Home Journals Progress in Chemistry
Progress in Chemistry

Abbreviation (ISO4): Prog Chem      Editor in chief: Jincai ZHAO

About  /  Aim & scope  /  Editorial board  /  Indexed  /  Contact  / 
Review

Research Progress and Application of Flexible Thermoelectric Materials

  • Dong Baokun ,
  • Zhang Ting , * ,
  • He Fan
Expand
  • College of New Materials and Chemical Engineering, Beijing Institute of Petrochemical Technology,Beijing 102617,China
* Corresponding author e-mail:

Received date: 2022-08-15

  Revised date: 2022-09-29

  Online published: 2023-02-20

Supported by

National Natural Science Foundation of China(51501014)

2021 Beijing Undergraduates Research Training Program(2021J00057)

2022 National Undergraduates Research Training Program of China

Abstract

Thermoelectric materials, as one new kind of energy materials, can realize the direct conversion of thermal and electrical energy, which have important applications in power generation and refrigeration. Compared with traditional thermoelectric materials, flexible thermoelectric materials demonstrate excellent application prospects in wearable devices and flexible electronics fields, due to the advantages of being bendable, a lightweight and environmentally friendly. At present, how to further improve the performance of flexible thermoelectric materials is the focus, especially the collaborative optimization of flexibility andthermoelectric properties. In this paper, we have reviewed the research progress of polymer-based flexible thermoelectric materials, carbon-based flexible thermoelectric materials and inorganic semiconductor flexible thermoelectric materials, introduced their characteristics, performance optimization and preparation methods, and summarized the applications of flexible thermoelectric materials in the fields of electronics, medicine and industry. Also, based on the shortcomings of flexible thermoelectric materials, the future research directions are prospected.

Cite this article

Dong Baokun , Zhang Ting , He Fan . Research Progress and Application of Flexible Thermoelectric Materials[J]. Progress in Chemistry, 2023 , 35(3) : 433 -444 . DOI: 10.7536/PC220812

Contents

1 Introduction
2 Types of flexible thermoelectric materials and their thermoelectric properties
2.1 Polymer-based flexible thermoelectric materials
2.2 Carbon-based flexible thermoelectric materials
2.3 Inorganic semiconductor flexible thermoelectric materials
3 Preparation method of flexible thermoelectric materials
3.1 Physical vapor deposition
3.2 In-situ polymerization
3.3 Electrospinning
3.4 High temperature melting method
4 Applications of flexible thermoelectric materials
5 Conclusion and outlook

1 Introduction

Energy and environmental crisis is an important challenge facing mankind in the 21st century. Improving energy efficiency and reducing the burden on the environment are of great significance to the development of human society today. Therefore, it is very necessary to develop new green, efficient and sustainable new energy and new energy materials[1~3]. Thermoelectric conversion materials are a new type of energy materials that use the transport of carriers and phonons in materials to realize the direct conversion between heat and electricity, which have important applications in thermoelectric power generation and semiconductor refrigeration[4~6]. In recent years, thermoelectric materials have attracted more and more research and attention because of their advantages such as noiselessness, pollution-free, simple structure, safety, reliability and easy maintenance. They are not only used in military, aerospace and other high-tech fields, but also in medical thermostats, microsensors and other civil fields[7,8].
The energy conversion efficiency of a thermoelectric material can be measured by the dimensionless thermoelectric figure of merit ZT=S2σT/κ, where, S, σ, κ and T are the Seeback coefficient, electrical conductivity, thermal conductivity and absolute temperature of the material, respectively[9~11]. Thermoelectric materials with excellent performance should have high electrical conductivity and Seebeck coefficient and low thermal conductivity, but because of the strong coupling correlation between these parameters, they often affect and restrict each other, thus limiting the improvement of ZT value[12~14]. Therefore, how to improve the thermoelectric figure of merit has become the key to promote the further application of thermoelectric materials. Traditional inorganic thermoelectric materials, such as Bi2, Te 3 , PbTe and SiGe, usually have higher Seebeck coefficients than those of metals. At the same time, the synergistic optimization of their thermal and electrical properties can be achieved to a certain extent by element doping, nanocrystallization and band engineering, so as to obtain relatively high thermoelectric figure of merit[15][16][17]. However, although these traditional inorganic thermoelectric materials have good electrical properties, their mechanical processing and deformability are poor, and they usually show intrinsic brittleness, which is easy to break under tensile, compressive or bending loads, and it is not easy to make bendable thermoelectric devices.
In recent years, with the acceleration of the process of social informatization and intellectualization, the flexible electronics industry has developed vigorously. Due to the increasing demand for multifunctional green energy harvesting, flexible thermoelectric materials have also received more and more attention. Compared with traditional thermoelectric materials, flexible thermoelectric materials usually show good mechanical properties and deformability, and have the advantages of bendable shape, light weight and environmental friendliness, which have good application prospects in wearable electronic devices and other flexible electronic fields. However, due to the influence of the structure and properties of the materials, the thermoelectric efficiency of the flexible thermoelectric materials reported at present is generally lower than that of the traditional thermoelectric materials, which limits their wide application in the field of flexible electronics[18~20].
At present, the research on flexible thermoelectric materials mainly focuses on material development and performance optimization. According to the properties of materials, common flexible thermoelectric materials mainly include polymer-based flexible thermoelectric materials, carbon-based flexible thermoelectric materials and inorganic semiconductor flexible thermoelectric material. Polymer-based flexible thermoelectric materials and carbon-based flexible thermoelectric materials have good intrinsic flexibility, but their thermoelectric properties are relatively low due to the generally low Seebeck coefficient. Inorganic semiconductor flexible thermoelectric materials mainly include inorganic thin films and bulk materials. Fig. 1 shows the ZT value range of flexible thermoelectric materials reported in recent years. It can be seen from the figure that compared with polymer-based and carbon-based flexible thermoelectric materials, inorganic semiconductor flexible thermoelectric materials have higher ZT values and better thermoelectric properties. However, these materials usually need to have a special layered structure or through thin film and nano-treatment to have better flexibility, which has the problems of less types of materials and complex preparation methods. Therefore, how to optimize the thermoelectric and mechanical properties of flexible thermoelectric materials is the focus of the whole field of flexible thermoelectric research. In this paper, the research progress of flexible thermoelectric materials in recent years is reviewed, and the optimization methods of the structure and properties of the above three flexible thermoelectric materials are emphatically introduced. The preparation methods and practical application fields of flexible thermoelectric materials are summarized. In addition, in view of the current problems in the performance optimization and preparation of flexible thermoelectric materials, we also put forward some solutions, and look forward to the development of flexible thermoelectric materials.
图1 近几年报道的柔性热电材料(导电聚合物[21~23]、碳纳米管[24~26]、无机薄膜[27~29]和块体无机材料[30~32])的ZT

Fig. 1 ZT values of different FTE materials (conductive polymers[21~23], carbon nanotubes[24~26], inorganic films[27~29], inorganic bulks[30~32]) reported in recent years

