- Email: [email protected]
MnTe2 as a novel promising thermoelectric material
Accepted Manuscript MnTe2 as a novel promissing thermoelectric material Yidong Xu, Wen Li, Chen Wang, Zhiwei Chen, Yixuan Wu, Xinyue Zhang, Juan Li, Siqi Lin, Yue Chen, Yanzhong Pei PII:
S2352-8478(18)30017-0
DOI:
10.1016/j.jmat.2018.04.001
Reference:
JMAT 131
To appear in:
Journal of Materiomics
Received Date: 28 February 2018 Revised Date:
1 April 2018
Accepted Date: 10 April 2018
Please cite this article as: Xu Y, Li W, Wang C, Chen Z, Wu Y, Zhang X, Li J, Lin S, Chen Y, Pei Y, MnTe2 as a novel promissing thermoelectric material, Journal of Materiomics (2018), doi: 10.1016/ j.jmat.2018.04.001. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Yidong Xu1, Wen Li1, Chen Wang2, Zhiwei Chen1, Yixuan Wu1, Xinyue Zhang1, Juan Li1, Siqi Lin1, Yue Chen2 and Yanzhong Pei*,1 1
Interdisciplinary Materials Research Center, School of Materials Science and Engineering, Tongji Univ., 4800 Caoan Rd., Shanghai 201804, China. 2 Department of Mechanical Engineering, The University of Hong Kong, Pokfulam Road, Hong Kong SAR, China
AC C
EP
TE D
M AN U
SC
RI PT
*Email: [email protected] (YP)
MnTe2 as a novel promissing thermoelectric material
RI PT
ACCEPTED MANUSCRIPT Semiconducting manganese ditelluride (MnTe 2) crystalizes in a high symmetry cubic structure with sufficient band gap and consists of nontoxic elements only, therefore is focused on in this work for its potential thermoelectric applications. This material intrinsically comes with a very low hole concentration of 1019 cm-3, which can be successfully increased to 4×1020 cm-3 through Ag-doping at Mn site. Such a broad carrier concentration enables an effective optimization on thermoelectric power factor, and the doping process effectively reduces the lattice thermal conductivity down to ~0.5 W/m-K due to the phonons scattered by additional point defects. As a result, a peak zT of ~0.7 is obtained in p-type conduction. Moreover, the SPB model with acoustic scattering estimates the electronic properties well, which also enables insight into the underlying physical parameters related to the thermoelectric performance. Importantly, band structure calculation suggests a potentially higher thermoelectric performance for n-type conduction due to both higher band degeneracy and lower band effective mass. This work reveals MnTe2 is a novel promising thermoelectric material. Keywords: MnTe2, Thermoelectric, SPB, Transport properties.
SC
cubic crystal structure is usually believed to be beneficial for their high electronic performance, semiconducting MnTe2 crystalizes in a cubic structure with all environment-friendly constituent elements and therefore motivates this work to focus on its electronic and thermal transport properties for thermoelectric applications.
Thermoelectric, which has been considered as the refrigeration or power generation based on Peltier or Seebeck effects, enables a conversion of the heat and electricity directly. However, the large-scale applications of thermoelectric materials is hindered due to their low conversion efficiency, which is determined by the dimensionless figure of merit zT, zT=S2σT/(κL+κE), where σ, T, S, κL and κE are the electrical conductivity, the absolute temperature, Seebeck coefficient, lattice thermal conductivity and electronic thermal conductivity, respectively. Numerous efforts are made on improving the figure of merit zT of material. Proven thermal strategies to reduce κL, which is the only independent parameter determining zT, have led to a significant zT-enhancement. This can be typified by a variety of approaches such as nanostructuring1-4, lattice anharmonicity5, 6, liquid-like ions7, 8, dislocations9-11, point defects including substitutional12, 13, interstitial14 and vacancy defects15, as well as low sound velocity16 and low cutoff frequency of acoustic phonons17, 18. Alternatively, recently developed electronic strategies of band engineering19-21 enable an increase in band degeneracy, therefore successfully decouple the correlation among S, σ and кE and eventually lead to an enhanced electronic performance (power factor, PF=S2σ). This has been demonstrated in many thermoelectric materials, for example, Te20, PbTe22-24, SnTe25-28, GeTe29, Mg2Si30, 31 and half-Heusler32 alloys. Recently, manganese monotelluride (MnTe), as an important additive in thermoelectric IV-VI alloys for manipulating the band structures22, 25, has been demonstrated as a promising thermoelectric material with an intrinsic low κL 33 . A peak zT up to unity is achieved in MnTe with optimized carrier concentrations and reduced lattice thermal conductivity33, 34. These results attract increasing interests to explore semiconducting manganese tellurides as thermoelectric materials with an environment-friendly composition. Being another important compound from the Mn-Te binary phase diagram35 36as shown in the Fig 1, manganese ditellurides (MnTe2) crystallizes in a pyrite cubic structure (Fig. 2a) and has been extensively studied for its magnetic properties37-44, electronic structure45, 46, optical properties37, 45, 47 and phase transition48-50. Its thermoelectric properties have been rarely investigated, and intrinsic MnTe2 was known to show a p-type semiconductor, which possesses a large Seebeck coefficient of ~400 µV/K due to its low carrier concentration51. Similar to many conventional thermoelectircs such as Si/Ge, Half-Heuslers, Skutterudites, PbTe, Clathrates, where the
