Project supported by the National Natural Science Foundation of China (Grant Nos. 51472052 and U1601213).
Project supported by the National Natural Science Foundation of China (Grant Nos. 51472052 and U1601213).
† Corresponding author. E-mail:
Project supported by the National Natural Science Foundation of China (Grant Nos. 51472052 and U1601213).
Pb-based group-IV chalcogenides including PbTe and PbSe have been extensively studied as high performance thermoelectric materials during the past few decades. However, the toxicity of Pb inhibits their applications in vast fields due to the serious harm to the environment. Recently the Pb-free group-IV chalcogenides have become an extensive research subject as promising thermoelectric materials because of their unique thermal and electronic transport properties as well as the enviromentally friendly advantage. This paper briefly summarizes the recent research advances in Sn-, Ge-, and Sichalcogenides thermoelectrics, showing the unexceptionally high thermoelectric performance in SnSe single crystal, and the significant improvement in thermoelectric performance for those polycrystalline materials by successfully modulating the electronic and thermal transport through using some well-developed strategies including band engineering, nanostructuring and defect engineering. In addition, some important issues for future device applications, including N-type doping and mechanical and chemical stabilities of the new thermoelectrics, are also discussed.
Thermoelectric (TE) materials, a kind of functional materials able to realize direct energy conversion between heat and electricity, have been extensively studied due to the great desire for clean energy.[1] The dimensionless figure of merit zT = (S2σ T/κ) is used to characterize the performance of TE material, where S is the Seebeck coefficient, σ is the electrical conductivity, T is the absolute temperature and κ is the thermal conductivity (κ = κe + κl, κe is the electronic thermal conductivity, κl is the lattice thermal conductivity). Therefore, the high performance TE material should have a low thermal conductivity (κ) and large S and σ.[2–4] Unfortunately the three parameters, i.e., S, σ, and κ, are strongly coupled, and thus preventing the thermoelectric performance from improving by singularly tuning one of them. In the last few decades, the zT values of thermoelectrics have been constantly improved through the successful utilizations of new concepts and mechanisms including defect engineering,[5] band engineering,[6,7] nanostructuring,[8,9] nanoinclusions,[10,11] and modulation doping.[12,13]
The strategies mentioned above have been applied to many TE material systems, such as Half-Heuslers,[14] Skutterudites,[11] Bi2Te3[10,15] and PbTe,[16] making great achievements. Among those material systems, PbTe has been well investigated for a long history, and most important concepts and strategies for enhancing the TE performances were developed or started in PbTe. The high TE performance for PbTe is related to its intrinsic electronic band structure possessing two valence bands (L and Σ) with a small energy offset that can be further reduced by band engineering to obtain a high band degeneracy, resulting in the improved zT.[17] Besides band engineering, defect engineering has also been utilized to tune the phonon transport in PbTe and obtain a high zT value of 2.2 by their synergistic effects.[18] Despite the fact that great achievements have been made in PbTe-based thermoelectrics, that Pb seriously harms the environment is still a fatal weakness that hinders its large-scale applications.
In 2014, SnSe single crystal was found to have an unprecedentedly high zT value of 2.6 along the crystallographic direction b axis, which uncovered and testified the great potential applications in non-Pb based group-IV chalcogenides thermoelectrics,[19] including tin-, germanium-, and silicon-chalcogenides, because they have the same crystal structure as and similar band structures to PbTe and SnSe.[6,20] As seen from Fig.
In this paper, we outline the research advances in the group-IV chalcogenides out of PbTe and PbSe in recent years, trying to give a rough guide to this rising topic of thermoelectricity. We first survey the basic crystal and electronic band structural parameters of the materials, showing their potential applications in thermoelectrics. Then we present some typical researches on Sn-, Ge-, and Si-based chalcogenide thermoelectrics, respectively, in which high TE performance has been achieved by utilizing some recently developed strategies or approaches in PbTe thermoelectrics to tune the thermal and electronic transport in those materials. Finally we briefly discuss the possible challenges in the N-type doping, mechanical and chemical stabilities of these TE materials for the future device applications.
The crystal structures and electronic band structures of IV-chalcogenides thermoelectrics are briefly summarized in Table
Considering the similarity in the crystalline and electronic band structures of SnTe and PbTe, it is reasonable to anticipate that the high thermoelectric performance can be achieved in SnTe by applying those successful strategies used in PbTe. Meanwhile, the layered structures of those orthorhombic Sn- and Ge-chalcogenides also imply their anisotropic thermoelectric transports, which might lead to their high thermoelectric performances. In the following sections, the detailed discussion is categorized according to the specific material systems that are tin-, germanium- and silicon-chalcogenides, respectively.
