Research Progress of Thermoelectric Materials—A Review

Abstract

Thermoelectric materials are functional materials that directly convert thermal energy into electrical energy or vice versa, and due to their inherent properties, they hold significant potential in the field of energy conversion. In this review, we examine several fundamental strategies aimed at enhancing the conversion efficiency, classification, preparation methods, and applications of thermoelectric materials. First, we introduce an important parameter for evaluating the performance of thermoelectric materials, the dimensionless quality factor ZT, and present the theory of electroacoustic transport in thermoelectric materials, which provides the foundation for enhancing the performance of thermoelectric materials. Second, strategies for optimizing electroacoustic transport properties, carrier concentration, energy band engineering, phonon engineering, and entropy engineering are summarized, emphasizing that energy band engineering presents numerous possibilities for enhancing thermoelectric material performance by tuning the carrier effective mass, energy band convergence, and energy band resonance. By analyzing the importance of various optimization strategies, it is concluded that co-optimization is the primary method for improving the performance of thermoelectric materials in the future. In addition, an overview of the currently available thermoelectric materials is provided, including two categories, classical thermoelectric materials and novel thermoelectric materials, along with a highlight of two thermoelectric material preparation techniques. Finally, the principles of thermoelectric technology are illustrated, its applications in various fields are discussed, problems in the current research are analyzed, and future trends are outlined. Overall, this paper provides a comprehensive summary of optimization strategies, material classifications, and applications, offering valuable references and insights for the researchers in this field, with the aim of further advancing the development of thermoelectric material science.

1. Introduction

In recent years, the growing concern over the energy crisis and environmental pollution has garnered increasing attention. Therefore, finding renewable energy sources or improving the efficiency of traditional fossil fuel usage is one of the effective solutions to the aforementioned challenges. Thermoelectric materials are functional materials that rely on the Seebeck effect and the Peltier effect, as shown in Figure 1, to directly convert thermal energy into electrical energy or vice versa [1,2]. In the Seebeck effect, a temperature gradient is established when P-type and N-type thermoelectric materials are connected in a closed loop with their ends maintained at different temperatures. Under thermal excitation, charge carriers (holes in P-type and electrons in N-type materials) migrate from the high-temperature region to the low-temperature region, resulting in the generation of a potential difference, referred to as the Seebeck voltage. This enables an electric current to circulate within the closed loop, facilitating the direct conversion of thermal energy into electrical energy for thermoelectric power generation. In the Peltier effect, when an external power supply provides a current, it flows through the junctions of the P-type and N-type materials, leading to heat absorption or release. On one side (T − ΔT), the system absorbs heat due to carrier redistribution, resulting in cooling; on the other side (T), heat is released, causing heating. Thermoelectric performance is commonly described by the dimensionless figure of merit (ZT), as shown in Equation (1), where S represents the Seebeck coefficient, ∂ denotes the electrical conductivity, K is the thermal conductivity, and T is the absolute temperature. The power factor is given by S²∂, and the total thermal conductivity (K) consists of electronic thermal conductivity and lattice thermal conductivity. In general, high-performance thermoelectric materials must demonstrate both a high power factor and low lattice thermal conductivity.

$$\mathrm{Z}\mathrm{T}={\displaystyle \frac{S^2\partial }{K}}T$$

(1)

In the 1950s, driven by rapid advancements in semiconductor physics, the researchers shifted from studying metals with low thermoelectric efficiency to focusing on semi-conductor materials. This led to the development of classic semiconductor thermoelectric materials, such as bismuth telluride (Bi₂Te₃) and lead antimony (PbTe). However, due to their relatively low performance and high cost, the practical applications of these materials remain limited. In 1998, Slack [3] introduced the concept of “phonon glass-electron crystal” at the Materials Research Society conference in Boston, proposing that high-performance thermoelectric materials should integrate the electronic properties of crystals with the thermal properties of glasses. Based on this concept and the rapid development of technologies and disciplines, such as materials science, manufacturing processes, and materials physics in recent years, thermoelectric materials have made significant progress in terms of the fundamental principles, performance enhancement of traditional materials, and development of new high-performance thermoelectric materials. To date, although the thermoelectric performance of materials has consistently achieved new highs, the strong coupling between electrical and thermal transport properties continues to limit the ZT value, which has not yet reached the ideal range. Therefore, the researchers have proposed a series of strategies based on the microscopic crystal structure of thermoelectric materials, including carrier concentration optimization, band engineering, defect engineering, and phonon engineering, to synergistically enhance the electrical and thermal transport properties, with the goal of producing materials with a high thermoelectric performance.

Based on the inherent properties of thermoelectric materials, thermoelectric devices are widely applied in the fields of waste heat recovery, solar thermoelectric power generation, and solid-state refrigeration. Compared to conventional power generation and refrigeration methods, thermoelectric conversion technologies based on thermoelectric materials provide advantages such as noiseless operation, lack of moving parts, and environmental friendliness [4,5]. Although thermoelectric conversion technologies offer numerous advantages in energy conversion, their efficiency still falls significantly short when compared to traditional technologies. As a result, efforts are needed in areas such as developing high-performance materials, optimizing device design, and selecting appropriate application scenarios.

