Use of Low Melting Point Metals and Alloys (T_m < 420 °C) as Phase Change Materials: A Review

Abstract

Phase Change Materials (PCMs) are materials that release or absorb sufficient latent heat at a constant temperature or a relatively narrow temperature range during their solid/liquid transformation to be used for heating or cooling purposes. Although the use of PCMs has increased significantly in recent years, their major applications are limited to Latent Heat Storage (LHS) applications, especially in solar energy systems and buildings. PCMs can be classified according to their composition, working temperature and application. Metallic PCMs appear to be the best alternative to salts and organic materials due to their high conductivity, high latent heat storage capacity and wide-ranging phase change temperature, i.e., melting temperature and chemical compatibility with their containers. This paper reviews the latest achievements in the field of low-melting point metallic PCMs (LMPM-PCMs), i.e., those with melting temperatures of less than 420 °C, based on Zn, Ga, Bi, In and Sn. Pure LMPM-PCMs, alloy LMPM-PCMs and Miscibility Gap Alloy (MGA) LMPM-PCMs are considered. Criteria for the selection of PCMs and their containers are evaluated. The physical properties and chemical stability of metallic PCMs, as well as their applications, are listed, and new application potentials are presented or suggested. In particular, the novel application of metallic PCMs in casting design is demonstrated and suggested.

1. Introduction

With the growing demand for the use of energy, the consumption of fossil fuels resulting in greenhouse gas emissions has increased. To reduce these environmentally unfriendly gases, renewable energy sources, such as solar and wind, are replacing fossil fuels. Solar energy systems, as thermal energy storage (TES) applications, use the stored thermal energy of the sun directly or even convert it into other kinds of energy, such as electricity [1]. The storage of both sensible and latent heat energy increases thermal efficiency in TES applications, where the stored energy of the latter is greater than that of the former [2]. Consequently, the most productive approach for saving the thermal energy of materials is by using the latent heat of different phase changes [3], such as solid–solid, solid–liquid, solid–gas, and liquid–gas transitions [4].

Researchers have proposed phase change materials (PCMs) with constant or narrow transition temperatures that can absorb/release a specific amount of heat during each phase transition. Among the abovementioned transitions, solid–liquid transitions have much more latent heat compared to solid–solid ones. In general, gas transitions are not practical for use as PCMs due to their large volume change and explosion hazard [3]. As a result, PCMs with a solid–liquid phase change, aimed at in the present paper, are more applicable.

To select the appropriate PCM, some crucial parameters, such as thermal conductivity, specific heat capacity, thermal stability, latent heat and melting temperature of PCM, should be considered [5]. In addition, the thermal stability of PCM containers is as vital as PCM to avoid corrosion, soldering and chemical reactions [6]; as a consequence, encapsulated PCMs have been extended [2,5,7].

There are various approaches to classifying PCMs, for instance, according to the chemical composition and melting temperature. In terms of the chemical composition, PCMs are divided into three different groups, namely organic materials (paraffin and non-paraffin) [8], inorganic materials (salts and metals) [2] and eutectic compositions (organic–organic, organic–inorganic and inorganic–inorganic) [9,10].

Organic PCMs are available at low temperatures (<200 °C) with sufficient latent heat [11]. They solidify with a small undercooling and show phase change stability with no corrosive behavior [2]. The main drawbacks of these PCMs are low thermal conductivity, high volume change and flammability, which restrict their applications.

Compared to organic PCMs, inorganic salt PCMs present higher thermal conductivity and enthalpy of fusion. Unfavorable points, which include high supercooling, corrosive nature and segregation-induced instability during various cycles, have limited their use as an ideal PCM group [3,12]. To improve the thermal conductivity of organic PCMs, porous materials with higher thermal conductivity, such as graphite [13], metallic foams [14,15] and nano/micron particles (carbon [16], Al₂O₃ [17]) can be added.

On the contrary, metallic materials (for example, Al-25%Si [18] or Al-Si eutectic alloys [19]) have been proven to be suitable alternatives. Therefore, metallic PCMs have been recognized as efficient PCMs to overcome the disadvantages of salt and organic ones.

The phase transition energy in the melting process of metallic PCMs is higher than that of non-metallic ones. High-energy storage density caused by the high latent heat of melting is a remarkable merit of metallic PCMs [6]. During the melting process, pure metallic PCMs store a great amount of thermal energy at a constant temperature. Due to the high heat diffusivity of metals, the storage/releasing of thermal energy occurs in a short time. Even though high latent heat per volume of metallic PCMs diminishes the need for large containers, the low latent heat per weight of other PCMs can limit their use and spread [6].

Currently, investigations of pure and alloy metallic PCMs have been extended. New applications of pure metals as PCMs in foundry have appeared to provide better control of the final macrostructure throughout the castings [20]. In addition, the use of miscibility gap alloys (MGAs) [21] is a promising approach to improve thermal energy storage.

PCMs can be classified according to their melting temperature (T_m) [6,22,23]. The range of melting temperatures for metals and metallic alloys is very wide, i.e., from −38.87 °C (Mercury) to 3400 °C (Tungsten) [24]. In this paper, metallic PCMs are divided into three groups of low (T_m < 420 °C), medium (420 °C < T_m < 1000 °C) and high (T_m > 1000 °C) melting temperature materials.

This paper aims to present a comprehensive and detailed account of metallic PCMs classified in the low melting temperature group (LMPM-PCMs). The following sections describe and critically review various LMPM-PCMs, such as pure metals, alloys, metal matrix composites and miscibility gap alloys. Furthermore, criteria for PCM and capsule selection are listed and PCM applications are investigated with the hope of widening the horizon in the research of metallic PCMs for innovative applications.

2. Criteria for PCM Selection

The following is a list of the most crucial parameters that may be considered for the selection of an appropriate PCM for a specific application [6,25,26,27]:

• Melting temperature of PCM driven by the required service temperature.
• High latent heat and suitable specific heat. The former controls the volume and thermal density storage of the PCM, and the latter controls the rate of temperature change of the PCM before reaching the transition temperature.
• High thermal conductivity enables high discharge capacity of the PCM. This characteristic led PCMs to be used in buildings to take less time to store/release energy.
• High density to reduce the required volume of the PCM.
• Low thermal expansion and low vapor pressure to avoid damaging the container and to avoid residual stress in the solid surroundings.
• Reversibility of the transition.
• Low solidification undercooling and high solidification rate as they affect the transition kinetic.
• Non-toxicity, non-flammability and non-reactivity with the container.
• Availability and affordability.

Due to the interdependence and interaction of these parameters, no straightforward approach for the selection of PCMs has been established so far. In fact, the specific types and details of the application and service conditions play a key role in selecting an appropriate PCM for a specific purpose. For instance, in concentrated solar power (CSP) systems, a number of factors consisting of desired melting temperature, low-cost materials and high energy density play a key role in the general performance of PCM [28]. Figure 1 illustrates different types of PCM candidates (metals, salts and eutectic compositions) for CSP applications, considering the melting temperature, energy density and material cost. In some critical applications, cost may not be a key issue, while high thermal conductivity and low thermal expansion, for example, may be more important factors.

Several thermophysical properties, such as thermal conductivity (k), specific heat capacity (C_P), density (ρ), latent heat (L_m) and melting temperature (T_m) determine the thermal characteristics of a PCM. In addition to these, thermal diffusivity (α = $\frac{\mathrm{k}}{{\mathsf{\rho }\mathrm{C}}_{\mathrm{P}}}$) defined in terms of k, C_P and ρ, is another important parameter affecting the effectiveness of a PCM. The effects of melting point on the thermal diffusivity and latent heat of some common pure elements are plotted in Figure 2a. An enlarged section of the diagram is shown in Figure 2b. It is clear from Figure 2a that the thermal diffusivity does not depend much on the melting point, but the latent heat increases almost linearly with the melting point. Figure 2c is an effort to display the interconnections of different parameters and their importance in PCM selection.

To determine the critical parameters for selecting a suitable PCM, a mathematical method based on multi-criteria decision making (MCDM) has been proposed [27]. This method includes two steps. In the first step, different techniques such as the Analytical Hierarchy Process (AHP), Simple Additive Weighted (SAW) and Weight Product Method (WPM) weight each parameter. In the second step, the Technique for Ordering Priority by Similarity to an Ideal Solution (TOPSIS), Multi-Objective Optimization by Ratio Analysis (MOORA), and Graph Theory and Matrix representation Approach (GTMA) methods rank the evaluated parameters [29,30].

The application of MCDM was carried out to select suitable PCMs for a solar domestic hot water system [27,31], where the most crucial factor was considered to be the melting temperature of the PCM, which had to be in the range of 30–60 °C. Other quantitative and qualitative criteria, including latent heat, thermal conductivity, specific heat at solid and liquid states, density, volume change, vapor pressure, supercooling, phase segregation, recycling, toxicity and flammability, were also considered. All of these parameters were weighted by the AHP method. Subsequently, the TOPSIS and fuzzy TOPSIS methods were applied to rank them. Latent heat was finally determined as the most important factor in solar domestic hot water system [27].

