foam matrix loading salt hydrate, which is beneficial for better dispersion of salt, and combining the matrix with salt.

2.3.2. Salt Content of the Composite Materials

The microstructure properties (specific surface area, pore volume, and pore size) of the pores decreased significantly during the preparation process with increasing salt content, and the adsorption and desorption properties were greatly affected. Consequently, an appropriate salt content should be identified to obtain excellent composites.

Aiming to maximize the salt content and optimize mass transport, some investigations have been proposed. The adsorption performance improved owing to the continuous porous structure preserving the vapor diffusion flow [196] and the good interfacial interaction between salt and matrix [197] when employing cellular silicone foam supporting salts to form the composites. Interestingly, a smaller pore size distribution region was found for the silicone foam-based composites compared with the silica gel-based composite owing to the better dispersion of salt molecules on pore walls for the former [198,199]. Furthermore, the silicone foam matrix may be effective for hygroscopic salts, such as LiCl, to limit swelling and agglomeration [196]. However, the sedimentation phenomenon was found during the preparation of the composites mainly because of the hydrophilic nature of the salt and the hydrophobic behavior of the foam leading to an immiscible mixture, which is unfavorable for application in the TCES system [164].

Zeolites have been proven to be an excellent host matrix for MgSO₄ in TCES applications owing to their strong adsorption capacity, higher temperature lift, and other advantages. Zeolite13X-MgSO₄ composites with a salt content of 8.05 wt.% MgSO₄ showed the best performance at a high relative humidity of 80% without extracting salt on the surface of the composites [106], while Hongois et al. [200] proposed an optimum percentage of 15 wt.% salt in composites with a 27% increase in energy density (166 kWh/m³) in comparison with pure zeolite 13X (131 kWh/m³). Rising salt content led to a serious pore-blocking effect, and the specific surface area, pore volume, and size decreased significantly from 336 to 187 m²/g, from 0.28 to 0.19 cm³/g, and from 3.36 to 4.16 nm, respectively. In addition, MgSO₄ molecules tended to deposit in the smaller pore diameter during the impregnation process [106]. However, the low energy density of 600 kJ·kg⁻¹ and cyclability may need to be further improved in 8.05 wt.% MgSO₄ composites [106]. Recently, the best configuration of mass ratio (zeolite-13X:CaCl₂:MgSO₄ = 10:54:36) was proposed with 20 good cycle stabilities (Figure 19a–c), an extremely high energy density of 1414.5 kJ·kg⁻¹ at 250 °C, and an adsorption capacity of 0.449 g·g⁻¹ [110]. This type of composite is preferred in medium- and low-temperature TCES systems. After a compromise between the largest amounts of salt and the reaction heat, the activated carbon-MgSO₄ (30 wt.% MgSO₄) composite in the latest research [201] was proven to be better for completing eight cycles of adsorption/desorption stably (Figure 19d) and had a higher reaction enthalpy of 1324 kJ·kg⁻¹ during adsorption. A new composite material hydroxyapatite-MgSO₄ containing two salt contents of 5 wt.% and 20 wt.% was prepared and characterized, and the results indicated that the porous matrix hydroxyapatite greatly improved the slow kinetics of MgSO₄ and considerably reduced the reaction time. The composite with 20 wt.% MgSO₄ achieved cyclability and stability for 20 adsorption/desorption cycles (Figure 19j–l) and the best water uptake capacity. The kinetic model equation to determine the rate-controlling adsorption mechanism was obtained and, thus, provided a reference for reactor design and system operation [150]. Diatomite, with its excellent hydrophilicity [115], allowed for a higher salt content of MgSO₄ of 60 wt.% to load, and an expected energy density (772.9 kJ·kg⁻¹) and better water adsorption capacity (0.37 g·g⁻¹) were proposed [202]. Some investigations have indicated that zeolite-MgCl₂ has better thermal performance than zeolite-MgSO₄ [203,204]. Additionally, Xu et al. [205] found the energy density of zeolite 13X-MgCl₂ to be as high as 1368 kJ·kg⁻¹, which is 2.26 times higher than that of pure zeolite 13X when the salt content is 22 wt.%. A novel composite employed as much as 97.3 wt.% MgCl₂ loading into a reduced graphene oxide aerogel (rGOA), resulting in a maximum water uptake of up to 1.16 g·g⁻¹ (Figure 20c), and only 5% was lost after five cycles compared with the initial energy storage capacity [206] (Figure 19e,f).

Zhang et al. [108] proposed that an activated alumina-LiCl composite with a maximum salt content of 14.68% exhibited the highest adsorption capacity of 0.41 g·g⁻¹ and energy density of 318.3 kWh/m3 (1041.5 kJ·kg⁻¹) despite the decrease in the microstructure parameters of the pores; then, they prepared four expanded vermiculite-based composite materials containing 10–40 wt.% SrBr₂ and compared their adsorption capacity and reaction kinetics [60]. Furthermore, they found that the composite with 40 wt.% SrBr₂ gave a better adsorption performance with a water uptake of 0.53 g·g⁻¹, as presented in Figure 20a, and related work on expanded vermiculite-SrBr₂ with 45 wt.% salt content reached 0.6 g·g⁻¹ at 30 °C and 60% RH accompanied by a desorption reaction enthalpy as high as 644.9 kJ·kg⁻¹ [215]. However, a lower energy density of 105.36 kWh/m³ in the composites was obtained due to the extremely low bulk density of expanded vermiculite (127.74 kg/m³), an energy density of only 2.7 kWh/m³, and the serious mass transfer barrier, swelling, and agglomeration [60]. The expanded vermiculite allowed for more K₂CO₃ loading, and the composites with 69 wt.% K₂CO₃ exhibited good hydrothermal stability for over 47 cycles (Figure 19g) and adsorbed 0.4–1.5 g·g⁻¹ H₂O (at 30 °C to 50 °C, 22–117 mbar) coupled with an energy density of 0.9 GJ/m³ (250 kWh/m³) [208]. Courbon et al. [216] prepared a silica gel-SrBr₂ composite with a high salt content of 58 wt.%, and its kinetic rate was faster than pure owing to the smaller particle size of the former (approximately 180 μm and 400 μm for the composites and pure SrBr₂, respectively) and good stability over 14 cycles (Figure 19m). A maximum specific thermal power of 200 W/kg was presented when placed into the reactors.

The salt content varies from approximately 30 to 60 wt.% for CaCl₂-based composites, and the amount of salt depends on the host matrix and reaction conditions. Courbon et al. [211] used CaCl₂ (40–43 wt.%) encapsulated into silica gel, in which the composite with 43 wt.% salt content showed a high energy density of 211 kWh/m³ and 350 kWh/m³ at the packed bed and material scale, respectively, and excellent stability for 10 cycles (Figure 19n). However, the thermal conductivity of the composites [217] (from 0.13 to 0.16 W/(m·K); Figure 21a) remains much lower than that of a simple combination of silica gel and salt (0.6 W/(m·K)), probably due to poor physical contact [218]. Another similar investigation [160] on silica gel-CaCl₂ composites with an optimized solution of 30 wt.% CaCl₂ led to a high energy storage efficiency of 78%. The adsorption capacity and adsorption rate of the adsorbent increased significantly with increasing CaCl₂ content. Thus, the equilibrium water adsorbed can reach up to 0.73 g·g⁻¹ for the composites with the highest salt content (40 wt.% CaCl₂). Recently, expanded graphite-based CaCl₂ composites with different salt contents in the range from 23.8 to 57.8 wt.% were proposed, and their adsorption capacity and multistep desorption performance were identified. The highest energy density of 1637.6 kJ·kg⁻¹ for the composite (48.1 wt.% CaCl₂) with an adsorption capacity up to 0.79 g·g⁻¹ was obtained, and deliquescence within the composite does not cause solution leakage when the water uptake of the composite is less than 1.0 g·g⁻¹. Interestingly, the deliquescence of salt contributed to improving energy storage performance [135].