This paper begins by discussing the performance optimization of thermoelectric materials and reviewing the impact of improving their electrical and thermal transport properties on the overall performance. Next, we summarize the classification and preparation methods of thermoelectric materials, briefly discussing the advantages and disadvantages of various preparation techniques and material types. Finally, we highlight the practical applications of thermoelectric materials and outline their future trends and development directions.

2. Optimization of Thermoelectric Material Performance

In this section, we summarize the key strategies proposed over the past few decades to optimize the performance of thermoelectric materials: carrier concentration optimization [6,7,8], band engineering [9,10,11], phonon engineering [12,13], and entropy engineering [14], as shown in Table 1. These strategies enhance the ZT value by improving the electrical and thermal transport properties of thermoelectric materials. Typically, the carrier concentration in thermoelectric materials directly influences the Seebeck coefficient (S), electrical conductivity (σ), and electronic thermal conductivity (K_E), as shown in Figure 2; the relationship among these parameters is described by the following equation [15]:

$$S={\displaystyle \frac{8{\pi }^2{k}_B^2}{3e{h}^2}}{m}_d^{\mathrm{*}}T{\left({\displaystyle \frac{\pi }{3n}}\right)}^{{\displaystyle \frac{2}{3}}}$$

(2)

$$\sigma =ne\mu$$

(3)

$$K=K_E+K_L=Lne\mu T+K_L$$

(4)

where m_d^* is the effective mass of the density of states, n is the carrier concentration, μ is the carrier mobility, K_B is the Boltzmann constant, and h is Planck’s constant. The thermal conductivity (K) of the material consists primarily of electronic thermal conductivity and lattice thermal conductivity (K_L). As shown in Figure 2, S, n, and σ are interrelated. This coupling indicates that for thermoelectrics with high S, σ, and low K, a compromise in carrier concentration (n) is required, with semiconductor thermoelectrics performing optimally when n is in the range of 10¹⁹~10²⁰ cm⁻³. As shown in Equations (3) and (4), both electrical conductivity and thermal conductivity increase with increasing the carrier concentration, which results in a reduction in the material’s performance. The primary reason is that high thermal conductivity facilitates heat transfer from the hot side to the cold side, thereby reducing the temperature gradient and diminishing the Seebeck effect’s effectiveness. Furthermore, according to Equation (2), as the carrier concentration increases, the Seebeck coefficient (S) decreases, consequently impacting the ZT value of the material. Therefore, the coordinated control of carrier concentration represents an effective strategy for optimizing thermoelectric materials.

Table 1. Strategies for enhancing the performance of thermoelectric materials.

Materials	Strategies	ZT	T(K)	Ref.
Cu2CoTi3S8	carrier concentration	0.2	670	[16]
Cu3Sb0.92Mn0.06Sn0.02Se4	carrier concentration	0.74	673	[17]
Ti0.3Zr0.35Hf0.35CoSb1−xSnx	carrier concentration	0.8	380	[18]
dithienothiophene	phonon engineering	1.48	300	[19]
CuxBi0.5Sb1.5Te3	phonon engineering	1.34	400	[20]
Nb0.8Ti0.2FeSb	phonon engineering	0.9	973	[21]
FeNb1−xTixSb	band engineering	1.1	1100	[22]
α-MgAgSb	band engineering	2.0	575	[23]
Al0.04Sn0.96Se	band engineering	0.84	823	[24]

Table 1. Strategies for enhancing the performance of thermoelectric materials.

Materials	Strategies	ZT	T(K)	Ref.
Cu2CoTi3S8	carrier concentration	0.2	670	[16]
Cu3Sb0.92Mn0.06Sn0.02Se4	carrier concentration	0.74	673	[17]
Ti0.3Zr0.35Hf0.35CoSb1−xSnx	carrier concentration	0.8	380	[18]
dithienothiophene	phonon engineering	1.48	300	[19]
CuxBi0.5Sb1.5Te3	phonon engineering	1.34	400	[20]
Nb0.8Ti0.2FeSb	phonon engineering	0.9	973	[21]
FeNb1−xTixSb	band engineering	1.1	1100	[22]
α-MgAgSb	band engineering	2.0	575	[23]
Al0.04Sn0.96Se	band engineering	0.84	823	[24]

2.1. Optimization of Carrier Concentration

Currently, carrier concentration is typically optimized through the doping of impurity atoms, as shown in Figure 3. Figure 3a,b,e depict the trends of carrier concentration and carrier mobility with temperature for different samples. The carrier concentration either increases slightly or remains stable with temperature, which may be related to the doping concentration of the material or the solid solution effect. The mobility remains generally stable with temperature, but the black curve shows an increase with temperature and tends to saturate within a specific temperature range, possibly due to high-quality crystals or special doping. As the concentration of dopant atoms increases, the carrier concentration of the material increases, while the Hall mobility decreases, as shown in Figure 3c. However, in contrast to Figure 3c, in Figure 3f, the Hall mobility of the material increases, suggesting that additional doping introduces a distinct scattering mechanism with differing results. The carrier concentration of the material influences the power factor of the thermoelectric material, as shown in Figure 3d. The typical parabolic shape indicates that the power factor peaks within a specific carrier concentration range, suggesting the presence of an optimal carrier concentration. For example, doping erbium (Er) atoms into GeTe [25] results in an intrinsic carrier concentration of approximately 10²¹ cm⁻³, which exceeds the optimal carrier concentration range (0.5–2 × 10²⁰ cm⁻³), leading to a lower ZT value. Doping results indicate that the carrier concentration in GeTe decreases with the increasing Er atom content. The primary reason is that in GeTe, holes are the dominant charge carriers, and Er atoms have one more valence electron than Ge atoms. Doping with Er atoms reduces the hole concentration in GeTe, resulting in a decrease in the overall carrier concentration as the Er content increases. Additionally, other experimental results have shown that doping is an effective method for adjusting the carrier concentration in thermoelectric materials [26,27,28,29].

