Synthesis of Mesoporous γ-Alumina Support for Water Composite Sorbents for Low Temperature Sorption Heat Storage

Abstract

The efficiency of thermochemical heat storage is crucially determined by the performance of the sorbent used, which includes a high sorption capacity and a low regeneration temperature. The thermochemical salt hydrate– γ-alumina composite sorbents are promising materials for this application but lack systematic study of the influence of γ-alumina structural properties on the final storage performance. In this study, mesoporous γ-Al₂O₃ supports were prepared by solvothermal and hydrothermal synthesis containing a block copolymer (F-127) surfactant to design thermochemical CaCl₂ and LiCl composite water sorbents. Altering the solvent in the synthesis has a significant effect on the structural properties of the γ-Al₂O₃ mesostructure, which was monitored by powder XRD, nitrogen physisorption, and SEM. Solvothermal synthesis led to a formation of mesoporous γ-Al₂O₃ with higher specific surface area (213 m²/g) and pore volume (0.542 g/cm³) than hydrothermal synthesis (147 m²/g; 0.414 g/cm³). The highest maximal water sorption capacity (2.87 g/g) and heat storage density (5.17 GJ/m³) was determined for W-46-LiCl containing 15 wt% LiCl for space heating, while the best storage performance in the sense of fast kinetics of sorption, without sorption hysteresis, low desorption temperature, very good cycling stability, and energy storage density of 1.26 GJ/m³ was achieved by W-46-CaCl₂.

1. Introduction

Thermochemical energy storage (TCES), as one of the three technologies of thermal energy storage (TES), can reduce fossil fuel consumption, which still dominates building space heating, by shifting some of the solar thermal energy collected in summer to winter [1]. Thermochemical heat storage uses the reversible chemical reaction (reaction of water and CaCl₂) and/or sorption processes of gases in solids or liquids (water sorption on porous solids). Water sorption heat storage consists of two phases. In the first phase, which is also called charging or desorption, gas or water vapour is desorbed from the material under solar radiation or waste heat. In the second phase or discharging (adsorption), the water vapour is adsorbed on the material with heat release, which can be used for space heating. The efficiency of the sorption technology, based on the alternating (ad)sorption (exothermic phenomenon) and desorption (endothermic phenomenon) of the working fluid on the sorption materials, is determined by the performance of the used sorbent, which should have a high sorption capacity, and consequently a high energy storage density, a low regeneration temperature of 80 to 120 °C, stability in humid conditions and also at temperatures up to 120 °C, and no hysteresis (of the sorption and desorption curves) during the sorption cycles. These requirements can be accomplished by the thermochemical (TCM) composite sorbents [2], formed of various porous supports and hygroscopic salt hydrates (halides, nitrates, sulphates) and are currently the most studied sorbents, especially when water is used as the working fluid. The process of water sorption in these composites can consist of three mechanisms: adsorption of water on the active support, chemical reaction between water and entrapped salt in the pores of the support, and absorption of water in salt solution in the pores [3]. The main advantage of these sorbents is tailoring their sorption properties by changing the structural properties and chemical composition of the support, and by entrapping of the particular hygroscopic salt hydrate nanoparticles in the support.

A brief review of the literature on the composites for thermochemical energy storage has shown that the increase in sorption capacity has mainly occurred by changing the salt hydrate and increasing its amount [4]. Fewer TCES studies have been carried out on the development of the active porous support, which can be involved in the adsorption process to further improve the storage performance in the context of increased sorption capacity by adjusting the chemical composition of the support and creating additional adsorption sites by using different synthesis routes [5,6,7,8]. Porous supports for TCM composites found in the literature are mainly commercial materials such as Zeolites [9], silica gels [10], activated alumina [11], and vermiculite [12], while recently, novel supports such as MOFs [13], polymeric foam [14], alginate [15], and porous carbon structures [16,17,18] have been used.