1200
L
1220
1149
800
740
MnTe
600 400
0
Mn
20
40
x%
MnTe2
M AN U
T (oC)
1000
60
448
80
100
Te
AC C
EP
TE D
Fig. 1. The binary phase diagram of Mn-Te system36
1
In this work, it is found that the hole concentration is effectively increased through Ag-doping from ~2×1019 cm-3 to ~4×1020 cm-3. Such an increase in carrier concentration at the whole temperature range successfully optimizes the electrical performance. Such a doping also introduces point defects and second phase (depending on the doping concentration), and results in a well reduced lattice thermal conductivity. The obtained lowest κL of ~0.5 W/m-K reaches its amorphous limit. Eventually, a peak zT of ~0.7 is obtained. The electronic transport properties are well assessed by SPB model with acoustic scattering in a such broad carrier concentration, which consists with the band structure calculation. It is further indicated that n-type MnTe2 might possess an even higher thermoelectric performance because of the existence of multiple conduction bands with small energy offsets and light band effective masses. This work demonstrates MnTe2 is a novel potential thermoelectric material. The synthesis, characterization, measurement and band structure calculation are given in the supplementary. The crystal structure of MnTe2 is shown in Fig. 2a. It crystalizes in a NaCl-like cubic structure, where Mn ions and Te2 molecules occupy in Na sites and Cl sites, respectively40, 41. The high purity for the samples were identified here, because peaks can be well indexed to the pyrite cubic structure from powder X-ray diffraction (XRD) patterns for Mn1-xAgxTe2 (x≤0.04), as shown in Fig. 2b. Scanning Electron Microscope (SEM) observations (Fig. S1a, for Mn1-xAgxTe2 with x≤0.04) and
x=0.02
Mn
Te
40
50
60
70
2θ (degree)
80
20
90 100
0.00
c
0.01
0.02
3
T=300K Edef~ 8eV
b
0.5
µΗ (cm2/V*s)
x=0 x=0.005 x=0.01 x=0.02 x=0.03 x=0.04
0.6
0.7
hν (eV))
, ,
Prediction Expt. data
m* ~ 5me
50 0K 400 3 00 K
300 200
0.03
0.04
0.05
d
100 1E19
T=500K T=800K Edef~ 6eV Edef~ 5eV
40
1E20
nH (cm
-3
1E21 )
10
35 30
2
1
Eg~0.72eV
TE D
(α /K*(hν ))2
Mn1-xAgxTe2
x
, ,
80 0K
500
1
Fig. 2. Crystal structure (a) for MnTe2 and the XRD patterns (b) for Mn1-xAgxTe2 (x≤0.04). a
2
M AN U
ICSD #643784
30
Mn1-xAgxTe2 Expected Expt.
3
nH (*10
x=0
20
4
cm-3)
x=0.01 x=0.005
b 600
5
x=0.03
25 20 15 10
Mn1-xAgxTe2
x=0 x=0.005 x=0.01 x=0.02 x=0.03 x=0.04
2.4E19 5.5E19 1.8E20 3.1E20 4.0E20 3.9E20
1
m* (me)
a
x=0.04
RI PT
Mn1-xAgxTe2
S (µ µ V/K)
b
Edef (eV)
a
suggests a high doping efficiency of Ag in MnTe2. The broad carrier concentration ranging from 2×1019 cm-3 to 4×1020 cm-3 MANUSCRIPT makes the assessment reliable on the electronic transport properties and insight into the elemental physical parameters determining the thermoelectric performance as discussed below. Temperature dependent Hall carrier concentration (nH) and Hall mobility is shown in Fig. S3. The nearly constant nH (a nearly constant of Hall coefficient) for heavily doped samples (x≥1%) in the whole temperature range indicates a degenerated semiconducting behavior by a single band, further suggesting the doping efficiency of Ag. The charge carrier is found to be dominantly scattered by acoustic phonons54 because the Hall mobility (µH) decreases with increasing temperature approximately via µH~T-1.5. A µH at room temperature as low as ~2 cm/V-s is obtained for the all samples, presumably resulting from the high band effective mass, magnetic scattering55 and huge electronegativity difference between Te and Mn.
SC
Energy Dispersive Spectrometer (EDS) compositional mapping analysis (Fig. S1a, for Mn0.96Ag0.04Te2) are carried ACCEPTED out for further identifying the phase composition. It is shown that when x≤0.03 the materials are single phased while small amount of Ag-rich precipitates (a few micron in size) are observed in the sample with x=0.04. These results suggest the solubility of Ag in MnTe2 is ~3%. The calculated band structures for MnTe2 with different on-site Coulomb (U) and Exchange (J) potentials are shown in Fig. 3a. A negligible influence on the band structure occurred when the U value changed from 4.0 to 6.0 eV. According to the solid atom method52, U of 5.0 eV and J of 0.8 eV for MnTe2 are determined. A direct Eg of 0.7 eV is observed. In Brillouin zone, the extrema for the valence band located at Γ point suggests a single band transport behavior and an effective band degeneracy (Nv) of 1 for p-type MnTe2.
5 1E19
1E20
nH (cm
-3
)
1E21
0
300 400 500 600
0.1 700 800 900
T (K)
0.8
EP
Fig. 3. Band structures of MnTe2 calculated with different on-site Coulomb (U) and Exchange (J) potentials (a) and photo energy dependent normalized optical absorption (b) for Mn1-xAgxTe2 (x≤0.04) at room temperature.
