Development of High-Entropy Shape-Memory Alloys: A Review

Abstract

High-entropy shape-memory alloys (HESMAs) exhibit unique properties, including high yield strength, a wide temperature range of phase transformation, and excellent thermal cycling stability, among others. Hence, they have broad prospects for research and application. Currently, various new high-entropy shape-memory alloys have been developed based on TiNi, Fe-based, Ti-based, and NiMn-based alloys. This article focuses on summarizing and analyzing the research status of various types of high-entropy shape-memory alloys, including their microstructure characteristics, martensitic phase transformation, and functional properties. The advantages and disadvantages of high-entropy shape-memory alloys compared to traditional shape-memory alloys are briefly discussed. Finally, based on the analysis of existing problems, future research prospects are proposed, hoping to provide assistance for the development and application of high-entropy shape-memory alloys in the future.

1. Introduction

Intelligent materials are a type of material that exhibit clear responses to external stimuli such as the changing of stress and temperature [1]. As an important smart material, the shape-memory alloy (SMA) has attracted much attention due to its unique shape-memory effect and superelasticity [2,3]. The shape-memory effect and superelasticity inherent in shape-memory alloys are derived from the martensitic phase transformation and its reverse transformation [4]. The shape-memory effect refers to the phenomenon where the shape-memory alloys deform at low temperatures (martensite) under external stress. As it is heated to a certain temperature (Af), it can recover its original shape owing to the reverse martensitic transformation. Superelasticity refers to a phenomenon in which, at a temperature above the austenitic end temperature (Af), a shape-memory alloy undergoes non-linear deformation and then experiences shape recovery upon unloading [5]. Due to the unique shape-memory effect and superelasticity, shape-memory alloys are widely used in many fields, such as automotive, aerospace, biomedical, and medical engineering [5,6,7].

In the research and application of shape-memory alloys, Ti-Ni shape-memory alloys show more excellent comprehensive properties. For example, the TiNi shape-memory alloy has an ultra-high strength of 2500 MPa via high-pressure torsion (HPT) treatment [8]; it can also withstand more than 20 million times (room temperature) of mechanical load stress without failure. Additionally, it has advantages such as good weldability and ductility (60–70%) [9], high corrosion resistance [10], high damping [11,12], and favorable biocompatibility [13], making it the most widely applied in various fields. However, conventional shape-memory alloys have a narrow range of phase transition temperatures, and the introduction of dislocations and precipitates at high temperatures leads to poor thermal cycling stability and is not suitable for extreme service environments such as deep-space exploration and deep-sea mining [14,15]. Therefore, in order to overcome the limitations of shape-memory alloys in stable operations under harsh environments, researchers have introduced the concept of high-entropy alloys (HEAs) into shape-memory alloys [16,17,18].

Compared with the conventional shape-memory alloys, HESMAs have a higher yield strength and a larger phase transition temperature [4,19,20] range to a certain extent due to the severe lattice distortion and slow diffusion [21,22,23]. Such excellent properties promote a wider range of applications for high-entropy shape-memory alloys. This present paper first introduces the design criteria of high-entropy shape-memory alloys. Then, the research progress on the microstructural features, phase transformation behavior, and functional properties of high-entropy shape-memory alloys, such as TiNi-based, Fe-based, and Ti-based alloys, etc., have been reviewed. The aim is to provide a theoretical basis for the composition design and performance optimization of high-entropy shape-memory alloys. Finally, the current problems and future research directions of high-entropy shape-memory alloys are presented.

2. Design Criteria for High-Entropy Shape-Memory Alloys

2.1. Selection Principles of Alloying Elements

HEAs are required to contain at least five different types of alloying elements. Moreover, the content of each atom lies between 5 and 35 atomic percent [24]. For HESMAs, two or more other alternative elements are introduced to partially replace the initial SMA composition. Therefore, it is essential to select substitute elements that possess good solubility and similar electronic configurations to the primary elements [25,26,27]. Figure 1 displays the equivalent elements for the five principal chemical elements in SMAs, which can promote the development of new types of shape-memory alloys in the future.

Furthermore, the martensitic transformation behaviors of shape-memory alloys are strongly influenced by their chemical composition [28,29,30]. For instance, Chang et al. [31] found that the martensitic transformation temperature decreases with the Cu content increasing in (HfTiZr)₅₀(CuNi)₅₀ high-entropy shape-memory alloys. To better understand the relationship between the chemical composition and the phase transition temperature, Peltier et al. [29] established a relationship between the forward martensitic transformation starting temperature (Ms) of HESMAs and their chemical composition, as shown in Equation (1). It can be seen from Equation (1) that Hf, Zr, Fe, and Al can lead to an increase in T_Ms, whereas Ti, Co, Cu, and Ni would cause a reduction in T_Ms. The relationship between the chemical composition and the martensitic transformation starting temperature can provide accurate predictions for designing high-temperature shape-memory alloys.

$$\begin{gathered} {\left(M_s^{calc}\right)}_i(℃)=-4.634{\left(Ti\right)}_i+16.146{\left(Hf\right)}_i+18.480{\left(Zr\right)}_i-46.399{\left(Nb\right)}_i-16.120{\left(Ta\right)}_i+15.775{\left(Fe\right)}_i \\ -27.146{\left(Co\right)}_i+35.265{\left(Al\right)}_i-24.485{\left(Cu\right)}_i-0.904{\left(Pb\right)}_i-13.966{\left(Ni\right)}_i\left(wt.\%\right) \end{gathered}$$

(1)

Lastly, the cost of original materials is also an important factor that should be considered during the process of designing high-entropy shape-memory alloys. For example, substitute elements such as Hf, Pd, Pt, and Au can achieve high-temperature shape-memory alloys [32,33], but their higher cost is not conducive to engineering applications. Therefore, it is necessary to develop a high-performance high-entropy shape-memory alloy, such as those with high phase transition temperatures, high strength, good shape-memory effects, superelasticity, etc., under the premise of low cost.