As an alternative analogue of PbTe, SnTe has received more attention from the thermoelectric field due to their high similarity in crystal and electronic band structures. However, an unexceptional breakthrough was first achieved in SnSe. As a narrow band gap semiconductor, SnSe was noticed because of its photo electronic potentials[31,32] instead of thermoelectric properties. The SnSe is an intrinsic P-type material with a carrier concentration in a range of 1017 cm−3–1018 cm−3 and an electrical resistivity between 101 Ω·cm–105 Ω · cm at room temperature.[33] These features make SnSe widely studied and used in solar cells,[34] optoelectronics,[35] and other electronic devices.[36] Recently, the unexceptionally high zT of 2.66 ± 0.3 at 923 K along the b-crystallographic direction was found in un-doped SnSe single crystal, which soon made SnSe a focus of the thermoelectric community.[19] Soon after the first report on un-doped SnSe single crystals,[19] some reports were presented by Zhao et al.[37] and Zhou et al.,[38] respectively. Na-doped SnSe single crystal with greatly increased hole concentration realized a record-high average zT of ∼1.34 from 300 K to 773 K[37] and an average zT of ∼ 1.2 from 300 K to 800 K,[38] respectively. Inspired by the surprising discovery in SnSe single crystal, some consequent researches on its polycrystalline form were carried out, but the much lower zT of ∼0.5 was obtained[39] as seen in Fig.
The SnS has the same crystal structure as that of SnSe at room temperature, and is considered as a promising new thermoelectric material that has an abundant resource and a better environmental compatibility.[47] The pristine SnS shows a high Seebeck coefficient and low thermal conductivity, but the low carrier concentration leads to low electrical conductivity that results in a low zT value. The pristine SnS shows a carrier concentration of about 1015 cm− 3 ∼ 1017 cm− 3 and a maximum zT value of 0.16 at 873 K.[48] The P-type SnS was doped with 0.5% Ag bulk material synthesized by a mechanical alloying method, and the derived powders were pressed into pellets by using spark plasma sintering (SPS) and thus improving the zT value from 0.16 to 0.6. The improvement arises from the significantly increased electrical conductivity by Ag doping on Sn sites which greatly increased the carrier concentration (nave ≈ 1018 cm− 3).[48] The electric conductivity and phonon scattering were also increased simultaneously in SnS0.2Se0.8 solid solution, and thus obtaining a high zT of 0.82 at 823 K, which is much higher than that in the pristine SnS compound.[46] Although the significant increase of zT for SnS-based material has not been achieved to date, it is worth further investigating as a potential thermoelectric material.
As shown in Table
The GeTe-based alloys have been known for a long time for their promising thermoelectric properties due to their superior structural and thermal properties.[52] The GeTe as a P-type semiconductor shows a narrow band gap with a high electrical conductivity and carrier concentration (1021 cm−3), which is due to a large number of Ge vacancies.[53] In spite of the high electrical conductivity, the low Seebeck coefficient and high thermal conductivity were also exhibited in the pristine GeTe. For these reasons, the pristine GeTe exhibited a maximum zT value of 0.88 at 712 K.[54] The doping, band engineering, and nanostructuring have been applied to GeTe for high TE performance. Recently, the GeTe–AgSbTe2-based alloys,[54] Ge–Sb–Te ternary alloys,[25,55] doped GeTe,[56,57] and Ge–Pb–Te alloys[22] each have shown a significant enhancement in zT value as seen in Fig.
It was found that in spite of the decrease of the electrical resistivity, the substitution of 2% In for Ge in GeTe enlarged the Seebeck coefficient, and remarkably reduced the total thermal conductivity, which finally achieved a high zT value of 1.3 at temperature near 628 K in In0.02Ge0.98Te.[56] The group-V elements Sb- and Bi-doped GeTe also achieved effective reduction in the carrier concentration and collectively increased the Seebeck coefficient from ∼ 30 μV·K− 1 to ∼100 μ V·K− 1, and ∼80 μ V·K− 1.[25,57] Importantly, Sb in GeTe not only reduced the carrier concentration, but also enhanced the valence band degeneracy that highly increased the Seebeck coefficient. Combining with the meso-structured significant decrease in the thermal conductivity, the high thermoelectric figure of merit of 1.85 at 725 K in Ge0.9Sb0.1Te[25] was realized. Similarly, Bi in the GeTe matrix also got a zT value of 1.8 at 700 K.[57] The rhombohedral GeTe and rock-salt type AgSbTe2 could form complete solid solution (Ge–Te)x(AgSbTe2)100 - x (named TAGS).[58] The AgSbTe2 increased density of states of TAGS, and the lead TAGS showed a high Seebeck coefficient.[53] The high thermoelectric figure of merit of 1.5 at 750 K was achieved in TGAS-85.[59] The new pseudo-binary solid solution (Ge-Te)80(AgSbSe2)20[54] underwent phase-separation and the formation of hierarchical nano/mesostructures, which caused extensive scattering of heat carrying phonons of different wavelengths, leading to ultra low klat ≈ 0.4 W·m−1·K−1, a high peak zT of 1.9 at 660 K, and an ultrahigh average zT of 1.4 in a temperature range of 300K–700 K.[54] The Ge0.9Sb0.1Te0.9Se0.05S0.05[55] also exhibited a low κlat of 0.7 W·m−1·K−1 at 730 K due to a broad set of multiple types of mass fluctuations such as Ge/Sb,Te/Se, Te/S, and Se/S that enhanced the phonon scattering. As a result, the SPS-processed Ge0.9Sb0.1Te0.9Se0.05S0.05 sample showed a remarkably high zT of 2.1 at 630 K.[55]
The GeSe and GeS have similar crystal structures to SnSe. The orthorhombic layered GeSe and GeS have been predicted as good thermoelectric materials via theoretical calculation,[20] but the experimental investigation indicated that the polycrystalline GeSe with a low carrier concentration showed a low zT value of only 0.2 at 700 K.[60] The elemental doping could be an effective route to optimizing the carrier concentration, but it is still challenging work to ahieve realiable doping experimentally in the GeSe compound.