2 Types and Thermoelectric Properties of Flexible Thermoelectric Materials

2.1 Polymer-based flexible thermoelectric material

Polymer-based flexible thermoelectric materials are mainly conductive polymer materials. Common polymer-based flexible thermoelectric materials include poly (3,4-ethylenedioxythiophene) (PEDOT), polyaniline (PANI), and polypyrrole (PPy)[22][33][34]. On the one hand, the molecular backbone of these materials usually has a certain degree of freedom of internal rotation, which can bend and show good flexibility. At the same time, the material has the advantages of easy processing, light weight and low thermal conductivity, and has a good application prospect in flexible wearable thermoelectric devices. On the other hand, polymer-based flexible thermoelectric materials, due to their highly disordered structure, exhibit excellent thermal properties, and their thermal conductivity is usually lower than 1 W·m-1·K-1, which is close to the lower limit of the thermal conductivity of inorganic thermoelectric materials[35]. However, these materials have poor electrical properties in the early stage of development, resulting in ZT values that are not comparable to inorganic thermoelectric materials. In recent years, with the in-depth study of polymer-based flexible thermoelectric materials, some methods to improve the electrical properties have been proposed, and the electrical properties and ZT values have been greatly improved. The thermoelectric performance optimization methods of this kind of materials mainly include the following two methods.
One is to improve the thermoelectric properties of materials by doping or de-doping. In general, doping and dedoping can strongly affect the carrier mobility and carrier concentration, and then affect the conductivity and Seebeck coefficient of conducting polymers. PEDOT is the most studied and best applied conductive polymer, which is usually mixed with polystyrene sulfonic acid (PSS). Due to the special conjugated molecular chain structure of PEDOT: PSS, the material has high electrical conductivity, low thermal conductivity and good flexibility. At present, the main methods to optimize the electrical properties of PEDOT polymers are doping or de-doping. For example, Mengistie et al. prepared a flexible PEDOT: PSS paper-like film (as shown in Figure 2a) by vacuum-assisted filtration, and they doped the film with ethylene glycol (EG), polyethylene glycol (PEG), methanol and formic acid, respectively.The conductivity and power factor of the doped film have been greatly improved, especially after doping with formic acid, the conductivity of the film at room temperature is as high as 1900 S·cm-1, and the power factor also reaches 80.6μW·m-1K-2 (Figure 2B)[21]. In 2013, Kim et al. Used polystyrene sulfonate to dope p-type PEDOT to reduce the volume of counterions and improve the carrier mobility of the material by removing part of the unit counterions[22]. Finally, the material obtained high electrical conductivity (900 S·cm-1) and high Seebeck coefficient (72μV·K-1) at the same time, and the power factor was improved to 460μW·m-1·K-2, which made the ZT value of the material reach 0.4 at room temperature, which belongs to a higher level among polymer-based flexible thermoelectric materials.
图2 (a)PEDOT:PSS类纸状薄膜的柔韧性;(b)室温下柔性PEDOT:PSS类纸状薄膜掺杂不同物质后的电学性能[21];SnSe纳米片/PEDOT:PSS复合材料在室温下的Seekbeck系数(c)和ZT值(d)随SnSe纳米片含量的变化[36]

Fig. 2 (a)Flexibility of PEDOT:PSS bulky papers;(b)Electrical properties of flexible PEDOT:PSS bulky paper doped with different substances[21];Seekbeck coefficient (c) and ZT value (d) of SnSe nanosheets/PEDOT:PSS composites with various of SnSe nanosheet contents at room temperature[36]

Another performance optimization method is to compound inorganic nanomaterials with polymer-based flexible thermoelectric materials, and use the high conductivity and Seebeck coefficient of inorganic nanomaterials and the low thermal conductivity of polymer-based flexible thermoelectric materials to achieve the performance improvement of composite flexible thermoelectric materials. For example, Ju et al. prepared SnSe nanosheet/PEDOT: PSS composite, and the Seekbeck coefficient of the composite was greatly improved after SnSe nanosheet was compounded (Fig. 2C)[36]. Especially when the content of SnSe nanosheets is 20%, the Seekbeck coefficient and power factor of the composite at room temperature reach 110μV·K-1 and 390μW·m-1·K-2, respectively, and the maximum ZT value also reaches 0.32 (Figure 2D). In addition, See et al. Prepared Te nanorods/PEDOT: PSS nanocomposites, and the ZT value of the composite reached 0.1 at room temperature by using the synergistic effect of the high Seebeck coefficient of Te nanorods (408 V·K-1 at room temperature) and the low thermal conductivity of PEDOT: PSS (0.24~0.29 W·m-1K-1 at room temperature)[37].
Although polymer-based thermoelectric materials have good intrinsic flexibility and their electrical properties have been greatly improved by appropriate methods, their Seebeck coefficient and power factor are still low compared with inorganic thermoelectric materials, which limits the further development and application of polymer-based flexible thermoelectric materials.

2.2 Carbon-based flexible thermoelectric material

Carbon-based flexible thermoelectric materials mainly include graphene-based flexible thermoelectric materials and carbon nanotube-based flexible thermoelectric materials. Graphene and carbon nanotubes (CNTs) are the most advanced carbon materials at present. In addition to their advantages of high mechanical strength, light weight, low cost and non-toxicity, they also have certain flexibility due to their special carbon-carbon bond structure, which shows certain application prospects in the field of flexible electronics[38]. Carbon atoms in graphene and carbon nanotubes are hybridized in a sp2 manner, so that the graphene and the carbon nanotubes have special two-dimensional or nanotubular structures and can generate high carrier mobility and conductivity; But at the same time, both of them have high thermal conductivity and low Seebeck coefficient, which limits the further improvement of the thermoelectric performance of these materials[39,40]. At present, to solve this problem, people mainly use the method of compounding conductive polymer with graphene or carbon nanotubes. On the one hand, graphene, carbon nanotubes and conductive polymers all have flexibility, and the composite material can still maintain good flexibility; On the other hand, the conductive polymer with low thermal conductivity can be used as a second phase to enhance phonon scattering, reduce lattice thermal conductivity, improve the Seebeck coefficient of the composite material, and greatly improve the thermoelectric performance of the carbon-based flexible thermoelectric material.
At present, the research of graphene-based composite flexible thermoelectric materials is mainly focused on PANI/graphene composites. Because of its large specific surface area, graphene can form more nano-interfaces, stronger π-π conjugation effect and more ordered PANI molecular chains when compounded with PANI. The nano-interface can effectively reduce the lattice thermal conductivity of PANI/graphene composites, while the π-π conjugation effect and the more ordered PANI molecular chain structure can improve the Seebeck coefficient, so that the thermal and electrical properties of the composites can be optimized synergistically. For example, Wang et al. Prepared PANI/graphene (PANI/GP-P) flexible composite thermoelectric film by in-situ polymerization using chemical vapor deposition, and the film was bent and deformed without fracture after bearing large stress, indicating that the film has good flexibility[41]; At the same time, the in-situ polymerization makes the graphene uniformly dispersed in the PANI matrix, which not only introduces more nano-interfaces, but also strengthens the π-π conjugation effect between PANI and graphene, and improves the Seekbeck coefficient of the composite material, so that the Seebeck coefficient of the PANI/GP-P flexible composite film containing 48% graphene is increased to 26μV·K-1 at room temperature, and the maximum power factor reaches 55μW·m-1·K-2. In addition, Xiang et al. Prepared PANI/graphene nanosheet (GNP) flexible nanocomposite film on polyvinylidene fluoride (PVDF) filter membrane by in-situ polymerization using vacuum-assisted self-assembly method. Because of the flexibility of PVDF filter membrane, the composite film showed better flexibility[42]. At the same time, the π-π conjugation effect between PANI and graphene makes PANI have a strong affinity for GNP, which can form a uniform nanocoating on the surface of GNP, which not only retains the high Seebeck coefficient of PANI, but also improves the conductivity of the composite film. Finally, the ZT value of the composite film with 40% PANI can reach up to 1.51×10-4 at room temperature.
PEDOT: PSS/CNT, PANI/CNT and PPy/CNT are relatively mature in the research of carbon nanotube-based composite flexible thermoelectric materials. The composite of CNTs and these conductive polymers not only maintains the original flexibility, but also reduces the lattice thermal conductivity of the materials and improves the Seekbeck coefficient. For example, Liang et al. Prepared PPy/SWCNT flexible composite film by vacuum filtration method, and with the increase of SWCNTs content, the conductivity and Seekbeck coefficient of the composite film were improved, and the conductivity, Seekbeck coefficient and power factor of 60% SWCNTs composite film reached 399±14 S·cm-1, 22.2±0.1μV·K-1 and 19.7±0.8μW·m-1·K-2 at room temperature, respectively, which were greatly improved compared with PPy without SWCNTs[43]; In addition, the research group also studied the thermoelectric properties of the composite film after multiple bending and stretching. As shown in Fig. 3A and B, the large-area, flexible and stretchable PPy/SWCNT composite film can still maintain stable thermoelectric properties when it is repeatedly bent for 1000 times and its elongation reaches 2.6% (Fig. 3C and d), indicating that its mechanical stability is very good. Du et al. Prepared a series of SWCNT/PEDOT: PSS (CNT/PP) flexible composite thermoelectric films with different SWCNT contents (20%, 40%, 80%) on porous nylon membranes by dilution-filtration method. These films can be bent without breaking at different curvature radii, showing good flexibility[44]. Finally, with the increase of SWCNTs content, the electrical and thermoelectric properties of the composite films are improved.
图3 PPy/SWCNT柔性复合薄膜弯曲(a)和拉伸(b)示意图;PPy/SWCNT柔性复合薄膜在弯曲(c)和拉伸(d)后的热电性能[43]