2.2. HEA Standard Verification

To develop high-entropy shape-memory alloys, in addition to selecting specific chemical compositions, three standards of high-entropy alloys also should be taken into consideration to avoid the formation of metallic glasses [34]. Baiz et al. [35] conducted a systematic investigation of the mixing entropy ($∆S_{mix}$) and the valence electron concentration (VEC) range of the existing high-entropy shape-memory alloys. It was found that the VEC value should be within the range of 6.8 to 8, and the $∆S_{mix}$ value should be 1.5R or higher, where R is the gas constant (R = 8.314 J K⁻¹ mol⁻¹), [25,26,27] in order to obtain the high-entropy shape-memory alloys. Hence, the values of $∆S_{mix}$ and VEC should be important considerations in the designing of HESMAs.

2.2.1. Mixing Entropy and Enthalpy

From a thermodynamic perspective, the formation of stable solid solutions in high-entropy alloys (HEAs) requires the system to have the lowest mixing Gibbs free energy. This is related to both the mixing entropy and the mixing enthalpy ($∆H_{mix}$) [36]. The higher the disorder, the greater the difference in the free energy of the solid–liquid two phases, making metal crystallization easier. Therefore, the phase stability of HEAs requires a high mixing entropy and zero or close to zero enthalpy of mixing [37]. For multi-component high-entropy shape-memory alloys, the mixing Gibbs free energy can be expressed as the following equation [38]:

$$∆G_{mix}=∆H_{mix}-T∆S_{mix}$$

(2)

where T is the absolute temperature.

The mixing entropy and mixing enthalpy of multi-principal components can be expressed as the following formula, respectively:

$$∆S_{mix}=-R{{\displaystyle \sum }}_{i=1}^n{C}_i\ln Ci$$

(3)

where R stands for gas constant, and $C_i$ stands for the atomic percentage of the ith element.

$$∆H_{mix}={{\displaystyle \sum }}_{i-1, i\ne j}^n{\mathsf{\Omega }}_{ij}{C}_i{C}_j$$

(4)

where ${\mathsf{\Omega }}_{ij}$ is the enthalpy of mixing of a binary liquid alloy composed of equal atoms of element $i$ and element $j$; $C_i$ and $C_j$ are the atomic fractions of the $i$-th element or $j$-th element.

2.2.2. Valence Electron Concentration

The valence electron concentration is the ratio of the total number of valence electrons of each component element to the total number of electrons. When the valence electron concentrations of the alloying elements are similar, the solid solubility of them is greater. This means that they can form a much more stable solid solution. The valence electron concentration can be calculated using the following equation [39]:

$$\mathrm{VEC}={{\displaystyle \sum }}_{i=1}^n{X}_i{\left(\mathrm{VEC}\right)}_i$$

(5)

where ${\left(\mathrm{VEC}\right)}_i$ is the number of valence electrons in the nucleus of the $i$th component.

3. TiNi-Based High-Entropy Shape-Memory Alloys

3.1. (TiZrHf)₅₀(NiCoCu)₅₀ Series High-Entropy Shape-Memory Alloy

Currently, the research on high-entropy shape-memory alloys is mainly focused on equiatomic Ti-Ni-based high-entropy shape-memory alloys [40]. In 2014, Firstov et al. [17] prepared a B2-structured (TiZrHf)₅₀(NiCoCu)₅₀ high-entropy shape-memory alloy by adopting arc-melting technology for the first time, on the basis of the Ti₅₀Ni₅₀ shape-memory alloy. In the present high-entropy shape-memory alloy, Hf and Zr of the same group replace Ti, whereas Co and Cu of neighboring groups replace Ni element. Although Firstov et al. observed that the microstructure of these alloys contains a typical dendritic structure, this microstructure has not been reported in the literature. Kosorukova et al. [41] explored the microstructural features of the casted (TiZrHf)₅₀Ni₂₅Co₁₀Cu₁₅ high-entropy shape-memory alloy. As shown in Figure 2a, the microstructure is characterized by a typical dendritic structure containing plate-like martensite in the dendritic region. Moreover, a high density of second phases appears between the dendrites. Hinte et al. [42] also reveal that the nano-scale precipitates in the as-casted (TiZrHf)₅₀Ni₂₅Co₁₀Cu₁₅ high-entropy shape-memory alloy, as shown in Figure 2b. These high-density nano-precipitates uniformly distribute in the interdendritic regions, which can significantly affect the martensitic transformation, mechanical properties, and functional properties. Subsequently, Chen et al. [43] investigated the microstructure, martensitic transformation, and mechanical/functional properties of the solution-treated (TiZrHf)₅₀Ni₂₅Co₁₀Cu₁₅ alloy at 1273 K. The dendritic microstructure disappears in the solution-treated (TiZrHf)₅₀Ni₂₅Co₁₀Cu₁₅ shape-memory alloy. Instead, the microstructure is composed of a gray matrix, a black Ti₂Ni-like phase, and white (Ti,Zr,Hf) carbide. After solution treatment, the martensitic transformation starting temperature (T_Ms) decreases to around 309 K. However, the solution treated (TiZrHf)₅₀Ni₂₅Co₁₀Cu₁₅ HESMAs exhibit a much higher yield strength under the same shape-memory effect compared with the Ti₅₀Ni₅₀ shape-memory alloy. Furthermore, the recoverable strain of solution-treated (TiZrHf)₅₀Ni₂₅Co₁₀Cu₁₅ HESMAs is significantly improved, reaching up to 4.8%, far higher than that of as-cast (TiZrHf)₅₀Ni₂₅Co₁₀Cu₁₅ HESMAs (1.63%). This indicates that solution treatment can significantly enhance the performances of (TiZrHf)₅₀Ni₂₅Co₁₀Cu₁₅ HESMAs. Yaacoub et al. [44] found that aging treatment results in a decrease in martensitic transformation temperatures of solution-treated (TiZrHf)₅₀Ni₂₅Co₁₀Cu₁₅ HESMAs. The martensitic transformation starting temperature (T_Ms) decreases from 265 K to 225 K, and the martensitic transformation finishing temperature T_Af decreases from 304 K to 269 K. Correspondingly, the aged (TiZrHf)₅₀Ni₂₅Co₁₀Cu₁₅ HESMAs exhibit excellent superelasticity over a wide temperature range of 175 K.