For these reasons above, the alloying strategy has been used in GeSe by adding the AgSbSe2,[61] which made a structural phase transition from the original orthorhombic structure to a high symmetry rhombohedral phase, and thus obtaining a higher Seebeck coefficient, a modified band structure, and a higher carrier concentration (1.2 × 1020 cm−3). As a result, a zT of 0.86 was achieved. The GeS had a larger band gap (about 1.25 eV) than those of SnSe and GeSe, leading to a low carrier concentration and large Seebeck coefficient. A large power factor of GeS could be obtained by alloying and reasonable doping strategy.[20]
Although SiS and SiSe are of an isostructure with SnSe, they have received less attention than other group-IV chalcogenides up to now. Thermoelectrics properties of SiS and SiSe monolayers have been prdicted via the first-principles calculation as we can see in Fig.
The above promising results indicate that they are all achieved in P-type materials. It is known that the intrisic metallic ion vacancies in group IV–chalcogenides result in their P-type conductivities and make it difficult to realize reliable N-type doping.[21,51,55] In order to achieve the P–N conduction transition and enhance the electrical conductivity, the effective heavy doping is necessary. However, it is diffcult to realize heavy doping, which is possibly due to the lower solubility of the dopants in group IV–chalcogenide lattices. Some studies demonstrated that extra group- IV metal and halogen elements were promising candidates. Figure
The mechanical and chemical stability is another important issue for the practical device applications of the group- IV chalcogenides thermoelectrics, but few researches have been performed to date. As shown in Table
The Berkovich nano-indentation and Vickers micro-indentation methods can be used to examine the mechanical properties of materials.[53] Figure
In addition, chemical stability of the material is also an improtant factor. Selenide is easily oxidized by long-term exposure to the air,[85] which would harm its thermoelectric performance. It was reported that the surface of polycrystalline SnSe sample is easily oxidized to form tin oxide, leading to much higher thermal conductivity.[86] The low chemical stablilities of Ge- and Si-tellurides would also be detrimental to their applications. Tellurium easily evaporates during heating and tends to form oxides on the surface, which is harmful to their thermoelectric properties.[87] Hence the fabrication and service stablity of the thermoelectric device based on the group-IV chalcogenides, which have not been well explored, need to be studied in depth.
Based on the above survey of the recent reports, it is reasonable to conclude that non-Pb based group-IV chalcogenides have emerged as a new promsing thermoelectric material with high performance comparable to (even beyond) that of PbTe besides the advantage of environmental friendship. Most strategies that have been applied to PbTe-based thermoelectrics have proven their availability and effectiveness, which inculde defect engineering, band engineering, nanostructuring, nanoinclusions and modulation doping. During the last several years, notable achievements have been made both experimentally and theoretically, thereby greatly deepening and broadening our knowledge of the thermoelectrics. It could be expected that more exciting and encouraging progress of material research in science and technology would be made in the near future.
[1] | |
[2] | |
[3] | |
[4] | |
[5] | |
[6] | |
[7] | |
[8] | |
[9] | |
[10] | |
[11] | |
[12] | |
[13] | |
[14] | |
[15] | |
[16] | |
[17] | |
[18] | |
[19] | |
[20] | |
[21] | |
[22] | |
[23] | |
[24] | |
[25] | |
[26] | |
[27] | |
[28] | |
[29] | |
[30] | |
[31] | |
[32] | |
[33] | |
[34] | |
[35] | |
[36] | |
[37] | |
[38] | |
[39] | |
[40] | |
[41] | |
[42] | |
[43] | |
[44] | |
[45] | |
[46] | |
[47] | |
[48] | |
[49] | |
[50] | |
[51] | |
[52] | |
[53] | |
[54] | |
[55] | |
[56] | |
[57] | |
[58] | |
[59] | |
[60] | |
[61] | |
[62] | |
[63] | |
[64] | |
[65] | |
[66] | |
[67] | |
[68] | |
[69] | |
[70] | |
[71] | |
[72] | |
[73] | |
[74] | |
[75] | |
[76] | |
[77] | |
[78] | |
[79] | |
[80] | |
[81] | |
[82] | |
[83] | |
[84] | |
[85] | |
[86] | |
[87] |