Fig. 3 (a)Schematic diagram of bending (a) and stretching (b) of PPy/SWCNT flexible composite films;thermoelectric properties of PPy/SWCNT flexible composite films after bending (c) and stretching (d)[43]

In recent years, in addition to graphene-based and carbon nanotube-based flexible thermoelectric materials, scholars have also studied the flexible composite thermoelectric materials composed of graphyne (GDY), Ta4SiTe4 one-dimensional whiskers and PVDF. In 2020, Shi et al. Prepared PVDF/Ta4SiTe4 flexible composite by wet chemical method, and the composite material showed good flexibility while maintaining high Seekbeck coefficient and power factor[45]. On this basis, the research group prepared Ta4SiTe4/PVDF/GDY flexible composites with graphyne (GDY) and PVDF/Ta4SiTe4 composites[46]. GDY acts as a bridge to promote the transport of carriers from one Ta4SiTe4 whisker to nearby Ta4SiTe4 whiskers, which significantly enhances the electrical conductivity, combines the high Seekbeck coefficient and low thermal conductivity of the material itself, and finally, the maximum ZT value of the 40 wt%Ta4SiTe4/PVDF/5.8 wt%GDY flexible composite reaches 0.2 at room temperature.
Generally speaking, the thermoelectric performance of composite carbon-based flexible thermoelectric materials has been improved while having high flexibility, but there is still a gap between the thermoelectric performance of composite carbon-based flexible thermoelectric materials and that of traditional inorganic semiconductor materials, which needs to be further optimized. In addition, the research on the design of graphene crystal structure is largely limited by experiments, and most of them remain in the stage of theoretical calculation, which also affects the practical application of graphene-based flexible thermoelectric materials[47~49].

2.3 Inorganic semiconductor flexible thermoelectric material

Inorganic semiconductor flexible thermoelectric materials can be divided into two categories according to their morphologies: one is flexible thermoelectric thin films prepared by depositing traditional inorganic semiconductor thermoelectric materials on flexible substrates, and the other is bulk inorganic semiconductor flexible thermoelectric materials based on metal chalcogenides discovered in recent years. Flexible thin films prepared by traditional inorganic semiconductor materials, on the one hand, low thermal conductivity can produce better thermoelectric properties; On the other hand, flexible materials such as polyimide can be used as substrates to make the film flexible. For example, Kong et al. Prepared flexible Bi2Te3 thermoelectric film on polyimide substrate by magnetron sputtering, and the film can still maintain good thermoelectric performance after 2000 times of bending[50]; At the same time, the power factor of the film can reach 21.7μW·cm-1·K-2 at room temperature and 3.0 Pa pressure by adjusting the sputtering pressure. However, these flexible thermoelectric thin films often have some problems, such as harsh preparation conditions, high cost, and difficult control of film thickness and composition uniformity, which limit their further application[51]. Compared with thin films, the thickness of bulk materials is usually larger, and because of its weak interlayer force, cleavage is easy to occur, and the deformation ability is relatively weak. However, in recent years, scholars have discovered a class of flexible bulk inorganic semiconductor thermoelectric materials based on metal chalcogenides. This kind of material usually has a special layered structure, and there are Van der Waals forces between the layers, so that the relative slip between the layers can occur when the material is deformed, but it does not break, resulting in better flexibility. At present, the bulk flexible thermoelectric materials that have been discovered mainly include Ag2S-based thermoelectric materials, InSe-based thermoelectric materials and the newly discovered AgCuSe-based thermoelectric materials. These materials have good thermoelectric properties, and have good application prospects in the field of flexible thermoelectricity because of their easy processing and batch synthesis.

2.3.1 Ag2S-based flexible thermoelectric materials

Ag2S is a typical fast ionic conductor with a relative density of 7.317 g/cm3 and a melting point of 825 ° C. Due to their excellent physical and chemical properties, Ag2S have attracted wide attention in the fields of catalysis, biology, optoelectronics and solar cells[52~55]. As early as 1951, Hebb et al. Have done related research on the electrical properties of Ag2S, but the mechanical properties of Ag2S, such as flexibility, have been neglected[56]. Until 2018, Shi et al. Made α-Ag2S polycrystal by high temperature melting combined with plasma sintering technology, and found that the α-Ag2S showed excellent ductility and high plasticity, and the deformation far exceeded that of known ceramics and semiconductor materials, which was comparable to the mechanical properties of some metals, which opened the research on Ag2S-based flexible thermoelectric materials[57].
Structure and Properties of 1)Ag2S-Based Compounds
Ag2S can undergo a structural phase transition at 455 K, and there are two different phase structures before and after the transition, which are room temperature α phase and intermediate temperature β phase, respectively[58]. As shown in Fig. 4, the room temperature α-Ag2S phase has a zigzag layered monoclinic structure (space group P21/c), each layer is composed of multiple 8-atom rings, each atom ring is composed of four S atoms and four Ag atoms, and the rings are connected by S atoms, so the folded layers are stacked along the a axis. There is a relatively low slip energy barrier between the atomic layers of this layered structure, which makes it easy to slip between the layers, but because there is a certain atomic interaction force between the sliding planes, the material can avoid cracking or dissociation in the process of sliding, which makes the α-Ag2S show good flexibility[57]. When the temperature exceeds 455 K, the α-Ag2S will transform into the body-centered cubic structure of the β-Ag2S, so that its layered structure is destroyed, and the β-Ag2S is no longer flexible.
图4 α-Ag2S的晶体结构[57]

Fig. 4 Crystal structures of α-Ag2S[57]

The Ag2S,N type direct band gap semiconductor, which has very low intrinsic lattice thermal conductivity, is a potential thermoelectric material. At room temperature, the theoretical band gap of α-Ag2S is about 1.0 eV, and the wide band gap makes its conductivity not high, only 0.1~0.5 S·m-1. However, in the intermediate temperature β-Ag2S, the carrier mobility and electrical properties of the β-Ag2S are improved due to the free migration of Ag ions in the unit cell gap due to its liquid-like characteristics, and the band gap is reduced to 0. 42 eV, which is comparable to that of traditional thermoelectric materials such as PbTe and SiGe[16][17]. The improvement of electrical properties makes the thermoelectric performance of β-Ag2S much higher than that of α-Ag2S. For example, Wang et al. Prepared single-phase Ag2S by high temperature melting combined with annealing, and the carrier concentration, carrier mobility and Seebeck coefficient of the Ag2S were greatly improved before and after the phase transition at 455 K, which increased the power factor from lower than 0.3μW·cm-1·K-2 to 5μW·cm-1·K-2[59]. At the same time, due to the lower thermal conductivity, the maximum ZT value of α-Ag2S is lower than that of 0.02,β-Ag2S, which reaches 0.55 at 580 K. It can be seen that although α-Ag2S has good flexibility, its thermoelectric performance is not high, which limits its application in flexible thermoelectric wearable devices. How to optimize the mechanical properties and thermoelectric properties of room temperature phase α-Ag2S has become the main research content of Ag2S-based flexible thermoelectric materials.
Performance optimization of 2)Ag2S-based flexible thermoelectric materials
In silver chalcogenide Ag2X(X=S,Se,Te), the thermoelectric properties of Ag2S are much lower than those of Ag2Se and Ag2Te due to their large band gap.But among the three, Ag2Se and Ag2Te present brittleness, and only Ag2S has flexibility, as shown in Fig. 5A[30,60,61]. Meanwhile, the existence of natural minerals such as Ag4SSe and Ag4STe indicates that Ag2S can form solid solution with Ag2Se and Ag2Te[62][63]. Therefore, the synergistic optimization of flexibility and thermoelectric properties of Ag2S based compounds is expected to be achieved through solid solution and the regulation of solid solution ratio.
图5 (a)银硫族化合物的柔韧性-ZT值相图;(b)Ag2S基柔性热电材料的ZT值随温度的变化;(c)Ag2S0.5Se0.5的柔韧性;(d)Ag2S0.5Se0.5制作的柔性热电器件[30]