Metal–organic frameworks (MOFs), as novel porous materials, have been successfully incorporated into salt hydrates owing to their high specific surface area and stable frame structure, and these excellent properties exceed those of conventional porous materials in water adsorption capacity. A relatively large amount of CaCl₂ with 58 wt.% was impregnated into aluminum fumarate (MOFs) to form the composites, and the best energy storage performance and adsorption capacity (1840 kJ·kg⁻¹ and 0.68 g·g⁻¹, respectively) allowed for the development of a TCES system, especially when a limited adsorption level is maintained (CaCl₂·xH₂O, x ≤ 4) to avoid deliquescence [219]. Furthermore, adding mesoporous alumina to the composite material mentioned above [217] offers favorable channels for water vapor diffusivity, which favors more quantities of salts. An enhancement in the energy density to 1938 kJ·kg⁻¹ is suggested for the highest salt content of 61% of the composites. The adsorption capacity is the same as that in the literature [219], along with the adsorption level of no more than 4.4 crystalline waters for CaCl₂ to limit the precipitation of salt [220]. A series of UiO (UiO-66) and MIL(MIL-100/101/160) exhibited good hydrothermal stability; see, for example, UiO-66-CaCl₂ (53 wt.% for CaCl₂) with a maximum water uptake of 0.6 g·g⁻¹ [221] and MIL-101-CaCl₂ (46.7 wt.% for CaCl₂ [222] and 62 wt.% for CaCl₂ [218]) with high water uptake of 0.47 g·g⁻¹ and 0.75 g·g⁻¹ and energy densities of 1118 kJ·kg⁻¹ and 1746 kJ·kg⁻¹, respectively. MIL-101-SrBr₂ (63 wt.% for SrBr₂) with an energy density of 233 kWh/m³ [223]. More detailed information is shown in Table 7.

In general, impregnation with a higher salt content is more conducive to the adsorption of more water vapor and further produces and stores larger amounts of energy. However, the microstructure properties (specific surface area, pore volume, and pore size) of the pores decreased significantly during the preparation process with increasing salt content and had a great influence on the adsorption and desorption performance. Consequently, a compromise must be reached between salt content and performance improvement.

2.3.3. Microstructure of the Host Matrix

According to Fisher et al. [224], the adsorption mechanisms of the pure salt impregnated in the host matrix can be driven by two successive processes: the water vapor reacts first with readily accessible salts deposited on the surface of the porous matrix, followed by the reaction of most salts found in the porous structure of the host matrix, which are more difficult to access for water vapor. Therefore, the pore size and other microstructure properties of the host matrices have a significant impact on the adsorption/desorption performance of the composites.

Xu et al. [225] studied the adsorption behavior of different zeolite composites containing various contents of MgSO₄ hydrate (3A-MgSO₄, 4A-MgSO₄, and 13X-MgSO₄ with 14.16, 10.17, and 9.27 wt.% salt, respectively) at 25 °C and 65% RH. The 13X-MgSO₄ composite containing the lowest MgSO₄ concentration exhibited the best water uptake of 0.21 g·g⁻¹ compared with the other composites. The opposite adsorption capacity and salt loading may be due to the microstructural variation in zeolites during the preparation process. Meanwhile, the adsorption performance decreased significantly once the adsorption temperature exceeded 50 °C. Another investigation [226] on 13X-MgSO₄ zeolite and activated alumina-MgSO₄ composites with a salt mass concentration of 15% indicated that the impregnation of MgSO₄ could effectively alleviate the leakage issue. The energy densities of zeolite 13X-MgSO₄ and activated alumina-MgSO4 are 123.4 kWh/m³ and 82.6 kWh/m³, respectively. As shown in Figure 20d,e, the water uptake amounts and rates of zeolite 13X sorbents (0.165 g·g⁻¹) are much higher than those of activated alumina sorbents (0.1 g·g⁻¹). According to Posern et al. [212], the pore size of the porous matrix has a significant effect on the adsorption characteristics of MgSO₄, especially for large pore sizes, and the existence of a kinetic barrier was consistent with the aforementioned results [225]. The activated alumina (5.53 nm [108]) has a larger pore size, embedding more MgSO4 molecules compared with zeolite 13X (2.08 nm [145]), but the latter has better performance.

Nevertheless, the smaller particle size also seems to be a significant obstacle to the improvement of the salt hydrate performance [212,214]. In contrast, large pores are more favorable for salt dispersion and accessibility for water vapor in the pores of the composites to improve the energy storage performance [132,214,227]. Zheng et al. [214] investigated the adsorption kinetics of various salts, including LiCl, LiBr, and CaCl₂, incorporated into silica gels with different pore sizes (2–3 nm (SGA), 7–8 nm (SGB), 9–10 nm (SGC)), employing the Dubinin-Astakhov (D-A) equation and the linear driving force (LDF) model. They found that silica gel-based composites with smaller pore sizes (2–3 nm) exhibited lower adsorption capacities and reaction rates due to a larger number of salt-clogged pores limiting passage for water vapor. In addition, silica gel-LiCl showed the best water uptake performance of 0.36 g·g⁻¹, followed by CaCl₂ and LiBr (Figure 20g–i). A similar trend was presented in the literature [226], in which a smaller pore size (2–3 nm) of the silica gel in the composites led to pore blockage and better energy storage capacity and efficiency (128.3 W/kg, and 27%, respectively) for the composites with pore sizes of 8–10 nm. By comparing the adsorption kinetics of three types of composites on CaCl₂-SBA-15, CaCl₂-KSK (silica gel), and CaCl₂-MCM-41, the higher adsorption capacity of 0.47 g·g⁻¹ belongs to CaCl₂-SBA-15, which can be attributed to the largest average pore diameter [132]. Additionally, the CaCl₂-SBA-15 composite modified by mesoporous silica [209,228] with a larger pore size of SBA-15 (8.2 nm [209], and 11.8 nm [209]) attained a better adsorption rate (7 ± 10−5/s [209]) and desorption performance (371 ± 6 kJ/mol [209]) for five stable cycles than that with a pore size of 7.2 nm for SBA-15 (Figure 19h,i). In particular, more salt tightly bound to water molecules in smaller pores leads to an increase in desorption temperature [228]. Whiting et al. [203] proved that zeolite Na-Y can impregnate the highest amount of MgSO₄ solution (15 wt.% MgSO₄) owing to its larger specific surface area (780 m²/g) and pore volume (0.32 cm³/g), therefore achieving the highest reaction enthalpy of 1090 kJ·kg⁻¹ compared with those of zeolite Na-X and H-Y.