According to the optical measurements53 at room temperature, the optical band gap is estimated ∼0.7 eV for Mn1-xAgxTe2 (x≤0.04) (Fig. 3b), which agrees well with the result from the calculated band structure (Fig. 3a). Furthermore, it is found that the band gap remains nearly unchanged with increasing Ag concentration (Fig. 3b). Fig. 4a shows the Ag concentration versus the Hall carrier concentration (nH) for Mn1-xAgxTe2. When x≤0.03, nH increases linearly with increasing x and then saturates. This further confirms the solubility of ~3% for Ag in MnTe2. It can
Fig. 4. Ag concentration versus measured and expected hole concentration (a) at 300K, Hall carrier concentration dependent Seebeck coefficient (b) and Hall mobility (c) with comparison to SPB model predictions at different temperatures, as well as temperature dependent density state of effective mass (m*) and deformation potential coefficient (Edef) (d) for Mn1-xAgxTe2 (x≤0.04).
AC C
According to the calculated band structure (Fig. 3a), the electronic transport is dominated by the valence band at Γ. In addition, a weak interaction between the conduction and valence bands due to the wide Eg of ~0.7 eV (Fig. 3a and 3b) leads to the approximated parabolic band. Therefore, it is a reliable for using the SPB model with acoustic phonon to estimate the fundamental material properties affecting the thermoelectric performance. It is shown that both the carrier concentration dependent Seebeck coefficient (Fig. 4b) and Hall mobility (Fig. 4c) can be well understood by the SPB model. A state effective mass (m*) of ~5me and a deformation potential coefficient (Edef) of ~8 eV at room temperature are estimated for Mn1-xAgxTe2. A rigid band behavior for Mn1-xAgxTe2 can be seen from the independence of both m* and Edef on carrier concentration. Fig. 4d shows temperature dependents deformation potential coefficient (Edef) and density of state effective mass (m*) for the Mn1-xAgxTe2 (x≤0.04). Fig. 5a and 5b show temperature versus electrical
be seen that the lattice parameter remains nearly unchanged due to Ag-doping (Fig. S2), which is presumably due to the similar ion radii between Mn and Ag and the low doping concentration. The precipitation of Ag-rich phases at x>0.03 would affect the electronic transport, but the low concentration leads their effect on carrier concentration to be negligible. It is found that the measured carrier concentration is in good agreement with the expected hole concentration, supposing each substitutional Ag atom releases one hole. This 2
a
ρ (mohm*cm)
100
x=0 x=0.005 x=0.01 x=0.02 x=0.03 x=0.04
b400 350
S (µ µ V/K)
Mn1-xAgxTe2
300 250
150 300
400
500
600
T (K)
700
800
900
T
-1
1.2
0.8
300
400
500
600
700
800
， ，
1.2
zT
0.5 0.4 0.3 0.2 0.1
T (K)
0.00
0.01
d 0.8
zT
0.6
T=300K T=500K T=850K
0.4
SC
0.7
Model Expt. data
1.6
300 400 500 600 700 800 900
c0.8
， ，
0.8
min κL
0.4
x=0 x=0.005 x=0.01 x=0.02 x=0.03 x=0.04
κ L (W/mK)
1.6
κ κL
RI PT
κ ,κ L (W/mK)
Mn1-xAgxTe2
0.0
300 400 500 600 700 800 900
0.7
T=850K Edef~5eV
0.6
m*~5me
0.02
0.03
0.04
0.05
x κLmin κL~0.5 κL~0.6
0.5
κL~0.7
0.4 0.3 0.2 1E+20
1E+21 nH (cm ) Fig. 6. Temperature dependent total thermal conductivity and lattice thermal conductivity (a), the reduction in lattice thermal conductivity due to Ag-doping along with the point defect scattering model predictions at different temperatures (b), temperature dependent zT (c) for Mn1-xAgxTe2 (x≤0.04) and Hall carrier concentration versus zT predicted by SPB model with different lattice thermal conductivity at 850 K (d).
T (K)
TE D
200 10
introduced. As a result, the lowest κL of ~0.5 W/m-K is achieved, which is closed to the κLmin of MnTe2. The MANUSCRIPT Debye-Callaway model57, 58 is used to understand the κL-reduction due to Ag-doping. Because of the large contrast on mass and strain between the host Mn and the substitutional Ag, the experimentally observed κL-reduction can be well expected according to the point defect scattering model, as shown in Fig. 6b at three different temperatures for Mn1-xAgxTe2 (x≤0.04). a2.0 b2.0
M AN U
resistivity and Seebeck coefficient for Mn1-xAgxTe2 (x≤0.04), respectively. Both of them decrease with increasing Ag-doping ACCEPTED concentration, because of the increased nH. The solubility of ~3% for Ag in MnTe2 limits the further decrease with an even higher Ag concentration (x=4%). All the samples are p-type conduction due to the positive value of Seebeck coefficient. Pristine MnTe2 shows an intrinsic p-type conduction with a hole concentration of 2×1019 cm-3, which suggests the existence of Mn vacancies as the major type of defects in this compound. Since the binary phase diagram of Mn-Te system (Fig. 1) shows no phase transition for MnTe2. The slight fluctuation in Seebeck coefficient (Fig. 5b) is believed to be due to measurement uncertainties. Due to the minority carriers exited thermally, electrical resistivity and Seebeck coefficient decrease with increasing temperature for low-nH sample. Majority of the samples obtained in this work shows a degenerated semiconducting behavior. Temperature versus thermoelectric power factor (PF=S2/ρ) for Mn1-xAgxTe2 (x≤0.04) is shown in Fig. S4. The PF for the Ag-doped samples is significantly enhanced in the entire temperature range, comparing to that of pristine MnTe2. Using the average Edef and m* as shown in Fig. 4d, the SPB model leads to a further prediction on power factor dependent carrier concentration at different temperatures. The model prediction indicates the obtained nH by Ag-doping is very close to the optimum, leading to a maximal PF of ∼5 µW/cm-K2 at 850 K.