Rehman et al. [45] systematically study the effects of (TiZrHf) content on the microstructure, phase transformation behavior, and superelasticity of TiZrHfNiCoCu high-entropy shape-memory alloys (HESMAs). The results show that the solution-treated TiZrHfNiCoCu HESMAs consisted of a (TiZrHf)(NiCoCu)-type matrix and a (TiZrHf)₂(NiCoCu)-type second phase, as shown in Figure 3. Moreover, as the (TiZrHf) content increases from 50 at.% to 52 at.%, the volume fraction of the (TiZrHf)₂(NiCoCu)-type second phase increases from 1.7% to 17.2%. In addition, the content of Ti, Hf, and Zr in the matrix increases, while the content of Ni, Co, and Cu decreases. In proportion, the martensitic transformation start temperature (T_Ms) increases from 326.6 K to 461.3 K, as shown in Figure 4a. By adjusting the chemical composition of TiZrHfNiCoCu HESMAs, the recoverable strain of superelasticity can be varied from 4.2% to 4.6%, as shown in Figure 4b. In summary, high-performance multicomponent HESMAs with high phase transformation temperatures and excellent superelasticity can be optimized via their chemical composition. Additionally, Resnina et al. [30] also explore the effects of a doping element on the microstructure and martensitic transformation of Ti_50−2xHf_xZr_xNi_50−2xCu_xCo_x(x = 1, 5, 10, 17) alloys under the condition of keeping the unchanged Ti/Ni ratio. The results reveal that the mixed entropy of the multi-principal element increases from 0.89 R (low entropy) to 1.79 R (high entropy), with the x value increasing from 1 to 17 in Ti_50−2xHf_xZr_xNi_50−2xCu_xCo_x(x = 1, 5, 10, 17) shape-memory alloys. More than 90% of the alloy organization is crystallized in the (Ti,Hf,Zr)₅₀(Ni,Cu,Co)₅₀ phase, but precipitated phases also appear, and the larger the concentration of doping elements, the larger the volume fraction of the precipitated phases. The chemical composition and volume fraction of the precipitated phase is related to the x parameter. When x equals 10, (Ti,Hf,Zr)₂(Ni,Cu,Co) particles are observed; when x = 17, (Ni,Cu,Co)₄(Ti,Hf,Zr)₃ precipitated phases are found. Correspondingly, the increase in the x parameter also decreases the temperature of the martensitic phase transformation.

3.2. (TiZrHf)₅₀(NiCu)₅₀ Series High-Entropy Shape-Memory Alloy

Although the addition of the Co element can increase the yield strength of the high entropy shape-memory alloy [46], it will also significantly lower the transformation temperature [29]. Therefore, in order to develop a high-entropy shape-memory with a high martensitic transformation temperature, the quinary alloying (TiZrHf)₅₀(NiCu)₅₀ high-entropy shape-memory alloy containing no Co is of more concern.