The kinetics of water uptake on two composite materials containing 33.7 and 28 wt.% CaCl₂ in silica gel and activated alumina were evaluated by Dawoud et al. [206], who observed that an increased diffusion resistance to water resulting from salt impregnation led to reduced kinetics of water adsorption, and the silica gel-CaCl₂ composite gave faster kinetics of water adsorption than the activated alumina-CaCl₂ composite. These findings are consistent with the results reported in the literature [213], where 15 wt.% CaCl₂ was the maximum salt-loaded composite prepared on silica gel to avoid blocking the diffusion of H₂O molecules through the pores of the support because of agglomeration on the surface of the host matrix. The composite of silica gel-CaCl₂(SCa15) had the optimal performance with the highest adsorption capacity and energy density of 0.27 g·g⁻¹ (Figure 20f) and 746 kJ·kg⁻¹, respectively, compared with alumina (ACa15) and bentonite (BCa15). Notably, the active carbon foam (ACF) appeared to be a much better matrix option for CaCl₂ in terms of adsorption properties and salt impregnation performance. In Figure 20b, ACF30 (30 wt.% CaCl₂) has the best adsorption performance with a water uptake of 1.7 g·g⁻¹, which is three times more than that of silica gel-CaCl₂ [98]. Composites with different natural mineral porous matrices, including sepiolite, diatomite, and expanded perlite loaded with CaCl₂, were prepared by Wei et al. [118]. Some results indicated that expanded perlite-CaCl₂ processed a larger pore volume with a 5.22 cm³/g benefit for the majority of salt dispersion to adsorb various water than others and, thus, achieved the highest energy density and adsorption capacity of 2166 kJ·kg⁻¹ and 1.2 g·g⁻¹, respectively. Mahon et al. [229] proposed zeolite 13X-MgSO₄ composites employing a novel ion exchange method where each Mg²⁺ ion was used instead of two Na+ ions to bond water molecules within 13X. The results showed better performance after the Mg²⁺ ion exchange process with an increase in energy density of approximately 14% compared with standard nontreated 13X pellets, which can be explained by the higher reaction enthalpy of Mg²⁺ ions bonding more tightly to water.

The pore size of the porous matrix has a significant effect on the adsorption characteristics of MgSO₄, especially for large pore sizes, with a kinetic barrier. That is, the smaller particle size of the matrix may be beneficial for the adsorption performance optimization of MgSO₄. Nevertheless, the smaller particle size also seems to be a significant obstacle to the salts. By contrast, large pores in the matrix are more favorable for salt dispersion and accessibility for water vapor in the pores of composites, such as CaCl₂ and LiCl hydrates, to improve the energy storage performance.

2.3.4. Thermal Conductivity of the Host Matrix

Several studies have examined the use of composites with higher thermal conductivity for improvements in the kinetics and energy storage performance of materials.

An example of this is the use of expanded natural graphite, which has a high thermal conductivity of ~100 W/(m·K) with a controllable porosity [109,230]. The conductivity of the composites gradually increased with increasing EG content, and the addition of ENG not only enhanced the thermal conductivity of the composites by ~84.8% but also improved the adsorption/desorption kinetics [231] (Figure 21b). Yu et al. [232] impregnated LiCl into an active carbon foam (ACF)-silica-expanded natural graphite (ENG) (ALi40:ENG-TSA:SS) composite, with LiCl for water uptake, ACF as a host matrix, silica as a binder to increase mechanical strength, and ENG for improving thermal conductivity. As shown in Figure 21c, the use of ENG improved the thermal conductivity to 2–2.8 W/(m·K), approximately 14 times that of pure ACF and LiCl. However, the use of ENG decreased the water uptake, and the reduced kinetics could be explained by the decreased water transport. Lahmidi and Mauran et al. [233,234] studied SrBr₂ composites containing ENG and observed an increased thermal conductivity and permeability but a significantly decreased energy density. To address this issue, Zhao et al. [27] treated ENG with sulfuric acid to formulate a SrBr₂ composite. Their study showed that the composite with 10 wt.% SrBr₂ had a thermal conductivity of 7.97 W/(m·K) (Figure 21d) without degradation of water uptake, but the permeability deteriorated. Consistent with the previous sulfuric acid treatment, zeolite 13X and activated alumina were selected as the host matrices of MgSO4-ENG to increase the thermal conductivity of the material and improve the heat transfer performance [226]. Salts loading into CNTs avoided agglomeration and benefitted a faster adsorption rate and increased energy storage performance owing to the best thermal conductivity (3000–4000 W/(m·K) [103]) for CNTs and the addition of hygroscopic groups. Additionally, MWCNT-CaCl₂ with a stable structure has been developed successfully [105], and for MWCNT-LiCl composites, a rather high energy density of 1700 kJ·kg⁻¹ was obtained, which appeared to be the priority candidates for the water adsorption cycles [235].

Figure 21. Thermal conductivity of composites: (a) silica gel-CaCl₂ [211], (b) expanded graphite-MgSO₄ [231], (c) (ALi40: ENG-TSA:SS) [232], and (d) ENG-SrBr₂ [27]: Reprinted with permission from Ref. [211] Copyright © 2017 Elsevier Ltd. All rights reserved; with permission from Ref. [231] Copyright © 2021 Elsevier Ltd. All rights reserved; with permission from Ref. [232] Copyright © 2015 Elsevier Ltd. All rights reserved; with permission from Ref. [27] Copyright © 2016 Elsevier Ltd. All rights reserved.

Figure 21. Thermal conductivity of composites: (a) silica gel-CaCl₂ [211], (b) expanded graphite-MgSO₄ [231], (c) (ALi40: ENG-TSA:SS) [232], and (d) ENG-SrBr₂ [27]: Reprinted with permission from Ref. [211] Copyright © 2017 Elsevier Ltd. All rights reserved; with permission from Ref. [231] Copyright © 2021 Elsevier Ltd. All rights reserved; with permission from Ref. [232] Copyright © 2015 Elsevier Ltd. All rights reserved; with permission from Ref. [27] Copyright © 2016 Elsevier Ltd. All rights reserved.

Table 7. Advancements of salt hydrate composite materials for thermochemical energy storage.