In general, the carrier concentration is proportional to T³/² and increases with temperature. Pei et al. [32] doped excess Ag atoms into PbTe/AgTe composite materials, and the results indicate that as the temperature increases from 300 K to 750 K, the carrier concentration increases by an order of magnitude. In the composite material, the solubility of Ag atoms increases with temperature, leading to the generation of additional charge carriers. As a result, the carrier concentration in the composite increases.

2.2. Band Engineering

The electrical and thermal transport properties of thermoelectric materials are governed by the material’s band structure. Therefore, modifying the band structure can enhance the material’s performance. Doping with atoms can increase the bandgap of the material’s band structure, raising the carrier transition temperature to higher values, which helps maintain the carrier concentration in the conduction band within the optimal range [33,34], as shown in Figure 4a,b. For a specific thermoelectric material, in addition to adjusting the bandgap width, increasing the degeneracy of the energy bands can also enhance the effective mass of the density of states, thereby improving the material’s Seebeck coefficient, as shown in Figure 4d,e. Previous studies have shown that, without inducing scattering between energy valleys, increasing the degeneracy of the energy bands significantly enhances the thermoelectric performance of the material [35,36,37]. For example, in the PbTe_1−xSe_x alloy (Nv = 6, degeneracy) [38], doping increases the degeneracy of the energy bands to 12. When the temperature reaches 850 K, the material achieves a ZT value of 1.8.

In semiconductor materials, the energy levels of impurity atoms introduced by doping are typically located within the bandgap of the semiconductor. In certain cases, the energy levels of impurity atoms are located within the conduction or valence band of the semiconductor, forming resonance levels. This results in changes in the effective mass of the energy bands, which can, in some cases, enhance the performance of thermoelectric materials [41]. In Figure 4f, the energy band diagram of Mg₂₄Sb₁₅Te₁ reveals a possible localized resonance energy level in a region where some energy bands cross or overlap, particularly between the conduction and valence bands. In this region, the energy bands may become denser or exhibit distinct peaks, suggesting that electrons may form resonance states near this energy level. In 2008, Joseph et al. [42] first demonstrated that Tl introduces resonance levels in the valence band of PbTe. Following this, resonance levels were also observed in other materials [43,44,45]. Currently, band engineering is a widely used strategy for enhancing the performance of thermoelectric materials.

2.3. Phonon Engineering

Phonon engineering involves reducing the lattice thermal conductivity of the crystal through phonon scattering, thereby enhancing the ZT value, as shown in Figure 5. Figure 5a illustrates a three-dimensional lattice structure featuring various microscopic defects and structural features, primarily nanoprecipitates, nanopores, and grain boundaries, which significantly affect phonon propagation. The figure illustrates how nanostructures and defects influence phonon propagation within a material and its thermal conductivity, with nanoprecipitates, pores, and grain boundaries scattering phonons at different frequencies, particularly at low- and mid-frequency ranges. This effect can significantly reduce the material’s thermal conductivity and enhance the properties of thermoelectric materials. Figure 5b–e show SEM images of the structural features. As shown in Equations (2)–(4), lattice thermal conductivity is the sole independent physical parameter. Currently, lattice thermal conductivity is primarily reduced by increasing the types and concentrations of defects in crystal. In thermoelectric materials, alloying to introduce point defects is an effective approach for reducing lattice thermal conductivity [46,47]. For example, in Sb₂Se₃, substituting Bi atoms for Sb atoms and Te atoms for Se atoms creates point defects at the atomic scale in the lattice. The ZT value of the resulting n-type thermoelectric material increased from 0.21 to 0.45 [4].

Dislocations, grain boundaries, phase boundaries, and nano-deposits are additional defects that can significantly reduce the lattice thermal conductivity of thermoelectric materials. These have been incorporated into materials, such as Bi₂Te₃, Sb₂Se₃, and Mg₃(Sb, Bi)₂ [49,50].

3. Classification and Fabrication Techniques of Thermoelectric Materials

Since the discovery of the thermoelectric effect in the 19th century, the research on thermoelectric materials has evolved from metals to semiconductors, with semiconductor thermoelectric materials undergoing rapid development in the early 21st century. Researchers have developed numerous high-performance thermoelectric materials to date, as shown in Table 2. Table 2 lists some of the high-performance thermoelectric materials (ZT > 1) for the last three years. These high-ZT-value thermoelectric materials provide significant advantages, including improved thermal energy conversion efficiency, reduced heat loss, and enhanced thermoelectric properties, thereby expanding their application potential across a wider range of scenarios. Based on their development stages, thermoelectric materials can be broadly classified into classical and novel types.