An important task in the preparation of salt hydrate TCM composites is the selection of the support for the salt inclusion. The support should meet certain requirements such as the ability to provide a stable nano-environment for salt confinement, hydrothermal stability, stability of the pore structure under working conditions, inertness to salt hydrates, and good thermal conductivity, to name a few, to enhance the performance of the composite. In addition, a simple and environmentally friendly synthesis that leads to the preparation of the composite and does not show hysteresis during sorption is of great importance [19].

To date, TCM composite sorbents based on mesoporous alumina (Al₂O₃) [20,21] or activated alumina [11,22,23,24] have been prepared from commercially available aluminas and combined with the following salt hydrates: MgSO₄, LiCl, and CaCl₂. These alumina composites have been investigated for thermochemical heat storage at low temperature (<100 °C). Recently, anodic alumina was prepared by anodizing of aluminium sheet and used for the salt composites [25,26]. In all these studies, only the physicochemical properties of the alumina support, such as specific surface area or pore volume and pore size, were shown, and it was not evident whether crystalline or amorphous alumina was used. Indeed, aluminas exist as amorphous or crystalline phases of various oxides and oxyhydroxides. The crystalline structures, which are mainly composed of γ, δ, ρ and α phases, exhibit different properties. All these structures can be prepared by irreversible thermal transformation of the crystalline aluminium hydroxides and salts or by colloidal gel precipitation. α-Al₂O₃ is the high-temperature end product of the thermal treatment [27]. For example, the γ-Al₂O₃ support is often described as robust, thermally stable, and easy to shape. In addition, it offers acid-base properties with polarity [28], favourable surface properties, such as high specific surface area (100–400 m²/g) and a tendency to disperse the active phase. Together with its low price, alumina is at the top of the list of the used supports in catalysis and adsorption [29]. The thermal and hydrothermal stability strongly depends on the synthetic methods and conditions.

Mesoporous alumina is most commonly obtained by sol-gel synthesis using soft or hard templates and by evaporation-induced self-assembly (EISA). Sol-gel synthesis in the presence of soft templates is one of the most common methods for preparing mesoporous alumina. The soft templates can be anionic (stearic acid, etc.), cationic (CTAB, etc.), nonionic (P123, etc.) surfactants, as well as non-surfactants (ionic liquids) [29]. The alumina precursor, usually aluminium isopropoxide, and the surfactant were dissolved in alcoholic solvent with the minimum amount of water added to promote hydrolysis of the isopropoxide and start the condensation reaction. Further evaporation of the solvent led to the formation of the ordered or disordered mesoporous inorganic substance [27,30,31]. Its final structural properties also depend significantly on the thermal treatment (e.g., calcination) during the removal of the template. Calcination is an important parameter in creating crystalline alumina with high surface area and has a great impact on the thermal, hydrothermal, and mechanical stability of alumina. The use of higher temperatures improves the crystallinity of porous materials but has a drawback, which can be seen in reduction in their mesostructural order as well as in the reduction in their pore volume and specific surface area [32].

The aim of this study is to design γ-alumina supports with mesoporous character, to determine the influence of the synthesis route using different solvents (water/ethanol/ethanolic solution) on the textural properties of the mesoporous γ-alumina support, to combine mesoporous γ-alumina supports with LiCl and CaCl₂ salt hydrates, and to determine water sorption properties of these composites. In addition, the stability for the composites of the water-based γ-alumina (W-46) with CaCl₂ and LiCl under humid conditions up to 100 °C has been evaluated. Crystalline γ-alumina has been selected as the support for the composite due to its thermal conductivity of 30 W/mK, which can further promote heat transfer in the system compared to silicate supports with a much lower thermal conductivity of 0.2 W/mK.

To the best of the author’s knowledge, this is the first report on the synthesis of crystalline mesoporous γ-alumina as a support for the TCM composite water sorbents.