Porous Matrixes	Salts	Salt Content (wt %)	Preparation Method	Adsorption Conditions	Desorption Conditions	Adsorption Capacity (g·g−1)	Energy Density	Reference
Mesoporous silicone foams	MgSO4	50	Wet impregnation	30 °C, 25 mbar	150 °C	-	-	[198]
Macroporous silicone foams	MgSO4	60	Foam synesis	-	-	-	-	[199]
Expanded graphite	MgSO4	60	Wet impregnation	25 °C, 85% RH	300 °C	-	576.4 kJ·kg−1	[231]
Zeolite	MgSO4	15	Wet impregnation	30 °C, 80% RH	150 °C	-	166 kWh/m3	[200]
Activated Carbon	MgSO4	30	Wet impregnation	30 °C, 60% RH	300 °C	-	1324 kJ·kg−1	[201]
Diatomite	MgSO4	60	Wet impregnation	25 °C, 85% RH	150 °C	0.37	772.9 kJ·kg−1	[202]
Aerogels	MgSO4	59.1	Vacuum impregnation	~20 to 25 °C, ~100% RH	--	-	-	[236]
Zeolite Na-Y	MgSO4	15	Wet impregnation	20 °C, 13 mbar	150 °C	-	1090 kJ·kg−1	[203]
Zeolite H-Y							867 kJ·kg−1
Zeolite Na-X							731 kJ·kg−1
Zeolite X-Y							507 kJ·kg−1
Zeolite 13X	MgSO4	9.27	Wet impregnation	25 °C, 65% RH	300 °C	0.26	-	[225]
Zeolite 3A		14.16				0.16
Zeolite 4A		10.17				0.18
Activated alumina	MgSO4	20.9	Wet impregnation	30 °C, 60% RH	250 °C	0.126	82.6 kWh/m3	[226]
Zeolite 13X		15			150 °C	0.205	123.4 kWh/m3
Zeolite 13X	MgSO4	12.9	Wet impregnation	20 °C, 56% RH	150 °C	-	479 kJ·kg−1	[229]
Zeolite Y							615 kJ·kg−1
Zeolite Na-X	MgSO4	15	Wet impregnation	30 °C, 85% RH	110 °C	-	270 kWh/m3	[237]
	MgCl2	15
Zeolite Na-Y	MgSO4	35	Wet impregnation	20 °C, 56% RH	150 °C	-	791 kJ·kg−1	[238]
Zeolite 13X	MgCl2:MgSO4 (50:50)	15	Wet impregnation	-	150 °C	-	400 kJ·kg−1	[239]
Reduced graphene oxide aerogel	MgCl2	97.3	Wet impregnation	25 °C, 95% RH	350 °C	1.16	2225.71 kJ·kg−1	[207]
Vermiculite	K2CO3	40	Wet impregnation	25 °C, 7.5 mbar	140 °C	-	-	[224]
	MgCl2	60
Zeolite	MgCl2	22	Wet impregnation	30 °C, 65% RH	200 °C	0.55	1368 kJ·kg−1	[205]
Expanded Vermiculite	K2CO3	69	Wet impregnation	25 °C, 14 mbar	160 °C	1.5	250 kWh/m3	[208]
MIL-100(Fe)	CaCl2	46	Spray-drying (Encapsulation)	30 °C, 12.5 mbar	80 °C	0.39	310 kWh/m3	[222]
MIL-101(Cr)		62				0.17	208 kWh/m3
MIL-127(Fe)		40				0.28	-
UiO-66(Zr)-NH2		43				0.33	-
UiO-125(Ti)-NH2		45				0.39	-
MIL-160(Al)		34				0.37	-
Aluminum fumarate	CaCl2	58	Wet impregnation	25 °C, 30% RH	150 °C	0.68	1840 kJ·kg−1	[171]
Aluminum fumarate-alumina	CaCl2	61	Wet impregnation	25 °C, 30% RH	150 °C	0.68	1938 kJ·kg−1	[219]
UiO-66(Zr) -NH2	CaCl2	38	Spray-drying (Encapsulation)	25 °C, 90% RH	-	0.6	367 kJ·kg−1	[220]
MIL-101(Cr)- SO3H	CaCl2	-	Vacuum Impregnation	30 °C, 32% RH	120 °C	0.6	1274 kJ·kg−1	[221]
Silica gel	LiBr	62.5	Wet impregnation	20 °C, 70% RH	-	~0.3	-	[214]
	LiCl	45.5				0.36
	CaCl2	42.7				~0.35
SBA-15	CaCl2	60	Wet impregnation	20 °C, 30% RH	150 °C	-	~200 kWh/m3	[209]
SBA-15	CaCl2	29.5	Wet impregnation	50 °C, 12–13 mbar	150 °C	0.059	-	[228]
Silica gel	CaCl2	15	Wet impregnation	20 °C, 30% RH	150 °C	0.27	746 kJ·kg−1	[213]
Alumina						0.17	576 kJ·kg−1
bentonite						0.23	719 kJ·kg−1
Vermiculite	CaCl2	86	Wet impregnation	30 °C, 13 mbar	150 °C	-	2000 kJ·kg−1	[240]
Silica gel	CaCl2	43	Wet impregnation	30 °C, 12.5 mbar	80 °C	-	211 kWh/m3	[211]
Silica gel	CaCl2	40	Wet impregnation	30 °C, 12.5 mbar	80 °C	0.14	-	[217]
SBA-15	CaCl2	43	Wet impregnation	50 °C, P/P0 < 0.3	80–100 °C	0.47	-	[216]
MCM-41		42				~0.4	-
Silica gel		33.7				~0.35	-
Expanded graphite	CaCl2	48.1	Wet impregnation	25 °C, 80% RH	130 °C	0.79	1637.6 kJ·kg−1	[135]
Wakkanai siliceous shale	CaCl2	22.4	Wet impregnation	25 °C, 22.18 mbar	120 °C	0.7	272 MJ/m3	[241]
Sepiolite	CaCl2	15	Wet impregnation	20 °C, 80% RH	200 °C	0.69	1026 kJ·kg−1	[118]
Diatomite		33				1.31	1520 kJ·kg−1
Expanded perlite		64				2.28	2166 kJ·kg−1
Multi-Wall Carbon NanoTubes	CaCl2	50	Mixing	25 °C, 50% RH	100–110 °C	0.76	-	[105]
Multi-Wall Carbon NanoTubes	LiCl	44	Wet impregnation	35 °C, 9 mbar	75 °C	0.57	1700 kJ·kg−1	[234]
	CaCl2	53				0.18	530 kJ·kg−1
MIL-101(Cr)	CaCl2	25	Wet impregnation	25 °C, 30% RH	-	0.437	-	[242]
	LiCl	15				0.446
Bentonite	CaCl2	20	Wet impregnation	30 °C, 60% RH	150 °C	-	854.5 kJ·kg−1	[243]
	LiCl	20					704.2 kJ·kg−1
	SrCl2	20					778.6 kJ·kg−1
Silica gel	SrBr2	58	Wet impregnation	30 °C, 12.5 mbar	80 °C	-	233 kWh/m3	[210]
MIL-101(Cr)	SrBr2	63	Encapsulation	30 °C, 1.27 kPa	30 °C	0.4	233 kWh/m3	[223]
Vermiculite	SrBr2	45	Wet impregnation	30 °C, 60% RH	300 °C	0.6	644.9 kJ·kg−1	[215]
Macroporous silicone foams	SrBr2	40	Foam synthesis	-	300 °C	-	-	[183]
Vermiculite	SrBr2	59	Wet impregnation	30 °C, 25.4mbar	100 °C	0.53	1656 kJ·kg−1	[60]
Silica gel	LiCl	30	Wet impregnation	32 °C, 23.6 mbar	-	0.74	-	[244]
Expanded Vermiculite		45		35 °C, 31.98 mbar	-	2.4	-
Expanded graphite	LiCl	40	Wet impregnation	30 °C, 60% RH	180 °C	0.527	-	[232]
Vermiculite	LiCl	59	Dry impregnation	35 °C, 12.3 mbar	75 °C	0.6	1800 kJ·kg−1	[245]
Vermiculite	LiCl	52	Wet Impregnation	35 °C	75–85 °C	0.75	2150 kJ·kg−1	[246]
Aluminum (Al2O3)	LiCl	25	Wet impregnation	20 °C, 80% RH	-	0.33	345.58 kWh/m3	[247]
Expanded Vermiculite	LiCl	20	Wet impregnation	35 °C, 60% RH	120 °C	1.41	171.61 kWh/m3	[248]
Expanded graphite	LiCl (LiOH)	60	Wet impregnation	-	100 °C	-	200 kWh/m3	[249]
Attapulgite	LiCl	33	Mixing	30 °C, 7.5–15 mbar	-	0.31–0.44	-	[250]
Silica gel	LiBr	57	Dry impregnation	35 °C	90 °C	~0.39	0.7 GJ/m3	[251]
Silica gel	LiCl	30				~0.45	0.6 GJ/m3
Expanded vermiculite	LiCl	59				0.52	0.4 GJ/m3
Silica gel	LiCl	30	Wet impregnation	30 °C, 16.6 mbar	80 °C	~0.6	163.6 kWh/m3	[252]
Macroporous silicone foam	LiCl	40	Foam synthesis	30 °C 15% RH	-	-	665 kWh/m3	[195]

Table 7. Advancements of salt hydrate composite materials for thermochemical energy storage.