Table 2. Selected thermoelectric materials with ZT > 1 in the last 3 years.

Samples	T (K)	ZT	Ref.
SnSeS	700	3.07	[51]
(Ca0.85Ba0.15)0.995Na0.005Mg1.85Cd0.15Bi2	873	1.30	[52]
Ge0.93Bi0.03Pb0.04Te	670	2.14	[31]
Ge0.93Ti0.01Bi0.06Te0.01Cu	623	2.30	[53]
CoGe2/Ge0.85Sb0.10Te	775	2.20	[54]
(Ge0.89Pb0.08Bi0.03Te)0.97(HgTe)0.03	650	2.30	[55]
Bi2(Te,Se)3	375	1.20	[56]
AgSbTe2	643	1.70	[57]
Cu3SbS4	773	1.30	[58]
Na0.99Cd0.995Ag0.005Sb	673	1.41	[59]
CuIn7Se11	873	1.23	[60]
PbSnS2	473	1.20	[61]
Sn0.71Ge0.2Mn0.07In0.02Te	873	1.64	[62]
EMIM:DCA	330	3.10	[63]
Sn0.78Sb0.16Te(MgB2)0.09	850	1.22	[64]
Mg3.2Bi1.998−xSbxTe0.002Cu0.005	348	1.10	[65]
GeSb2Te4	673	1.00	[66]
Mg3(Sb,Bi)2	773	1.82	[67]
AgMnGePbSbTe5	750	2.64	[68]
Nb0.75Ti0.25FeSb	973	1.21	[69]
(Nb, Hf)FeSb	973	1.47	[70]

Table 2. Selected thermoelectric materials with ZT > 1 in the last 3 years.

Samples	T (K)	ZT	Ref.
SnSeS	700	3.07	[51]
(Ca0.85Ba0.15)0.995Na0.005Mg1.85Cd0.15Bi2	873	1.30	[52]
Ge0.93Bi0.03Pb0.04Te	670	2.14	[31]
Ge0.93Ti0.01Bi0.06Te0.01Cu	623	2.30	[53]
CoGe2/Ge0.85Sb0.10Te	775	2.20	[54]
(Ge0.89Pb0.08Bi0.03Te)0.97(HgTe)0.03	650	2.30	[55]
Bi2(Te,Se)3	375	1.20	[56]
AgSbTe2	643	1.70	[57]
Cu3SbS4	773	1.30	[58]
Na0.99Cd0.995Ag0.005Sb	673	1.41	[59]
CuIn7Se11	873	1.23	[60]
PbSnS2	473	1.20	[61]
Sn0.71Ge0.2Mn0.07In0.02Te	873	1.64	[62]
EMIM:DCA	330	3.10	[63]
Sn0.78Sb0.16Te(MgB2)0.09	850	1.22	[64]
Mg3.2Bi1.998−xSbxTe0.002Cu0.005	348	1.10	[65]
GeSb2Te4	673	1.00	[66]
Mg3(Sb,Bi)2	773	1.82	[67]
AgMnGePbSbTe5	750	2.64	[68]
Nb0.75Ti0.25FeSb	973	1.21	[69]
(Nb, Hf)FeSb	973	1.47	[70]

3.1. Classical Thermoelectric Materials

Classical thermoelectric materials are the most widely used and longest-researched types, primarily including bismuth telluride (Bi₂Te₃), lead telluride (PbTe), and germanium-silicon (GeSi).

Bismuth telluride materials are among the earliest-studied thermoelectric materials, with the research dating back to the mid-1950s. Due to their excellent thermoelectric performance near room temperature, they have been widely industrialized, particularly in solid-state refrigeration and temperature control applications. Bismuth telluride has a rhombohedral crystal structure, with lattice constants a = b = 0.4385 nm and c = 3.0497 nm. Since the fundamental unit of the bismuth telluride crystal structure consists of a layered arrangement of ionic and covalent bonds, it exhibits relatively low thermal conductivity. Bismuth telluride possesses a complex band structure, and due to the strong orbital coupling effect of Bi atoms, it exhibits a small bandgap and high band degeneracy [71,72]. N-type and P-type thermoelectric materials based on bismuth telluride have been extensively studied, with ZT values typically ranging from 1.0 to 1.6 [73,74], as shown in Figure 6. Recently, Shi et al. [75] proposed a doping method using NaBiS₂ to increase the solubility of Na in p-type BST alloys. The BST + 4.0 wt% NaBiS₂ sample achieved a peak ZT of 1.44 at 373 K. However, the extremely low abundance of tellurium in the Earth’s crust is one of the main factors limiting the widespread application of bismuth telluride alloys.