2. Materials and Methods

2.1. Materials Synthesis

Pluronic F-127 (EO₁₀₆PO₇₀EO₂₁₀₆) and aluminium isopropoxide (98%) were purchased from Aldrich. Nitric acid (67%) and absolute anhydrous ethanol (99.5%) were purchased from Sigma-Aldrich (Darmstadt, Germany) and CARLO ERBA Reagents (Milan, Italy), respectively. Calcium chloride hexahydrate (98%) and lithium chloride hydrate (99%) were purchased from Sigma Aldrich (Darmstadt, Germany).

2.1.1. The γ-Alumina Support Synthesis

Mesoporous γ-alumina samples were synthesised in a two-step procedure. First, the evaporation-induced self-assembly (EISA) method was used. The following procedure is an adaptation of the synthesis described in [31]. The molar ratio of the reactants was: 1 AIP: 0.006 F-127: 2.3 HNO₃: 68 ethanol or water and 1 AIP: 0.006 F-127: 2.3 HNO₃: 34 ethanol: 34 water.

In all syntheses, two solutions were prepared: solution A containing solvent (an ethanol or a water or an ethanolic solution) and aluminium isopropoxide (AIP), and solution B containing a triblock copolymer non-ionic surfactant F-127, the same solvent as in solution A and nitric acid to aid solubility and hydrolysis.

Both types of solutions (A and B) were stirred at room temperature until completely dissolved and then mixed together and stirred at room temperature for 5 h. After solvent evaporation at 60 °C for 48 h in an oven, a semi-transparent solid was obtained. In the second step, the obtained solids were further solvothermally (ethanol) or hydrothermally (water, ethanolic solution) treated in Teflon-lined stainless steel autoclaves at 160 °C for 12 h.

The as-prepared samples were calcined in the furnace at 800 °C for 3 h under air flow with a temperature ramp of 1.6 °C/min to remove the surfactant and achieve the transformation of the precursor phase to crystalline γ-alumina.

The prepared support samples were named after the solvent used: water (W-46), ethanol (E-37), and ethanolic solution (ES-47).

2.1.2. Composite Preparation

The mesoporous γ-Al₂O₃ prepared with water and ethanolic solution was preheated in the oven at 120 °C for 2 h. A predetermined quantity of CaCl₂·2H₂O and LiCl·H₂O was dissolved in 1 mL of deionized water in order to reach the required salt amount. The prepared salt solutions were then spread to the dehydrated mesoporous γ-Al₂O₃ using the incipient wetness impregnation method. The composites were dried over night at room temperature.

2.2. Methods

XRPD patterns were collected on a PANalytical X’Pert PRO high-resolution diffractometer (Almelo, The Netherlands) from 10 to 70° 2θ, and a fully opened X’Celerator detector.

Nitrogen physisorption was performed at 77 K on a Tristar3020 volumetric analyzer (Micromeritics, Norcross, GA, USA). The nonporous salt does not contribute to nitrogen adsorption to a large extent, so the amount of the salt was considered, and nitrogen isotherms, specific surface area, and pore volume values were corrected. The BET specific surface area was calculated from adsorption data ranging from 0.05 to 0.2 of relative pressure [33]. The total pore volume was evaluated at a relative pressure of 0.98. The KJS method [34] was used to determine the pore size distribution (PSD), and the maximum of the PSD was used for the pore diameter determination [35].

SEM images were recorded on Verios 4G HP field-emission gun microscope (Thermo Fisher Scientific, Waltham, MA, USA).

Energy dispersive X-ray analysis (EDAX) using Zeiss SupraTM 3VP microscope (Jena, Germany) and ICP-OES using spectrometer ES 715 (Varian, Palo Alto, Santa Clara, CA, USA) were performed for elemental analysis.

