Synergistic Strategy to Enhance the Thermoelectric Properties of

Subscriber access provided by University of Newcastle, Australia

Article

A synergistic strategy to enhance the thermoelectric properties of CoSbS1-xSex compounds via solid solution Wei Yao, Dingfeng Yang, Yanci Yan, Kunling Peng, Heng Zhan, Anping LIu, Xu Lu, Guoyu Wang, and Xiaoyuan Zhou ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b12796 • Publication Date (Web): 10 Mar 2017 Downloaded from http://pubs.acs.org on March 12, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 22

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

ACS Applied Materials & Interfaces

A synergistic strategy to enhance the thermoelectric properties of CoSbS1-xSex compounds via solid solution Wei Yao,‡ Dingfeng Yang,§ Yanci Yan,‡, § Kunling Peng,‡, § Heng Zhan,‡ Anping Liu,‡ Xu Lu,‡ Guoyu Wang*, § Xiaoyuan Zhou*,‡ ‡

College of Physics, Chongqing University, Chongqing 401331, People’s Republic of China.

§

Chongqing Institute of Green and Intelligent Technology, Chinese Academy of Sciences Chongqing, 400714, People’s Republic of China.

Corresponding author: Tel: +86-23-6593-5603; Fax: +86-23-6567-8362 Email: [email protected]; [email protected]

Abstract High thermal conductivity of CoSbS-based limited its own prospect application in thermoelectric energy conversion. Solid solution is an effective approach to optimize the performance of thermoelectric materials with high lattice thermal conductivity because of the enhanced phonons scattering from disorder atoms. In this paper, we have

synthesized and measured the thermoelectric properties of solid solution

CoSbS1-xSex (x=0, 0.05, 0.10, 0.15, 0.20, 0.30) series samples. The collaborative optimization (enhancing the power factors and reducing the thermal conductivities) to add zT values were realized via substitution of S atoms with the isoelectronic Se atoms in the matrix. Meanwhile, the lowest room temperature lattice thermal conductivity in CoSbS-based materials is obtained (4.72 W m-1 K-1) at present. Benefiting from the results of synergistic strategy, a zT of 0.35 was achieved at 923 K for sample CoSbS0.85Se0.15, a 59% improvement as compared with that of the pristine CoSbS. Band calculation demonstrated that CoSbS0.85Se0.15 present a similar band dispersion with CoSbS. The mechanism of point defect scattering for reducing the 1

ACS Paragon Plus Environment


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

lattice thermal conductivity at room temperature, was also analyzed by the Callaway model. The contributions to decrease the room temperature lattice thermal conductivity from the mass and the strain fluctuation in the crystal are comparable. These results can also be extended to other high efficiency thermoelectric material with stiff bond and smaller Gruneisen parameters.

Key words: CoSbS; solid solution; thermal conductivity; point defect; zT

1. Introduction Thermoelectric material, a type of solid-state heat pump, is able to convert thermal energy into electric energy.1,2 The performance of thermoelectric material is evaluated by a figure of merit zT, defined as ZT =

, where S is the Seebeck coefficient,

σ is the electrical conductivity, is the lattice thermal conductivity, is the electrical thermal conductivity and T is the absolute temperature.3 Generally, there are two strategies to optimize the zT of a thermoelectric material: increasing the power factor (S2σ) or reducing the thermal conductivity. With regards to the first strategy, many approaches such as introducing band convergence,4-6 electron energy barrier filtering, 7quantum confinement effects,8 and using the resonant level9 have been proved fruitfully. As for reducing the thermal conductivity, equally impressive gains have also been obtained, such as introducing the ‘rattler’ concept proposed by Slack; 10-11

for example, using filling atoms in the skutterudite12 to reduce the mean free path

of phonon.

13-14

Another important and effective approach to decrease the thermal

conductivity is point defect scattering by alloying with other elements,15 which has

2


Page 2 of 22

Page 3 of 22

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


been successfully demonstrated for thermoelectric materials in lead telluride,

16

Heusler compounds, 17and SiGe.18 Cobalt-based compounds, such as CoSb3,19-20 Ca3Co4O9,21-22 and NaxCoO2,23-24 have been investigated as promising thermoelectric materials for a long time due to their excellent power factor and stability. Recently, a new potential thermoelectric cobalt-based intermetallic compound, CoSbS (paracostibite), exhibiting a high power factor of 1.5 mWm-1K-2 and a thermal conductivity of 3.9 Wm-1K-1 at 773 K, has been investigated by Parker et al.25 Previous band calculation demonstrated that the excellent power factor was determined from its multi-valley characteristics on the top of valence band and the bottom of the conduction band. Besides, CoSbS possessed a suitable band gap (0.5 eV). Therefore, the optimization of the performance of these materials has been conducted, in spite of their high thermal conductivity.26,