900
T (K)
Fig. 5. Temperature versus electrical resistivity (a) and Seebeck coefficient (b).
AC C
EP
Fig. 6a shows temperature dependent total (κ) and lattice (κL) thermal conductivities for Mn1-xAgxTe2 (x≤0.04). The total thermal conductivity mainly consists of the two parts: lattice thermal conductivity (κL) and electronic thermal conductivity (κE) calculated by the Wiedemann-Franz law, κE=LT/ρ, where L is the Lorenz factor, which is estimated by the SPB model with acoustic phonon scattering. And κL can be estimated by subtracting the κE from κ. κL decreases with increasing temperature by T-1, showing a dominant phonon scattering in MnTe2 by Umklapp process. In order to understand the low κL in MnTe2, longitudinal (vl) and transverse (vt) sound velocities are measured and listed in Table S1. The corresponding physical parameters, such as Debye temperature (θD), Poisson ratio (ε), bulk modulus (B) and Grüeneisen parameter (γ), are estimated based on the measured vl and vt and listed in Table S1. It is shown that the change in sound velocities due to Ag-doping does not exceed 5%. The amorphous limit (κLmin) (dashed curve in Fig. 6a) is estimated by the Cahill model56. It is also found that κL reduces with increasing Ag-doping level, which mainly results from the additional scattering of phonon by the point defects
3
-3
Fig. 6c shows the temperature dependent zT for Mn1-xAgxTe2 (x≤0.04). zT increases with increasing Ag concentration and temperature. A peak zT of 0.7 is achieved, resulting from the synergistic effects of optimization-nH and reduction-κL by Ag-substitution. Using the measured κL and the minimum predicted by Cahill model, the SPB model makes a further a prediction on zT versus nH for each composition at 850K, as shown in Fig. 6d. The experimental results are in good agreement with the prediction. It is found that the hole concentration obtained in this work is close to the optimum for maximizing zT at 850 K for Mn1-xAgxTe2 with x≥0.04). It should be noted that the maximum PF of ~5 µW/cm-K2 here is much lower than those of conventional thermoelectrics59-61 due to their high band degeneracy (Nv of 4 or higher32, 60, 62, 63). The low PF of MnTe2 in p-type is presumably due to the high m* and low Nv of 1 (Fig. 3a). According to the calculated band structure (Fig. 3a), it is found that the band effective mass for conduction band at Γ point is lower than of the valence band, indicating higher carrier mobility for n-type MnTe2. Moreover, the Γ and R conduction bands show a small band energy offset, therefore, a heavily doped MnTe2 would show a multiple-band transport behavior with a higher effective Nv in n-type. Both low band effective mass and high band degeneracy suggest a great potential for a higher thermoelectric performance in n-type MnTe2. Unfortunately, attempts on doping quite a few
SC
RI PT
29, 1606768. 12. Jiawen Shen, Xinyue Zhang, Zhiwei Chen, Siqi Lin, Juan Lin, MANUSCRIPT Wen Li, Shasha Lin, Yue Chen and Y. Pei, Journal of Materials Chemistry A, 2017, 5, 5314-5320. 13. S. D. Mahanti and D. Bilc, J Phys-Condens Mat, 2004, 16, S5277-S5288. 14. Y. Pei, L. Zheng, W. Li, S. Lin, Z. Chen, Y. Wang, X. Xu, H. Yu, Y. Chen and B. Ge, Advanced Electronic Materials, 2016, 2, 1600019. 15. Wen Li, Siqi Lin, Xinyue Zhang, Zhiwei Chen, Xiangfan Xu and Y. Pei, Chem Mater, 2016, 28, 6227-6232. 16. W. Li, S. Lin, B. Ge, J. Yang, W. Zhang and Y. Pei, Advanced Science, 2016, 3, 1600196. 17. S. Lin, W. Li, S. Li, X. Zhang, Z. Chen, Y. Xu, Y. Chen and Y. Pei, Joule, 2017, 1, 816-830. 18. Z. Chen, X. Zhang and Y. Pei, Advanced materials, 2018, 1705617. 19. G. CHEN, M. S. DRESSELHAUS, J.-P. FLEURIAL and T. CAILLAT, Anglais, 2003, 48, 45-66 20. S. Lin, W. Li, Z. Chen, J. Shen, B. Ge and Y. Pei, Nature communications, 2016, 7, 10287. 21.Y. Pei, X. Shi, A. LaLonde, H. Wang, L. Chen and G. J. Snyder, Nature, 2011, 473, 66-69. 22.Y. Pei, H. Wang, Z. M. Gibbs, A. D. LaLonde and G. J. Snyder, NPG Asia Materials 2012, 4, e28. 23. Y. Pei, A. D. LaLonde, N. A. Heinz, X. Shi, S. Iwanaga, H. Wang, L. Chen and G. J. Snyder, Advanced materials, 2011, 23, 5674-5678. 