Fig. 5 (a)The flexibility-ZT value phase diagram of silver chalcogenides;(b)Temperature dependence of ZT value of Ag2S-based FTE materials;(c)Flexibility of Ag2S0.5Se0.5;(d)FTE device fabricated by Ag2S0.5 Se 0.5 30

Shi et al. Made a detailed study on solid solution Ag2S based flexible thermoelectric materials, prepared a series of Se and Te solid solution Ag2S1-xSex, Ag2S0.8Te0.2 and ternary solid solution Ag2S0.5Se0.45Te0.05 by high temperature melting and plasma sintering technology, and studied the effect of solid solution on the performance of Ag2S based flexible thermoelectric materials[30,31,57]. The results show that the conductivity of Ag2S0.5Se0.5 and Ag2S0.5Se0.45Te0.05 increases to 3.0×104S·m-1 and 2.7×104S·m-1, respectively, after solution treatment.5 orders of magnitude higher than that of the room temperature phase α-Ag2S (about 0.1 S·m-1); At the same time, the solid solution of Se and Te can introduce point defects to scatter phonons and reduce the lattice thermal conductivity. Finally, the ZT values of Ag2S0.5Se0.5 and Ag2S0.5Se0.45Te0.05 reach 0.26 and 0.44 at room temperature, respectively, which are greatly improved by four orders of magnitude compared with α-Ag2S (about 8×10-5), as shown in Fig. 5B. While the thermoelectric properties were improved, they also studied the flexibility of the sample after solid solution, and found that the transition boundary between monoclinic structure and orthorhombic structure in Ag2 Se 1 - x Sx solid solution was about X = 0.3, and when X ≤ 0.2, the transition boundary between monoclinic structure and orthorhombic structure was about X = 0.3.The Ag2 Se 1 - x Sx samples crystallize in the orthorhombic structure, while they crystallize in the monoclinic structure for X ≥ 0.4,Only the monoclinic Ag2 Se 1 - x Sx samples with X ≥ 0.4 have good flexibility. For example, when X = 0.5, the flexibility of the Ag2S0.5Se0.5 sample is comparable to that of α-Ag2S (Figure 5C), and the thermoelectric properties of the sample are not greatly affected after repeated bending, so the Ag2S0.5Se0.5 sample is also used to fabricate flexible thermoelectric devices (Figure 5D). Similarly, the sample Ag2S0.8Te0.2, in which Te is solubilized with Ag2S, has similar properties. This indicates that an appropriate amount of solid solution can not only control the thermoelectric properties of Ag2S-based compounds, but also keep them flexible. In addition, He et al. Prepared Ag2Te1-xSx-based flexible thermoelectric materials by high temperature melting combined with planar hot pressing, and found that the Ag2Te0.6S0.4 sample showed good flexibility and thermoelectric properties.The carrier concentration of the sample at room temperature is remarkably improved, and meanwhile, the microstructure shows that amorphous Ag2Te0.6S0.4 with a disordered structure exists in the sample, so that the lattice thermal conductivity of the material is remarkably reduced, and finally, the ZT value of the material at room temperature reaches 0.22[32]; Meanwhile, the mechanical property test results of different solid solutions show that the flexibility of the Ag2Te0.6S0.4 sample is the best, which is close to α-Ag2S. To sum up, for Ag2S-based flexible thermoelectric materials, solid solution is an effective way to improve their thermoelectric performance and maintain good flexibility.

2.3.2 InSe-based flexible thermoelectric materials

Inspired by the α-Ag2S, Shi et al. focused the development of new semiconductor flexible thermoelectric materials on two-dimensional materials containing van der Waals force, and found that the InSe single crystal with layered structure not only has the excellent physical properties of traditional inorganic semiconductors, but also has extraordinary plasticity and deformability like metals (Fig. 6C)[64]. This brings new opportunities for the development and application of new semiconductor flexible thermoelectric materials.
图6 β-InSe(a)和γ-InSe(b)的晶体结构[68];(c)不同材料的可变形能力;(d)β-InSe单晶的柔韧性[64]

Fig. 6 Crystal structures of β-InSe (a) and γ-InSe (b)[68];(c)deformability of different materials;(d)flexibility of β-InSe single crystals[64]

1) Structure and properties of InSe-based compounds
InSe mainly includes hexagonal β phase and rhombohedral γ phase, both of which show layered structures, as shown In Fig. 6a, B. Each layer In these two structures is composed of covalently bonded Se-In-In-Se, and the layers are stacked along the C axis. The stacking sequences of β-InSe and γ-InSe are ABAB and ABCABC, respectively, and the layers are connected by weak van der Waals forces[65][66~68]. Similar to α-Ag2S, InSe single crystal also has good flexibility, and related experiments also verify the flexibility of β-InSe and γ-InSe. For example, in 2002, Mosca et al. measured the hardness and elastic modulus of γ-InSe single crystal by nanoindentation test, which were about 1.6 and 28 GPa, respectively, proving that γ-InSe has good flexibility[69]. Shi et al. also studied the flexibility of two-dimensional β-InSe single crystal, and found that the material can be bent, twisted, and even deformed into various shapes without fracture in the bulk form, showing good flexibility macroscopically (Fig. 6d)[64]; The related mechanical test results also show that the compressive engineering strain of the material can reach 80%, and the bending and tensile engineering strains in specific directions are also higher than 10%, which further verifies the good mechanical properties of the material.
InSe is an n-type semiconductor, and the theoretical band gaps of β-InSe and γ-InSe are 1.2 and 0.52 eV, respectively, both of which are direct gap semiconductors. Shi et al. Showed that InSe can obtain a low thermal conductivity of about 1.6 W·m-1K-1 at room temperature[70]; But at the same time, the carrier concentration of InSe is also very low, only 3.2×1013cm-3, which is several orders of magnitude lower than that of traditional thermoelectric materials with better performance (such as Bi2Te3 with a carrier concentration of about 1.1×1019cm-3), which largely limits the improvement of InSe thermoelectric performance[71]. Therefore, how to improve the carrier concentration of InSe to further optimize its electrical properties is the key to the research of InSe-based flexible thermoelectric materials.
2) Performance optimization of InSe-based flexible thermoelectric materials
At present, the research on InSe-based flexible thermoelectric materials is mainly focused on InSe single crystals. Zhang et al. made a detailed study on the mechanical properties and thermoelectric properties of γ-InSe single crystal flexible thermoelectric materials. They found that the bending deformation process of γ-InSe single crystal not only did not affect the thermoelectric properties of the material, but also induced more defects, enhanced phonon scattering, thereby reducing the lattice thermal conductivity, which was conducive to the improvement of thermoelectric properties[72]. However, the conductivity of γ-InSe single crystal is only 00027 at 280 K due to the low carrier concentration, which is only 0.00027. This indicates that the thermoelectric properties of γ-InSe single crystal need to be further optimized and improved, although it has good flexibility. Zhang et al also pointed out that the electrical properties of γ-InSe single crystal can be further optimized by doping or solid solution, so as to obtain better InSe-based flexible thermoelectric materials.