From the review of the literature and PCM applications, it appears that the melting point and latent heat of fusion are usually the most important criteria in PCM selection. The remaining listed parameters normally take the second and third priority levels, depending on the requirement of the application. In specific applications, where the importance of these parameters is more significant, the use of analytical methods such as those described above may be necessary for optimum selection of the PCM.

3. Application of Metallic PCMs

Since metallic PCMs have a higher energy storage capacity, as well as a higher thermal conductivity and smaller volume change, compared to organic PCMs, they have been used or are among the potential candidates in a wide range of applications, including thermal management and reduction of energy usage in buildings [32,33], hot water systems [34], fluid metal coolants in electronic devices [35,36] bioengineering [22] and concentrated solar power to produce high-temperature steam for electricity production [37]. In addition, in solid–solid PCMs, crystalline to amorphous transition of the structure can be utilized to increase the electrical resistance needed in hard disks and flash memory sticks [38,39,40,41].

4. Encapsulation of PCMs

In all known applications of metallic PCMs, they have been encapsulated in a higher melting point metal or ceramic container. Encapsulation is a practical way to save thermal energy. It improves the heat transfer between the PCM and the heat source and, as a result, reduces the required time for melting or solidification of the PCM. Furthermore, solidification of PCM can occur without any significant undercooling, which diminishes the segregation phenomenon. Encapsulation may also serve to increase the specific contact area of a PCM. It improves the heat transfer in low thermal conductivity PCMs, e.g., in organic PCMs, and encourages heterogeneous nucleation in the liquid matrix material, leading to reduced segregation phenomena during solidification [5].

An appropriate capsule must fulfill the following criteria [5,42,43,44,45]:

• Chemical stability. The capsule should react with neither the metallic PCMs nor the surrounding matrix. It should have close to zero solubility with the PCM and the matrix. Such reactions may result in corrosion of the capsule, contamination of the PCM or the matrix, leakage of PCM or formation of undesirable intermetallic phases at the interface. These will affect such thermal properties as the melting point, heat transfer coefficient and latent heat of the PCM.
• Thermal and mechanical stability. The capsule should have high thermal resistance, appropriate coefficient of thermal expansion and be dense and stable during solid/liquid transition to overcome the stresses induced by repeated expansion/contraction of the PCM. This will prevent capsule cracking and PCM leakage during thermal cycling.
• Large coefficient of heat transfer and small heat capacity.
• Small coefficient of thermal expansion (CTE) and enough space for melting and solidification of the PCM.
• Oxidation resistance for high-temperature applications.
• High specific contact area.
• Nontoxicity, availability and affordability.

The most widely used capsule materials for metallic PCMs are different types of stainless steels [46], nickel [47,48] and different metal oxides [49].

In some cases, nano-encapsulated PCMs (nePCMs) have been added to the fluid surroundings. Nanofluids contain nanoparticles in which the nePCM is located at their core. There is a high meting point shell that surrounds the core and keeps it during the melting and solidification phases. By adding the nanoparticles into the fluid, the thermal conductivity and heat capacity of the nanofluid increases by absorbing the heat to melt the PCM [50]. With a focus on the melting and solidification of encapsulated metals, the melting and solidification temperatures vary from the equilibrium temperatures; therefore, it is necessary to apply superheating and supercooling, respectively. This means melting and solidification can occur at higher and lower melting points, respectively [51].

Core/Shell PCMs and shape-stabilized PCMs are the main types of encapsulation employed to embed PCMs in metal matrices [7]. Encapsulated PCMs are manufactured in various shapes, such as spherical, cylindrical and non-spherical shapes in macro, micro and nano scales with sizes from more than 1 mm, 0–1000 μm and 0–1000 nm, respectively [5].

By considering all listed criteria, after stability, which is the most vital parameter in capsule selection, the coefficient of thermal expansion and specific heat capacity are the next most important factors in the selection of capsule materials. A coefficient of thermal expansion matching those of the surrounding and the PCM is necessary to avoid excessive capsule deformation or cracking during repeated thermal cycling. A low specific heat capacity is required to induce a high heat transfer rate.

5. Pure Metal-Based PCMs

High- and commercial-purity metals melt and solidify at a constant temperature or a very narrow temperature range, respectively. Pure metals are ideal candidates for use as PCM materials at a variety of service temperatures. The latent heat from the melting and solidification of pure metals is typically in the range of about 23–1877 J/g [52], and their melting point ranges from about −39 to 3400 °C [24].

Zinc, gallium and tin seem to be the favorite low-melting point pure metallic PCMs. Thermophysical properties, i.e., melting temperature, latent heat of fusion, thermal conductivity and specific heat capacity of Ga [53,54,55], Sn [49] and Zn [47,48] are reported in

Table 1. Physical properties of pure gallium, indium and zinc. Data from [24].

	Tm (°C)	Lm (J/g)	k (W/mK) @ 0–100 °C	CP (J/gK) @ 0–100 °C	ρ (g/cm3)
Ga	29.7	80.23	41	0.377	5.91
Sn	231.9	59.64	73.2	0.226	7.3
Zn	419.5	110.12	119.5	0.394	7.14

. This section has been divided into three sub-sections: pure zinc, pure gallium and pure tin.

5.1. Pure Zinc

Sabol et al. [47] used pure zinc encapsulated into an Ni container as PCM for solar energy system application. By using a metallic PCM such as pure zinc, the thermal energy storage (TES) capacity of the solar energy system increased in comparison with conventional organic PCMs. Pure metals have higher thermal conductivity and latent heat than organic materials. Moreover, they do not suffer from segregation problems and supercooling of salts.

Sabol et al. [47] also investigated the formation of intermetallic compounds between the Ni capsule and Zn. They noted that if the encapsulated PCM reacted with the capsule material during melting and solidification, its composition would change, resulting in a change in TES capacity.

According to the Ni-Zn phase diagram, four different intermetallic phases, namely β₁, β, γ and δ, can be formed between Ni and Zn. Ni has an FCC structure and high solubility in Zn, while Zn has an HCP structure and an insignificant solubility in Ni. EPMA analyses indicated the formation of γ (NiZn₃) on the Ni side and δ (Ni₃Zn₂₂) on the Zn side of the Ni/Zn interface. The thickness of both intermetallic phases increased over time and after 16 h reached 134 and 201 μm, respectively. A schematic representation of Ni/Zn interfaces at 1 and 16 h is shown in Figure 3. In this figure, IPL1 and IPL2 mark the γ and δ phases, respectively [47].

Diffusion constants are 1.09 × 10⁻⁹ and 2.2 × 10⁻⁹ cm²/s for γ and δ phases, respectively. As a result, the thickness of δ is more than that of γ. Another reason for the higher thickness of the δ structure is the fact that zinc atoms are in a molten state, so their diffusion rate is more than that of Ni atoms. Moreover, to produce the γ structure, Zn atoms should diffuse across δ thickness and reach the γ structure (near the Ni wall). Since diffusion in the solid state is slower than in the liquid state, and Zn has a longer distance to reach and produce the γ structure, the thickness of γ structure decreases (Figure 3). These results demonstrated that while Zn-encapsulated PCM had an appropriate TES for the mentioned application, it did not have sufficient repeatability for extended usage [47].

Zhao et al. [43] demonstrated that Zn encapsulated in Ni had suitable heat transfers above 400 °C in solar energy systems. The schematic of the encapsulated PCMs is shown in Figure 4, where R1 and R2 are the outer radius of the container and the radius of the PCM. r = s(t) is the location of liquid/solid interface because it depends on the solubility of the two elements. The outer radius of encapsulated Zn (R1) and the shell thickness were chosen as 10–25 mm and 1.6 mm, respectively. Two different fluids, i.e., air and liquid Therminol/VP-1, were used. Therminol/VP-1 is an ultra-high temperature synthetic heat transfer fluid with a density of 0.654 g/cm³, thermal conductivity of 0.07 W/mK and heat capacity of 2.76 J/gK.

The results indicated that by decreasing the capsule size from 50 to 20 mm diameter, the required time for melting the PCM in air decreased from 66.1 to 13.6 min. Further, the type of fluid can affect the total time for conduction and phase transition. For example, the required time for charging, i.e., melting of pure zinc changed from 66.1 to 28.1 min. Therefore, the amount of heat transfers at a constant time was increased by decreasing the particle dimension. Additionally, the required time to melt the PCMs decreased by using the VP-1 liquid as fluid due to its higher heat capacity and density, but the use of air as a commonly available heat transfer fluid can also be considered useful. In total, both PCMs are appropriate materials as PCMs in solar energy systems for the continuous production of electricity [43].