24. J. Shen, Z. Chen, L. Zheng, W. Li and Y. Pei, Journal of Materials Chemistry C, 2016, 4, 209-214. 25. W. Li, Z. Chen, S. Lin, Y. Chang, B. Ge, Y. Chen and Y. Pei, Journal of Materiomics, 2015, 1, 307-315. 26. W. Li, Y. Wu, S. Lin, Z. Chen, J. Li, X. Zhang, L. Zheng and Y. Pei, ACS Energy Letters, 2017, 2, 2349-2355. 27.Linglang Zheng, Wen Li, Siqi Lin, Juan Li, Zhiwei Chen and Y. Pei, ACS Energy Letters, 2017, 2, 563-568. 28. Wen Li, Linglang Zheng, Binghui Ge, Siqi Lin, Xinyue Zhang, Zhiwei Chen, Yunjie Chang and Y. Pei, Adv. Mater., 2017, 29, 1605887. 29. Juan Li, Zhiwei Chen, Xinyue Zhang, Yongxing Sun, Jiong Yang and Y. Pei, NPG Asia Mater, 2017, 9, e353. 30. W. Liu, X. J. Tan, K. Yin, H. J. Liu, X. F. Tang, J. Shi, Q. J. Zhang and C. Uher, Phys Rev Lett, 2012, 108, 166601. 31. M. Inoue, Y. Takai, K. Fukui, H. Yagi and T. Tatsukawa, J Magn Magn Mater, 1983, 36, 255-258. 32. C. G. Fu, T. J. Zhu, Y. Z. Pei, H. H. Xie, H. Wang, G. J. Snyder, Y. Liu, Y. T. Liu and X. B. Zhao, Adv. Energy Mater., 2014, 4, 1400600. 33.Y. Xu, W. Li, C. Wang, J. Li, Z. Chen, S. Lin, Y. Chen and Y. Pei, Journal of Materials Chemistry A, 2017, 5, 19143-19150. 34. Y. Ren, J. Yang, Q. Jiang, D. Zhang, Z. Zhou, X. Li, J. Xin and X. He, Journal of Materials Chemistry, 2017, 00, 1-9. 35. G. S. Mann and L. H. V. Vlack, Physical Chemistry, 1977, 8, 53-57. 36. M. N. Mamonntov, Inorganic materials, 1996, 32, 1190-1194. 37. Y. Mita, Y. Ishida, M. Kobayashi and S. Endo, Acta Physica Polonica A, 2008, 113, 617-620. 38. M. T. Heinilä and A. S. Oja, Physical Review B, 1994, 49, 11995-12002.
AC C
EP
TE D
M AN U
elements, such as Fe, Al, Ga, In, La, Co, Ni, Sn, Pb at Mn site and I at Te site, turn out to be unsuccessful for a sufficient ACCEPTED high electron concentration in this work. This opens a question to the field for a realization of possibly high thermoelectric performance in n-type MnTe2. In summary, this work reveals a new promissing thermoelectric material MnTe2. The broad carrier concentration ranging from ~2×1019 cm-3 to ~4×1020 cm-3, enabled by Ag-doping, not only optimizes the hole carrier concentration but also reduces the lattice thermal conductivity, both of which contribute to the peak zT as high as 0.7. The broad carrier concentration further enables a systematical analysis of elemental material properties determining thermoelectric performance by the SPB model with acoustic scattering. The band structure calculations indicate an even higher thermoelectric performance in n-type MnTe2 due to its high band degeneracy and low band mass, which deserves further investigation. This work demonstrates MnTe2 as a promising matrix material for high-efficient thermoelectric applications. Acknowledgements This work is supported by the National Natural Science Foundation of China (Grant No. 11474219 and 51772215), and the national Recruitment Program of Global Youth Experts (1000 Plan) and the Fok Ying Tung Education Foundation (Grant No. 20170072210001). CW and YC acknowledge the financial support from the Early Career Scheme of RGC under Project Number 27202516 and the research computing facilities offered by ITS, HKU. References 1. K. F. Hsu, S. Loo, F. Guo, W. Chen, J. S. Dyck, C. Uher, T. Hogan, E. K. Polychroniadis and M. G. Kanatzidis, Science, 2004, 303, 818-821. 2. K. Biswas, J. He, I. D. Blum, C.-I. Wu, T. P. Hogan, D. N. Seidman, V. P. Dravid and M. G. Kanatzidis, Nature, 2012, 489, 414-418. 3. B. Poudel, Q. Hao, Y. Ma, Y. C. Lan, A. Minnich, B. Yu, X. A. Yan, D. Z. Wang, A. Muto, D. Vashaee, X. Y. Chen, J. M. Liu, M. S. Dresselhaus, G. Chen and Z. F. Ren, Science, 2008, 320, 634-638. 4. Y. Pei, J. Lensch-Falk, E. S. Toberer, D. L. Medlin and G. J. Snyder, Advanced Functional Materials, 2011, 21, 241-249. 5. L.-D. Zhao, S.-H. Lo, Y. Zhang, H. Sun, G. Tan, C. Uher, C. Wolverton, V. P. Dravid and M. G. Kanatzidis, Nature, 2014, 508, 373-377. 6. D. T. Morelli, V. Jovovic and J. P. Heremans, Phys Rev Lett, 2008, 101, 035901. 