2.3.3 AgCuSe-based flexible thermoelectric material

The fabrication of traditional cross-plane π-type flexible thermoelectric devices often requires both P-type and N-type thermoelectric materials with good flexibility. Although the discovered Ag2S-based flexible thermoelectric materials and InSe-based flexible thermoelectric materials have good flexibility,However, both of them are N-type semiconductors, and there is a lack of P-type semiconductor thermoelectric materials with both flexibility and good thermoelectric performance, which greatly limits the development of flexible thermoelectric devices. Based on this, in 2022, Shi et al. Discovered a new type of p-type AgCuSe-based flexible thermoelectric material, namely AgCu (Se, S, Te) pseudo-ternary solid solution[73].
Similar to Ag2S and InSe compounds, AgCuSe also has two different phase structures, which can undergo a structural phase transition at 504 K from the low temperature β phase of the orthorhombic structure to the high temperature α phase of the face-centered cubic structure. Shi et al. Found that AgCuSe was brittle and not flexible. They used the AgCu (Se, S, Te) pseudo-ternary solid solution composition-performance phase diagram (Fig. 7 a) to dissolve S and Te into AgCuSe, which not only converted AgCuSe from a brittle N-type semiconductor to a P-type semiconductor with better flexibility, but also improved its thermoelectric performance, in which AgCuSe0.22S0.08Te0.7 had the best performance. In addition, the mechanical property experiment of AgCuSe0.22S0.08Te0.7 by the research group shows that the material is easy to bend without fracture (Fig. 7 C), and its better flexibility mainly comes from the S element. Although the content of S element is very low, in the three-point bending experiment, the engineering strain of AgCuSe0.3Te0.7 without S is only 3%, while the AgCuSe0.22S0.08Te0.7 is significantly improved to 18%, which is comparable to that of Ag2S-based flexible thermoelectric materials. Meanwhile, the electrical properties of AgCuSe are improved due to the solid solution of Te, which makes the ZT value of AgCuSe0.22S0.08Te0.7 reach 0.45 at room temperature (Fig. 7 B). Based on the good flexibility and thermoelectric performance of AgCuSe0.22S0.08Te0.7, the research group fabricated a cross-plane π-type flexible thermoelectric device with N-type Ag20S7Te3 flexible semiconductor (Fig. 7d).
图7 (a)AgCu(Se,S,Te)伪三元固溶体成分-性能相图;(b)AgCuSe基柔性热电材料的ZT值随温度的变化;(c)AgCuSe0.22S0.08Te0.7的柔韧性;(d)AgCuSe0.22S0.08Te0.7制作的柔性热电器件[73]

Fig. 7 (a)Compositionperformance phase diagram of AgCu(Se,S,Te)pseudoternary solid solutions;(b)temperature dependence of ZT value of AgCuSe-based FTE materials;(c)flexibility of AgCuSe0.22S0.08Te0.7;(d)FTE device fabricated by AgCuSe0.22S0.08 Te 0.7 73

3 Preparation method of flexible thermoelectric material

Different types of flexible thermoelectric materials can be prepared by different methods, including physical vapor deposition, in-situ polymerization, electrospinning and high temperature melting.

3.1 Physical vapor deposition

Physical vapor deposition (PVD) is a method in which a solid source is converted into a vapor phase material by evaporation or sputtering under vacuum conditions, and then these energy-carrying evaporation particles are deposited on the substrate to form a thin film[18]. Physical vapor deposition (PVD) for the preparation of flexible thermoelectric materials mainly includes magnetron sputtering and thermal evaporation deposition, both of which directly deposit inorganic thermoelectric materials on flexible substrates to prepare flexible thermoelectric films.
Magnetron sputtering is a process in which positive ions running at high speed bombard the target material to make the surface atoms of the target material emit and deposit on the surface of the substrate[74]. The method can accurately control the thickness, crystallinity and the like of the film by regulating and controlling process parameters such as sputtering time, pressure, annealing temperature and the like, so that the flexible film can obtain the best thermoelectric performance. Higher sputtering pressure will result in films with better crystallinity and more ordered crystalline structure, Shen et al. Prepared flexible Sb2Te3 thermoelectric films by DC magnetron sputtering on polyimide substrates at deposition temperature of 473 K,The conductivity and Seekbeck coefficient of the films are improved due to the higher sputtering pressure, which increases the carrier mobility and optimizes the carrier concentration[75]. At the same time, the increase of sputtering pressure will also reduce the grain size and increase the grain size, thus obtaining more grain boundaries to enhance phonon scattering and reduce the lattice thermal conductivity. Finally, the ZT value of the film is the best at the sputtering pressure of 4. 0 Pa. In addition, annealing the sputtered film can reduce the defects and enhance the crystallinity and grain size of the film, and improve the conductivity. Singkaselit et al. Prepared flexible BixTey thermoelectric films on polyimide substrates by radio frequency magnetron sputtering.The annealing temperature of the film was optimized, and the carrier concentration and carrier mobility were significantly increased. When the annealing temperature was 400 ℃, the power factor of the film reached a 11.45×10-4W·m-1·K-2 at 300 ℃[76]. Magnetron sputtering is one of the most commonly used physical methods to prepare flexible thermoelectric thin films, but this method also has the problems of high production cost and relatively low preparation efficiency, and the internal strain of the sputtered film usually affects the improvement of the performance of the flexible thin film.
Thermal evaporation deposition is a method in which a source substance to be evaporated is heated under vacuum conditions to evaporate and deposit on the surface of a substrate to form a thin film[77]. In the process of thermal evaporation deposition, the parameters such as evaporation rate and substrate temperature will affect the properties of the film. The increase of substrate temperature will make the adatoms gain more energy, thus improving the surface mobility and aggregation ability of adatoms, increasing the grain size and improving the conductivity of the film. However, with the increase of substrate temperature, the deposition elements will re-evaporate to affect the composition of the film. In addition, it is necessary to control the evaporation ratio between the deposition elements to obtain the best stoichiometric ratio of the film. Goncalves et al. Prepared P-type Sb2Te3 flexible thermoelectric thin films on 25 μm-thick polyimide substrates by thermal evaporation deposition method, and the carrier mobility and conductivity of the films were improved as the substrate temperature increased from 150 ℃ to 220 ℃, but Sb and Te would re-evaporate at this temperature, so it was necessary to optimize the evaporation rate of both to control the evaporation ratio, so as to obtain the best stoichiometric ratio of Sb2Te3[78]. Finally, the power factor of the film can reach 2.8×10-3W·K-2·m-1 at a substrate temperature of 220 ° C and Sb and Te evaporation rates of 2 and 6.4Å·s-1, respectively. In addition, the same method was used to prepare the N-type Bi2Te3 flexible thermoelectric film, and the evaporation rate and substrate temperature of the film were optimized to make the maximum power factor of the Bi2Te3 film reach 4.87×10-3W·m-1·K-2[79]. Compared with magnetron sputtering, thermal evaporation deposition has the advantages of simple equipment and high preparation efficiency, but the thin films prepared by thermal evaporation deposition are easy to produce impurity phases due to the different evaporation rates of the two evaporation sources, and the composition of the thin films is easy to deviate.