In another application [48], pure Zn encapsulated in Ni and stainless steel containers was used as PCM to produce electric power (~1000 MWh) in a solar energy system. Zinc is used at 300–500 °C, which has sufficient thermal energy. Zn encapsulated pellets with diameters ranging from 10 to 100 mm were used. Two different fluids, i.e., air and VP-1, were used to transfer heat from the sun to encapsulated PCMs. It was found that at a given pellet diameter, by increasing the number of thermal cycling cycles, the heat transfer of PCM decreased from 11.3 to 6.4 kJ. The results showed that by increasing the diameter of zinc-encapsulated pellets, the heat conduction decreased and the time for the heat transfer and melting process increased due to the reduced heat transfer area. It was illustrated that at a given diameter, the required time for charging and discharging of heat was more for air than for VP-1 due to the higher heat capacity of the latter (2.76 J/gK at 427 °C compared to 1.1 J/gK for air at the same temperature). Overall, VP-1 was found to be a more suitable fluid for heat transfer due to its better thermal diffusivity. Moreover, it was shown that in Ni encapsulated zinc, molten zinc could diffuse into the solid Ni shell and make a high melting point Ni-Zn intermetallic phase with δ structure after several cycles. This would change the initial composition of the PCM in such a way that after 7 cycles, the Zn content of the PCM reached from 100 to 55.2%, which is not desirable.

Rizvi et al. [56] used a mixture of Li₂CO₃-K₂CO₃ eutectic salt and sol-gel silica-coated zinc nanoparticles as PCM in a solar energy system. Calorimetry performed on the coated nanoparticles and the mixture of nanoparticles and salt proved that pure zinc and salt melted at 420 and 470 °C, respectively. By adding the nanoparticles into the molten salt, the heat capacity was increased from 1.567 to 2.31 J/gK due to the high surface area and high thermal conductivity of the nanoparticles. No rupture was detected on the coated surface of the zinc nanoparticles, resulting in their oxidation. Furthermore, unlike Ni-encapsulated Zn, no intermetallic compound was formed between Zn and the silica coating. It should be noted that the shape, size and roughness of the particles can affect the heat transfer properties of the capsules.

Blaney et al. [46] investigated the stresses produced on Ni and stainless steel encapsulated zinc surfaces at 450 °C using Abaqus software. Ni capsules with 250 μm thickness were produced by the electroless method. Cylindrical hallow 316 L steel capsules were filled with 70 to 86 vol% of Zn. The results indicate that during melting of the pure zinc, a large amount of stress is exerted from the PCM on the inner surface of the Ni shell, reaching to more than the critical value (~100 MPa), which causes plastic deformation or failing of the capsule. The stress is intensified at local defects such as shell thinning, dents or cracks, resulting in fracture initiation from these areas. While the cylindrical stainless steel containers showed high stability against PCM expansion, the induced stresses increased with increasing zinc content of the capsules and reached about 18 MPa at 86 vol% Zn.

Cingarapu et al. [57] used pure zinc as PCM, Trioctyl phosphine oxide (TOPO) as the zinc coating and alkali chloride salt as the fluid in CSP application to produce electricity within the 350–500 °C temperature range. Different particle sizes, i.e., 0.6 and 5 μm, were utilized to investigate their influence on the thermal properties. A DSC test (200 cycles) was carried out on the coated particles in N₂ and air atmospheres. The results demonstrated that the thermal properties of both samples in both atmospheres were stable and acceptable. In other words, the TOPO coating could protect Zn particles from oxidation. The measured latent heat was 70 and 82 J/g for 0.6 μm particles in N₂ and air atmospheres, respectively. It was 97 and 103 J/g for 5 μm coated particles in N₂ and air atmospheres, respectively. Moreover, the specific fusion heat for the coated 0.6 μm Zn particles was 15% lower than that for the uncoated particles. This difference can be related to the effect of particle size on the mass fraction of coating and oxidation of the surface. However, it was similar for both coated and uncoated 5 μm particles.

By adding 21 vol% of 5 μm Zn-coated particles in an N₂ atmosphere and 15 vol% of 5 μm Zn-coated particles in an air atmosphere to the salt, the latent heat reached 41 and 27 J/g, respectively. According to Figure 5, there are no obvious changes in the morphology of particles during various cycles. In addition, the increase in the latent heat was 20 J/g for 0.6 μm Zn-coated particles for both 11.5 vol% in N₂ and 13 vol% in air atmospheres. It was concluded that both types of particles were thermally and chemically stable in the salt. The thermal properties of the salt improved by the addition of both particles. For example, the thermal conductivity of pure salt (1–2 W/mK) was increased 30% by adding 10 vol% Zn-coated particles and will further increase with more Zn particle addition. However, the viscosity of salt should also be considered. Overall, the TES efficiency of the instrument increased about 45% by the addition of 10 vol% Zn-coated particle [57]. It seemed that the use of 5 μm Zn particles was more beneficial for TES applications.

Hsu et al. [51] used Zn/TiO₂ and Zn/Al₂O₃ core/shell PCMs in salt at 350–450 °C. Zn/TiO₂ and Zn/Al₂O₃ core shells were produced by the sol-gel technique and thermal deposition of an aluminum nitrate nonahydrate precursor, respectively. Thermal Hysteresis (TH) obtained by DSC test showed the difference between melting and crystallization temperature of the material, i.e., it demonstrated the lost energy during melting/solidification process which could reduce the efficiency of TES in the solar energy system. Different parameters, such as heating rate, shell thickness and nucleation mode (homogenous or heterogeneous), influence the TH.

Heat transfer in DSC is determined by Equation (1).

$$\mathrm{q}=\frac{\Delta {\mathrm{T}}_{\mathrm{DSC}}}{{\mathrm{R}}_{\mathrm{core}}+{\mathrm{R}}_{\mathrm{shell}}},$$

(1)

where ΔT_DSC, R_core and Rs_hell are the increase of temperature in DSC and the thermal resistances of the core and shell, respectively. Thermal resistance depends on shell thickness, shell thermal conductivity and heat transfer area. By increasing the ramping rate from 20 to 60 °C/min, q increases. Subsequently, ΔT_DSC rises, which leads to the growth of TH. By increasing the thickness of TiO₂ from 165 to 1312 nm at a constant heating rate of 20 °C/min, the slope of phase changing in DSC becomes lower. This means that the q value becomes smaller due to the higher thermal resistance of the shell. Consequently, TH increases from 11.7 °C for a shell thickness of 165 nm to 16.8 °C for a shell thickness of 1312 nm. Similar results were obtained for the Al₂O₃ shell. By changing the shell thickness from 485 to 1028 nm, TH increased from 8.2 to 11 °C. In addition, the thermal conductivity of the shell plays a key role in heat transfer. Thermal conductivity is 11.7 W/mK for TiO₂ and 35 W/mK for Al₂O₃. The shell with a higher thermal conductivity has a lower TH. It was observed that thinner shells had a more heterogeneous nucleation with no need for more supercooling, resulting in smaller TH. Between these two core/shell pairs, Zn/Al₂O₃ was found to be more suitable. By using 1–20 wt% Zn-Al₂O₃ core-shell particles in the salt, the heat capacity at the working temperature increased by 0.68–13.6%. It should be noted that increasing the core-shell content in the salts increases the salt viscosity, a fact that should be considered during the design of the salt [51].

5.2. Pure Gallium

Nowadays, the use of smaller and higher power density electronic chips in electronic devices is a very challenging topic. One main challenge is the increased temperature of electronic devices, which diminish their lifetime. Different methods, such as air or water cooling, have been used to decrease the temperature. By advancing PCM technology, researchers have started to use encapsulated PCMs to control the temperature of electronic devices [58]. Materials such as paraffin, Na₂SO₄-10H₂O, and N-eicosane have been used as PCMs. However, organic materials, such as paraffin, suffer from low thermal conductivity (0.21 and 0.29 W/mK in solid and liquid states, respectively). In addition, the use of salts results in segregation. It has been shown that by using low melting point pure metals, such as gallium, it is possible to overcome these problems. Pure gallium, for instance, has been used as PCM in electronic devices, such as mobile phones, to reduce CPU temperature [53].

Gallium has the following merits in comparison to others:

• High thermal conductivity, which improves the heat transfer required for cooling systems;
• Low specific heat capacity, which reduces the delay time for melting of the PCM, which in turn shortens the time required for keeping the CPU at constant temperatures;
• Low volume change.

Figure 6a schematically shows how PCMs can be used to cool the CPU of a smart phone [53]. In this work, different PCMs were contained in stainless steel containers. To simulate CPU-operating conditions, each container was warmed up with a 1.6 V and 1.77 DC current (Figure 6b) for 30 min to melt the PCMs and then the power was turned off to let the PCMs solidify at natural air convection. The heating–cooling curves with and without the PCM in the containers are shown in Figure 6c. As shown, the time to keep the temperature of the container under 45 °C for paraffin, N-eicosane, sodium sulfate decahydrate and gallium was 500, 600, 800 and 1000 s, respectively, indicating how gallium could shorten the overheating time in comparison to others with a similar volume. In addition, it was shown that shaking the system could increase the driving force to start solidification. As a practical consequence, touching the screen or keyboard of the smart phone could encourage solidification and heat absorption of the PCM [53].