7. H. Liu, X. Shi, F. Xu, L. Zhang and W. Zhang, Nature materials, 2012, 11, 422-425. 8. H. Liu, X. Yuan, P. Lu, X. Shi, F. Xu, Y. He, Y. Tang, S. Bai, W. Zhang, L. Chen, Y. Lin, L. Shi, H. Lin, X. Gao, X. Zhang, H. Chi and C. Uher, Advanced materials, 2013, 25, 6607-6612. 9. S. I. Kim, K. H. Lee, H. A. Mun, H. S. Kim, S. W. Hwang, J. W. Roh, D. J. Yang, W. H. Shin, X. S. Li and Y. H. Lee, Science, 2015, 348, 109-114. 10. Zhiwei Chen, Binghui Ge, Wen Li, Siqi Lin, Jiawen Shen, Yunjie Chang, Riley Hanus, G. Jeffrey Snyder and Y. Pei, Nature Communications, 2017, 8, 13828. 11. Zhiwei Chen, Zhengzhong Jian, Wen Li, Yunjie Chang, Binghui Ge, Riley Hanus, Jiong Yang, Yue Chen, Mingxin Huang, Gerald Jeffrey Snyder and Y. Pei, Adv. Mater., 2017,
4
SC
RI PT
the Less Common Metals, 1965, 8, 272-287. 52. S. L. Dudarev, G. A. Botton, S. Y. Savrasov, C. J. Humphreys MANUSCRIPT and A. P. Sutton, Physical Review B, 1998, 57, 1505-1509. 53. Z. M. Gibbs;, A. Lalonde; and G. J. Snyder, New Journal of Physics, 2013, 15, 075020. 54. Y. I. Ravich, B. A. Efimova and I. A. Smirnov, Semiconducting Lead Chalcogenides, Plenum Press, New York, 1970. 55.W. Xie, S. Populoh, K. Gałązka, X. Xiao, L. Sagarna, Y. Liu, M. Trottmann, J. He and A. Weidenkaff, Journal of Applied Physics, 2014, 115, 103707. 56. D. G. Cahill and R. O. Pohl, Annu. Rev. Phys. Chem., 1988, 39, 93-121. 57. J. Callaway and H. C. Vonbaeyer, Physical Review, 1960, 120, 1149-1154. 58. P. G. Klemens, Physical Review, 1960, 119, 507-509. 59. C. G. Fu, S. Q. Bai, Y. T. Liu, Y. S. Tang, L. D. Chen, X. B. Zhao and T. J. Zhu, Nature Communications, 2015, 6, 7. 60. A. D. LaLonde, Y. Pei and G. J. Snyder, Energy & Environmental Science, 2011, 4, 2090–2096. 61. L.-D. Zhao, G. Tan, S. Hao, J. He, Y. Pei, H. Chi, H. Wang, S. Gong, H. Xu and V. P. Dravid, Science, 2016, 351, 141-144. 62.Y. Pei, X. Shi, A. LaLonde, H. Wang, L. Chen and G. J. Snyder, Nature, 2011, 473, 66-69. 63. Y. Tang, Z. M. Gibbs, L. A. Agapito, G. Li, H.-S. Kim, M. B. Nardelli, S. Curtarolo and J. Snyder, Nat Mater, 2015, 14, 1223-1228.
AC C
EP
TE D
M AN U
39. O. O. and M. T., Journal of the Physical Society of Japan, 1977, 43, 343-344. 40. J. M. Hastings, L. M. Corliss, M. Blume andACCEPTED M. Pasternak, Physical Review B, 1970, 1, 3209-3210. 41. M. Pasternak and A. L. Spijkervet, Physical Review, 1969, 181, 574-579. 42. P. Burlet, E. Ressouche, B. Malaman, R. Welter, J. P. Sanchez and P. Vulliet, Physical Review B, 1997, 56, 14013-14018. 43. T. Chattopadhyay and H. Fjellvag, Physics Letters A, 1987, 120, 44-46. 44. J. M. Hastings, L. M. Corliss, W. Kunnmann and D. Mukamel, Physical Review B, 1986, 33, 6326-6330. 45. K. V. Kaznacheyev, T. Muro, T. Matsushita, T. Iwasaki, Y. Kuwata, H. Harada, S. Suga, H. Ishii, T. Miyahara and T. Mizokawa, Physical Review B, 1998, 58, 13491-13497. 46. A. Sawaoka and S. Miyahara, Journal of the Physical Society of Japan, 1965, 20, 2087. 47.B. Muller and H. D. Lutz, Solid State Communications, 1991, 78, 469-471. 48. P. Vulliet, J. P. Sanchez, D. Braithwaite, M. Amanowicz and B. Malaman, Physical Review B, 2001, 63, 184403. 49. H. Fjellvag, W. A. Grosshans, W. Honle and A. Kjekshus, Journal of magnetism and Magnetic Materials, 1995, 145, 118-124. 50. H. Fjellvag and A. Kjekshus, Physics Letters A, 1985, 112, 411-413. 51. W. D. Johnston, R. C. Miller and D. H. Hamon, Journal of
5
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Yanzhong Pei has been working on advanced thermoelectric semiconductors for more than a decade, from synthesizing the materials to understanding the underlying physics and chemistry. He holds a B. E. from Central South University in China, a Ph. D. from the Shanghai Institute of Ceramics, CAS, and postdoctoral research experience of around 5 years from Michigan State University and the California
AC C
EP
TE D
M AN U
SC
RI PT
Institute of Technology. He became a National 1000 Plan junior professor in 2012 at the School of Materials Science and Engineering, Tongji University, Shanghai, China. His interests are focused on materials physics and chemistry for energy applications.