27

Compared to other state of art TE materials, the lattice thermal conductivity of CoSbS compounds is relatively high (8 Wm-1K-1 at room temperature) due to the nature stiff bond in these compounds. In previous studies, aliovalent ion doping on Co or Sb has been studied. 26, 27 However, the thermal conductivities of these compounds are not significantly decreased, leaving much room for further decreasing the thermal conductivity. The object of our work is thus to reduce the thermal conductivity effectively while enhancing the power factor via substituting S of Se in CoSbS compounds, and thereby their improved thermoelectric performance would be expected. First, it will enhance point defect scattering for phonons, leading to the reduction of lattice thermal 3



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

conductivity due to mass field fluctuations and strain field fluctuations. Second, considering the similar electronegativity but different radius size(the energy level of and S and Se is different) between S and Se, the solid solution of CoSbS compounds would appear increased band convergence near the Fermi level, which will optimize the Seebeck coefficient and electrical conductivity synergistically, resulting in a high power factor, such as in Mg2Si1-xSnx,6, 26 PbTe1-xSex.4, 27-28 In this work, we experimentally demonstrate that the thermoelectric transport properties of CoSbS could be optimized synergistically through introducing the isoelectronic Se in CoSbS matrix. The decreased thermal conductivity and the optimized power factor result in an improved zT of 0.35 for CoSbS0.85Se0.15 at 923 K, which is about 59% higher than that of parent compound CoSbS.

2. Experimental Section 2.1 Synthesis Bulk samples of polycrystalline with nominal compositions of CoSbS1-xSex (x=0, 0.05, 0.10, 0.15, 0.20, 0.30) were synthesized by the conventional solid state chemical reaction method, starting with elements of Co (pieces, 99.999%), Sb (pieces, 99.9999%), S (pieces, 99.9999%) and Se (sphere, 99.999%). The starting materials were weighed in a stoichiometric ratio. Then, sealed in a quartz tube under the vacuum of 10-4 mbar. Subsequently, the materials were heated slowly up to 1273 K for 15 h and kept 1273K for 20h, and then naturally cooled down to room temperature. The obtained reaction product were reground into powders and sintering at 1053 K for 4


Page 4 of 22

Page 5 of 22

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


30 h to achieve pure phase CoSbS-based compounds. Afterwards, the as-prepared ingots were crushed and grounded into a fine powder. Densified bulk samples (~95%, 6.63g/cm3) were obtained by the spark plasma sintering at 953 K for approximately 5 min under a pressure of 65 MPa.

2.2 Structure and characterization Powder x-ray diffraction (PXRD) was performed on a PANalytical X-ray diffractometer. The Cu Kα monochromator was used and operated at 45 kV and 40 mA. The dense samples were cut into 1.2-1.5 mm thick, 10 mm diameter disks and 2.5×2.5×9 mm cuboids for thermal and electrical performance, respectively. The high temperature electrical transport behavior measurements, including electrical conductivity and the Seebeck coefficient, were tested on the LSR-3 system (LINSEIS, LSR-3, Germany) via a four-probe method from 300 K to 923 K in He gas atmosphere. The high temperature thermal transport performance, thermal diffusivity parameter (λ), was determined by the laser flash method under a flowing argon atmosphere in a Netzsch LFA 457 (NETZSCH, LFA457, Germany). The samples for the thermal diffusivity measurements were spray-coated with a thin layer of graphite to decrease errors in the emissivity measurements. The heat capacity (Cp) was obtained by a differential scanning calorimetry thermal analyzer Netzsch, 404 F3 (NETZSCH, 404 F3, Germany). The heat capacity for the solid solution samples were close to the theoretical Dulong-Petit value (0.35 J/g K).The real density (d) of the samples were measured on commercial densimeter using the Archimedes method. The thermal conductivity was calculated from the formula κ=λCpd. The estimated error of 5



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

Page 6 of 22

the Seebeck coefficient, electrical conductivity, thermal conductivity and ZT values are approximately 7%, 10%, 5% and 20%, respectively. The room temperature hall carrier concentration of all samples was measured using a home-made Hall apparatus under a magnetic field of ±1 T.