Despite all the advantages of gallium as a PCM, there is one big drawback that should be solved. Ge and Liu [53] showed that during cooling, gallium could not solidify at its nominal melting temperature of 30 °C. It needed 10 °C undercooling to start the solidification. As a result, gallium could not maintain the required holding temperature range of 22–45 °C during cooling. To tackle this problem, gallium was stir mixed with 10 wt%. paraffin at 50 °C. Earlier solidification of paraffin results in temperature reduction and earlier solidification of gallium. Hence, the holding time between 22–45 °C decreases from 1000 to 400 s, which is undesirable if the thermal performance of the PCM is concerned. As a second solution, 1 wt%. SiO₂ was similarly added to gallium. The SiO₂ particles acted as nucleating agents for the solidification of gallium and increased its solidification temperature to 28 °C.

In another work, Ge and Liu [59] used encapsulated pure gallium as a PCM around a smartphone to reduce its temperature during calling for 100 min. The dimensions of the phone and the PCM container are 12.38 × 5.86 × 0.76 cm³ and 13 × 6.5 × 1 cm³, respectively.

Table 2. Variation of temperature with time for phones with and without PCM shells. Data from [59].

t (min)	0	20	40	60	80	100
T Without PCM (°C)	29.2	47	50	50.2	50.2	50.2
T With PCM (°C)	29	34	34	34	34.2	34.5

presents the variation of temperature with time for phones with and without PCM. As observed, by using the PCM, the phone temperature was kept at 34.6 °C for 100 min during calling, while it was at 50.7 °C without PCM. During a call, as the phone temperature increases, solid PCM absorbs the heat and its temperature reaches gallium’s melting temperature. After calling, phone temperatures start to decrease, and PCM is solidified, keeping the temperature constant. The results suggested pure gallium as an appropriate material to control the temperature of electrical appliances during usage.

5.3. Pure Tin

One to five vol% of Sn-SiO₂ core-shell nanoparticles were used in Therminol 66 (TH66) fluid in a CSP application [49]. TH66 is a kind of high-temperature application (−3 to 345 °C) heat transfer fluid with 2.122 J/gK specific heat and 0.107 W/mK thermal conductivity [60]. Sn-SiO₂ core-shell nanoparticles, with 50–100 nm Sn particles and 5 nm thick SiO₂ shell with a dense and crystalline structure, were produced by the reduction of modified polyole and sol-gel methods. By the addition of 1 to 5 vol% nanoparticles into the fluid, the thermal conductivity of fluid increased from 0.124 to 0.14 W/mK. Generally, by addition of nanoparticles into the fluid, fluid viscosity increases, an issue that can be rectified by rising the temperature. For example, while the addition of 5 vol% Sn-SiO₂ nanoparticles increased the viscosity at an ambient temperature, by increasing the temperature to 125 °C, the viscosity of the fluid containing nanoparticles became about similar to that of the fluid without nanoparticles. Consequently, depending on the fluid viscosity, the use of nanoparticles can be beneficial in high-temperature applications. The latent heat, specific heat (at 100–280 °C), and TES per volume of Sn/SiO₂ with 80% Sn were measured to be 48 J/g, 2 J/gK and 408 J/cc, respectively. Interestingly, temperature has an effective influence on TES, i.e., the presence of latent heat is more significant around the melting point. For example, for 5 vol% Sn-SiO₂ in TH66, within 50 °C of the melting temperature, the storage heat capacity increases 20%, while at higher than 150 °C, melting energy to total heat energy storage decreases to 5%. After 20 cycles, no obvious changes in the structure and final thermal properties were observed. The results indicated that solidification of Sn needed a large supercooling, so it could not be a perfect choice for CSP applications. However, it can be useful because of its high latent heat and thermal stability if there is not a better alternative [49].

The “encapsulation ratio,” i.e., the ratio between the enthalpy per unit mass of nePCM during phase change and the bulk of PCM, is important when nePCM is used. This ratio should be high enough to melt the PCM. Navarrete et al. [50] used pure Sn and SnO_X as a PCM and shell, respectively, in TH66 fluid. To investigate the size effect on TES, fine (60 to 80 nm) and coarse (less than 300 nm) nanoparticles were used. The melting temperatures of fine nanoparticles, coarse nanoparticles and bulk Sn were measured by the DSC test to be 225.5, 232.1 and 229.5 °C, respectively. While differences in the melting points are not remarkable, the difference in the crystallization temperatures was about 118.9, 90.3 and 2.3 °C for fine nanoparticles, coarse nanoparticles and bulk Sn, respectively. Heterogeneous nucleation of nanoparticles was not observed during solidification due to the small size and purity of the PCM. As a result, homogenous nucleation should occur at lower temperatures. By decreasing the particle size, the latent heat was also reduced. DSC tests were carried out on both nanoparticle samples. It was found that the enthalpy value was reduced by 15% for fine nanoparticles and by 9% for coarse nanoparticles during 50 cycles; subsequently, it became stable. It was reported that by increasing the temperature to 400 °C, the shell on the fine nanoparticles was fractured and pure Sn was exposed and oxidized, but the shell of coarse nanoparticles maintained its stability. This will in turn affect the enthalpy of the phase change in the nanoparticles.

In summary, the addition of nanoparticles to the fluid improves its thermal behavior (Figure 7). It is possible to add nanoparticles to fluids to increase their TES. Another proposed option is to use an off-eutectic alloy, such as Sn/Pb alloy, to have some primary pro-eutectic solid particles in its mushy zone to encourage heterogeneous nucleation of the PCM, leading to smaller supercoolings [50].

Sheng et al. [61] used pure tin-encapsulated spheres as core/shell PCMs. The Sn particle diameter and thickness of SnO₂ shells were 45–50 and 1–2 μm, respectively. They used a two-step procedure, including pretreating of Sn particles and oxidation heat treatment. The pretreating step included two various processes of vapor and water pretreatments. In vapor treatment, Sn particles were distributed in deionized boiling water, and the water was completely vaporized in 30 or 90 min. In water treatment, Sn particles were dispersed in 80 °C deionized warm water for 20 min and then dried. During the oxidation stage, pretreatment particles were placed in a furnace for 30 min at 500 °C under an oxygen atmosphere.

XRD results after the first step indicated the Sn peaks, but after the second step, both Sn and SnO₂ peaks, related to the core and the shell, were detected. SnO₂ peaks were undetectable after the first step due to the amorphous nature of the shell. In addition, according to Figure 8, the melting point of Sn was measured to be 232 °C during heating and 143 °C during the cooling cycle, indicating that the 87 °C supercooling required for its solidification. The latent heat of solidification was 53 J/g [61].

By increasing the heating/cooling cycles to 100 cycles, no significant changes were observed in the thermal properties. Moreover, no agglomeration of capsules was observed during the DSC tests, proving suitable thermal stability of the PCM. The thermal conductivity of the encapsulated PCM was 4.2 W/mK which was lower than that of pure Sn. However, pure Sn did not show acceptable stability during different cycles because of its oxidation during heating [61].

Lai et al. [62] used a 145 nm nanoencapsulated Sn-SiO_X core shell as PCM in a salt (7 wt% NaNO₃ + 53 wt% KNO₃ + 40 wt% NaNO₂) with 142–535 °C meting temperature range. This kind of composition is appropriate for solar energy applications. By increasing the Sn-SiO_X nePCM content of the salt to 5 wt%, the heat capacity of the fluid increased from 1.57 to 2.03 J/gK at 225 °C, and it remained stable for several cycles. If the thickness of SiO_X is less than 5 nm, pure Sn can be oxidized during heating, resulting in SnO formation. A 10 nm thickness was sufficient to protect the tin from oxidation. Moreover, the addition of more Sn-SiO_X into the salt resulted in excessive salt viscosity and deterioration of the nanofluid performance.

As mentioned before, the use of microencapsulated PCMs (MEPCMs) has expanded in TES applications. The most important limitation of MEPCMs is their rupture during the phase change process due to the expansion of PCMs. An innovative method called the “double-layer coating, sacrificial inner layer” procedure has been proposed recently to overcome this difficulty [45]. In this method, PCM particles are ultrasonically coated with two different agents and are then heat treated. The role of the first layer is to decompose into a gas layer during heating and to permit the expansion of metals during melting. Bao et al. [45] used pure tin as PCM, which was coated by polymethylmethacrylate (PMMA) and SiO₂ as the first and second layers, respectively. Heat treatment was performed at 400 °C at N₂ atmosphere to decompose the first layer into gas. Simultaneous with gas generation, pure tin melts, and volume expansion occurs. The CO₂ gas released from the MEPCMs does not allow for penetration of pure tin outside the MEPCMs due to Sn’s high surface tension and size. In fact, a “thermal expansion void” is produced to control the core expansion during phase changes. In this work, the melting and freezing temperatures of MEPCMs during the heating and cooling processes were 235 and 141.7 °C, respectively, indicating increased supercooling. Latent heat reached about 57.5 J/g which was close to 96% of that of pure tin. By increasing the thermal cycles to 200, the MEPCMs with voids withstood more than the EMPCMs without voids. Moreover, no leakage was observed on Tin particles and no structural change occurred during various cycles [62].