3.2 In situ polymerization

In situ polymerization is a process in which nanoparticles are dispersed in a polymer matrix and then polymerized with a catalyst under certain conditions to form polymer nanocomposites[80]. Generally, the nanomaterials added during in situ polymerization are cross-linked with the polymer matrix through covalent bonds and form a stable second phase in the matrix. The method is mainly used for preparing the polymer-based flexible composite thermoelectric material, and compared with simple mechanical blending, the in-situ polymerization can easily and uniformly disperse the inorganic nano thermoelectric material into the polymer matrix,To enhance the thermodynamic compatibility of the composite interface and obtain a strong interfacial interaction, which will be beneficial to the material to obtain better flexibility and thermoelectric properties. Chatterjee et al. Synthesized PANI/Bi2Te3 composite flexible thermoelectric materials by in situ polymerization using solvothermal method and chemical oxidative polymerization technique[81]. On the one hand, the composite retains the good flexibility of PANI, on the other hand, the addition of Bi2Te3 nanorods makes PANI obtain a more ordered structure on the surface of PANI, so that the carrier mobility of the composite is increased, and the conductivity and power factor are also improved; Meanwhile, the polymerization generated more nanointerfaces to scatter phonons and reduce the lattice thermal conductivity, and finally, the ZT value of the PANI/Bi2Te3 composite flexible thermoelectric material obtained by in-situ polymerization reached 0.0043 at room temperature.

3.3 Electrospinning

Electrospinning is a method for preparing nanofibers by jet spinning of a melt or polymer solution under a strong electrostatic field. Fig. 8A is a schematic diagram of electrospinning[82][83]. The method has the advantages of simple operation, low cost, wide application range and the like, and the flexible thermoelectric nanofiber mat can be prepared on the premise of not affecting the thermoelectric performance. Li et al. prepared a flexible thermoelectric nanofiber mat of N-type conjugated polymer (N2200) by putting N2200, dimethylamine (N-DMBI) and polyethylene oxide into a chloroform/chlorobenzene mixed solvent at a mass ratio of 100:5:25 by electrospinning under the action of a strong electrostatic field, as shown in Fig. 8B and C[84]. The flexible thermoelectric nanofiber mat can be easily bent, twisted and stretched without breaking, and shows good flexibility. At the same time, in the process of electrospinning, the polymer chain in the nanofiber can be oriented and straightened under the action of electrostatic force, which may contribute to carrier transport and improve conductivity. And there are a lot of pores between the nanofibers, which may inhibit the thermal transport and reduce the thermal conductivity. Finally, the maximum conductivity and Seekbeck coefficient of the nanofiber mat at room temperature are 7.06×10-4S·cm-1 and -346μV·K-1, respectively, and the maximum power factor is 0.0085μW·m-1·K-2 under the action of dopant N-DMBI. In addition, by cold pressing the nanofibers obtained by electrospinning, the fibers can be arranged more closely and their conductivity can be improved. For example, Akram et al. Carried out cold pressing treatment at 283 K on Bi2Te3/ polyvinylpyrrolidone (PVP) nanofiber composites prepared by electrospinning, which significantly increased the ZT value of the materials from 10-12 to 10-2 at room temperature[85].
图8 (a)静电纺丝法原理图[83];(b,c)静电纺丝法制备的柔性热电纳米纤维材料[84]

Fig. 8 (a)Schematic diagram of electrospinning[83];(b,c)FTE nanofibers prepared by electrospinning[84]

3.4 High temperature melting method

High temperature melting method is a method of preparing bulk materials by vacuum packaging raw materials in a quartz tube according to a certain proportion for high temperature melting, combined with sintering technology. The preparation process of this method is simple, and most of the Ag2S-based flexible thermoelectric materials mentioned above are prepared by this method. For example, Shi et al. Packaged Ag powder and S powder in a ratio of 2 ∶ 1 in a quartz tube in vacuum, held the sample at 1000 ° C for 12 H and then slowly cooled to 100 ° C, and then annealed the sample at 450 ° C for 5 days to obtain a α-Ag2S ingot, which was crushed and ground before SPS sintering to obtain a more compact α-Ag2S bulk material[57]. Similarly, Te or Se solid solution Ag2S based flexible thermoelectric materials can also be prepared by high temperature melting method combined with SPS[30]. The solid solution of elements further optimizes both the mechanical and thermoelectric properties of Ag2S-based flexible thermoelectric materials.
In addition, screen printing, drop coating, chemical bath deposition and other methods can also be used to prepare flexible thermoelectric materials[86][87][88].

4 Applications of Flexible Thermoelectric Materials

At present, although the thermoelectric properties and practical applications of flexible thermoelectric materials can not be completely compared with those of traditional inorganic thermoelectric materials, due to their good flexibility,The flexible thermoelectric device prepared by the method has the advantages of small size, light weight and wearability, can achieve the purpose of continuous power generation or refrigeration by using the thermoelectric effect, and has a good application prospect in the field of intelligent wearable and medical implantable equipment.
Wearable devices made of flexible thermoelectric materials can not only effectively fit and contact with the surface of the human body, but also achieve maximum thermoelectric power generation and refrigeration.It can also reduce the damage of deformation or vibration impact to internal parts, and at the same time, it can also reduce the contact resistance between different interfaces to reduce the internal resistance of the device, thereby improving the output power of the device. There are mainly two kinds of flexible thermoelectric wearable devices: thin film flexible thermoelectric devices and fiber flexible thermoelectric devices, most of which are made of conductive polymers or nanofibers together with traditional inorganic thermoelectric materials with better performance. At present, wearable thermoelectric devices in power generation mainly include smart glasses, small pedometers, small heart rate monitors and smart watches. In terms of refrigeration, the main equipment is wearable thermoelectric cooler to solve the problems of equipment overheating and human body cooling. For example, Zhang et al. Used hot stretching technology to make traditional inorganic thermoelectric materials (P-type Bi0.5Sb1.5Te3 and N-type Bi2Se3) into flexible thermoelectric fiber materials, and made them into thermoelectric wearable cooling fabrics to achieve the effect of cooling the human body[89].
In addition, flexible thermoelectric materials have certain application prospects in medical implantable devices and industrial waste heat recovery. Implantable medical sensors, such as cochlear implants and pacemakers, can solve the problem of limited working life of implanted devices by collecting heat from human activities to generate electricity for a long time.

5 Conclusion and prospect

Based on the research status of flexible thermoelectric materials in recent years, this paper reviews the research progress of polymer-based flexible thermoelectric materials, carbon-based flexible thermoelectric materials and inorganic semiconductor flexible thermoelectric materials.Emphasis is placed on the characteristics and performance optimization methods of these flexible thermoelectric materials. At present, the overall performance of flexible thermoelectric materials has been improved by some performance optimization methods, showing good application prospects in smart wearable and medical implantable devices. However, there are still some problems in the performance optimization of flexible thermoelectric materials and the practical application of flexible thermoelectric devices, which need to be further studied and explored.
(1) Develop N-type polymer-based flexible thermoelectric materials with better performance, improve the thermoelectric performance of carbon-based flexible thermoelectric materials, and optimize the thermal performance test methods. Due to the low dopant concentration, poor air stability and operational stability of N-type polymer-based thermoelectric materials, their semiconductor properties are easily changed, resulting in the lack of N-type polymer-based flexible thermoelectric materials. To solve this problem, we can reasonably design the molecular structure of the polymer and dopant or select the appropriate dopant to obtain N-type polymer-based flexible thermoelectric materials with better performance. For carbon-based flexible thermoelectric materials, the main problem is that the Seekbeck coefficient of the material itself is too low, which leads to the low Seekbeck coefficient of the composite material. We can choose appropriate dopants to dope or compound with inorganic semiconductors with better thermoelectric properties to improve its thermoelectric properties and make up for the lack of low Seekbeck coefficient. In addition, these two kinds of materials are mainly thin films, which are difficult to measure the thermal properties of thin films and have limitations in measurement methods. It is necessary to optimize the test methods of thermal properties to improve the test efficiency.
(2) Synergistically optimize the flexibility and thermoelectric properties of inorganic semiconductor flexible thermoelectric materials. For Ag2S-based semiconductor flexible thermoelectric materials, previous studies have shown that the solid solution of Se or Te can indeed improve the thermoelectric performance of Ag2S, but when the solid solution of Se or Te is more, the thermoelectric performance will be improved while the flexibility will be reduced.Next, some methods that do not destroy the flexibility of the Ag2S but also improve its thermoelectric performance can be selected, such as doping the Ag2S with non-isoelectronic elements to optimize its carrier concentration or properly introducing Ag vacancies while dissolving Se and Te, which can improve the thermoelectric performance of the Ag2S without greatly destroying its flexibility. For InSe-based semiconductor flexible thermoelectric materials, InSe single crystal is mainly studied at present, but the preparation of single crystal materials is relatively difficult, and the performance optimization method is relatively single.In the next step, we can try to prepare flexible InSe polycrystalline materials, and improve their flexibility and thermoelectric properties by adjusting the microstructure of polycrystalline materials and doping high-valence cations at In site to optimize their carrier concentration.
(3) Continuously develop the application field of flexible thermoelectric materials. In the application of flexible thermoelectric materials, the current research focuses on wearable devices, while the practical application in industrial and medical fields is relatively small. In the future, we should continue to develop the application fields of flexible thermoelectric materials, such as various micro-sensors that can be used in the medical field to help human damaged organs maintain normal state.
To sum up, although flexible thermoelectric materials have received extensive attention and made some progress in the field of thermoelectric research and application, the research on them is still in its infancy and faces many problems. We believe that with the continuous improvement of preparation technology and performance optimization methods, flexible thermoelectric materials will be applied to more fields and play a greater role.
[1]
Bell L E. Science, 2008, 321(5895): 1457.