The desorption temperatures of W-46 and ES-47 composites were determined using a Q5000IR thermogravimetric analyser (the accuracy ±0.01 µg; ±0.1 °C) (TA Instruments, Inc., New Castle, DE, USA) where the maximum (Tpeak) of the DTG curve was considered the desorption temperature [36]. The samples were heated in air at a heating rate of 10 °C/min. Before thermogravimetric measurements, the samples were kept over a saturated potassium sulphate solution for 5 days.

The surface charge of the W-46 and ES-47 supports was measured using Zetasizer nano ZS (Malvern, UK) instrument.

Water sorption isotherms were measured using the IGA-100 gravimetric analyser (the accuracy ±0.1 µg) (Hiden Isochema Ltd., Warrington, UK) at 25 and 35 °C and used to the calculate the value of the differential enthalpy of sorption according to the well-known Clausius–Clapeyron equation. The equal pressure intervals of 1.6 mbar between vacuum and 40 mbar (saturation vapor pressure of 56.3 mbar) were set with an equilibrium time of 80 s. The samples were degassed at <10⁻⁵ mbar at 25 °C (the accuracy ±0.1 °C) for 4 h to a constant mass. The definition of the thermodynamic heat cycle is shown in [37], and the calculation of the amount of heat involved can be found in [38]. The integral enthalpy of sorption Q_sor can be considered as the achievable heat storage density at the material level and was calculated by using the following equation: Q_sor = ΔH_sor (w_sor − w_des) [kJ/kg_sor]. ΔH_sor [kJ/kg_water] is the differential enthalpy of sorption according to the sorbed water amount, and w_sor and w_des [kg_water/kg_sor] are the maximum and minimum sorbed water amount over the sorbent at the given boundary conditions. The specific heat of the γ-alumina is 94.6 kJ/mol K at 25 °C [39]. The integral heat of sorption was calculated for the application of heat storage.

The sorption/desorption cycling performance [38] of the W-46-CaCl₂ and W-46-LiCl composites were carried out on IGAsorp-XT automated water sorption gravimetrical analyser (the accuracy ±0.1 µg) (Hiden Isochema Ltd., Warrington, UK.) with the sequential procedure of water uptake measurements at 35 °C (the accuracy ±0.1 °C) under moist nitrogen gas flow (flow rate of 250 mL/min and 80% relative humidity) for 3 h, followed by drying at 100 °C with dry nitrogen flow for 2 h. The sequence was repeated 25 times according to the literature [40] at a desorption temperature of 100 °C, which can be attained by solar thermal collectors. The sorption temperature was fixed to 35 °C, which is sufficient for space heating applications. The water vapor pressure during desorption and sorption of the samples was set to 12.3 mbar (a dew point temperature of 10 °C). The difference in the amount of sorbed water at 35 °C and 100 °C at 12.5 mbar is the cycle (water) loading lift of the composite.

3. Results and Discussion

3.1. Structural Properties of All γ-Alumina Supports and Composites

All mesoporous γ-alumina supports were prepared by the same synthesis procedure, which differs only in the use of different solvents (i.e., ethanol, water, and ethanolic solution), while the other conditions remained the same. The calcined samples exhibited mesostructures of all γ-alumina with low crystallinity, which can be due to the high calcination temperature (800 °C) [29] after the second step of hydrothermal or solvothermal treatment. Grant et al. [31] obtained cage-like ordered mesoporous γ-alumina by one-step synthesis, i.e., evaporation-induced self-assembly synthesis and calcination at 400 °C. Powder XRD was used to identify the crystal structure of the products after thermal treatment at 800 °C. The calcined powders in Figure 1a exhibited diffraction peaks corresponding to the γ-Al₂O₃ phase, which is consistent with JCPDS 00-050-0741. It can be observed that the intensity of the characteristic peak at 2θ 66.5° decreases in the sample prepared by hydrothermal synthesis compared to the sample prepared by solvothermal synthesis, indicating higher crystallinity of the solvothermal product. All the diffraction peaks are quite broad, which can be due to nanosized particles of the γ-Al₂O₃ products. The diffraction peak at 2θ 44.2° in the W-46 sample belongs to the XRD sample holder.