Firstov et al. [17] investigate the performances of as-cast (TiZrHf)₅₀(NiCu)₅₀ high-entropy shape-memory alloy. It is found that its forward martensitic transition starting temperatures T_Ms and reverse martensitic transition finishing temperatures T_Af are 500 K and 610 K, respectively. The recoverable strain of the present high-entropy shape-memory alloy is only 0.4%, which is much lower than that of the as-cast (TiZrHf)₅₀(NiCoCu)₅₀ alloy. Chang et al. [31] found that the martensitic transformation temperature can be adjusted by changing the Cu content in (TiZrHf)₅₀(CuNi)₅₀ high-entropy shape-memory alloy. The martensitic transition temperature is reduced by about 20 K/at.%. Although the addition of Cu lowers the martensitic transformation temperature, it can effectively enhance the stability of B19′ martensite during the phase transition as a result of the strong d–d electron interactions between Cu atoms. Li et al. [47] designed a new type of Ti₂₀Hf₁₅Zr₁₅Cu₂₅Ni₂₅ high-temperature HESMAs by optimizing the chemical composition. The present Ti₂₀Hf₁₅Zr₁₅Cu₂₅Ni₂₅ high-temperature HESMAs have a martensitic transformation start temperature T_Ms of 374 K and an austenitic finish temperature T_Af of 452 K, which can be categorized as high-temperature shape-memory alloys. Meanwhile, the present HESMAs exhibit larger superelasticity at a wide temperature ranging from 458 K to 558 K, with a completely recoverable strain of 4.0%. Li et al. [4] also reduced the content of Zr and successfully prepared Ti₂₅Zr₁₀Hf₁₅Ni₂₅Cu₂₅ HESMAs. Although the phase transition onset temperatures of T_Ms and T_Af were only 203 K and 254 K, the superelastic temperature window of the alloy reached 200 K. It is noteworthy that the yield strength of the alloy exceeded 1680 MPa even at 443 K, and the strain was still fully recovered upon unloading. For comparison, the yield strength of cast Ti-Ni SMAs was only 850 MPa [48]. Studies have shown that the high yield strength of HEA materials is due to the significant solid solution hardening caused by lattice distortion. The reversible strain of HESMAs reported above is only 3–5%, and there is still much room for improvement. Zhao et al. [19] prepare a series of Ti₃₅Hf₁₀X₅Ni₄₄Cu₆ (X = Y, Zr, Nb, Cr, Co, Al, V, Mn, Ta) high-entropy shape-memory alloys. In comparison, it is found that the Ti₃₅Hf₁₀Ta₅Ni₄₄Cu₆ high-entropy shape-memory alloy has a good overall performance with a maximum recoverable strain, fracture strain, and compressive strength of 9.4%, 15%, and 2100 MPa, respectively, where the recoverable strain is comparable to that of conventional Ti-Ni SMA (8–12%) [49,50].

The main challenges hindering the application of high-temperature shape-memory alloys are the deterioration of mechanical strength and thermal cycling stability [50]. Peltier et al. [26] compare the evolution of phase transformation and shape-memory effects of Ni₂₇Cu₂₃Ti₁₆Hf₁₉Zr₁₅ and Ni₄₉Ti₃₃Zr₁₈ alloys during thermal fatigue. It is found that both the phase transition temperature and transformation heat of the two types of shape-memory alloys decrease with an increasing number of cycles. Nevertheless, the martensitic transformation temperature of ternary Ni₄₉Ti₃₃Zr₁₈ shape-memory alloy decreases more rapidly, and the transformation heat almost approaches zero after 200 cycles. It is also observed that the maximum recoverable strain of the Ni₂₇Cu₂₃Ti₁₆Hf₁₉Zr₁₅ high-entropy shape-memory alloy could reach 2%, but no irreversible strain is observed after 200 cycles. Although the maximum recoverable strain of the Ni₄₉Ti₃₃Zr₁₈ shape-memory alloy is twice that of the Ni₂₇Cu₂₃Ti₁₆Hf₁₉Zr₁₅ high-entropy shape-memory alloy, the irreversible strain of 3.5% can be observed after 200 cycles. In contrast, the Ni₂₇Cu₂₃Ti₁₆Hf₁₉Zr₁₅ high-entropy shape-memory alloy has better temperature cycling stability. Therefore, the present NiCuTiHfZr high-entropy shape-memory alloy can be used as a driving material for aerospace applications owing to its excellent stability. Chang et al. [51] prepare the Cu₁₅Ni₃₅Ti₂₅Hf₁₂.₅Zr₁₂.₅ high-entropy shape-memory alloy. It is found that the present high-entropy shape-memory alloy also possesses superior cycling stability. After 10 thermal cycles, the irrecoverable strain of the alloy was only 0.14%, and the transformation temperature remained basically unchanged. Moreover, the forward martensitic transformation starting temperatures (T_Ms) and reverse martensitic transformation finishing temperatures (T_Af) are 365.7 K and 442.3 K, respectively. Meanwhile, the present alloy also exhibited excellent mechanical properties and shape-memory effects, with a fracture strength of 1670 MPa and a shape recovery rate of 88.7% under a pre-stress of 1400 MPa.

Canadinc et al. [52] investigate the distinction among Ni₃₅Pd₁₅Ti₃₀Hf₂₀, Ni₂₅Pd₂₅Ti₂₅Hf₂₅, and (TiZrHf)₅₀(NiPd)₅₀ high-entropy shape-memory alloys. It is found that high-entropy shape-memory alloys containing quaternary or quinary alloying elements have a higher martensitic transformation temperature (≥773 K) compared with TiNi-based high-temperature shape-memory alloys. Among them, the transformation temperatures of (TiZrHf)₅₀(NiPd)₅₀ high-entropy shape-memory alloys exceed 973 K, as shown in Figure 5. Both Ni₃₅Pd₁₅Ti₃₀Hf₂₀ and Ni₂₅Pd₂₅Ti₂₅Hf₂₅ shape-memory alloys show a certain high-temperature superelasticity. As shown in Figure 6a,b, the loading stress and strain recovery rate of Ni₂₅Pd₂₅Ti₂₅Hf₂₅ alloy (400 MPa and 3.5%) are higher than those of the Ni₃₅Pd₁₅Ti₃₀Hf₂₀ alloy (210 MPa and 3%). In summary, (TiZrHf)₅₀(NiPd)₅₀ high-entropy shape-memory alloy has a higher martensitic transformation temperature, larger transformation stress, and higher strain recovery rate at high temperatures. However, the high cost of a large amount of Pd elements is another key issue that needs to be addressed.