Porous Matrixes	Salts	Salt Content (wt %)	Preparation Method	Adsorption Conditions	Desorption Conditions	Adsorption Capacity (g·g−1)	Energy Density	Reference
Mesoporous silicone foams	MgSO4	50	Wet impregnation	30 °C, 25 mbar	150 °C	-	-	[198]
Macroporous silicone foams	MgSO4	60	Foam synesis	-	-	-	-	[199]
Expanded graphite	MgSO4	60	Wet impregnation	25 °C, 85% RH	300 °C	-	576.4 kJ·kg−1	[231]
Zeolite	MgSO4	15	Wet impregnation	30 °C, 80% RH	150 °C	-	166 kWh/m3	[200]
Activated Carbon	MgSO4	30	Wet impregnation	30 °C, 60% RH	300 °C	-	1324 kJ·kg−1	[201]
Diatomite	MgSO4	60	Wet impregnation	25 °C, 85% RH	150 °C	0.37	772.9 kJ·kg−1	[202]
Aerogels	MgSO4	59.1	Vacuum impregnation	~20 to 25 °C, ~100% RH	--	-	-	[236]
Zeolite Na-Y	MgSO4	15	Wet impregnation	20 °C, 13 mbar	150 °C	-	1090 kJ·kg−1	[203]
Zeolite H-Y							867 kJ·kg−1
Zeolite Na-X							731 kJ·kg−1
Zeolite X-Y							507 kJ·kg−1
Zeolite 13X	MgSO4	9.27	Wet impregnation	25 °C, 65% RH	300 °C	0.26	-	[225]
Zeolite 3A		14.16				0.16
Zeolite 4A		10.17				0.18
Activated alumina	MgSO4	20.9	Wet impregnation	30 °C, 60% RH	250 °C	0.126	82.6 kWh/m3	[226]
Zeolite 13X		15			150 °C	0.205	123.4 kWh/m3
Zeolite 13X	MgSO4	12.9	Wet impregnation	20 °C, 56% RH	150 °C	-	479 kJ·kg−1	[229]
Zeolite Y							615 kJ·kg−1
Zeolite Na-X	MgSO4	15	Wet impregnation	30 °C, 85% RH	110 °C	-	270 kWh/m3	[237]
	MgCl2	15
Zeolite Na-Y	MgSO4	35	Wet impregnation	20 °C, 56% RH	150 °C	-	791 kJ·kg−1	[238]
Zeolite 13X	MgCl2:MgSO4 (50:50)	15	Wet impregnation	-	150 °C	-	400 kJ·kg−1	[239]
Reduced graphene oxide aerogel	MgCl2	97.3	Wet impregnation	25 °C, 95% RH	350 °C	1.16	2225.71 kJ·kg−1	[207]
Vermiculite	K2CO3	40	Wet impregnation	25 °C, 7.5 mbar	140 °C	-	-	[224]
	MgCl2	60
Zeolite	MgCl2	22	Wet impregnation	30 °C, 65% RH	200 °C	0.55	1368 kJ·kg−1	[205]
Expanded Vermiculite	K2CO3	69	Wet impregnation	25 °C, 14 mbar	160 °C	1.5	250 kWh/m3	[208]
MIL-100(Fe)	CaCl2	46	Spray-drying (Encapsulation)	30 °C, 12.5 mbar	80 °C	0.39	310 kWh/m3	[222]
MIL-101(Cr)		62				0.17	208 kWh/m3
MIL-127(Fe)		40				0.28	-
UiO-66(Zr)-NH2		43				0.33	-
UiO-125(Ti)-NH2		45				0.39	-
MIL-160(Al)		34				0.37	-
Aluminum fumarate	CaCl2	58	Wet impregnation	25 °C, 30% RH	150 °C	0.68	1840 kJ·kg−1	[171]
Aluminum fumarate-alumina	CaCl2	61	Wet impregnation	25 °C, 30% RH	150 °C	0.68	1938 kJ·kg−1	[219]
UiO-66(Zr) -NH2	CaCl2	38	Spray-drying (Encapsulation)	25 °C, 90% RH	-	0.6	367 kJ·kg−1	[220]
MIL-101(Cr)- SO3H	CaCl2	-	Vacuum Impregnation	30 °C, 32% RH	120 °C	0.6	1274 kJ·kg−1	[221]
Silica gel	LiBr	62.5	Wet impregnation	20 °C, 70% RH	-	~0.3	-	[214]
	LiCl	45.5				0.36
	CaCl2	42.7				~0.35
SBA-15	CaCl2	60	Wet impregnation	20 °C, 30% RH	150 °C	-	~200 kWh/m3	[209]
SBA-15	CaCl2	29.5	Wet impregnation	50 °C, 12–13 mbar	150 °C	0.059	-	[228]
Silica gel	CaCl2	15	Wet impregnation	20 °C, 30% RH	150 °C	0.27	746 kJ·kg−1	[213]
Alumina						0.17	576 kJ·kg−1
bentonite						0.23	719 kJ·kg−1
Vermiculite	CaCl2	86	Wet impregnation	30 °C, 13 mbar	150 °C	-	2000 kJ·kg−1	[240]
Silica gel	CaCl2	43	Wet impregnation	30 °C, 12.5 mbar	80 °C	-	211 kWh/m3	[211]
Silica gel	CaCl2	40	Wet impregnation	30 °C, 12.5 mbar	80 °C	0.14	-	[217]
SBA-15	CaCl2	43	Wet impregnation	50 °C, P/P0 < 0.3	80–100 °C	0.47	-	[216]
MCM-41		42				~0.4	-
Silica gel		33.7				~0.35	-
Expanded graphite	CaCl2	48.1	Wet impregnation	25 °C, 80% RH	130 °C	0.79	1637.6 kJ·kg−1	[135]
Wakkanai siliceous shale	CaCl2	22.4	Wet impregnation	25 °C, 22.18 mbar	120 °C	0.7	272 MJ/m3	[241]
Sepiolite	CaCl2	15	Wet impregnation	20 °C, 80% RH	200 °C	0.69	1026 kJ·kg−1	[118]
Diatomite		33				1.31	1520 kJ·kg−1
Expanded perlite		64				2.28	2166 kJ·kg−1
Multi-Wall Carbon NanoTubes	CaCl2	50	Mixing	25 °C, 50% RH	100–110 °C	0.76	-	[105]
Multi-Wall Carbon NanoTubes	LiCl	44	Wet impregnation	35 °C, 9 mbar	75 °C	0.57	1700 kJ·kg−1	[234]
	CaCl2	53				0.18	530 kJ·kg−1
MIL-101(Cr)	CaCl2	25	Wet impregnation	25 °C, 30% RH	-	0.437	-	[242]
	LiCl	15				0.446
Bentonite	CaCl2	20	Wet impregnation	30 °C, 60% RH	150 °C	-	854.5 kJ·kg−1	[243]
	LiCl	20					704.2 kJ·kg−1
	SrCl2	20					778.6 kJ·kg−1
Silica gel	SrBr2	58	Wet impregnation	30 °C, 12.5 mbar	80 °C	-	233 kWh/m3	[210]
MIL-101(Cr)	SrBr2	63	Encapsulation	30 °C, 1.27 kPa	30 °C	0.4	233 kWh/m3	[223]
Vermiculite	SrBr2	45	Wet impregnation	30 °C, 60% RH	300 °C	0.6	644.9 kJ·kg−1	[215]
Macroporous silicone foams	SrBr2	40	Foam synthesis	-	300 °C	-	-	[183]
Vermiculite	SrBr2	59	Wet impregnation	30 °C, 25.4mbar	100 °C	0.53	1656 kJ·kg−1	[60]
Silica gel	LiCl	30	Wet impregnation	32 °C, 23.6 mbar	-	0.74	-	[244]
Expanded Vermiculite		45		35 °C, 31.98 mbar	-	2.4	-
Expanded graphite	LiCl	40	Wet impregnation	30 °C, 60% RH	180 °C	0.527	-	[232]
Vermiculite	LiCl	59	Dry impregnation	35 °C, 12.3 mbar	75 °C	0.6	1800 kJ·kg−1	[245]
Vermiculite	LiCl	52	Wet Impregnation	35 °C	75–85 °C	0.75	2150 kJ·kg−1	[246]
Aluminum (Al2O3)	LiCl	25	Wet impregnation	20 °C, 80% RH	-	0.33	345.58 kWh/m3	[247]
Expanded Vermiculite	LiCl	20	Wet impregnation	35 °C, 60% RH	120 °C	1.41	171.61 kWh/m3	[248]
Expanded graphite	LiCl (LiOH)	60	Wet impregnation	-	100 °C	-	200 kWh/m3	[249]
Attapulgite	LiCl	33	Mixing	30 °C, 7.5–15 mbar	-	0.31–0.44	-	[250]
Silica gel	LiBr	57	Dry impregnation	35 °C	90 °C	~0.39	0.7 GJ/m3	[251]
Silica gel	LiCl	30				~0.45	0.6 GJ/m3
Expanded vermiculite	LiCl	59				0.52	0.4 GJ/m3
Silica gel	LiCl	30	Wet impregnation	30 °C, 16.6 mbar	80 °C	~0.6	163.6 kWh/m3	[252]
Macroporous silicone foam	LiCl	40	Foam synthesis	30 °C 15% RH	-	-	665 kWh/m3	[195]