Three different mixtures, i.e., EP-01 to EP-03, of different core/shell particles, i.e., A to E, as shown in

Table 3. Compositions of initial and mixed powders of Sn_XZn_1-X/SiO_y core/shell particles. Adapted with permission from [63]. 2016, Elsevier.

Initial Powders	wt% Zn	wt% Sn	EP-01	EP-02	EP-03
A	100	0	22.11 wt% A	-	-
B	79.7	20.3	54.31 wt% B	23.58 wt% B	-
C	62	38	23.58 wt% C	54.31 wt% C	11.11 wt% C
D	34.8	65.2	-	22.11 wt% D	34.58 wt% D
E	14.2	85.8	-	-	54.31 wt% E

, were used as PCM. By using EP-01 to EP-03, the heat capacity changed from 85.9 to 67.4 J/g at 321–429 °C and 278–392 °C temperature ranges, respectively. Consequently, the EP-01 core/shell was selected for solar energy application due to its temperature range and heat capacity. By adding 5 wt% EP-01 into the salt, the eutectic temperature appeared at 370–407 °C and melting temperature at 420 °C. By time, the total power output reached 12.39 mW for pure salt and 14.96 mW for salt with 5 wt% core/shell particles at liquidus temperature. This proved that using the core/shell PCM particles could improve the thermal capacity of the salt [63].

Zhu et al. [64] used Al₂O₃ as capsule material and pure Sn as PCM for rapid TES application at 100–300 °C temperature range. Alumina has higher thermal conductivity than SnO₂; therefore, it is more effective for the heat transfer and melting of the PCM. After preparation of SnO₂ from the synthesis of SnCl₂, it was stir mixed with Al(NO₃)₃ solution, ammonia and deionized water, a procedure called boehmite treatment. After drying, the calcination process was done at 1000 °C in the air; then, the hydrogen reduction was performed at 560 °C, and, finally, Sn-Al₂O₃ was produced. By penetrating Al into the SnO₂ particle, they react to make a porous SnO₂/A_l2O₃ structure. After reduction, the Sn-Al₂O₃ core-shell, i.e., an alumina shell around Sn nanoparticles, is produced.

Figure 9a shows the effect of initial SnO₂ size on the formation and size of Sn particles. For all 60–1600 nm particles, both SnO₂/Al₂O₃ and Sn/Al₂O₃ particles coexist. It appears that for larger SnO₂-Al₂O₃ particles, Al cannot diffuse into the deeper part of SnO₂, and as a result, larger Sn particles are observed at the center, and some very smaller ones around the central region. By decreasing the SnO₂ dimension, Al can diffuse into deeper parts to make smaller alumina-encapsulated Sn particles. For Sn-Al₂O₃ with 1600 nm SnO₂ particles, the melting and solidification temperatures of Sn particles were measured to be 229.1 and 137.1 °C, respectively. The decrease in the melting point of core-shell particles in comparison with that of commercially pure Sn (231.6 °C) is related to the size effect. The thermal stability test at 100–250 °C temperature range until 100 cycles did not show any difference in the microstructure before and after thermal cycling (Figure 9b). However, SAED patterns illustrated that while small Sn particles could oxidize during different cycles, affecting the thermal properties of the components, larger particles retained their purity. Consequently, smaller particles have more potential for oxidation and their solidification needs larger supercoolings. Therefore, 1600 nm core–shell particles were found to be more suitable for TES applications. No rupture was observed on the alumina surface of these particles, and it was stable during thermal cycling. The thermal storage ability of 1600 nm Sn-Al₂O₃ was 54 J/g which was suitable for the mentioned application [64].

6. Metallic Alloys as PCM

Since high-purity metals are not readily accessible, use of their alloys has been extensively practiced. Metallic alloy PCMs have been considered PCMs due to their large heat of fusion and sensible heat during solidification/melting processes [65]. Earlier researchers worked on different binary and ternary eutectic alloys, as well as other metallic alloys, to find their melting point and latent heat of fusion by both thermodynamic calculations and thermal analyses [66,67]. This section is divided into three subsections corresponding to the binary, ternary and quarterly metallic alloys used as LMPM-PCM.

6.1. Binary Alloys

Adinberg et al. [68] utilized Zn₇₀Sn₃₀ alloy as PCM and the eutectic composition of 26.5% bephenyl/73.5% diphenyl oxide as a heat transfer fluid (HTF). These materials have good applications in solar energy systems for producing electricity from superheated (350–400 °C) water steam. The PCM starts to melt by absorbing the heat and its temperature reaches 400 °C. According to the Zn-Sn phase diagram [68], the liquidus and solidus temperatures of the alloy are 370 and 198.5 °C, respectively. Upon decreasing the alloy temperature to 270 °C, the total heat reaches about 107 J/g, which includes the sensible heat (37 J/g) and the fusion heat (70 J/g) of zinc. Alloys with higher zinc concentrations have higher storage heat capacity. It seems that Zn₇₀Sn₃₀ is a perfect choice for applications with a working temperature of about 400 °C.

The large difference between the liquidus and solidus temperatures of Zn₇₀Sn₃₀ alloy results in a decrease in steam temperature, which may significantly affect the efficiency of the system. To overcome this problem, a secondary heat source can be used to keep the temperature between 340–360 °C. When the steam temperature starts to decrease, the secondary heat source gives the required energy to the steam to keep the temperature within 340–360 °C range. By melting the PCM, extra heat is given to the secondary source by steam to prevent the increase of steam temperature more than 400 °C. This heat can be used again in subsequent cycles [68].

The use of Mg-51 wt% Zn as PCM is not only very important for LHTES applications, but it is also very valuable in Direct System Generation (DSG) in solar energy systems [69]. According to the Mg-Zn phase diagram, five different types of intermetallic phases can be formed in the alloy, depending on the cooling/heating rates. When the cooling and heating rates are 0.1 °C/min, the final composition is a mixture of Mg₂₁Zn₂₅ + α-Mg, which is compatible with the Mg-Zn phase diagram. Upon increasing the cooling rates up to 5–10 °C/min, only Mg₅₁Zn₂₀, as a metastable phase, and a few α-Mg phase are observed at room temperature. Mg₅₁Zn₂₀ can convert into an eutectoid composition at 325 °C according to the phase diagram. By increasing the cooling rate, the system does not have enough time to form an eutectoid composition [69].

Thermal analyses demonstrated two peaks during the heating process, which were related to the eutectoid transition (at 329 °C) and the melting temperature (at 342 °C), with a total energy of 155 J/g. Only one peak was observed during the cooling process at 337 °C with 127 J/g solidification enthalpy; this peak was related to precipitation of Mg₅₁Zn₂₀ phase. Sometimes, a mixture of Mg₄Zn₇ and α-Mg with 153 J/g solidification enthalpy and melting temperature of 332 °C is formed [69]. The reason is not so clear, but it can be related to nucleation difficulties.

The experiments showed that after 20 cycles, there were no significant changes in the PCM weight, indicating the appropriate thermal stability of the PCM during repeated melting and solidification cycles. The most important properties of Mg-51 wt% Zn alloy are high thermal conductivity, thermal diffusivity and energy density in comparison with the molten salts typically used in this application [69]. According to the thermal properties and stability of Mg-Zn eutectic alloy, it can be considered a suitable candidate for DSG application.

Dey et al. [70] used a hypereutectic ZA12 (Zn-12 wt% Al) alloy as a PCM for Latent Heat Thermal Energy Storage (LHTES). Solidus and liquidus temperatures were determined to be 370 °C and 440 °C, respectively. Thermal diffusivity at the solid and liquid state, C_P and latent heat of fusion were determined to be 3.04 × 10⁻¹ and 5.04 × 10⁻¹ cm²/s, 0.531 J/gK and 130.68 J/g, respectively. Despite the large solidification range of the alloy, it was reported as a suitable PCM for LHTES applications in solar energy systems, considering its high thermal conductivity [70].

Kawaguchi et al. [71] used Zn-30 wt% Al alloy as a PCM in TES applications. To produce an MEPCM, atomized Zn-30 wt% Al powder was boiled in distilled water and dried. Then, pure oxygen was applied to the dried powder while it was heated to 800 °C at 10 °C/min heating rate and held for 3 h at 800 °C. The powder was then cooled down to ambient temperature at 50 °C/min cooling rate and heat oxidation was performed on them. This would result in the formation of a ZnO external shell around PCM particles, while an internal Al₂O₃ coating was between ZnO and the PCM (MEPCM sample). Glass Frit (GF) was also added to the MEPCM as a sintering agent to synthesize a composite MEPCM. In this process, a 50–50 mixture of MEPCM and GF was shaped to cylindrical pellets under pressure. The samples were heated to 800 °C and held at this temperature under an oxygen atmosphere for 1 h (MEPCM-GF sample) [71]. DTG and DSC tests were used to determine the repeatability and thermal properties of the MEPCM and MEPCM-GF samples. The results indicated that within the 25–415 °C temperature range, the weight of the composite sample decreased because of AlOOH dehydration (2AlOOH→Al₂O₃ + H₂O). Within the 415–800 °C range, the weight increased because of the formation of ZnO. During the initial heating process, AlOOH is dehydrated into Al₂O₃ close to the Zn-Al alloy and cracked. Then, the weight of the sample increased due to the solid–liquid transition of PCM. According to Figure 10, during holding samples at 800 °C, Zn could penetrate out of the cracks and on the Al₂O₃ coating, where it was oxidized and converted to a ZnO coating. Consequently, two different types of coatings, i.e., a 300 nm Al₂O₃ layer and a 500 nm ZnO layer, were detected after oxidation treatment. After 100 heating and cooling cycles at 10 °C/min, composite samples showed suitable thermal durability between 400–500 °C with 48 J/g heat capacity. The heat capacities of Zn-30 wt% Al and MEPCM were also 166 and 177 J/g, respectively. Clearly, the heat capacity of the composite PCM was less than that of MEPCM, but it showed better stability during repeated cooling and heating cycles [71].