Aging treatment can adjust the microstructure and phase transformation of Ti-Ni-based shape-memory alloy (SMA), thus further affecting the shape-memory effect and superelasticity. For example, in Ni-rich binary TiNi and TiNiHf/TiNiZr alloys, the nano-scale Ti₃Ni₄ and H-phase precipitates are induced, respectively [53,54]. Lee et al. [55] and Yaacoub et al. [45] have also shown that solution treatment and aging treatment significantly affect the microstructure and phase transformation behavior of (TiZrHf)₅₀Ni₂₅Co₁₀Cu₁₅. After solid solution and aging treatments, (TiZrHf)₅₀Ni₂₅Co₁₀Cu₁₅ HESMAs’ dendritic organization dissolved in the matrix, resulting in a more homogeneous alloy organization; however, the phase transition temperature of the alloy was significantly lower after solid solution and aging, which was related to the change in elemental content in the matrix. In addition, Chang et al. [56] investigated the microstructure and phase composition of Ti₂₀Zr₁₅Hf₁₅Ni₃₅Cu₁₅ high-entropy shape-memory alloys under different aging conditions. As shown in Figure 7, a large amount of H-phase was formed during aging at temperatures below 773 K. In addition, the strain field around the H-phase inhibits the martensitic transformation. However, as the aging temperature increases, the size of the H-phase gradually increases, and its inhibitory effect on the martensitic transformation decreases, resulting in an increase in the transformation temperature (T_Ms and T_Af) within the range of 573–773 K. When the aging temperature exceeds 773 K, T_Ms and T_Af increase. At the aging treatment of 873 K, the eutectoid reaction led to the formation of laminated structures containing (Zr,Hf)₇Cu₁₀ and Ti₂Cu second phases. When the aging temperature reaches 973 K, the laminated structures disappear. The (Zr,Hf)₇Cu₁₀ phase and the newly formed Ti₂Ni phase are formed around the original Ti₂Ni phase. In proportion, the martensitic transformation temperature (T_Ms and T_Af) increases to the largest value, reaching 7 K and 86 K, respectively. By adjusting the aging temperature, the type, density, and distribution of precipitated phases can be optimized, thereby obtaining perfect compressive performances.

It is well known that the comprehensive performances of shape-memory alloys are closely related to the crystalline orientation [57]. Li et al. [58] employ the directional solidification technology to fabricate Ti₃₀Ni₃₀Fe₁₀Hf₁₀Nb₂₀ high-entropy shape-memory alloy with a specific orientation. Figure 8a–d shows micrographs of the directional solidification Ti₃₀Ni₃₀Fe₁₀Hf₁₀Nb₂₀ high-entropy shape-memory alloys with different solidification rates. When the solidification rates are 1 mm/h and 180 mm/h, the microstructure is featured with a wavy texture and equiaxed eutectic clusters. And large BCC phases are formed in the eutectic clusters. However, at a solidification rate of 60 mm/h, the microstructure is characterized by a columnar eutectic structure parallel to the direction of directional solidification. The columnar eutectic structure consists of a central eutectic containing Nb fibers and an outer willow-like eutectic structure. The Ti₃₀Ni₃₀Fe₁₀Hf₁₀Nb₂₀ high-entropy shape-memory alloy prepared at a rate of 60mm/h has a yield strength of 1409 MPa and relatively excellent superelasticity with a recoverable strain of 3.1%, which is attributed to the synergistic effects of BCC/B2, eutectic, and preferred orientation. This means that the directional solidification method provides an effective way to develop high-performance shape-memory alloys.

4. Fe-like High-Entropy Shape-Memory Alloy

Compared with Ni-Ti-based shape-memory alloys, Fe-based shape-memory alloys have been paid more attention due to their perfect combination of lower costs, excellent workability, and suitability for large-scale industrial applications [59]. To date, Fe-based high-entropy shape-memory alloys can be classified into two categories on the basis of the type of martensitic transformation: (1) Fe-based HESMAs with a face-centered cubic γ-austenite phase → hexagonal close-packed ε-martensite phase; (2) Fe-based HESMAs with a face-centered cubic γ-austenite phase → body-centered cubic α or body-centered cubic α′-martensite phase [60].

4.1. γ → ε Martensitic Phase Transition

Fe-Mn-Si-based shape-memory alloys with a γ → ε martensitic transformation have attracted considerable interest due to their low costs, good mechanical properties, and cutting performances. However, the shape-memory effect of Fe-Mn-Si-based shape-memory alloy is poor, possessing a recoverable strain of only 2~3%. Therefore, the improvement of the shape-memory effect has become a topic [61]. The shape-memory effect of the Fe-Mn-Si-based shape-memory alloy can be improved by optimizing the chemical composition, enhancing the matrix strength, tailoring the heat treatment parameters, etc. Although the shape-memory effect is improved to some extent, there exist some other questions that should be enhanced. In recent years, the concept of high-entropy has been introduced into Fe-Mn-Si-based shape-memory alloys to obtain high performances. For instance, Liao et al. [62] designed and prepared Fe₄₇.₇Mn₁₅.₅Co₉.₈Cr₁₀.₈Ni₅.₁Si₁₁.₁ high-entropy shape-memory alloys. The results show that the hot-rolled Fe₄₇.₇Mn₁₅.₅Co₉.₈Cr₁₀.₈Ni₅.₁Si₁₁.₁ high-entropy shape-memory alloy with annealing temperatures of 773 K and 1273 K is a single FCC phase. π phase precipitates in the matrix, as it is annealed at 973~1173 K, as shown in Figure 9. After annealing at 773 K, the Fe₄₇.₇Mn₁₅.₅Co₉.₈Cr₁₀.₈Ni₅.₁Si₁₁.₁ high-entropy shape-memory alloy has good shape-memory effects and mechanical properties with a maximum reversion strain of 5.7%, which indicates that the high-entropy shape-memory alloy is expected to solve the problem of its generally low shape-memory reversion rate and its ultimate tensile strength and fracture strain reach 888 MPa and 36.2%, respectively, which are better than most of the reported CoCrFeNiMn-based and FCC-based HEAs. This suggests that high-entropy shape-memory alloys with both excellent shape-memory effects and mechanical properties will be a promising direction for the development of new shape-memory alloys.