[2]
Pei Y Z, Shi X Y, LaLonde A, Wang H, Chen L D, Snyder G J. Nature, 2011, 473(7345): 66.

[3]
Thirugnanasambandam M, Iniyan S, Goic R. Renew. Sustain. Energy Rev., 2010, 14(1): 312.

[4]
Swarnkar N. JETIR, 2019, 6(5): 131.

[5]
Vining C B. Nat. Mater., 2009, 8(2): 83.

[6]
Shi X, Chen L, Uher C. Int. Mater. Rev., 2016, 61(6): 379.

[7]
He Y, Day T, Zhang T S, Liu H L, Shi X, Chen L D, Snyder G J. Adv. Mater., 2014, 26(23): 3974.

[8]
Zheng X F, Liu C X, Yan Y Y, Wang Q. Renew. Sustain. Energy Rev., 2014, 32: 486.

[9]
Gaultois M W, Sparks T D, Borg C K H, Seshadri R, Bonificio W D, Clarke D R. Chem. Mater., 2013, 25(15): 2911.

[10]
Du Y, Xu J Y, Paul B, Eklund P. Appl. Mater. Today, 2018, 12: 366.

[11]
Zhang X, Zhao L D. J. Materiomics, 2015, 1(2): 92.

[12]
Zeier W G, Zevalkink A, Gibbs Z M, Hautier G, Kanatzidis M G, Snyder G J. Angew. Chem. Int. Ed., 2016, 55(24): 6826.

[13]
Yang J, Xi L L, Qiu W J, Wu L H, Shi X, Chen L D, Yang J H, Zhang W Q, Uher C, Singh D J. Npj Comput. Mater., 2016, 2: 15015.

[14]
Hasan M N, Wahid H, Nayan N, Mohamed Ali M S. Int. J. Energy Res., 2020, 44(8): 6170.

[15]
Zhang G Q, Kirk B, Jauregui L A, Yang H R, Xu X F, Chen Y P, Wu Y. Nano Lett., 2012, 12(1): 56.

[16]
Gelbstein Y, Dashevsky Z, Dariel M P. Phys. B Condens. Matter, 2005, 363(1/4): 196.

[17]
Bathula S, Jayasimhadri M, Dhar A. Mater. Des., 2015, 87: 414.

[18]
Wang Y, Yang L, Shi X L, Shi X, Chen L D, Dargusch M S, Zou J, Chen Z G. Adv. Mater., 2019, 31(29): 1807916.

[19]
Zhang L, Shi X L, Yang Y L, Chen Z G. Mater. Today, 2021, 46: 62.

[20]
Wu P Q, He Z M, Yang M, Xu J H, Li N, Wang Z M, Li J, Ma T, Lu X, Zhang H, Zhang T. Int. J. Thermophys., 2021, 42(8): 111.

[21]
Mengistie D A, Chen C H, Boopathi K M, Pranoto F W, Li L J, Chu C W. ACS Appl. Mater. Interfaces, 2015, 7(1): 94.

[22]
Kim G H, Shao L, Zhang K, Pipe K P. Nat. Mater., 2013, 12(8): 719.

[23]
Huang D Z, Yao H Y, Cui Y T, Zou Y, Zhang F J, Wang C, Shen H G, Jin W L, Zhu J, Diao Y, Xu W, Di C A, Zhu D B. J. Am. Chem. Soc., 2017, 139(37): 13013.

[24]
Zhao W Y, Fan S F, Xiao N, Liu D Y, Tay Y Y, Yu C, Sim D, Hng H H, Zhang Q C, Boey F, Ma J, Zhao X B, Zhang H, Yan Q Y. Energy Environ. Sci., 2012, 5(1): 5364.

[25]
Wang H, Hsu J H, Yi S I, Lae Kim S, Choi K, Yang G, Yu C. Adv. Mater., 2015, 27(43): 6855.

[26]
MacLeod B A, Stanton N J, Gould I E, Wesenberg D, Ihly R, Owczarczyk Z R, Hurst K E, Fewox C S, Folmar C N, Holman Hughes K, Zink B L, Blackburn J L, Ferguson A J. Energy Environ. Sci., 2017, 10(10): 2168.

[27]
Varghese T, Hollar C, Richardson J, Kempf N, Han C, Gamarachchi P, Estrada D, Mehta R J, Zhang Y L. Sci. Rep., 2016, 6: 33135.

[28]
Jin Q, Jiang S, Zhao Y, Wang D, Qiu J H, Tang D M, Tan J, Sun D M, Hou P X, Chen X Q, Tai K P, Gao N, Liu C, Cheng H M, Jiang X. Nat. Mater., 2019, 18(1): 62.

[29]
Jiang C, Ding Y F, Cai K F, Tong L, Lu Y, Zhao W Y, Wei P. ACS Appl. Mater. Interfaces, 2020, 12(8): 9646.

[30]
Liang J S, Wang T, Qiu P F, Yang S Q, Ming C, Chen H Y, Song Q F, Zhao K P, Wei T R, Ren D D, Sun Y Y, Shi X, He J, Chen L D. Energy Environ. Sci., 2019, 12(10): 2983.

[31]
Liang J S, Qiu P F, Zhu Y, Huang H, Gao Z Q, Zhang Z, Shi X, Chen L D. Research, 2020, 2020: 6591981.

[32]
He S Y, Li Y B, Liu L, Jiang Y, Feng J J, Zhu W, Zhang J Y, Dong Z R, Deng Y, Luo J, Zhang W Q, Chen G. Sci. Adv., 2020, 6(15): eaaz8423.

[33]
Yao Q, Wang Q, Wang L M, Wang Y, Sun J, Zeng H R, Jin Z Y, Huang X L, Chen L D. J. Mater. Chem. A, 2014, 2(8): 2634.

[34]
Park J, Lee Y R, Kim M, Kim Y, Tripathi A, Kwon Y W, Kwak J, Woo H Y. ACS Appl. Mater. Interfaces, 2020, 12(1): 1110.

[35]
Zhang Q, Sun Y M, Xu W, Zhu D B. Adv. Mater., 2014, 26(40): 6829.

[36]
Ju H, Kim J. ACS Nano, 2016, 10(6): 5730.

[37]
See K C, Feser J P, Chen C E, Majumdar A, Urban J J, Segalman R A. Nano Lett., 2010, 10(11): 4664.