2.3.5. Structure Design of the Device

Numerous studies related to reactor design and operation conditions to improve the poor heat and mass transfer during the adsorption/desorption of salt hydrate have been presented in TCES systems. In general, the laboratory bench is built to obtain relevant data on the real reaction, but the processes inside the reactor are difficult to investigate experimentally because of the limited access inside the bed and the costly laboratory-scale experimental setups. At this point, numerical modeling is also a reliable approach to deeply understand and analyze complex physical and chemical processes [253,254].

Here, the optimum TCES reactor design was considered, as it was generally suitable for a broad range of space heating applications and aimed at high performance with respect to energy density and thermal power. Recently, Hawwash et al. [255] examined two shapes of the reactor (cylindrical and conical), which they investigated numerically at different sizes by using the inlet-to-outlet area ratio (AR) as an evaluation indicator during the desorption process for MgCl₂·6H₂O. Notably, the reactor design has a significant effect on the concentration of products of MgCl₂·6H₂O, and the lower AR has a shorter charging time and higher pressure drop and temperature difference in the reactor. Some work embedded the heat exchanger into the bed to improve the heat and mass transfer performance, and the results showed that optimizing the total plate area can improve the mass transfer, such as the larger plate radius and larger distance between the plates in the fin plate heat exchanger [256]. Applying and developing a three-dimensional numerical model (Figure 22) for the adsorption reaction of K₂CO₃, Kant et al. [257] demonstrated that the geometrical parameters of the reactor seriously affected the reaction rate and heat transport in the bed, and better heat transport was obtained by reducing the size of the honeycomb heat exchanger up to a certain level and improving the reaction rate. An open system was investigated experimentally with a mesoporous honeycomb element based on TCES composite materials by CaCl₂/LiCl filling into Wakkanai siliceous shale (WSS) (Figure 23). Some results provided further proof that the storage element exhibited high energy density, and the adsorption rate was observed to significantly affect the outlet air temperature during the adsorption process. The composite WSS + 9.6 wt.% LiCl showed potential for storing and releasing solar energy with a higher energy density of 180 MJ/m³ and stability when it was regenerated at 80 °C and a higher adsorption rate compared with that of WSS + 22.4 wt.% CaCl₂ [258].

Under partial vacuum conditions, mass transfer limitations within the solid volume were an issue for two reasons: a higher gas concentration (or gas pressure) in the solid phase led to a higher local reaction temperature, and a large pressure drop within the solid bulk phase might slow down the chemical reaction. Thus, the use of heat exchangers in closed TCES systems is required to improve mass transfer and further promote heat transfer. Fopah Lele et al. [259] focused on closed system development with a cylindrical reactor in which a plate tube honeycomb heat exchanger was inserted, and a three-dimensional model consisted of the reactor (Figure 24). A comparison simulation experiment was discussed, and the optimization results indicated an energy storage density of 115 kWh/m³, a storage capacity of 61 kWh, a thermal efficiency of 78%, and a COP of 0.97, which led to some significant recommendations for future prototype development. As shown in Figure 25, Stengler et al. [260] set up a TCES closed system based on SrBr₂·6H₂O, and heat-conducting fins were chosen to enhance the mass and transfer performance inside the reactor. The results revealed a high thermal power of approximately 1.2 kW for charging, 0.38 kWh storage capacity for the module containing 4.7 kg of SrBr₂, and a high energy density of 291 kJ·kg⁻¹ owing to its scalable design for the given operation parameters.

Other optimal design examples included a small-scale closed TCES system containing WSS-40 wt.% LiCl for air cooling [261] with different fin arrangements, e.g., a fin-tube heat exchanger filled with EG-SrBr₂ between fins [27], composite salt hydrate coated on the surface of corrugated fin aluminum heat exchangers [262], CaCl₂ in mesoporous silica gel on a heat exchanger [263], and parallel finned flat tube aluminum heat exchangers coated with vermiculite-LiCl [246]. Recently, a corrugated-shaped heat storage unit with a porous metal bracket embedded (Figure 26) was developed to improve heat transfer between the composite material and the heat fluid transfer. Compared with the storage unit of a straight external channel without a metal bracket inside, the optimum heat storage module saves 34% and 23% in charging and discharging, respectively [264].

Figure 22. Schematic of the geometry of the bed: (a) TCM energy storage system, (b) side view of honeycomb heat exchanger filled with TCM, (c) computational domain [257]: Reprinted with permission from Ref. [257] Copyright © 2021 Elsevier Ltd. All rights reserved.

Figure 22. Schematic of the geometry of the bed: (a) TCM energy storage system, (b) side view of honeycomb heat exchanger filled with TCM, (c) computational domain [257]: Reprinted with permission from Ref. [257] Copyright © 2021 Elsevier Ltd. All rights reserved.

Figure 23. Schematic of the open adsorption TCES system [258]: Reprinted with permission from Ref. [258] Copyright © 2014 Elsevier Ltd. All rights reserved.

Figure 23. Schematic of the open adsorption TCES system [258]: Reprinted with permission from Ref. [258] Copyright © 2014 Elsevier Ltd. All rights reserved.

Figure 24. Reactor geometry model in (a) charging phase and (b) discharging phase for simulations [259]: Reprinted with permission from Ref. [259] Copyright © 2015 Elsevier Ltd. All rights reserved.

Figure 24. Reactor geometry model in (a) charging phase and (b) discharging phase for simulations [259]: Reprinted with permission from Ref. [259] Copyright © 2015 Elsevier Ltd. All rights reserved.

Figure 25. Prototype of the scalable thermochemical storage module: (a) Lab-scale module consisting of two cells I and II, with in total 1.25 m of finned tube filled with 4.7 kg of the reactive material strontium bromide and wrapped with stainless steel filter tissue, (b) cross-section of the extruded aluminum fins prepared to be mounted on the heat transfer fluid (HTF) tube with steel clips, and (c) positions of the temperature sensors in the porous bulk (A–D) and in the HTF (E) at the height level indicated by the yellow arrows [260]: Reprinted with permission from Ref. [260] Copyright © 2020 Elsevier Ltd. All rights reserved.

Figure 25. Prototype of the scalable thermochemical storage module: (a) Lab-scale module consisting of two cells I and II, with in total 1.25 m of finned tube filled with 4.7 kg of the reactive material strontium bromide and wrapped with stainless steel filter tissue, (b) cross-section of the extruded aluminum fins prepared to be mounted on the heat transfer fluid (HTF) tube with steel clips, and (c) positions of the temperature sensors in the porous bulk (A–D) and in the HTF (E) at the height level indicated by the yellow arrows [260]: Reprinted with permission from Ref. [260] Copyright © 2020 Elsevier Ltd. All rights reserved.