6.2. Ternary Alloys

Mg-Zn-Al ternary alloys have frequently been studied as PCMs. Risueno et al. [72] used two eutectic alloys of Mg₇₀Zn_24.9Al_5.1 and Zn_85.8Al_8.2Mg₆ compositions to investigate their thermal properties and chemical stability of Mg₇₀Zn_24.9Al_5.1 in different containers, such as 304, 304 L, 316 and 316 L stainless steel for CSP application. The thermophysical properties of both alloys are shown in

Table 4. Thermophysical properties of Mg₇₀Zn_24.9Al_5.1 and Zn_85.8Al_8.2Mg₆ alloys. Adapted with permission from [72]. 2015, Elsevier.

PCMs	Tm (°C)	Lm (J/g)	ρ (g/cm3)	CTE (10−6 °C−1)	CP (J/gK)
					25 °C	300 °C
Mg70Zn24.9Al5.1	340	157	0.282	24	0.690	0.830
Zn85.8Al8.2Mg6	344	104	0.619	34	0.410	0.530

. According to Table 4, melting points of both alloys are very similar, but because of the higher Mg content of Mg₇₀Zn_24.9Al_5.1 alloy, its enthalpy and C_P are higher and its density is less than those of Zn_85.8Al_8.2Mg₆ alloy. Linear thermal expansion of both alloys is constant at solid state but becomes higher at liquid state for Zn_85.8Al_8.2Mg₆ alloy due to the presence of Zn. In addition, there was no corrosion at PCM/stainless steel container interfaces [72].

Thermophysical properties of Zn₈₄Al_8.7Mg_7.3, Zn_88.7Al_11.3 and Zn_92.2Mg_7.8 alloys as PCM for LHS applications have been investigated [73]. Physical properties of the alloys are illustrated in

Table 5. Thermophysical properties of Zn₈₄Al_8.7Mg_7.3, Zn_88.7Al_11.3 and Zn_92.2Mg_7.8 alloys. Adapted with permission from [73]. 2017, Elsevier.

Alloy	CP (Solid) J/gK	k (Solid) W/mK	k (Liquid) W/mK	Transitions	TTransition (°C)	Q (J/g)
Zn84Al8.7Mg7.3	0.457	71	45	α-Al + η-Zn → β-Al (Zn rich Al solid solution)	282	4
				β-Al + η-Zn + Mg2Zn11 → L	344	132
Zn88.7Al11.3	0.449	133	53	α-Al + η-Zn → β-Al	285	5
				β-Al + η-Zn → L	377 = TS 382 = TL	118
Zn92.2Mg7.8	0.466	87	35	η-Zn + MgZn2 → L (metastable phases)	362 = TS 371 = TL	106

. Compositions of the first two alloys are compatible with the eutectic compositions in the corresponding phase diagrams, while there is metastable Zn and MgZn₂ with a trace of Mg₂Zn₁₁ in the third alloy. This metastable eutectic is formed under normal solidification conditions. The stable eutectic phases in Zn_92.2Mg_7.8 alloy are Zn and Mg₂Zn₁₁ which are formed under rapid solidification conditions. As MgZn₂ is a metastable phase, it can be converted to Mg₂Zn₁₁ by a peritectic transition (MgZn₂ + L → Mg₂Zn₁₁). For this reason, there is a trace of the Mg₂Zn₁₁ stable phase between the metastable phases. Comparison of these alloys with Mg-based alloys and KOH and KNO₃ salts demonstrated that the heat density of Zn-rich alloys is higher than that of the Mg-based alloys and is much higher than those of the salts. As a result, due to high thermal conductivity, the required area for heat exchange decreases, as does the required volume and cost [73].

In another work [65], short-term (until 100 cycles) and long-term (until 500 cycles) thermal stability of Zn₈₄Al_8.7Mg_7.3, Zn_88.7Al_11.3, Zn_92.2Mg_7.8, Mg₇₂Zn₂₈ and Mg₇₀Zn_24.9Al_5.1 alloys was investigated by DSC test. Changes in the melting point and latent heat of the alloys with short-term thermal cycling are shown in

Table 6. Melting temperature and latent heat of Al-Mg-Zn eutectic alloys at different cycles. Data from [65].

Alloys	Tm (1st Cycle) °C	Tm (100th Cycles) °C	Lm (1st Cycle) J/g	Lm (100th Cycles) J/g
Zn84Al8.7Mg7.3	343.69	343.88	131.5	129
Zn88.7Al11.3	381.72	381.72	118.4	119.9
Zn92.2Mg7.8	370.36	370.48	106.4	85.39
Mg72Zn28	341.07	341.22	152.7	150.1
Mg70Zn24.9Al5.1	339.98	340.37	159.6	158.9

. The large difference in heat of fusion between the first and the 100th cycles of Zn_92.2Mg_7.8 alloy is related to the formation of two metastable components, i.e., MgZn₂ and Mg₂Zn₁₁, which proves that this alloy is not a suitable option as a PCM. Moreover, there is a very small change at melting temperature and latent heat of other alloys, which is associated with oxidation of alloys during the tests [65].

Table 7. Long-term repeatability of Mg₇₂Zn₂₈ and Mg₇₀Zn_24.9Al_5.1 samples. Data from [65].

No. of Cycles	Mg72Zn28		Mg70Zn24.9Al5.1
	Tm (°C)	Lm (J/g)	Tm (°C)	Lm (J/g)
1	340.38	154.9	341.09	152.9
50	340.84	153.3	340.28	155.2
100	340.15	155.4	339.29	156.8
300	341.09	153.6	340.80	154.9
500	340.64	154.3	340.62	156.1

illustrates the results of long-term thermal cycling of Mg₇₂Zn₂₈ and Mg₇₀Zn_24.9Al_5.1 samples. As shown, there is very little variation in the thermal properties of both alloys after 500 cycles. In total, the Zn₈₄Al_8.7Mg_7.3, Zn_88.7Al_11.3, Mg₇₂Zn₂₈ and Mg₇₀Zn_24.9Al_5.1 specimens are appropriate samples for TES applications [65]. Unfortunately, the reasons for some inconsistencies between the results of the first and 100^th cycles in short- and long-term cycling tests have not been clarified in this paper.

It was found that Mg-Zn-Al ternary alloys had a more suitable performance in TES applications in comparison with Mg-Zn binary alloy [74]. Thermophysical properties of Mg₇₀Zn_24.9Al_5.1, Mg₇₁Zn_28.9Al_0.1 and Mg₇₀Zn_24.4Al_5.6 alloys were explored to distinguish their potential as PCM. The results indicated that, unlike the Mg-51 wt% Zn binary alloy [69], neither cooling nor heating rates had any influence on the ternary alloy transitions. During the heating of Mg₇₀Zn_24.4Al_5.6, only a peak at 341 °C temperature with 157 J/g latent heat. During cooling, two separate peaks at 359 °C (solidification of primary crystals) and 333 °C (eutectic transition) were observed. For Mg₇₀An_24.9Al_5.1 composition, there was one peak at 340 °C in heating and one peak at 330 °C in cooling with 157 J/g total energy. For Mg₇₁Zn_28.9Al_0.1 composition, two eutectoid and eutectic transitions at 335 °C and 343 °C, respectively, with 153 J/g total enthalpy in heating, and one overlapped peak at 332 °C with 154 J/g transition enthalpy in cooling were detected. The thermal properties of all the samples are represented in

Table 8. Thermal properties of some Mg-Zn-Al alloys. Data from [74].

Composition	CP (J/gK) at 25 °C	α (cm2/s) at 25 °C	k (W/mK) at 50 °C
Mg71Zn28.9Al0.1	0.7	26.2 × 10−2	56.5
Mg70An24.9Al5.1	0.72	22.5 × 10−2	46.8
Mg70Zn24.4Al5.6	0.73	20.4 × 10−2	40.8

[74].