[38]
Zhang Y H, Heo Y J, Park M, Park S J. Polymers, 2019, 11(1): 167.

[39]
Yun J S, Choi S, Im S H. Carbon Energy, 2021, 3(5): 667.

[40]
Dey A, Bajpai O P, Sikder A K, Chattopadhyay S, Shafeeuulla Khan M A. Renew. Sustain. Energy Rev., 2016, 53: 653.

[41]
Wang L M, Yao Q, Bi H, Huang F Q, Wang Q, Chen L D. J. Mater. Chem. A, 2015, 3(13): 7086.

[42]
Xiang J L, Drzal L T. Polymer, 2012, 53(19): 4202.

[43]
Liang L R, Gao C Y, Chen G M, Guo C Y. J. Mater. Chem. C, 2016, 4(3): 526.

[44]
Du Y, Shi Y L, Meng Q F, Shen S Z. Synth. Met., 2020, 261: 116318.

[45]
Xu Q, Qu S Y, Ming C, Qiu P F, Yao Q, Zhu C X, Wei T R, He J, Shi X, Chen L D. Energy Environ. Sci., 2020, 13(2): 511.

[46]
Qu S Y, Ming C, Qiu P F, Xu K Q, Xu Q, Yao Q, Lu P, Zeng H R, Shi X, Chen L D. Energy Environ. Sci., 2021, 14(12): 6586.

[47]
Sevinçli H, Cuniberti G. Phys. Rev. B, 2010, 81(11): 113401.

[48]
Ni X X, Liang G, Wang J S, Li B W. Appl. Phys. Lett., 2009, 95(19): 192114.

[49]
Chang P H, Bahramy M S, Nagaosa N, Nikoli©#263; B K. Nano Lett., 2014, 14(7): 3779.

[50]
Kong D Y, Zhu W, Guo Z P, Deng Y. Energy, 2019, 175: 292.

[51]
Hu H P, Xia K Y, Zhu T J, Zhao X B. Chinese Journal of Rare Metals, 2020, 45(5): 513.

(胡惠平, 夏凯阳, 朱铁军, 赵新兵. 稀有金属, 2020, 45(5): 513.).

[52]
You J C, Zhan S B, Wen J, Ma Y W, Zhu Z S. Optik, 2020, 217: 164900.

[53]
Hong G S, Robinson J T, Zhang Y J, Diao S, Antaris A L, Wang Q B, Dai H J. Angew. Chem. Int. Ed., 2012, 51(39): 9818.

[54]
Alharthi S S, Alzahrani A, Razvi M A N, Badawi A, Althobaiti M G. J. Inorg. Organomet. Polym. Mater., 2020, 30(10): 3878.

[55]
Hwang I, Seol M, Kim H, Yong K. Appl. Phys. Lett., 2013, 103(2): 023902.

[56]
Hebb M H. J. Chem. Phys., 1952, 20(1): 185.

[57]
Shi X, Chen H Y, Hao F, Liu R H, Wang T, Qiu P F, Burkhardt U, Grin Y, Chen L D. Nat. Mater., 2018, 17(5): 421.

[58]
Jin M, Bai X D, Zhang R L, Zhou L N, Li R B. J. Inorg. Mater., 2022, 37(1): 101.

[59]
Wang T, Chen H Y, Qiu P F, Shi X, Chen L D. Acta Phys. Sin., 2019, 68(9): 090201.

[60]
Ferhat M, Nagao J. J. Appl. Phys., 2000, 88(2): 813.

[61]
Pei Y Z, Heinz N A, Snyder G J. J. Mater. Chem., 2011, 21(45): 18256.

[62]
Bindi L, Pingitore N E. Mineral. Mag., 2013, 77(1): 21.

[63]
Bindi L, Stanley C J, Spry P G. Mineral. Petrol., 2015, 109(4): 413.

[64]
Wei T R, Jin M, Wang Y C, Chen H Y, Gao Z Q, Zhao K P, Qiu P F, Shan Z W, Jiang J, Li R B, Chen L D, He J, Shi X. Science, 2020, 369(6503): 542.

[65]
Han G, Chen Z G, Drennan J, Zou J. Small, 2014, 10(14): 2747.

[66]
Zheng Q, Liang C Y, Jiang J Y, Li S F. Phys. Status Solidi RRL Rapid Res. Lett., 2022, 16(3): 2100418.

[67]
Dai Y J, Zhao S X, Han H, Yan Y F, Liu W H, Zhu H, Li L, Tang X, Li Y, Li H, Zhang C J. Front. Mater., 2022, 8: 816821.

[68]
Grimaldi I, Gerace T, Pipita M M, Perrotta I D, Ciuchi F, Berger H, Papagno M, Castriota M, PacilÉ D. Solid State Commun., 2020, 311: 113855.

[69]
Mosca D H, Mattoso N, Lepienski C M, Veiga W, Mazzaro I, Etgens V H, Eddrief M. J. Appl. Phys., 2002, 91(1): 140.

[70]
Shi H N, Wang D Y, Xiao Y, Zhao L D. Aggregate, 2021, 2(4): e92.

[71]
Hong M, Chen Z G, Zou J. Chin. Phys. B, 2018, 27(4): 048403.

[72]
Zhang B, Wu H, Peng K L, Shen X C, Gong X N, Zheng S K, Lu X, Wang G Y, Zhou X Y. Chin. Phys. B, 2021, 30(7): 078101.

[73]
Yang Q Y, Yang S Q, Qiu P F, Peng L M, Wei T R, Zhang Z, Shi X, Chen L D. Science, 2022, 377(6608): 854.

[74]
Shi D T, Wang R P, Wang G X, Li C, Shen X, Nie Q H. Vacuum, 2017, 145: 347.

[75]
Shen S F, Zhu W, Deng Y, Zhao H Z, Peng Y C, Wang C J. Appl. Surf. Sci., 2017, 414: 197.

[76]
Singkaselit K, Sakulkalavek A, Sakdanuphab R. Adv. Nat. Sci: Nanosci. Nanotechnol., 2017, 8(3): 035002.

[77]
Goncalves L M, Alpuim P, Min G, Rowe D M, Couto C, Correia J H. Vacuum, 2008, 82(12): 1499.

[78]
Goncalves L M, Alpuim P, Rolo A G, Correia J H. Thin Solid Films, 2011, 519(13): 4152.

[79]
Goncalves L M, Couto C, Alpuim P, Rolo A G, Völklein F, Correia J H. Thin Solid Films, 2010, 518(10): 2816.

[80]
Wang Y Y, Cai K F, Shen S, Yao X. Synth. Met., 2015, 209: 480.

[81]
Chatterjee K, Mitra M, Kargupta K, Ganguly S, Banerjee D. Nanotechnology, 2013, 24(21): 215703.

[82]
Bhardwaj N, Kundu S C. Biotechnol. Adv., 2010, 28(3): 325.

[83]
Masoumi S, O’Shaughnessy S, Pakdel A. Nano Energy, 2022, 92: 106774.

[84]
Li J Y, Dong C S, Hu J L, Liu J, Liu Y C. ACS Appl. Electron. Mater., 2021, 3(8): 3641.

[85]
Akram R, Khan J S, Qamar Z, Rafique S, Hussain M, Kayani F B. J. Mater. Sci., 2022, 57(5): 3309.

[86]
Kim S J, We J H, Cho B J. Energy Environ. Sci., 2014, 7(6): 1959.

[87]
Yabuki H, Yonezawa S, Eguchi R, Takashiri M. Sci. Rep., 2020, 10: 17031.

[88]
Patil N S, Sargar A M, Mane S R, Bhosale P N. Mater. Chem. Phys., 2009, 115(1): 47.

[89]
Zhang T, Li K W, Zhang J, Chen M, Wang Z, Ma S Y, Zhang N, Wei L. Nano Energy, 2017, 41: 35.

Outlines

/