Figure 26. (a) Schematic diagram of the heat storage unit; (b) a quarter profile of the reaction unit [264]: Reprinted with permission from Ref. [264] Copyright © 2022 Elsevier Ltd. All rights reserved.

Figure 26. (a) Schematic diagram of the heat storage unit; (b) a quarter profile of the reaction unit [264]: Reprinted with permission from Ref. [264] Copyright © 2022 Elsevier Ltd. All rights reserved.

A lower inlet and outlet area ratio (AR) can shorten the charging time, but a higher pressure drop and temperature difference in the reactor are unfavorable for heat and mass transfer performance. To improve the poor mass transfer problem between the composites and transfer fluid in the closed system under vacuum conditions, scholars have proposed a variety of optimized designs. The novel heat exchanger with a fin-tube, plate tube honeycomb, and corrugated-shaped heat storage unit was proven to be effective in developing advanced TCES systems.

2.3.6. Operating Conditions

To better elucidate the optimization performance under suitable operating conditions in reactors for seasonal energy storage, many researchers have focused on open and closed TCES systems. A moving bed reactor in an open system with falling solid flow and humid air cross-flow was investigated focusing on the influence of operation conditions, such as pressure drop of the air on the reaction behavior, and the reactor performance (power, outlet air temperature) by applying a two-dimensional model (Figure 27). This phenomenon indicates that the pressure drop is mainly located in the hydrated zone in the reactor, and the unreacted solid occurred at the entrance of the reactor because of the pressure drop in the bed [265]. An experimental work on packed bed reactors with MgCl₂·6H₂O stored 17 L of adsorption material with a pressure drop of approximately 220 Pa over the entire system. The pressure drop increased during adsorption, which could be related to the swelling of MgCl₂ upon the uptake of water vapor. Thus, pressure drop was proven to be a focus in designing and optimizing the TCES system [266]. Xu et al. [267] developed a thermal non-equilibrium and pressure-driven adsorption model to improve the heat and mass transfer performance between the porous bed and air during the desorption process of a reactor filled with MgCl₂·6H₂O. The findings showed that the full desorption time decreases with increasing charging mass flow rate, while its decreasing rate gradually decreases. Li et al. [268] designed a novel three-dimensional energy storage reactor with two shaped columnar sorbent reactive beds (Figure 28), and the charging (desorption) and discharging (adsorption) processes under different operating conditions were numerically investigated and optimized. The best findings indicate that the temperature and conversion degree evolutions during charging and discharging exhibit a distinct uniformity in the bed owing to the excellent heat and mass transferability of the materials and obtained an overall thermal coefficient of performance of 80.9% and an exergy coefficient of performance of 27.7%. Walayat et al. [269] investigated the effect of pressure, particle size, and packing arrangement on the heat behavior with a three-dimensional packed bed model (Figure 29) by considering the effective thermal conductivity of packed beds with K₂CO₃ salt hydrates. Better heat transfer of a packed reactor is observed in numerical experiments at higher vapor pressures because of natural convection, while reaching a certain high pressure, the effect is no longer significant.

Several laboratory-scale experiments on the improvement of reactor performance combined with the different operation conditions are as follows. Aydin et al. [114] constructed a modular TCES system consisting of a novel ‘open adsorption pipe’ reactor with an internal perforated diffuser pipe network employing vermiculite-CaCl₂ as the composite material in the reactor, and its thermal performance was investigated under different inlet air humidity conditions. According to the testing results, the optimum system total storage capacity was 25.5 kWh, and the energy density was 290 kWh/m³. Energetic and exergetic cyclic efficiencies varied in the range of 0.61–0.69 and 0.14–0.21 in the cycles. Wang et al. [270] combined a three-dimensional numerical model with experimental (open system) findings to investigate operation conditions, such as temperature, flow rate, and relative humidity, affecting the thermal performance of the reactor (Figure 30). The maximum temperature lift significantly increased by 79.94% and 80.81% for air and water, respectively, when the inlet temperature of air and water increased from 23 to 38 °C. In addition, the optimum total heat uptake increased by 107.44%. Similar to Wang et al., the mass transfer performance was improved by designing an open experimental setup to validate the numerical model under different conditions. Some main results indicated that the average temperature of the outlet air and peak temperature significantly decreased with increasing air flow rate, and the adsorption rate was the opposite [21].

Figure 27. Geometry of the moving bed reactor in the model [265]: Reprinted with permission from Ref. [265] Copyright © 2018 Elsevier Ltd. All rights reserved.

Figure 27. Geometry of the moving bed reactor in the model [265]: Reprinted with permission from Ref. [265] Copyright © 2018 Elsevier Ltd. All rights reserved.

Figure 28. Reactor geometry model in charging and discharging phase for simulations: (a) Reactor geometry model in charging and discharging phase for simulations; (b) structures of reaction bed unit; (c) schematic diagram of hydration reaction model from top view [268]: Reprinted with permission from Ref. [268] Copyright © 2020 Elsevier Ltd. All rights reserved.

Figure 28. Reactor geometry model in charging and discharging phase for simulations: (a) Reactor geometry model in charging and discharging phase for simulations; (b) structures of reaction bed unit; (c) schematic diagram of hydration reaction model from top view [268]: Reprinted with permission from Ref. [268] Copyright © 2020 Elsevier Ltd. All rights reserved.

Figure 29. Simulation setup for the 3-D packed beds [269]: Reprinted with permission from Ref. [269] Copyright © 2021 Elsevier Ltd. All rights reserved.

Figure 29. Simulation setup for the 3-D packed beds [269]: Reprinted with permission from Ref. [269] Copyright © 2021 Elsevier Ltd. All rights reserved.

Figure 30. The illustration of shell-and-tube thermochemical reactor: (a) real reactor, (b) internal structure and flow direction [270]: Reprinted with permission from Ref. [270] Copyright © 2021 Elsevier Ltd. All rights reserved.

Figure 30. The illustration of shell-and-tube thermochemical reactor: (a) real reactor, (b) internal structure and flow direction [270]: Reprinted with permission from Ref. [270] Copyright © 2021 Elsevier Ltd. All rights reserved.

With respect to previous studies on the host matrices, higher thermal conductivity, thermal and hydrothermal stability, and hydrophilic and mechanical strength are requisite for incorporating salt hydrates to achieve the maximum adsorption and desorption performance. In general, the host matrices with strong hydrophilic can provide higher temperature lift during discharging process, such as CNTs (modified by hygroscopic groups), natural mineral materials and most MOFs, silica gel, and so on, but require higher desorption temperatures to improve energy storage performance, especially for zeolites 13X. Therefore, a balance between desorption temperature and temperature lift should be sought considering the combination of the host matrix with hydrated salts. The better thermal and hydrothermal stability for SBA-15, SAPO-5, and MIL-101(Cr) is beneficial for the long-term TCES cycles. All things considered, MIL-101(Cr)(MOFs) are the most promising host matrices for the preparation of the composites along with salt hydrates.

The composite materials analyzed in this section showed very promising results and might be instructive for applications in respect of a reactor. The preparation method, salt content, host matrix type, microstructure, and thermal conductivity are the key factors affecting the adsorption and desorption performance of the composites. The appropriate preparation process was selected mainly according to the host matrix type. Theoretically, the higher the salt concentration, the more favorable the absorption of more water vapor, and adsorption performance is better. In fact, as the salt concentration increased, the pores of the porous carrier were gradually blocked, which made it difficult for water vapor to access. As a consequence, a compromise must be made between salt content and performance improvement. In general, larger pores in the matrix are more favorable for salt dispersion and accessibility for water vapor in the pores of the composites, thus optimizing the energy storage performance. However, on the contrary, for some salts (MgSO₄) embedded in the matrix with larger pores, there was a kinetic obstacle. In addition, higher thermal conductivity with the host matrix was beneficial for improvements in the kinetic and energy storage performance of materials. Especially for the addition of ENG and CNTs (modified by hygroscopic groups,) a faster adsorption rate, as well as increased energy storage performance in the composites, was displayed.