By increasing the aluminum content, thermal diffusivity and thermal conductivity are reduced. In addition, the thermal conductivity of all alloys is smaller than that of Mg₇₂Zn₂₈ eutectic composition. These differences are related to the presence of aluminum in both pure Mg and Mg₂₁Zn₂₅ intermetallic phases that can diminish their unit cells and the average free path of electrons in the structure. Moreover, the thermal stability and properties of ternary alloys, especially of Mg₇₀An_24.9Al_5.1 alloy which is more stable than other alloys, are vital in their selection as a PCM for TES applications [74].

Nieto-Maestre et al. [75] used Thermo-Calc software to determine the melting temperature and latent heat of different ternary and quaternary eutectic metallic alloys, which were suitable for DSG applications to produce high pressure steam (about 100 bar) with 285–330 °C melting temperature. It was demonstrated that Mg-Zn-Cu and Mg-Zn-Ni were the best choices as PCM for DSG applications.

6.3. Quarterly Alloys

Bi-21 wt% In-18 wt% Pb-12 wt% Sn alloy was investigated as PCM for TES applications in electronic devices to reduce their temperature [76,77,78]. The melting temperature and thermal conductivity of the alloy are 58.2 °C and 7.143 W/mK, respectively. Due to the higher thermal conductivity and lower specific heat capacity of the alloy than those of the organic PCMs, it can adopt itself with temperature changes more easily, making the heat distribution in the electronic devices more homogenous and the temperature fluctuations during phase change less noticeable [76,78].

Another application of Bi-21 wt% In-18 wt% Pb-12 wt% Sn and Ga-13.5 wt% Sn alloys with 58 °C and 20 °C melting temperatures, respectively, is temperature control in aerospace [77]. Physical properties of these PCMs were investigated in the 30–50 °C temperature range. To investigate the thermal stability of PCMs, up to 500 heating and cooling cycles were applied. The results indicated that the Ga-13.5 wt% Sn alloy had desirable thermal durability in different cycles. For the Bi-In-Pb-Sn alloy, the required time to reach 70 °C was about 1150 s, and the required time to reach 40 °C was about 1055 s for the Ga-Sn alloy. Therefore, the holding time increases with the use of metallic PCMs. The results also demonstrated that the aluminum container was corroded after 10 h of contact with the Ga-Sn alloy. An Al₂O₃ coating on the inner surface of the container could protect the container from corrosion with the alloy.

In conclusion, a survey of the literature shows that both low melting point pure metals and alloys have been used as a PCM in different applications. However, the main application of low melting metallic PCMs is in solar energy systems (TES, DSG, CSP and electricity production) and cooling of electronic devices. Bi base alloys have been used extensively for the latter applications. Reactions (corrosion/dissolution) between the metal PCMs and their container, and the container and the surrounding material, must be controlled by careful selection of the container material or the coatings applied on the container surfaces. Pure PCMs usually display good thermal stability during various heating and cooling cycles (up to 200 cycles). However, pure metal PCMs need higher supercooling/overheating for the transformation process to initiate. To overcome this issue, the addition of proper nucleants to the PCM has been proposed.

Since metallic alloys are much more diverse and available than pure metals, the use of metallic alloy PCMs has progressed more rapidly. However, solidification of binary alloys is sensitive to the cooling rate. By increasing the cooling rate, some metastable compounds could be produced, which may affect the thermal properties and reversibility of the transitions. It seems that binary alloys are more sensitive to cooling rates than ternary alloys. Moreover, the energy storage density of Zn base ternary alloys is 199–240 kWhm⁻³ which it is more than that of Mg base alloys (122–128 kWhm⁻³). Customarily, the composition of the alloy PCMs is so selected to have a constant transformation temperature, e.g., eutectic composition. However, some off-eutectic alloys such as Zn70Sn30, ZA12 and Zn-30 wt% Al are also being extensively used as PCM owing to their high latent heat and thermal diffusivity.

7. Metal Matrix Composite PCMs

Metal matrix composites (MMCs) are extensively used in many industries. They are made by physical mixing of some reinforcing particles, whiskers, fibers or sheets in a metallic matrix, provided good bonding between the reinforcement and the matrix is formed. In structural applications, reinforcing of the metallic matrix is often carried out to improve some mechanical properties of the produced composite. In PCM-reinforced composites, however, the main purpose is to improve some thermal characteristics of the PCM. Reinforcing metals with particles, whiskers, fibers or sheets of a PCM is still among the new frontiers in the development of MMCs. These types of PCMs consist of a high thermal conductivity material as a matrix and PCM as a reinforcement.

Very few studies have investigated the effects of the addition of metal nanoparticles to metallic matrices on their thermal properties. Liu et al. [79] used Bi nanoparticles as a reinforcement and PCM in an Ag matrix. Since the solubility of Ag and Bi in each other is very low, no solution or intermetallic phase is formed between them during the heating and cooling cycles. DSC results indicated an endothermic peak at 246 °C related to melting of Bi nanoparticles during the heating cycle. Moreover, three exothermic peaks were detected during the cooling cycle, proving that nucleation of Bi occurred in three different stages. While the melting temperature of pure bulk Bi is about 271 °C, DSC results demonstrated melting of particles at 246 °C due to the particle size effect. The melting temperature and enthalpy of fusion were reduced from 252 to 236 °C and 37.6 to 20.1 J/g for Bi particle sizes from 14.9 nm to 8.1 nm, respectively. After 100 heating and cooling cycles, no significant changes in the melting temperature or fusion enthalpy were noticed. By increasing the nanoparticle volume to 34 vol%, the volumetric fusion enthalpy rose. By increasing the fraction of 13 nm Bi particles from 0–10 vol%, the thermal conductivity of the nanocomposite was reduced from 270 to 128 W/mK. This is believed to be due to the increased interfaces between the matrix and nanoparticles, which may act as a barrier for heat transfer. Since the thermal conductivity of Ag is higher than that of Bi, the addition of the reinforcement into the matrix reduces the continuity of the matrix and, as a result, the thermal conductivity of the nanocomposite. Consequently, two important parameters, including the particle size, which affects the melting temperature of the PCM, and the volume fraction of the nanoparticles, which influences the enthalpy of fusion, can influence the thermal properties of the nanocomposite.

In the previous section, the physical properties of Bi-In-Pb-Sn alloy were studied. In order to improve the thermal conductivity of Bi-In-Pb-Sn alloy [78], a porous carbon foam was placed in a 3 mm thick steel container and put into the furnace, where the Bi-In-Pb-Sn alloy was injected into and filled the foam completely. The steel container was then removed and a 10,000, 15,000, 20,000 and 25,000 W/m² heat flux was applied to the heat exchanger on Bi-In-Pb-Sn and Bi-In-Pb-Sn/carbon foam (matrix), separately. Figure 11 shows a T-t diagram of both samples. Due to the higher thermal diffusivity of the composite in comparison with pure Bi-In-Pb-Sn, there is a delay in the composite thermal sink, so the sample can absorb more sensible heat and hold the temperature at a constant range without much fluctuations. By increasing the heat flux, the composite thermal sink has more control over temperature changes, and it is possible to use the composite at higher heat flux [78].

8. Miscibility Gap Alloys (MGAs) as a New Generation of PCMs

Despite the growing use of metallic PCMs in different applications [57,70], potential leakage of the molten phase at high temperatures [80] and formation of intermetallic components [81] restricts their application. To overcome these issues, immiscibility gap alloys (MGAs) have been recently introduced in the PCM field as an innovative solution [21]. Immiscible alloys have been known for centuries. Cu–Pb, Cu–Fe, Cu–Cr, Cu–Co, Al–Pb and Al–Bi systems are among the prominent immiscible alloys used as high electric conductive, magnetoresistive or bearing materials [82].

In immiscibility gap phase change alloys, two thermodynamically stable immiscible metals are mixed using powder metallurgy techniques [83]. The desired microstructure of an MGA comprises a rather uniform dispersion of discrete active particles (phases) within a continuous passive matrix phase [84]. For instance, the “inverse microstructures” of Al-35% Sn and Fe-35% Cu MGAs are shown in Figure 12 [84]. This inverse microstructure, which is a converse of the microstructure of solidified two-phase alloys from the liquid region, has turned MGA into an efficient high thermal energy storage system. As the melting temperature of the dispersed phase is lower than that of the matrix, the dispersed phase plays a PCM role in each thermal cycle while still trapped within a continuous solid state matrix. During the solidification process, the latent heat of the molten phase is transferred away by the dense matrix phase [83].

Any alloy with immiscible phases at a solid state, such as Fe-Mg, Fe-Cu and Al-Sn, can be used as an MGA system [3]. Furthermore, MGAs can be recyclable, have higher thermal conductivity and energy density than most other thermal storage materials, can be applied in a wide temperature range, and have good durability during thermal cycling [3,84,85].

MGAs can be classified according to the type of materials, i.e., all metals, metal–semimetal or metal–nonmetal components, according to the number of units (binary, ternary, etc.), according to the required thermal properties, e.g., thermal density and thermal conductivity and according to their phase change temperature [83,86,87].