Lee et al. [63] reported the effect of Ni content on the solution-treated Cr₂₀Mn₂₀Fe₂₀Co_40−xNi_x high-entropy shape-memory alloy. It was found that superior properties can be achieved in Cr₂₀Mn₂₀Fe₂₀Co₃₅Ni₅ high-entropy shape-memory alloys via the optimization of its chemical composition. The present high-entropy shape-memory alloy has a single FCC phase. The present solution-treated Cr₂₀Mn₂₀Fe₂₀Co₃₅Ni₅ high-entropy shape-memory alloy exhibits a γ-ε martensitic transformation. Moreover, the martensitic transformation temperature ranges from 276 K to 456 K. It exhibits a maximum strain of 2% at room temperature, which is almost comparable to that of conventional Fe-Mn-Si-based high-entropy shape-memory alloys [60]. Park et al. [64] investigate the influence of annealing temperature on the microstructure, phase transition behavior, and shape-memory properties of the rolled Cr₂₀Mn₂₀Fe₂₀Co₃₅Ni₅ high-entropy shape-memory alloy. It is seen that the grain size increases from 3.04 μm to 82.2 μm with increasing annealing temperature. The grain refinement is favored to improve the recovery stress. The maximum recovery stress of 302 MPa can be obtained in the Cr₂₀Mn₂₀Fe₂₀Co₃₅Ni₅ high-entropy shape-memory alloy with a small grain size of 7.1 μm. Moreover, the rising annealing temperature leads to a reduction in the density of deformation-induced martensite, grain boundary, and annealing twin boundaries, as shown in Figure 10. As a result, the maximum recoverable strain of 1.2% is obtained at an annealing temperature of 1473 K. Clearly, the thermomechanical treated Cr₂₀Mn₂₀Fe₂₀Co₃₅Ni₅ high-entropy shape-memory alloy cannot simultaneously achieve a high yield strength and a good shape-memory effect via the thermomechanical treatment. In addition, Kireeva et al. [65] developed the $\left[\overline{1}11\right]$ single-oriented Cr₂₀Mn₂₀Fe₂₀Co₃₅Ni₅ high-entropy shape-memory alloy using the Bridgman method. The single-orientation high-entropy shape-memory alloy showed excellent shape-memory performance, and the recoverable strain was 6.8% under the 7.4% pre-strain condition.

4.2. γ → α′(α) Martensitic Phase Transition

The above-mentioned research is focused on Fe-based high-entropy shape-memory alloys exhibiting γ → ε martensitic transformation. However, there is also an extensive investigation on Fe-based high-entropy shape-memory alloys with a γ → α′ (α) martensitic transformation. A B-doped FeNi₂₇.₅Co₁₆.₅Al₁₀Ta₂.₂B₀.₀₄(NCATB) high-entropy shape-memory alloy was prepared by Zhang et al. [66]. In addition, the effects of the aging treatment on the organization and functional properties of the alloy were investigated. The results showed that the aging treatment significantly changed the morphology of the martensite and precipitated phases, which led to changes in the properties. With an aging time of less than 10 h, the NCTAB alloy formed a thermoelastic thin-plate martensite with a yield strength of up to 800 MPa at room temperature and a superelastic strain of 2.5%. Notably, at a 4% tensile strain, the NCATB alloy absorbed more than 15 MJ/m³ in one hyperelastic cycle, which is equivalent to applying a 6% deformation to the NiTi alloy, nearly four times that of the Fe-Mn-Al-Ni alloy and nearly ten times that of the Cu-Al-Ni alloy (Figure 11b). After 96 h of aging, the NCTAB alloy consisted of high-density plate-like martensite and butterfly-like martensite, and the yield strength of the alloy reached a maximum of 1.1 GPa (Figure 11a). At the same time, the phase transformation temperature T_Ms increased from 172 K to 513 K, which was caused by the loss of the inhibition of the martensite formation due to the oversized γ′ precipitates and the growing Ni-rich precipitates that made the martensite matrix Ni-poor.

This demonstrates that the NCATB high-entropy shape-memory alloy also has great potential as a high damping material. To expand the temperature range of damping materials, Peltier et al. [27] develop a Cu_xFe_yNi_zMn₂₀V₁₁ high-entropy shape-memory alloy. By optimizing the chemical composition, the Cu₁₁Fe₂₉Ni₂₉Mn₂₀V₁₁ high-entropy shape-memory alloy is featured with a single austenitic phase, which possesses the highest plasticity. When the present high-entropy shape memory is subjected to the rolling reduction of 80% in thickness, no cracks are observed. Meanwhile, it shows superior damping performance after aging for 800 h at 523 K. Moreover, within a wide temperature range of 233–473 K, it can maintain a high damping performance. In summary, the excellent combinations of excellent workability, remarkable high-temperature stability, and high damping characteristics can be obtained in the Cu₁₁Fe₂₉Ni₂₉Mn₂₀V₁₁ high-entropy shape-memory alloy.