Lead telluride compounds are among the earliest-studied thermoelectric materials for the mid-temperature range. In the 1950s, international institutions, primarily NASA, studied its thermoelectric properties and applied it in radioactive isotope thermoelectric power-generation devices, Figure 7. Lead telluride adopts a NaCl-type crystal structure, with lattice constants a = b = c = 0.6446 nm, and is classified as a direct-bandgap semiconductor [95]. Recently, significant progress has been made in the research of lead telluride thermoelectric materials, owing to the advances in material physics theory and improvements in fabrication techniques [96,97]. Currently, the main research direction focuses on adjusting the electrical and thermal transport properties of lead telluride-based thermoelectric materials by combining methods such as element doping, solid solution, and nanocomposites. Maxim et al. [98] reported that nanostructured lead telluride outperforms bulk lead telluride, with a thermoelectric conversion efficiency 10–14% higher and a ZT value of 1.34. The presence of nanostructures increases phonon scattering, thereby reducing thermal conductivity and enhancing conversion efficiency.

Silicon and germanium are elements in the same group, both exhibiting a diamond crystal structure and showing good solubility in one another. The lattice constant, band structure, and other physical properties of germanium–silicon semiconductor materials can be tuned by adjusting the elemental ratio. The research on germanium–silicon thermoelectric materials began in the 1960s. Over the years, the researchers have optimized their thermoelectric performance through doping, nanostructuring, and the fabrication of nanocomposites. Currently, the ZT values of N-type and P-type germanium–silicon thermoelectric materials have increased from their initial values of 1.0 and 0.7, respectively, to 1.5 [100] and 1.3 [101]. Lee et al. [102] proposed that the minimum thermal conductivity of germanium–silicon thermoelectric materials has not yet reached the theoretical limit of the alloy, as scattering effects can only be fully realized when silicon and germanium atoms are randomly distributed in space, which allows the material’s thermal conductivity to achieve its minimum value. Therefore, there is still potential for the further enhancement of the ZT value of germanium–silicon thermoelectric materials.

3.2. Novel Thermoelectric Materials

Since the turn of the 21st century, the research on thermoelectric materials has made remarkable progress, owing to the rapid advancements in science and technology. Not only have the thermoelectric properties of traditional materials greatly improved, but new classes of thermoelectric materials, such as tellurides, half-Heusler alloys, and skutterudite, have also been identified.

Telluride thermoelectric materials are considered promising alternatives to traditional lead telluride thermoelectric materials due to their non-toxicity, environmental friendliness, and alignment with sustainable development principles. These materials include tin telluride (SnTe), germanium telluride (GeTe), and others. Tin telluride and germanium telluride thermoelectric materials, like lead telluride, belong to the IV-VI compound family. Both exhibit a rhombohedral crystal structure at room temperature. In tin telluride, the high intrinsic Sn vacancy concentration results in a higher carrier concentration, leading to a poorer thermoelectric performance. Consequently, the current research primarily focuses on optimizing the material’s performance. Tanveer et al. [103] significantly enhanced the Seebeck coefficient and power factor of tin telluride thermoelectric materials by coordinating the control of band degeneracy and resonance levels. Consequently, the ZT value of the material reached 1.85 at 823 K. Due to its numerous similarities to lead telluride, germanium telluride has been the subject of thermoelectric property research since the 1960s. Owing to the rapid advancements in science and technology, the thermoelectric performance of germanium telluride has improved significantly in recent years. By employing various optimization strategies and methods, thermoelectric materials with a ZT value exceeding 2 have been achieved [93,104]. The existing reports show that the research on P-type germanium telluride is advancing rapidly, and the development of high-performance N-type germanium telluride thermoelectric materials is one of the main research directions for the future. Manisha et al. [105] synthesized N-type germanium telluride via the solid solution method, but its ZT value was limited to 0.6. Therefore, understanding the defects in germanium telluride and applying effective strategies are crucial for developing high-performance N-type germanium telluride.

Skutterudite has attracted significant attention from the researchers due to its “electron crystal–phonon glass” crystal structure, making it a natural high-performance thermoelectric material. For example, CoSb₃ has a body-centered cubic crystal structure, with each unit cell containing 8 CoSb₃ units, composed of 32 atoms. The unit cell also contains two Sb atoms, creating voids that can be occupied by other elements to form filled Skutterudite materials [106,107]. Due to the unique crystal structure of CoSb₃, its thermoelectric performance can be enhanced through methods such as impurity atom doping, solid solution alloys, and nanostructuring. Han et al. [108] synthesized N-type In_xCe_γCo₄Sb₁₂ thermoelectric materials by combining solution spinning with spark plasma sintering (SPS) techniques. The lattice thermal conductivity was reduced to 0.56 W/(m·K), while the ZT value reached 1.45. Xun et al. [109] doped three different atoms into the voids of Skutterudite, thus introducing multiphonon scattering and reducing the material’s thermal conductivity. At 850 K, the ZT value of the material reached 1.7.

Half-Heusler alloys are widely used in high-temperature applications due to their unique structure and properties. They typically have a ZnS-type crystal structure, with transition metals filling the octahedral voids, or a NaCl-type structure, with transition metals occupying the tetrahedral voids. The performance of half-Heusler alloys is influenced by the valence electron count of their constituent elements. When the valence electron count is 8 or 18, the material demonstrates a good thermoelectric performance. In the half-Heusler thermoelectric material system, MNiSn (M = Ti, Zr, Hf) is considered the best N-type material. Although it has a high Seebeck coefficient, its high thermal conductivity limits further development. Alloying, grain refinement, and doping are effective techniques for reducing thermal conductivity [110,111]. RFeSb (R = V, Nb) is the best P-type thermoelectric material, and band engineering improves its thermoelectric performance, providing a foundation for the development of half-Heusler thermoelectric devices [112,113].