S2352-8478(18)30017-0
DOI:
10.1016/j.jmat.2018.04.001
Reference:
JMAT 131
To appear in:
Journal of Materiomics
Received Date: 28 February 2018 Revised Date:
1 April 2018
Accepted Date: 10 April 2018
Please cite this article as: Xu Y, Li W, Wang C, Chen Z, Wu Y, Zhang X, Li J, Lin S, Chen Y, Pei Y, MnTe2 as a novel promissing thermoelectric material, Journal of Materiomics (2018), doi: 10.1016/ j.jmat.2018.04.001. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Yidong Xu1, Wen Li1, Chen Wang2, Zhiwei Chen1, Yixuan Wu1, Xinyue Zhang1, Juan Li1, Siqi Lin1, Yue Chen2 and Yanzhong Pei*,1 1
Interdisciplinary Materials Research Center, School of Materials Science and Engineering, Tongji Univ., 4800 Caoan Rd., Shanghai 201804, China. 2 Department of Mechanical Engineering, The University of Hong Kong, Pokfulam Road, Hong Kong SAR, China
AC C
EP
TE D
M AN U
SC
RI PT
*Email: [email protected] (YP)
MnTe2 as a novel promissing thermoelectric material
RI PT
ACCEPTED MANUSCRIPT Semiconducting manganese ditelluride (MnTe 2) crystalizes in a high symmetry cubic structure with sufficient band gap and consists of nontoxic elements only, therefore is focused on in this work for its potential thermoelectric applications. This material intrinsically comes with a very low hole concentration of 1019 cm-3, which can be successfully increased to 4×1020 cm-3 through Ag-doping at Mn site. Such a broad carrier concentration enables an effective optimization on thermoelectric power factor, and the doping process effectively reduces the lattice thermal conductivity down to ~0.5 W/m-K due to the phonons scattered by additional point defects. As a result, a peak zT of ~0.7 is obtained in p-type conduction. Moreover, the SPB model with acoustic scattering estimates the electronic properties well, which also enables insight into the underlying physical parameters related to the thermoelectric performance. Importantly, band structure calculation suggests a potentially higher thermoelectric performance for n-type conduction due to both higher band degeneracy and lower band effective mass. This work reveals MnTe2 is a novel promising thermoelectric material. Keywords: MnTe2, Thermoelectric, SPB, Transport properties.
SC
cubic crystal structure is usually believed to be beneficial for their high electronic performance, semiconducting MnTe2 crystalizes in a cubic structure with all environment-friendly constituent elements and therefore motivates this work to focus on its electronic and thermal transport properties for thermoelectric applications.
Thermoelectric, which has been considered as the refrigeration or power generation based on Peltier or Seebeck effects, enables a conversion of the heat and electricity directly. However, the large-scale applications of thermoelectric materials is hindered due to their low conversion efficiency, which is determined by the dimensionless figure of merit zT, zT=S2σT/(κL+κE), where σ, T, S, κL and κE are the electrical conductivity, the absolute temperature, Seebeck coefficient, lattice thermal conductivity and electronic thermal conductivity, respectively. Numerous efforts are made on improving the figure of merit zT of material. Proven thermal strategies to reduce κL, which is the only independent parameter determining zT, have led to a significant zT-enhancement. This can be typified by a variety of approaches such as nanostructuring1-4, lattice anharmonicity5, 6, liquid-like ions7, 8, dislocations9-11, point defects including substitutional12, 13, interstitial14 and vacancy defects15, as well as low sound velocity16 and low cutoff frequency of acoustic phonons17, 18. Alternatively, recently developed electronic strategies of band engineering19-21 enable an increase in band degeneracy, therefore successfully decouple the correlation among S, σ and кE and eventually lead to an enhanced electronic performance (power factor, PF=S2σ). This has been demonstrated in many thermoelectric materials, for example, Te20, PbTe22-24, SnTe25-28, GeTe29, Mg2Si30, 31 and half-Heusler32 alloys. Recently, manganese monotelluride (MnTe), as an important additive in thermoelectric IV-VI alloys for manipulating the band structures22, 25, has been demonstrated as a promising thermoelectric material with an intrinsic low κL 33 . A peak zT up to unity is achieved in MnTe with optimized carrier concentrations and reduced lattice thermal conductivity33, 34. These results attract increasing interests to explore semiconducting manganese tellurides as thermoelectric materials with an environment-friendly composition. Being another important compound from the Mn-Te binary phase diagram35 36as shown in the Fig 1, manganese ditellurides (MnTe2) crystallizes in a pyrite cubic structure (Fig. 2a) and has been extensively studied for its magnetic properties37-44, electronic structure45, 46, optical properties37, 45, 47 and phase transition48-50. Its thermoelectric properties have been rarely investigated, and intrinsic MnTe2 was known to show a p-type semiconductor, which possesses a large Seebeck coefficient of ~400 µV/K due to its low carrier concentration51. Similar to many conventional thermoelectircs such as Si/Ge, Half-Heuslers, Skutterudites, PbTe, Clathrates, where the
1200
L
1220
1149
800
740
MnTe
600 400
0
Mn
20
40
x%
MnTe2
M AN U
T (oC)
1000
60
448
80
100
Te
AC C
EP
TE D
Fig. 1. The binary phase diagram of Mn-Te system36
1
In this work, it is found that the hole concentration is effectively increased through Ag-doping from ~2×1019 cm-3 to ~4×1020 cm-3. Such an increase in carrier concentration at the whole temperature range successfully optimizes the electrical performance. Such a doping also introduces point defects and second phase (depending on the doping concentration), and results in a well reduced lattice thermal conductivity. The obtained lowest κL of ~0.5 W/m-K reaches its amorphous limit. Eventually, a peak zT of ~0.7 is obtained. The electronic transport properties are well assessed by SPB model with acoustic scattering in a such broad carrier concentration, which consists with the band structure calculation. It is further indicated that n-type MnTe2 might possess an even higher thermoelectric performance because of the existence of multiple conduction bands with small energy offsets and light band effective masses. This work demonstrates MnTe2 is a novel potential thermoelectric material. The synthesis, characterization, measurement and band structure calculation are given in the supplementary. The crystal structure of MnTe2 is shown in Fig. 2a. It crystalizes in a NaCl-like cubic structure, where Mn ions and Te2 molecules occupy in Na sites and Cl sites, respectively40, 41. The high purity for the samples were identified here, because peaks can be well indexed to the pyrite cubic structure from powder X-ray diffraction (XRD) patterns for Mn1-xAgxTe2 (x≤0.04), as shown in Fig. 2b. Scanning Electron Microscope (SEM) observations (Fig. S1a, for Mn1-xAgxTe2 with x≤0.04) and
x=0.02
Mn
Te
40
50
60
70
2θ (degree)
80
20
90 100
0.00
c
0.01
0.02
3
T=300K Edef~ 8eV
b
0.5
µΗ (cm2/V*s)
x=0 x=0.005 x=0.01 x=0.02 x=0.03 x=0.04
0.6
0.7
hν (eV))
, ,
Prediction Expt. data
m* ~ 5me
50 0K 400 3 00 K
300 200
0.03
0.04
0.05
d
100 1E19
T=500K T=800K Edef~ 6eV Edef~ 5eV
40
1E20
nH (cm
-3
1E21 )
10
35 30
2
1
Eg~0.72eV
TE D
(α /K*(hν ))2
Mn1-xAgxTe2
x
, ,
80 0K
500
1
Fig. 2. Crystal structure (a) for MnTe2 and the XRD patterns (b) for Mn1-xAgxTe2 (x≤0.04). a
2
M AN U
ICSD #643784
30
Mn1-xAgxTe2 Expected Expt.