2.3. Density function theory (DFT) calculations The Perdew Burke Ernzerhof (PBE) generalized gradient approximation (GGA) and projector augmented wave (PAW) potentials of Blochl, implementing in

the

Vienna ab initio simulation code (VASP), was employed.29,30 the cutoff value 300eV of the plane wave is chosen and the Brillouin zone integration was adopted under a 5×5×5 Monkhorst Pack k meshes.31 The convergence criteria of 1×10-6 eV/A is used for the structural relaxation and self-consistent calculations. To better under the electronic band change through Se doping on S, Supercell 1× 2 × 1 was chosen. Limited by the computation ability, the chosen theoretical doping concentration (x=0, 0.0625, 0.1250, 0.1875, 0.2500) is slightly different from that of the experiment (x=0, 0.05, 0.10, 0.15, 0.20, 0.25). The band energies Ek were calculated on a dense k mesh along the following high-symmetry directions, i.e., from the Brillouin zone center with the coordinates G (0.0, 0.0, 0.0) to the F point (0.0, 0.5, 0.0), Q point (0.0, 0.5, 0.5), Z point (0.0, 0.0, 0.5), G (0.0, 0.0, 0.0), B point (0.5, 0.0, 0.0), K point (0.0, 0.0, 0.0) and G (0.0, 0.0, 0.0) in units of 2π/a, 2π/b, and 2π/c.

3. Results and Discussion 3.1 Crystal Structure and PXRD Phase Analysis Figure 1(a) shows the crystal structure of CoSbS, which belongs to the orthorhombic family and crystallize in space group Pbca. The lattice constants of

6


Page 7 of 22

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


CoSbS along a, b, and c axes are 5.842 Å, 5.951 Å, and 11.666 Å, respectively. In Paracostibite, each Co octahedron has one shared edge, and all corners are shared with two other octahedron, each Co atom is coordinated to 3× Sb and 3× S atoms, and each S atom is coordinated to 1× Sb and 3× Co atoms. Similarly, each Sb atom is coordinated to 1× S and 3× Co atoms. The room temperature PXRD patterns of CoSbS1-xSex (x=0, 0.05, 0.10, 0.15, 0.20, 0.30, 0.35) bulk obtained after SPS are shown in Figure 1(b). All these peaks can be indexed to the standard CoSbS card (PDF #86-2272, space group Pbca) without the presence of any second phase. However, when the amount of Se substitution reaches 35%, some minor second phase appear, indicating the solubility of Se on S site is 30%. Figure 1(c) displays the lattice parameters and volume of the unit cell with respect to the Se content of CoSbS1-xSex solid solutions (x=0, 0.05, 0.10, 0.15, 0.20, and 0.30). Compared to the pristine compound CoSbS, the a, b, c and V of solid solutions increase linearly in the investigated range, where a continuous Vegard-like behavior is observed, which indicates that the substitution of Se on S site symmetrically expands the framework size and the overall volume. Meanwhile, the linear behavior of the observed increase in lattice parameters and volume demonstrates the successful substitution and solubility of Se for S is 30%.

7



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

Figure 1. (a) Crystal structure of CoSbS. Blue, Red and Yellow balls are Co, Sb, and S atoms, respectively. (b) Powder XRD patterns of CoSbS1-xSex (x=0, 0.05, 0.10, 0.15, 0.20, 0.30, 0.35). (c) Lattice parameters of the unit cell of CoSbS1-xSex (x=0, 0.05, 0.10, 0.15, 0.20, 0.30) via Rietveld Refinement

3.2 Electronic transport properties The temperature dependence of electronic transport properties (electrical conductivity and Seebeck coefficient) for the solid solution CoSbS1-xSex (x=0, 0.05, 0.10, 0.15, 0.20, 0.30) are drawed in Figure 2(a) and (b). For all samples, it is obviously observed that the electrical conductivities increased with increasing the temperature, which indicates a typical semiconductor transport behavior. After 600 K, an abrupt increment is observed, which may be attributed to the large amount of minor carriers thermal excitation. In addition, the room temperature electrical conductivity increased with the Se content from 433 S/m for pristine CoSbS to 1120 S/m for the sample with x=0.15, while it decreased to 151 S/m for the sample x=0.30. The increase of the electrical conductivity at room temperature is mainly caused by the increased carrier concentration (shown in Figure 2(c)). Whereas the electrical conductivity is decreased with higher Se contents at room temperature. This trend can be ascribed to the reduced mobility yielded by the strengthened alloy scattering for 8


Page 8 of 22

Page 9 of 22

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


electrons.