Conventional vacuum arc melting can result in uneven dendrites and the formation of carbides, which would have a negative impact on the shape-memory effect of Fe-based shape-memory alloys [67,68]. For comparison, mechanical alloying (MA) can produce a very uniform solid solution structure, avoiding the formation of any adverse intermetallic phases. Gülera et al. [69] first adopt the MA method to prepare TiHfZrFeMnSi high-entropy shape-memory alloys. It is found that the TiHfZrFeMnSi high-entropy shape-memory alloy powders show relatively uniform agglomerated particles after being ball milled for 60 h or more. Moreover, the present alloy powders heat-treated at 673 K, 873 K, and 1073 K also exhibit spherical and irregularly shaped particles. The heat-treated TiHfZrFeMnSi high-entropy shape-memory alloy powders undergo an FCC → BCC martensitic phase transformation, and its forward martensitic transformation starting temperature (T_Ms) varies in the range of 593–674 K.

5. Ti-Based High-Entropy Shape-Memory Alloy

The Ni-free Ti-based shape-memory alloys with excellent performances have received increasing attention due to the consideration of the hypersensitivity and cytotoxicity of the nickel element [70]. Moreover, the ideas of high entropy also are introduced into Ti-based shape-memory alloys, which promotes the achievement of high-performance shape-memory alloys. Wang et al. [71] developed Ti_50−xZr₂₀Hf₁₅Al₁₀Nb_5+x series high-entropy shape-memory alloys. It is found that the microstructure of the present high-entropy shape-memory alloy exhibits an equiaxed β phase. Moreover, a large number of plate-like α′-martensite phases with various sizes distributes within the grains. The designed Ti_50−xZr₂₀Hf₁₅Al₁₀Nb_5+x high-entropy shape-memory alloys possess excellent superelasticity with the maximum superelastic strain of 4.0% and the highest superelastic stress values of 900 MPa. Its remarkable superelasticity is derived from the reversible stress-induced martensitic phase transformation between β and α′. Meanwhile, the forward martensitic transformation starting temperatures T_Ms and reverse martensitic transformation starting temperatures T_Af reaches 872 K and 1044 K, respectively. Moreover, the multifunctional β-type (TiZrHf)_90−xNb₅Ta₅Al_x alloy is also designed by Hashimoto et al. [72]. The results reveal that the present high-entropy shape-memory alloys are dominated by a single β phase with an equiaxed crystal structure. In contrast, the present (TiZrHf)₈₂Nb₅Ta₅Al₈ alloy exhibited a poor superelastic strain of 3.2%. Nevertheless, they have low magnetization, excellent corrosion resistance, and biocompatibility, which makes them highly applicable in the biomedical field.

In order to avoid the potentially neurotoxic effects of the Al element in biomedical high-entropy shape-memory alloys, new types of high-entropy shape-memory alloys (HESMAs) are developed. The solution-treated Ti₃₀Hf₁₉Zr₂₅(NbTa)₂₆ high-entropy shape-memory alloy [25], free of the Al element, is characterized by a single equiaxed β phase. This means that the martensitic transformation temperature of the present high-entropy shape-memory alloy is lower than room temperature. The forward martensitic transformation starting temperatures T_Ms and reverse martensitic transformation starting temperatures T_Af are only 279 K and 319 K, respectively. The superelasticity of the present shape-memory alloy is related to the reversible stress-induced β to α′′ martensitic phase transformation. As shown in Figure 12, the attenuation degree of superelasticity during cyclic fatigue testing in the temperature range of 233–473 K was much smaller than that of Ti-Ni shape-memory alloy. This indicates that the present high-entropy shape-memory alloy has better thermal stability over a wide temperature range, which can be attributed to the high-entropy effect.

In addition, a new type of high-entropy (TiZrHf)-8.5Nb-3Sn shape-memory alloy with non-toxic and low-magnetization was developed by Tasaki et al. [73]. The results show that an increase in the content of Nb or Sn promotes the stability of the β phase. Moreover, the more stable the β parent phase is, the greater the driving force of stress-induced martensitic transformation is. The highest superelastic strain of 2.4% can be found in the high -entropy (TiZrHf)-7.5Nb-4Sn shape-memory alloy. Meanwhile, the present (TiZrHf)-8.5Nb-3Sn high-entropy shape-memory alloy has a lower magnetization of 1.48 × 10⁻⁶ cm³g⁻¹ compared with the conventional biomedical shape-memory alloys. The single-phase Ti₃₅Zr₃₅Hf₁₅Nb₅Ta₅Sn₅ high-entropy shape-memory alloy with a BCC structure is also reported [71]. The Ti₃₅Zr₃₅Hf₁₅Nb₅Ta₅Sn₅ high-entropy shape-memory alloy not only contains non-toxic elements but also exhibits a relatively superior superelasticity behavior with a strain recovery rate of 3.8%. In addition, owing to the higher cost of the Hf element, the lower cost Ti₃₅Zr₃₅Nb_15+XSn_10−XMo₅(X = 0, 2.5, 5) high-entropy shape-memory alloys are also investigated [74]. The results show that the main microstructure of the present shape-memory alloy at room temperature is a martensitic ${\alpha }^{″}$ phase, as X = 0. It has higher strength and hardness, while the ductility is relatively poor. Nevertheless, the present high-entropy shape-memory alloy shows the perfect shape-memory effect with the maximum recoverable strain of 4.5%. When X = 2.5, the present high-entropy shape-memory alloy is composed of the β parent phase, which is featured with lower strength and hardness. However, it exhibits good plasticity and a maximum recoverable strain of 2.7%. As the value of X is up to 5, the microstructure of the present high-entropy shape-memory alloy has a mixture of the β-Ti parent phase and ${\alpha }^{″}$ martensite phase. For comparison, its recoverable strain characteristics are much superior, exhibiting a maximum recoverable strain of up to 5%. Compared with Ni-Ti-based high-entropy shape-memory alloys, β type Ti-based high-entropy shape-memory alloys have the advantages of low cost and good biocompatibility, making them widely applicable in biomedical fields.