3.3. Techniques for Material Preparation

Various methods have been developed for the synthesis and preparation of thermoelectric materials, such as solid-state reaction, solution methods [114,115], high-temperature and high-pressure methods [116], melt spinning [117,118], and chemical vapor deposition. This paper specifically discusses the preparation techniques for bulk materials (solid-state reaction) and flexible materials (vapor deposition).

3.3.1. Solid-State Reaction Technique

The solid-state reaction method involves a chemical reaction between solid materials to form new solid phases. It is one of the most commonly used methods for synthesizing thermoelectric materials. This method generally involves mixing two or more solid raw materials and reacting them at high temperatures to synthesize the desired material. It primarily involves raw material selection and mixing, high-temperature reactions, and cooling. The method offers several advantages, such as not requiring complex equipment and being relatively simple to operate. Chaithanya et al. [119]. synthesized Cu2-xBixSe material using the solid-phase reaction method, and the experimental results demonstrate that, at room temperature, Cu2-xBixSe exhibits a monoclinic crystal structure. The highest power factor of 1474 µW·m⁻¹·K⁻² was achieved at 700 K for the Cu_1.988Bi_0.012Se sample, demonstrating great potential for the development of high-performance Cu₂Se-based thermoelectric materials. Qin et al. [120] synthesized SbxFeTe₂ (x = 0.05, 0.1, 0.3, 0.5, 0.7, 0.9) using the high-temperature solid-state method. The study showed that excess Sb led to the formation of an Sb₂Te₃ secondary phase, resulting in a multiphase composite structure consisting of FeTe₂ and Sb₂Te₃. These structures significantly reduced the resistivity of the sample. At 873 K, the sample with 0.9 Sb content exhibited a resistivity of 0.86 mΩ·cm. Additionally, at 473 K, the ZT value of the Sb_0.7FeTe₂ sample was 0.25. Bai et al. [104] employed a vacuum solid-state reaction method to synthesize the compound XTe₂ (x = Fe, Co, Ni), which exhibits abundant micropores and a complex crystal structure. In this material, FeTe₂ exhibited a maximum Seebeck coefficient of 47.24 μV/K at 305 K. NiTe₂ exhibited the lowest resistivity of 1.45 mΩ·cm at 567 K. The CoTe₂ sample achieved a maximum power factor of 334.05 μW·m⁻¹·K⁻² at 570 K. The thermal conductivity of the pyrite samples followed the trend: NiTe₂ > CoTe₂ > FeTe₂, with FeTe₂ exhibiting the lowest thermal conductivity of 2.29 W/(m·K) at 599 K.

Bulk materials synthesized by the solid-state method typically exhibit a high Seebeck coefficient and good electrical properties, but their mechanical processability and deformability are limited. These materials are inherently brittle and prone to cracking under tensile, compressive, or bending stresses, which makes fabricating flexible thermoelectric devices challenging. Flexible materials prepared by vapor deposition typically exhibit good mechanical properties and deformability, offering advantages such as flexibility, light weight, and environmental friendliness. Therefore, vapor deposition plays a key role in the development of thermoelectric materials.

3.3.2. Vapor Deposition Technique

The vapor deposition method involves the reaction of vapor-phase materials at the gas–solid interface to form thin film materials. It is primarily classified into chemical vapor deposition (CVD) and physical vapor deposition (PVD). This method produces high-purity material films with uniform thickness, and the strong adhesion between the film and the substrate ensures stability in diverse applications. Jin et al. [121] designed and synthesized SnSe single crystals using the vertical vapor deposition method. The crystals were found to have a stoichiometric ratio of 1:1 and a space group of Pnma at room temperature. The characterization results demonstrated that SnSe single crystals exhibited an electrical conductivity σ = 39.6 S cm⁻¹ and a Seebeck coefficient of S = 566 μVK⁻¹ at around 580 K. By calculating the figure of merit, ZT has the largest value of ∼1.0 at around 800 K. Chuai et al. [122] prepared bismuth telluride thermoelectric thin film materials using plasma-enhanced chemical vapor deposition. The experimental results demonstrated that the material exhibited a 90% increase in transparency in the mid-infrared region and room-temperature mobility of 2094 cm²/V·s. Furthermore, the thermoelectric thin film material exhibited a decrease in room temperature conductivity and a significant increase in the Seebeck coefficient with increasing selenium content.

Although flexible thermoelectric materials prepared by vapor deposition offer many advantages, their thermoelectric efficiency is lower than that of bulk materials synthesized by the solid-state reaction method, due to the effects of their structure and properties. This limits their broader application.