Figure 2. Electrical transport properties for the solid solution CoSbS1-xSex (x=0, 0.05, 0.10, 0.15, 0.20, 0.30) as a function of temperature. (a) The electrical conductivity (b) Seebeck coefficient. (c) The carrier concentration and carrier mobility at room temperature. (d) Power factor. The insert shows the electrical properties of CoSbS reported in literatures.

As shown in Figure 2(b), negative values of Seebeck coefficient for CoSbS1-xSex (x=0, 0.05, 0.10, 0.15, 0.20 and 0.30) are observed over the whole temperature range 300 K to 923 K, which demonstrates that the electrons are major charge carriers in these samples. The Seebeck coefficient of the whole samples are decreased with the increasing substitution of Se content in accordance with the increase of the carrier concentration (shown in Figure 2(c)). Moreover, the absolute values of Seebeck coefficient of CoSbS1-xSex compounds slightly increase with the increasing 9



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

temperature, and then reach a maximum value over 500-600 K, and then drop off as the temperature further increases, which should be attributed to the thermal excitation of minor carriers at a higher temperature. We also compare the thermoelectric properties of pristine CoSbS with that reported in literatures.26,32 As shown in the insets in Figures 2 (a) and (b), the electrical conductivity of our sample is comparable to that reported by Chmielowski et al, and significantly larger than that reported by Liu et al. Like other complex intermetallic compounds, different synthesis conditions for CoSbS give rise to different electrical transport properties, probably due to the difference in S deficiency in CoSbS compounds. The carrier concentration and mobility of CoSbS1-xSex (x=0, 0.05, 0.10, 0.15, 0.20 and 0.30) are shown in Figure 2 (c). It is noted that the carrier concentration rises with Se substitution by a factor of nearly 14 times, specifically from 2.7e18 cm-3 for the pristine CoSbS to 3.9e19 cm-3 for the CoSbS0.7Se0.3 sample. Correspondingly, as demonstrated by the electrical conductivity shown above, the calculated mobility first sharply decreases from 9.8 cm2 V-1S-1 for the pristine CoSbS to 5.18 cm2 V-1 S-1 for the CoSbS0.85Se0.15 sample, then slowly decreases to 1.45 cm2 V-1S-1 for CoSbS0.7Se0.3 sample. Such reduction of the mobility in isoelectronic CoSbS1-xSex solid solutions is because of the enhanced carriers scattering due to the electron-phonon interaction and point defects, which shorten the electrical mean free path. Besides, it is worthwhile to mention that the mobility of our pristine CoSbS sample is larger than that reported by Liu et al (4.4 cm2 V-1S-1)

33

and by Chmielowski et al (5.3 cm2 V-1S-1) 32. The power

factor (PF) is shown in Figure 2(d). For the samples with x ≤ 0.15, the PF 10


Page 10 of 22

Page 11 of 22

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


significantly improves as compared to that of our pristine CoSbS, which is owing to the increased electrical conductivity. Meanwhile, further increase of Se content, the PF decreases but still larger than that of the pure CoSbS. A maximum power factor of 1.2 mWm-1K-2 is reached at 923 K for the CoSbS0.85Se0.15 sample, 37% enhancement as compared to the pristine CoSbS.

Figure 3. (a), (b), (c) are the band structures of CoSbS, CoSbS0.875Se0.125 and CoSbS0.75Se0.25, respectively. (d) The band gap from DFT calculations and experimental estimations.

To better explain the electrical transport properties of these CoSbS compounds, we calculated the band structures of the CoSbS, CoSbS0.875Se0.125 and CoSbS0.75Se0.25 as demonstrated in Figure 3. The a, b, c and V of Super-cell solid solutions also increase linearly and follow the Vegard law, which are same with the experiment results. Figure 3(a) shows the band structures of the pure CoSbS. There are two lowest conduction bands, which have several pockets near the band edge in the first 11