γ-Al₂O₃ supports prepared from water (W-46) and ethanolic solution (ES-47) were used to prepare the composites, as their crystal structure was preserved after treatment at 100 °C for 5 h in vacuum. Introduction of CaCl₂ and LiCl salts into mesoporous γ-Al₂O₃ prepared in water and ethanolic solution revealed that the supports were inert to the salts, meaning that the crystal structure of the supports did not change or react with these salts (Figure 1b). Additional peaks due to the presence of the salts were not observed, indicating that the highly dispersed salts were successfully confined of in the mesopores of these supports and/or that the salts were dispersed in particles less than 5 nm in size, which are not detected by XRD.

The morphology of the hydrothermally prepared γ-Al₂O₃ supports (E-37, W-46, and ES-47) was investigated using SEM (Figure 2). The results showed that the particular solvent used in the synthesis had a significant effect on the morphology of the samples. In water, larger 3 µm γ-Al₂O₃ agglomerates were formed, while 100 nm and 25–50 nm round nanoparticles were obtained from the ethanol and ethanolic solution-containing, respectively.

The amount of impregnated CaCl₂ after the incipient wetness impregnation procedure on W-46 and ES-47 γ-Al₂O₃ supports was lower than expected, considering the amounts of inorganic salt used in the composites with silica [8]. This is reported to be influenced by the high polarity of the alumina [28] and pH. The dependence of surface charge of γ-alumina (W-46 and ES-47) on pH is illustrated in Figure 3. It can be seen that the solvents affect the surface charge of the prepared supports. At the pH value of CaCl₂ and the LiCl solutions (pH = 5.5), the W-46 γ-alumina support exhibits a positively charged surface (29 mV), as does the ES-47 γ-alumina support (28 mV).

Thus, the introduction of calcium chloride or lithium chloride solution into the mesopores of the W-46 or ES-47 support during incipient wetness impregantion was hindered by similarly charged species, compared to silica mesostructures with negatively charged surface (−20 mV) [41]. The negatively charged surface (−10 mV) of the γ-alumina support W-46 can be obtained at a lower pH than ES-47. In general, the surface charge can be changed by modifying the synthesis procedure.

EDX analysis determined 10 wt% CaCl₂ in both supports, while determination of LiCl was performed using ICP-OES. In the W-46 and ES-47 supports, 15 wt% and 10 wt% LiCl were determined, respectively.

Nitrogen physisorption measurements were used to evaluate the textural properties of γ-Al₂O₃ supports prepared from different solvents. The nitrogen adsorption isotherms of the γ-Al₂O₃ supports and their pore size distributions (PSDs) are shown in Figure 4a, the composites W-46 in Figure 4b, and the composites of ES-47 in Figure 4c. The textural parameters determined based on these isotherms are listed in

Table 1. Textural properties of the mesoporous γ-Al₂O₃ supports.

Sample Support	Solvent	SBET (m2/g)	Vp (cm3/g)	Average Pore Size (nm)
E-37-γ-Al2O3	Ethanol	213 ± 1	0.542	10.7
W-46-γ-Al2O3	Water	147 ± 1	0.414	9.1
ES-47-γ-Al2O3	Ethanolic solution	148 ± 1	0.367	11.3

and

Table 2. Textural properties of the of the W-46 and ES-47 composites.