4. Principles and Applications of Thermoelectric Conversion Technologies

4.1. Thermoelectric Power Generation

The principle of thermoelectric power generation is based on the direct conversion of the thermoelectric effect, which primarily relies on the temperature difference to generate voltage in the thermoelectric material. The charge carriers within the material migrate from the hot side to the cold side, generating a potential difference that becomes the output voltage [123,124], as shown in Figure 8. A thermoelectric generator mainly consists of thermoelectric materials, heat sources, and coolers. Thermoelectric power generation technology holds broad application potential, including waste heat recovery [125], space exploration [126,127], and portable power sources [128,129], among others. For thermoelectric devices, the power generation efficiency is defined as the ratio of the output power to the heat absorbed at the hot end, as shown in Equation (6). The refrigeration efficiency is defined as the ratio of heat absorbed to input electric power, with the maximum coefficient of performance (COP_max) shown in Equations (5) and (6) [130].

$$\mathrm{\varnothing }={\displaystyle \frac{T_{h-T_c}}{T_h}}{\displaystyle \frac{\sqrt{1+zT}-1}{\sqrt{1+zT}+\frac{T_h}{T_c}}}$$

(5)

$${COP}_{max}={\displaystyle \frac{T_c}{T_h-T_c}}{\displaystyle \frac{\sqrt{1+zT}-\frac{T_h}{T_c}}{\sqrt{1+zT}+1}}$$

(6)

where T_h is the hot-end temperature, T_c is the cold-end temperature, and ZT is the average dimensionless thermoelectric figure of merit.

4.2. Thermoelectric Refrigeration

Thermoelectric refrigeration is based on the Peltier effect, which describes the relationship between an electric current and the temperature change in the material. When an electric current flows through the junction of two conductors or semiconductors, heat is either absorbed or emitted. This means that at one contact point, heat is absorbed, causing a temperature drop, while at the other contact point, heat is emitted, causing a temperature rise [132,133]. A thermoelectric cooling system typically consists of four components: thermoelectric elements, a power supply, heat exchangers, and cooling surfaces. In contrast to traditional cooling methods, thermoelectric refrigeration offers advantages such as the absence of moving parts, environmental friendliness, and precise control. Thermoelectric cooling has broad applications in various fields, including electronic device cooling, portable refrigeration, and automobiles and other transportation vehicles.

4.3. Applications

System integration is a critical aspect of applying thermoelectric technology, involving the efficient combination of thermoelectric modules with other system components. Proper system integration can significantly improve the overall efficiency and reliability of thermoelectric materials. Selecting an appropriate heat source is a critical aspect of system integration. The ideal heat source should maintain a stable temperature gradient to ensure the efficient operation of the thermoelectric module. The design of the thermoelectric module is critical for ensuring the system operates efficiently. The design process includes material selection [134,135], geometric design [136], thermal interface materials [137,138], and heat dissipation design. In some systems, phase change materials (PCMs) can be integrated with thermoelectric modules. This integration stores excess thermal energy during peak load periods and releases it when demand increases, thereby enhancing the energy efficiency and stability of system [139].

4.3.1. Waste Heat Recovery

Studies have reported on converting industrial waste heat, geothermal energy, and waste heat from civilian sectors into electrical energy based on thermoelectric power generation principles, as shown in Figure 9. In the 1990s, Matsuura et al. [140] studied a project that generated megawatt-scale electrical power from industrial waste heat. The project used 98 °C circulating water as the heat source and 25 °C cooling water as the cold reservoir, achieving a power generation efficiency in the range of 8–10%. In 2017, the researchers from the Naval Engineering University of China developed an industrial wastewater heat recovery device, achieving a recovery efficiency of 1.28% and an investment payback period of approximately 8 years [141]. In the civilian sector, for example, waste heat from waste incineration, Japanese scholars developed kilowatt-scale waste heat power generation systems, resulting in the effective utilization of waste.

Solar energy and other natural heat sources are renewable and environmentally friendly sources of energy. Harnessing these natural heat sources for power generation to promote the transformation of energy utilization is in line with the principles of sustainable development. Among these, solar thermal power generation is the earliest-studied method. In 1954, Maria [142] studied the energy conversion efficiency of a solar thermoelectric generator using 25 pairs of ZnSb thermocouples, achieving an efficiency of 3.35%. Since then, the research on solar thermal power generation has gained increasing attention from scholars. After years of development, the research on solar thermal power generation now primarily focuses on integrating it with solar photovoltaic power generation to create a photovoltaic–thermoelectric hybrid power system [143,144]. Although the photovoltaic–thermoelectric hybrid system is an ideal energy conversion technology, it faces several key challenges that hinder its development, such as temperature matching between photovoltaic and thermoelectric devices, heat transfer at the interface, device selection, and energy conversion under non-steady-state conditions. Ge et al. [145] investigated the impact of TEG structural parameters on the efficiency of hybrid systems through mathematical modeling of PV–thermoelectric hybrid systems. The results showed that an optimal structural parameter (the ratio of the cross-sectional ratio to the height) with a value of 0.36 mm⁻¹ existed for the TEG, enabling the system efficiency to reach a maximum of 41.73% under simulated conditions.

Figure 9. (a) Assembling of the Te-free TE module; (b) performance of the Te-free segmented modules [146]. Schematic of an AEATEG with the counter-current flow: (c) overall structure, (d) sectional drawing in the axial direction, (e) sectional drawing in the radial direction, and (f) the structure of a single ATEG [147].

Figure 9. (a) Assembling of the Te-free TE module; (b) performance of the Te-free segmented modules [146]. Schematic of an AEATEG with the counter-current flow: (c) overall structure, (d) sectional drawing in the axial direction, (e) sectional drawing in the radial direction, and (f) the structure of a single ATEG [147].