A higher pressure drop provided more resistance and resulted in higher amounts of energy consumption, which was unfavorable for improvement in the performance of the system. Therefore, it should be minimized during the design and operation of the systems by means of a structure optimization design mainly involving a heat exchanger with a fin–tube, plate tube honeycomb, and corrugated-shaped heat storage unit. Suitable operation conditions such as flow rate, vapor pressure, temperature, and relative humidity helped to achieve higher thermal performance.

Table 8. Performance of salt hydrate adsorption materials for thermochemical energy storage [48].

Materials	Energy Density/kJ·kg−1	Theoretical Thermal Efficiency/%	Cost/EUR/MJ
MgSO4·7H2O	1360.8	20.3	0.07
MgCl2·6H2O	1252.8	30.1	0.09
SrCl2·6H2O	281.6	21.8	12.63
SrBr2·6H2O	946.8	33.8	4.13
Al2(SO4)3·18H2O	831.6	19.6	0.21
CaBr2·6H2O	1148.4	28.2	1.12
CaCl2·6H2O	1263.6	19.5	0.10
K2CO3·1.5H2O	579.6	30.0	1.64
KOH·2H2O	522	26	1.56
LaCl3·7H2O	957.6	24.7	2.34
La(NO3)3·6H2O	601.2	22.9	2.94
LiCl·H2O	1029.6	28.3	9.11
LiNO3·3H2O	1346.4	19.3	7.88
Na2S2O3·5H2O	1126.8	20.3	0.15
Zn(NO3)2·6H2O	1422	28.1	0.77

shows an overall comparison among different salt hydrate adsorption materials including energy density, thermal efficiency, and cost. Among the materials, MgSO₄·7H₂O, MgCl₂·6H₂O, and CaCl₂·6H₂O have high energy density and low cost, and SrBr₂·6H₂O has high thermal efficiency, though its cost is high. Generally, several materials mentioned above are preferred in the TCES system.

3. Prospects of Developments

Salt hydrate adsorption thermochemical energy storage materials use off-peak electricity to drive a desorption reaction to store heat, and the loss is nearly zero, which is conducive to the valley electricity consumption. In the process of improving the formation mechanism of the peak and valley electricity price. While improving the formation mechanism of the peak and valley electricity price, the corresponding energy storage system is formed for the off-peak electricity to effectively improve the energy utilization rate. In addition, renewables can also be used as a heat source. When energy is used, chemical energy can be converted into heat, so that renewables can replace traditional natural gas or electric heating, promote the application of green energy and clean heating in buildings, accelerate the diversified development of user-side energy storage, and help to reduce carbon emissions in the building field. It is a critical technical means to achieve “carbon peak” and “carbon neutrality”.

Therefore, further research on composites at both the material and reactor levels is clearly needed. Therefore, the following aspects are proposed:

(1)

Composites can overcome the technical challenges of pure materials and have great potential for the design and optimization of TCES reactors and their system integration. The study of the water transport mechanism in composites is expected to be a breakthrough for mass transfer limitation and regulation of microporous structures. Therefore, the water molecule migration behavior, binding form, and phase transition path on the surface or interior of the composite particles and the correlations between the reaction control factors (diffusion, adsorption, and chemical reaction) and the structural characteristics should be identified to accelerate further development in advanced composites.

(2)

Aiming at different host matrices, the corresponding micromechanical experimental model of the composite interface can be constructed to detect the degree of bonding between salt and porous substrates and the mechanical behavior under an external load to reasonably evaluate the possible problems (change in interfacial shear strength or location of debonding because of temperature, humidity, chemical composition, and other relevant factors) in the preparation and experiment of composite materials. This approach will play a key role in improving the cycle stability of composites and provide a critical reference for improving operating conditions.

(3)

More convenient structural design of packing and material replacement is necessary to improve the operational efficiency of the reactor. In addition, the effect of the composite arrangement, such as the particle size distribution, on the energy storage performance needs to be given more attention to provide opportunities for developing large-scale TCES systems.

4. Conclusions

The role of thermal energy storage (TES) in future net-zero carbon energy systems has been widely recognized. Salt hydrate-based thermochemical energy storage has the potential to meet seasonal storage requirements owing to the proper adoption of KPIs (e.g., high energy density and zero loss during storage). Therefore, in this study, a review covering the technical challenges of pure salts, the key indicators of host matrices, and the methods to improve adsorption and desorption performance on composites are provided. The following conclusions can be drawn:

(1)

Most pure salts, such as SrCl₂·6H₂O, K₂CO₃·1.5H₂O, Na₂S·9H₂O, MgSO₄·7H₂O, MgCl₂·6H₂O, LiCl·H₂O, CaCl₂·6H₂O, and SrBr₂·6H₂O, can achieve complete adsorption in the range of 25 to 45 °C and 15–25 mbar. For CaCl₂·6H₂O, its very low melting temperature (29 °C) makes its use in a thermochemical reactor challenging. MgSO₄·7H₂O, MgCl₂·6H₂O, and SrBr₂·6H₂O can remove most of the water molecules by reducing the heating rate to 0.2–1 K/min without melting. Pure salts loaded into pores to form composites are the best avenue to solve technical barriers, such as the mass transport limitation of water vapor within salt hydrate grains and ineffective heat transport because of the low thermal conductivity of salt hydrates.

(2)

The molecular-level investigation of pure salts, such as hydrogen bond networks affecting the coordination of water molecules during adsorption, provided the basis for water transport in the microstructure. In addition, models, including the ReaxFF force field and molecular dynamic simulations, calculated the crystalline structures of the materials and the diffusion coefficient of water, making them useful tools to gain insight into the adsorption and desorption reaction mechanisms.

(3)

The host matrices with strong hydrophilicity can provide a higher temperature increase during the adsorption process, such as CNTs (modified by hygroscopic groups), natural mineral materials, and most MOFs, such as silica gel, but require a higher desorption temperature to improve the energy storage performance, especially for zeolites 13X. Therefore, a balance between desorption temperature, temperature increase, adsorption capacity, and hydrothermal stability characteristics should be sought. All things considered, MIL-101(Cr)(MOFs) is the most promising host matrix and is especially suitable for hygroscopic salt and beneficial for long-term TCES cycles.

(4)

The appropriate preparation process is selected mainly according to the host matrix type and specific functional structure requirements. The wet impregnation method is highly suitable for most of the salt loaded into the matrix, while encapsulation by using the spray-drying method is better for MOFs.

(5)

The composite materials presented great potential for applications in reactors. Higher salt content clogs pores, while extremely low salt content reduces performance. Consequently, a compromise must be reached between salt content and performance improvement. The MgCl₂/rGOA composite is obtained with the highest salt loading of 97.3 wt.% and an optimal adsorption capacity and energy density of 1.16 g·g⁻¹ and 2225.71 kJ·kg⁻¹, respectively, compared with any other composite.

(6)

Larger pores with an 8–10 nm interior for the matrix are more favorable for salt dispersion and accessibility for water vapor in the pores of the composites, thus optimizing the adsorption and energy storage performance. However, some salts embedded in matrices with larger pores have a kinetic barrier. Notably, the smaller pores (2–3 nm) of MgSO₄-based composites are beneficial for faster reaction kinetics.