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

Brillouin Zone. In particular, multiple electronic pockets are found around points G, Z and B, which favor higher thermoelectric performance. Figure 3(b) and (c) shows the band structures of CoSbS0.875Se0.125 and CoSbS0.75Se0.25 samples. We compare the bottom band dispersion of the CoSbS and its solid solution because of its n-type electrical behavior, only a small energy splitting at the high symmetrical point Z and B was found. No clear band dispersion difference is observed. Since S(2.5) and Se(2.4) has the similar electronegativity and the small difference bond distance of Co-S(2.30Å) and Co-Se(2.37Å) after optimization. The band gap is slightly reduced as a result of Se substation, which may lead to the increase of carrier density in Se subtitled samples. The electrical conductivity will be enhanced. However, due to the increase of minority carrier concentration, the Seebeck coefficient reduced. This trade-off between Seebeck coefficient and electrical conductivity contribute the highest power factor when the experimental doping concentration at x=0.15. Figure 3 (d) shows the band gap from DFT calculations and experimental estimations using the Goldsmid-Sharp formula = 2 , where e, and are the electron charge, the maximum Seebeck coefficient and the corresponding temperature, respectively. A slight linear reduction curve is also observed in the experimental estimation from 0.52 eV for CoSbS to 0.42 eV for CoSbS0.7Se0.3, which is in agreement with our theoretical calculation. On the other hand, there exists a subtle difference between the experimental results and calculated values due to the deficiency of the PBE exchange correlation function.

3.3 Thermal transport properties 12


Page 12 of 22

Page 13 of 22

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


The temperature dependences of the total thermal conductivity κ, and κL+κbip for CoSbS1-xSex (x=0, 0.05, 0.10, 0.15, 0.20, 0.30) compounds are presented in Figure 4 (a) and (b), respectively. The inset of Figure 4 (a) compares the thermal conductivity of CoSbS based compounds reported in literatures and our experimental results. The total thermal conductivity in the solid solution shows a significant decrease as the fraction of Se atoms increases. At room temperature, κ decreases from 8.21 W m-1 K-1 for CoSbS to 5.97 W m-1 K-1 for CoSbS0.75Se0.25 and is dramatically reduced to less than 4.72 W m-1 K-1 at x=0.3. Besides, the total thermal conductivity κ can be obtained by the sum of the electronic thermal conductivity κe, the lattice thermal conductivity κL and the bipolar thermal conductivity κbip. κe is estimated using the Wiedemann-Franz law, which is expressed as κe = L0σT (here L0 and σ are the Lorenz number and the electrical conductivity, respectively) with a constant Lorentz number L0= 2.0 × 10-8 V2 K-2. The sum of κL and κbip can be obtained by directly subtracting

κe from κ. As presented in Figure 4 (b), the lattice thermal conductivity decreases as the temperature increases below 650 K, and then increases because of the bipolar effect. The lattice thermal conductivity κL also shows a decreasing trend as the fraction of Se atoms increases. The room temperature lattice thermal conductivity decreases from 8.2 W m-1 K-1 for CoSbS to 4.7 W m-1 K-1 for CoSbS0.7Se0.3. The reduction of the lattice thermal conduction indicates that the alloying Se atoms can effectively scatter phonons. The minimum lattice thermal conductivity of CoSbS compound was estimated based on the Einstein model.

34,35

The calculated κmin is

~1.12 W m-1 K-1, which is much lower than the lattice thermal conductivity obtained 13



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

in our study, leaving a large room to further reduce the lattice thermal conductivity in CoSbS based compounds.

Figure 4. (a) The total thermal conductivity. (b) The sum of lattice thermal conductivity and bipolar thermal conductivity. (c) The imperfection scaling parameter. (d) The ZT for the CoSbS1-xSex compounds. The insert shows the minimum thermal conductivity of CoSbS doped compounds reported in literatures. Black, Red, Blue, green are the thermal conductivity of Co0.99Ni0.01SbS (Ref.25), Co0.92Ni0.08SbS (Ref.26), CoSb0.96Te0.04S (Ref.27), CoSbS0.70Se0.30 (our work), respectively.

To better understand the point defect scattering mechanism including the mass difference

36,37

and strain difference

38

between Se and S, we used the Callaway

model39 for lattice thermal conductivity analysis in CoSbS1-xSex compounds. If we only consider U-process phonon scattering and point defect phonon scattering, the ratio of lattice thermal conductivity of Se substituted samples to that of pure CoSbS can be expressed as: 38-40 14


Page 14 of 22

Page 15 of 22

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


=

%& =

! (#)

(1)

#

' () * +,-

Γ

(2)

Where u,/ , Ω , ℎ, and Γ are the scaling parameter, the mean acoustic velocity, the average volume per atom, the Planck constant, and the imperfection scaling parameter, respectively. And the mean acoustic velocity v7 , which could be obtained as follows 41 9

9

/7 =8 ; > + : AB

(3)

Synergistic Strategy to Enhance the Thermoelectric Properties of

Recommend Documents