According to the phase change temperature, MGAs are classified into the following groups:

• In low-temperature systems, the phase change temperature is less than 300 °C. These MGAs can be used in waste heat recovery and space heating. For example, Sn 50%-Al with a 230 °C phase change temperature (the melting point of Sn as the active phase) is in this category [87].
• In medium-temperature systems, the phase change temperature is from 300 to 500 °C. Zn 50% -C with a 420 °C melting temperature is one of the most important systems of this group [87]. Zn-C systems are used as concentrator systems in solar thermal power (CSP) plants [86].
• High-temperature systems where the phase change temperature is between 500 and 750 °C. Such systems are appropriate for the generation of steam in dish CSPs. A suitable PCM for this application is Mg 50%-Fe with a 650 melting temperature [87].

It should be noted that the higher the phase change temperature of a system, the higher its energy density. Therefore, high melting temperature systems, i.e., those with stronger atomic binding energy, have the maximum energy density and vice versa. Research has shown that thermal features of MGAs can be influenced by mixing technique, compressing and sintering temperature and concentration of PCM particles. With regard to the mixing technique and compression temperature, Confalonieri et al. [3] produced Al-40Sn wt% MGA by powder metallurgy using ball milling. Ball milling features a more homogenous dispersion of PCM particles and results in a finer microstructure compared to more conventional mixing processes. The acquired fine microstructure associated with ball-milling prevents percolation of the active phase and enhances the thermal response. In this work, samples were compressed at 220 °C, a temperature lower than the melting point of Sn (232 °C), or at 240 °C where Sn is molten. DSC analyses demonstrated that for the samples compressed at 240 °C and exposed to 100 thermal cycles, solidification peaks changed toward a broader range. It confirmed that this Al-Sn alloy stored the heat quickly but released the heat more slowly in a wide temperature range, which may be very beneficial for thermal management.

In terms of sintering temperature, it was found that in Al-Sn, with a low thermal conductivity PCM phase (i.e., Sn), sintering temperature below PCM melting temperature improved the overall thermal conductivity of MGA. However, in Fe-Cu alloy, sintering above the melting temperature of the high thermal conductivity PCM (i.e., Cu) was needed to improve the thermal conductivity [84,88].

Investigations of the effects of volume fraction [88] and concentration of the PCM phase [21] on thermal conductivity of MGA have shown that in the Al-Sn system, since thermal conductivity of Al (matrix phase) is more than that of Sn (dispersed phase), an increase in volume fraction of the PCM (Sn) results in reduction of the alloy effective thermal conductivity [21]. However, in the Fe-Cu system where the thermal conductivity of the PCM phase (Cu) is higher than that of the matrix (Fe), the rise of the Cu volume fraction improves the overall thermal conductivity.

9. Innovative Application of Metallic PCMs

Low melting point metallic PCMs are still in their infancy and very few new applications are being publicly developed. Innovative applications of composite LMPM-PCMs and MGA-PCMs are still emerging for thermal energy storage, as described in previous sections. Another innovative application of PCMs is the use of multiple PCMs (M-PCMs) with different thermophysical properties to increase TES efficiency and reduce the thermal energy lost. In this method, every PCM has a different melting temperature and latent heat and is activated at a corresponding temperature range absorbing/releasing a certain amount of heat [89,90,91]. Figure 13, for example, an M-LMPM-PCM system is proposed to manage the temperature of a heat transfer fluid (HTF). In this system, four LMPM-PCMs are in a descending melting temperature arrangement from the input heat. During the charging process, the temperature of the HTF drops as it flows in from the left-hand side. However, all the PCMs can still be melted to release their latent heat and keep the HTF at the desired temperature range. During the discharging process, the temperature of the cold HTF entering the channel from the right increases gradually as it absorbs the latent heat of the solidification of the lower melting point PCMs. This provides for the melting of the higher melting point PCMs at the left side of the channel.

Some innovative applications of LMPM-PCMs in casting design were recently introduced by the authors. Many of our everyday items, as well as some of the most sophisticated engineering components, are made by casting. Control of the solidification microstructure during casting is vital for achieving the required mechanical properties and performance of the cast parts. Some of the traditional methods developed to control the solidification structure include controlling the cooling rate [92,93] and pouring temperature [94], grain refinement [93,95], dynamic nucleation [96], application of pressure [97], semisolid casting [98], application of magnetic and electric fields [99], superheat treatment [100] and directional solidification [101].

Recently, Noohi et al. [20] introduced an innovative method to control the solidification microstructure of an aluminum alloy. They used pure zinc as PCM embedded in a steel chill to affect the cast macrostructure of an Al-4.5 wt% Cu alloy. A schematic of the experimental setup is shown in Figure 14. They poured two identical castings, one chilled with a solid steel block and the other chilled with a steel container filled with pure zinc with the same overall cooling capacity. They showed that zinc in the latter melted after a given period of time and absorption of its latent heat of melting from the aluminum melt affected the cooling conditions and macrostructure of the casting.

Figure 15 depicts the macrostructures of two samples cast under similar conditions, except for the use of Zn PCM. As it is clear, the macrostructure, as well as the length of the columnar grains, columnar-equiaxed transition and size of the equiaxed grains, have been affected by the presence of PCM. Using experimental and computer simulation results, Noohi et al. [20] discussed that by using the zinc PCM, the thermal diffusivity of the new chiller increases, especially when the pure zinc starts to melt. At this time, it absorbs its latent heat of melting from the aluminum melt and increases the cooling rate and temperature gradient in front of the solidification front.

In another work on an Al-Cu alloy, Noohi [102] demonstrated the change in morphology of feathery grains using zinc PCM. In other works by this group, Fathi and Niroumand [103] have examined the effects of aluminum, zinc and tin PCMs on the macrostructure, feedability and defect formation in some brass and aluminum alloys. They also studied the effects of PCM on the structure of a transparent model material by in situ observation of solidification [104]. These emerging results indicate the potential of metallic PCMs in smart tailoring of the solidification structure of castings by judicious selection of the type, dimensions and location of PCMs in casting molds.

Using PCM-fitted chillers is believed to open new horizons in smart control of the casting structure. The structure of conventional castings is typically comprised of three distinct regions of chill, columnar and central zone of large equiaxed grains due to the inevitable gradual decrease in the cooling rate and thermal gradient in front of the solidification front. Employing properly selected and designed PCM-fitted chiller(s) in the mold, one may produce completely uniform macro- and microstructures across the castings, or castings that have finer microstructure in the center, or even periodic or functionally graded microstructures across the castings. This requires design and strategic planting of a number PCM (or M-PCMs) fitted chillers that become active at predetermined stages during solidification of the casting. The concept can also be employed for promoting directional solidification, increasing the feeding distance, removing hard-to-feed hot spots, and changing the location of porosity formation in the castings.

10. Conclusions and Outlook

A review of the published literature shows that interest in Phase Change Materials (PCMs) has increased significantly in recent years. While PCMs have great thermal management potential in many different science and engineering fields, their major applications are still limited to solar energy systems and buildings. While a number of very good reviews can be found in the literature that highlight the characteristics and applications as well as the advantages and disadvantages of organic and inorganic PCMs, less attention has been paid to metallic PCMs. Metallic PCMs appear to be the best alternative to organic and inorganic PCMs due to their high conductivity, high latent heat storage capacity, wide-ranging melting temperature, chemical compatibility with the containers, and non-toxicity and availability. This paper aimed to review the latest achievements in the field of low melting point metallic PCMs (LMPM-PCMs). Pure, alloy, composite and miscibility gap LMPM-PCMs based on Zn, Ga, Bi, In and Sn elements were surveyed. The encapsulation and selection criteria for LMPM-PCMs and their containers were accounted for. Core/shell and shape-stabilized encapsulation methods can be employed for effective arrangement of nano, micro and macro encapsulated LMPA-PCMs in metal matrices.

Metallic PCMs offer great potential for new applications. In this paper, some emerging applications were presented and a novel application of LMPM-PCMs in casting design enabling the smart control of solidification microstructure was reported and suggested. This is an area that needs further investigation. Additionally, along this line, metallic PCMs may be integrated into the feeder design to increase the feeder life. The concept is that after pouring the molten metal in the mold, a molten metallic PCM is poured into specially designed channels around the feeder neck. The latent heat released by solidification of the molten PCM is transferred to the neck and hinders its solidification to increase the feeding time and casting yield. The incorporation of M-LMPM-PCMs into chillers activated after predetermined times from the casting can be utilized to control the cooling rate and the resulting microstructure during different stages of cooling.

In another potentially important application, LMPM-PCMs can be used for safety promotion in explosion-prone applications. For example, the auto-ignition temperature of gasoline used in vehicles is about 415–530 °C [105]. Fitting the fuel tank, fuel pump or fuel lines with LMPM-PCM filled containers can substantially decrease the risk of fuel overheating and its explosion in case of any malfunction.

Advancement of metal matrix composite PCMs by addition of metallic/non-metallic macro/nano-particles into the metallic PCMs to customize their heat transfer rates or supercoolings based on the required working conditions is another emerging area.

Finally, very few works on computer and numerical modeling of the behavior of metallic PCMs can be found in the literature. The development of computer software for the selection of metallic PCMs and their containers based on the service conditions is also needed.