Sample Support	SBET (m2/g)	Vp (cm3/g)	Average Pore Size (nm)
W-46-LiCl	68 ± 1	0.196	8.7
W-46-CaCl2	81 ± 1	0.252	8.3
ES-47-LiCl	96 ± 1	0.204	9.3
ES-47-CaCl2	83 ± 1	0.207	10.2

. All isotherms are of Type IV, which is typical of mesoporous materials [42]. The isotherms exhibit a capillary condensation step, indicating the presence of mesoporosity. The sample E-37 shows a steeper capillary condensation, which is typical of a narrow pore size distribution. It can be clearly seen that the hysteresis loops change from the H1 loop for samples E-37 and ES-47 to the H3 loop for W-46, indicating different γ-Al₂O₃ support mesostructures due to the solvent used. The H3 loop is typical of materials with slit-like mesopores [42]. Macroporosity is found in the W-46 sample (Figure 4a) but not in the ES-47 and E-37 samples. The capillary condensation step for the ES-47 sample is shifted towards higher relative pressures, indicating an increase in the size of the mesopores due to solvothermal synthesis [43].

The E-37 sample has the highest specific surface area and pore volume (Table 1), while samples ES-47 and W-46 have lower specific surface areas and pore volumes. The largest size of mesopores (11.3 nm) was determined from the maximum of pore size distribution (Figure 4a), for sample, ES-47 prepared from ethanolic solution. On the other hand, the smallest size of mesopores (9.1 nm) was determined for the W-46 sample prepared hydrothermally.

The inclusion of salts in both supports resulted in less intense hysteresis loops (Figure 4b,c) and pore size distributions (Figure 4b,c), accompanied by a decrease in pore sizes, pore volumes, and specific surface areas (Table 2). It can be concluded that the salt nanoparticles been dispersed inside of mesopores of the supports. For example, after the introduction of LiCl into the W-46 support, the specific surface area (68 m²/g) and pore volume (0.196 cm³/g) of the composite decreased significantly (Table 2). The pore diameter (8.7 nm) of this composite was larger than the pore diameter (8.3 nm) of the composite with CaCl₂.

Desorption temperatures for the composites were determined from the DTG curves for composites W-46 and ES-47, which are shown in Figure 5. The desorption temperature of water was completed for the W-46 samples in the following order: the W-46-CaCl₂ sample (98 °C), the W-46-LiCl (88 °C), and the W-46 support (29 °C) (Figure 5a). Additionally, 88.0% of the water was desorbed from the composite containing LiCl, while a smaller amount of water (74.9%) was removed from the composite containing CaCl₂. The DTG curves show one water loss for all samples prepared in water and ethanolic solution. Water can be removed from W-46 composites at 120 °C. For ES-47 composites, water desorption can be completed even at a lower temperature (100 °C). Desorption temperature for the ES-47-LiCl composite was 68 °C, while a slightly lower desorption temperature (62 °C) was observed for ES-47-CaCl₂ (Figure 5b). Additionally, 75.0% and 60.8% of water was removed from the ES-47-LiCl and ES-47-CaCl₂ composites, respectively.

3.2. Water Sorption Properties

The mesoporous γ-Al₂O₃ and composites exhibit Type V water sorption isotherms (Figure 6a), measured at 35 °C. The γ-Al₂O₃ supports showed weak hydrophilic properties, which are also characteristic of mesoporous silica materials [8]. It was observed that the water uptake increased slowly up to 0.08 g/g at a relative pressure of 0.7. Thereafter, the water sorption capacity increased suddenly from 0.08 g/g to 0.31 g/g, which is due to the mechanism of capillary condensation and shows the active role of the supports. Jabbari-Hichri [21] reported that mesoporous alumina impregnated with 14 wt% CaCl₂ showed a water sorption capacity of 0.17 g/g. The composite W-46-LiCl (2.89 g/g) exhibited the highest maximal water sorption capacity of all composites, as determined from water sorption isotherms measured at 35 °C. The ES-47-LiCl sample showed the maximal water uptake of 1.00 g/g. Composites containing CaCl₂ showed lower maximal water sorption capacities: 0.76 g/g for W-46-CaCl₂ and 0.46 g/g for ES-47-CaCl₂. At a relative pressure of 0.4, the W-46 composite with 10 wt% CaCl_2, showed 4.5 times greater water uptake (4% versus 18%) than the support W-46 (not shown), while the composite with 15 wt% LiCl revealed 18 times greater water uptake than its support (4% versus 71%). It can be concluded that the presence of the salts in the γ-Al₂O₃ support increases the water sorption capacity of the composites, and the salt content affects the sorption performance of these composites. It should be emphasized that, among all the composites, only the composite W-46-CaCl₂ exhibits a water sorption isotherm without hysteresis loop (Figure 6a) up a relative pressure of 0.4, which may further improve the heat storage performance. Small hysteresis is seen in W-46-LiCl, while both ES-47 composites show larger hysteresis of the water sorption isotherm (Figure 6a inset).