4.3.2. Solid-State Refrigeration

Solid-state refrigeration is a cooling method that relies on thermoelectric materials and thermoelectric conversion technology, as shown in Figure 10. Compared to traditional compression-based systems, solid-state refrigeration technology offers numerous distinct advantages and application potential. At the core of solid-state refrigeration technology lies high-performance thermoelectric materials. Bi₂Te₃ is one of the most extensively studied thermoelectric refrigeration materials. Thermoelectric refrigeration devices and systems based on Bi₂Te₃ have been commercialized, including thermoelectric air conditioners and refrigerators. In the future, developing new, efficient thermoelectric materials to improve the efficiency and performance of solid-state refrigeration will be one of the primary research directions. Conversely, combining solid-state refrigeration technology with other cooling technologies could result in more efficient and flexible refrigeration solutions.

In recent years, with the advent of new preparation technologies and optimization strategies, the thermoelectric performance of Bi₂Te₃-based refrigeration materials has consequently improved. Wu et al. [150] incorporated Bi₂Fe₄O₉ magnetic nanoparticles into Bi₂Te₃-based thermoelectric refrigeration materials, where these nanoparticles effectively scatter low-energy charge carriers, resulting in enhanced thermoelectric properties. At 393 K, the thermoelectric figure of merit of the sample reached 1.1, representing a 13% improvement compared to its unmodified state. Liu et al. [151] optimized the thermoelectric properties at room temperature by constructing multiscale interfacial phases in LaFeSi (LFS)/BiSbT(BST) thermo-electromagnetically cooled materials. The results indicate that, at the atomic scale, the interfacial reaction between LFS and BST leads to the formation of (Fe,Co)(Sb,Te)₂ micrograins and LaTe₂ nanograins, which create a low-mismatch phase boundary with the LFS matrix. Compared to other thermoelectric materials, a favorable trade-off TE performance is achieved in LFS/BST composites, simultaneously demonstrating a high-TE performance. The 20% LFS/BST composite exhibits a room-temperature ZT of 0.46, along with a large maximum magnetic entropy change and relative cooling power of 0.81 J kg⁻¹ K⁻¹ and 44.83 J kg⁻¹, respectively.

5. Conclusions and Challenges

From the above discussion, it can be concluded that the current priority in the field of thermoelectric materials remains the development of thermoelectric materials with high ZT values to overcome low energy-conversion efficiencies, thereby establishing thermoelectric technology as a promising approach for energy utilization with extensive applications. The main strategies for enhancing the ZT value of thermoelectric materials include optimizing the carrier concentration, energy band engineering, phonon engineering, and entropy engineering, through which the electron and phonon transport properties of thermoelectric materials are optimized to achieve an enhanced thermoelectric performance. In the past decade, the ZT values of conventional thermoelectric materials have not only been increased to approximately 1.5, but new thermoelectric materials with ZT values exceeding 2 have also been developed.

Although there has been a rapid development of thermoelectric materials, the current research in this field continues to face several challenges, which not only limit their practical applications but also hinder the advancement of thermoelectric technology. The primary challenges include:

• Lower conversion efficiency. The conversion efficiency of thermoelectric materials is a critical issue, and the relatively low efficiency exhibited by current thermoelectric materials significantly restricts their broader application. Currently, although some thermoelectric materials exhibit an excellent performance within specific temperature ranges, the overall conversion efficiency has not yet reached a practical level.
• Poor high-temperature performance. Thermoelectric materials for high-temperature use typically have larger bandgaps, which complicates the balance between electrical conductivity and thermal conductivity. Most existing materials exhibit a poor performance in high-temperature environments, limiting their application in high-temperature conditions. However, a wider bandgap also results in a more complicated coupling effect between electrical transport properties (electrical conductivity and carrier mobility) and thermal transport properties (electronic and lattice thermal conductivities). This interaction makes the optimization of material properties more challenging; furthermore, most existing materials exhibit an inadequate performance at elevated temperatures due to insufficient thermal stability, decreased carrier mobility, and degradation of lattice structures, thus restricting their practical applications.
• The challenge of multicriteria optimization. The parameters governing the performance of thermoelectric materials are interdependent, requiring that multiple criteria (thermal conductivity, electrical conductivity, and Seebeck coefficient) be simultaneously satisfied during the design process. Optimizing material properties while simultaneously satisfying these criteria remains a complex theoretical and practical challenge, and the existing theoretical models and experimental approaches have not yet provided comprehensive solutions.
• Difficulty in converting technology. Although progress has been made in the basic research, translating these findings into practical applications remains challenging. Many promising thermoelectric materials perform well under laboratory conditions, but ensuring their consistency and repeatability in industrial production remains a challenge.

Although the research on thermoelectric materials is confronted with numerous challenges, advances may be achieved in overcoming these issues as new theoretical approaches, materials, and technologies continue to emerge. In particular, significant breakthroughs may be achieved through advancements in the multidisciplinary research, automated material design, and computational materials science. In summary, the development of high-performance thermoelectric materials is confronted with numerous significant challenges; however, with continued research and innovation, these materials are anticipated to achieve greater efficiency and broader applications in the future.