6. Other Types of High-Entropy Shape-Memory Alloys

The current research on high-entropy shape-memory alloys is mainly focused on Ni-Ti-based high-entropy shape-memory alloys. However, in order to meet various requirements, other types of high-entropy shape-memory alloys also have been paid increasing attention. For example, Gerstein et al. [75] successfully developed a new Co₃₁.₂₂Ni₂₉.₂₆Cu₁₁.₉₅Al₁₆.₆₄Ga₁₀.₃₉In₀.₅₅ high-entropy shape-memory alloy on the basis of Ni-Al-based shape-memory alloys. The present high-entropy shape-memory alloy has a mixture of B2 austenite and L1₀ martensite structures, as shown in Figure 13a–b. The forward phase transition starting temperature (T_Ms) and the reverse martensitic transformation temperature (T_Af) are 875 K and 890 K, respectively. Notably, the present CoNiCuAlGaIn high-entropy shape-memory alloy has a matrix yield strength of 1450 MPa, as displayed in Figure 13c. In addition, there appears the original research on high-entropy ferromagnetic shape-memory alloys. Ju et al. [76] designed Ni-Mn-Ga-Co-Gd high-entropy ferromagnetic shape-memory alloys. The results show that β, γ, and martensitic phases exist in the microstructure of A1-A4 alloy, with the change in Co and Gd elements, the γ phase gets smaller, and the martensitic phase gradually replaces the matrix β phase; secondly, with the change in Co and Gd elements, the valence electron concentration of the alloy gradually increases, which leads to a gradual increase in the martensitic phase transition temperature to 300 °C, which increases the possibility of ferromagnetic shape-memory alloys for high-temperature applications. As shown in Figure 14a, the morphology of the alloy martensite exhibits typical L1₀-type twinning characteristics, and the driving force for rearrangement strengthening of these twinned martensites decreases, so the magnetic-field-induced strain of the alloy shows an increasing trend with the change in Co and Gd elements. Ju et al. [77] also investigated the effect of thermomagnetic drawing on the performances of Ni₂₀Mn₂₀Ga₂₀Gd₂₀Co₂₀ high-entropy ferromagnetic shape-memory alloy Compared with the original Ni₂₀Mn₂₀Ga₂₀Gd₂₀Co₂₀ high-entropy ferromagnetic shape-memory alloy prepared via arc melting, the phase transition temperature is increased to 873 K under a magnetic field of 0.5 T as a result of thermomagnetic drawing. The present high-entropy shape-memory alloy subjected to thermomagnetic drawing shows a typical L1₀-type twin martensite, and the internal twins are (111) twinning related, as shown in Figure 14b. Due to the evolution of the twin martensite and the enhancement of the driving force, the magnetic field-induced strain of Ni₂₀Mn₂₀Ga₂₀Gd₂₀Co₂₀ high-entropy ferromagnetic shape-memory alloy can be up to 4.47% by thermomagnetic drawing under the conditions of 0.5 T and 8007 Oe.

7. Summary

To date, multiple varieties of high-entropy shape-memory alloys have been developed on the basis of traditional shape-memory alloys. In contrast, high-entropy shape-memory alloys bring great developmental prospects. The present high-entropy shape-memory alloys are mainly focused on the design and optimization of chemical composition, microstructure, martensitic transformation behavior, shape-memory effect, and superelasticity. Nevertheless, various high-entropy shape-memory alloys have been developed in order to meet the different applications. There exist some problems that should be solved, including the following:

(1)

Alloy design depends on the best chemical composition to improve performance. Therefore, in order to improve the efficiency of alloy research, computational simulations such as first principles and machine learning can be used to quickly screen out outstanding alloy compositions based on existing research data. Meanwhile, attention should be paid to selecting low-cost, non-toxic, and harmless elemental compositions that are more conducive to engineering application promotion.

(2)

It is necessary to investigate the correlation mechanism among chemical composition, microstructure, martensitic transformation, and functional properties of high-entropy shape-memory alloys in order to provide a theoretical basis for the design and development of high-entropy shape-memory alloys.

(3)

Currently developed alloy systems are mainly focused on TiNi-based and Fe-based high-entropy shape-memory alloys. In the future, the advantages of high-entropy alloy designs can be utilized to develop more alloy systems, with a focus on improving the thermal–mechanical cyclic fatigue properties of shape-memory alloys.