The comparison of the water sorption isotherms of W-46-CaCl₂ and ES-47-CaCl₂ revealed differences in the range of 0 < p/p₀ < 0.4, with the W-46-CaCl₂ sample showing slightly higher uptake, which is most likely due to the preparation procedure of the support in water. The water uptake at 0.4 relative pressure of the composite ES-47-CaCl₂, which has uniform mesopores with an average pore size of 10.2 nm and a total pore volume (0.207 cm³/g), was 1.7% lower than the water uptake of the composite W-46-CaCl₂ with a smaller pore diameter (8.3 nm) and a larger total pore volume (0.252 cm³/g). The water uptake curve of the composite with W-46-γ-Al₂O₃ and 15 wt% LiCl showed a plateau at 0.038 p/p_o due to the formation of lithium chloride hydrate [44], while this plateau was not observed for the composites with lower salt content.

The kinetic curves (Figure 6b) show that the maximal water uptake on the composite ES-47-LiCl was reached in the shortest time. The composite ES-47-CaCl₂ showed a slow sorption rate in the first two hours, increasing with a slow rate after 2 h. The water sorption on the composites W-46-LiCl and ES-47-LiCl was also slow in the first two hours and then gradually increased at a higher rate than that of the composites containing CaCl₂. Water uptake occurred in the following order: ES-47-LiCl, W-46-CaCl₂, ES-47-CaCl_2, W-46-LiCl.

The integral heat of sorption is one of the most relevant parameters for evaluating the effectiveness of sorbent at the material level for TES. For sorption heat storage the adsorbents with high water loading lift between adsorption and desorption can reach high storage densities. The integral heat of sorption was calculated for the given boundary conditions for space heating [8]: sorption/condensation temperature of 30 °C, desorption temperature of 100 °C, and evaporation temperature of 10 °C. The integral heat of sorption Q_sor of all composites is listed in

Table 3. Water loading lift and the integral heat of adsorption for the composites.

Sample	∆w (kg/kg)	Qsor (kWh/m3)	Qsor (GJ/m3)
W-46-CaCl2	0.127	350	1.26
W-46-LiCl	0.514	1437	5.17
ES-47-LiCl	0.137	380	1.37
ES-47-CaCl2	0.135	377	1.36

. Composites containing similar amounts of LiCl and CaCl₂ showed similar theoretical energy storage density values at material level ranging from 1.26 to 1.37 GJ/m³, while lower values were obtained with composites containing mesoporous silica and the same amount of CaCl₂ [8].

The increased amount of the salt (15 wt% LiCl) in the composite increased the calculated water loading lifts, and consequently, theoretical energy storage density up to 5.17 GJ/m³ at material level. The physical and structural properties of the mesoporous γ-Al₂O₃ support, such as high density and the suitable mesostructure with the pore size of up to 10 nm, are advantageous for the sorption heat storage performance, which can be further enhanced by the high thermal conductivity of the γ-alumina support. A direct comparison of the energy storage density of these composites with other alumina-containing composites was difficult, because it strongly depends on the boundary conditions. The value of energy storage density of the composite with LiCl (14.7 wt.%) and activated alumina prepared by wet impregnation equalled 1.14 GJ/m³, considering a desorption temperature of 120 °C, sorption temperature of 20 °C, and 80% relative humidity [11]. This composite was also tested in the open sorption TES system for space heating at a desorption temperature of 110 °C and reached 0.69 GJ/m³ [22]. The influence of different boundary conditions on the integral heat of sorption was evaluated (Figure 7) for the conditions of the space heating storage cycle defined in [1] for the Central European climate: sorption temperature of 35 °C, evaporation temperature of 5 °C, desorption temperature of 90 °C, and condensation temperature of 30 °C. Frazzica [1] used the condensation temperature of 30 °C as the average ambient temperature during the day in summer in Central Europe. The evaporation temperature was set at 5 °C, which was explained by the situation of some rigid climatic conditions where geothermal boreholes were used to provide ambient heat for the evaporator and satisfactory heat transfer inside the component.

It can be seen that lower desorption temperature decreases the energy storage density of the composites and still promotes their use for low-temperature thermal energy storage.

The cycling stability of the W-46-CaCl₂ (without hysteresis during the sorption cycle up to a relative pressure of 0.4) and W-46-LiCl (small hysteresis during the sorption cycle up to relative pressure of 0.4) composites was tested during 25 sorption and desorption cycles between 35 and 100 °C at 12 mbar, showing a small reduction in water uptake of 2.0% for W-46-CaCl₂ and 1.5% for W-46-LiCl after the last cycle (Figure 8). After the cycle tests, no salt leakage was observed for these samples. This confirms that the mesoporous γ-Al₂O₃ support can be used as a stable nano-environment for the entrapment of calcium and lithium chloride. These composites are promising candidates for low-temperature thermal energy storage.

4. Conclusions

Mesoporous γ-alumina supports were synthesized from gels containing aluminium isopropoxide, the surfactant Pluronic F-127, and various solvents: ethanol, water, or ethanolic solution. This was undertaken by the two-step synthesis procedure, which included the evaporation-induced self-assembly method followed by solvothermal or hydrothermal treatment at 160 °C for 12 h and calcination at 800 °C to obtain the crystalline phase of γ-Al₂O₃. The solvents affected the uniformity of the mesopores, the specific surface area, and the morphology of the γ-Al₂O₃ supports. The highest degree of uniformity of mesopores and surface area was found in the sample E-37 prepared in etanol. The uniformity of the samples ranged from the E-37 (ethanol) and ES-47 (etanolic solution) to W-46 (water). The solvent also influenced the surface charge of the γ-Al₂O₃ support and, consequently, the amount of the salt inclusion in the support performed by incipient wetness impregnation. The desorption temperatures of all composites varied from 62 to 98 °C, indicating low temperature heat storage utilization.

Among all composites, the highest heat storage density (5.17 GJ/m³) was achieved for W-46-LiCl at boundary conditions for space heating at the sorption and condensation temperature of 30 °C, the desorption temperature of 100 °C, and the evaporation temperature of 10 °C. The composite W-46-CaCl₂ showed the best heat storage performance in terms of fast kinetics of water sorption, no sorption–desorption hysteresis, low desorption temperature, and very good cycling stability and energy storage density of 1.26 GJ/m³. The fastest rate of water sorption was found for ES-47-LiCl composite, where the sorption of water reached its maximum after 9 h. Moreover, after 25 cycles between 35 °C and 100 °C at 12.5 mbar, small decreases in water uptake of 1.5 and 2.0 % were observed for the W-46-CaCl₂ and W-46-LiCl composites, respectively, indicating the very good hydrothermal stability of these composites.

The properties of mesoporous γ-Al₂O₃ supports such as high density, particular mesostructure, high thermal conductivity, and structure stability are advantageous characteristics for the preparation and use of these TCM composites as low-temperature sorption heat storage materials. This study opens up several possibilities to tailor the porosity of mesostructured oxide supports by proper selection of solvents, reagents, and synthesis routes. Moreover, it represents an important step towards the development of mesostructured metal oxides with designed properties for TCM materials, which are of great importance for thermochemical energy storage.