3
nH (*10
x=0
20
4
cm-3)
x=0.01 x=0.005
b 600
5
x=0.03
25 20 15 10
Mn1-xAgxTe2
x=0 x=0.005 x=0.01 x=0.02 x=0.03 x=0.04
2.4E19 5.5E19 1.8E20 3.1E20 4.0E20 3.9E20
1
m* (me)
a
x=0.04
RI PT
Mn1-xAgxTe2
S (µ µ V/K)
b
Edef (eV)
a
suggests a high doping efficiency of Ag in MnTe2. The broad carrier concentration ranging from 2×1019 cm-3 to 4×1020 cm-3 MANUSCRIPT makes the assessment reliable on the electronic transport properties and insight into the elemental physical parameters determining the thermoelectric performance as discussed below. Temperature dependent Hall carrier concentration (nH) and Hall mobility is shown in Fig. S3. The nearly constant nH (a nearly constant of Hall coefficient) for heavily doped samples (x≥1%) in the whole temperature range indicates a degenerated semiconducting behavior by a single band, further suggesting the doping efficiency of Ag. The charge carrier is found to be dominantly scattered by acoustic phonons54 because the Hall mobility (µH) decreases with increasing temperature approximately via µH~T-1.5. A µH at room temperature as low as ~2 cm/V-s is obtained for the all samples, presumably resulting from the high band effective mass, magnetic scattering55 and huge electronegativity difference between Te and Mn.
SC
Energy Dispersive Spectrometer (EDS) compositional mapping analysis (Fig. S1a, for Mn0.96Ag0.04Te2) are carried ACCEPTED out for further identifying the phase composition. It is shown that when x≤0.03 the materials are single phased while small amount of Ag-rich precipitates (a few micron in size) are observed in the sample with x=0.04. These results suggest the solubility of Ag in MnTe2 is ~3%. The calculated band structures for MnTe2 with different on-site Coulomb (U) and Exchange (J) potentials are shown in Fig. 3a. A negligible influence on the band structure occurred when the U value changed from 4.0 to 6.0 eV. According to the solid atom method52, U of 5.0 eV and J of 0.8 eV for MnTe2 are determined. A direct Eg of 0.7 eV is observed. In Brillouin zone, the extrema for the valence band located at Γ point suggests a single band transport behavior and an effective band degeneracy (Nv) of 1 for p-type MnTe2.
5 1E19
1E20
nH (cm
-3
)
1E21
0
300 400 500 600
0.1 700 800 900
T (K)
0.8
EP
Fig. 3. Band structures of MnTe2 calculated with different on-site Coulomb (U) and Exchange (J) potentials (a) and photo energy dependent normalized optical absorption (b) for Mn1-xAgxTe2 (x≤0.04) at room temperature.
According to the optical measurements53 at room temperature, the optical band gap is estimated ∼0.7 eV for Mn1-xAgxTe2 (x≤0.04) (Fig. 3b), which agrees well with the result from the calculated band structure (Fig. 3a). Furthermore, it is found that the band gap remains nearly unchanged with increasing Ag concentration (Fig. 3b). Fig. 4a shows the Ag concentration versus the Hall carrier concentration (nH) for Mn1-xAgxTe2. When x≤0.03, nH increases linearly with increasing x and then saturates. This further confirms the solubility of ~3% for Ag in MnTe2. It can
Fig. 4. Ag concentration versus measured and expected hole concentration (a) at 300K, Hall carrier concentration dependent Seebeck coefficient (b) and Hall mobility (c) with comparison to SPB model predictions at different temperatures, as well as temperature dependent density state of effective mass (m*) and deformation potential coefficient (Edef) (d) for Mn1-xAgxTe2 (x≤0.04).
AC C
According to the calculated band structure (Fig. 3a), the electronic transport is dominated by the valence band at Γ. In addition, a weak interaction between the conduction and valence bands due to the wide Eg of ~0.7 eV (Fig. 3a and 3b) leads to the approximated parabolic band. Therefore, it is a reliable for using the SPB model with acoustic phonon to estimate the fundamental material properties affecting the thermoelectric performance. It is shown that both the carrier concentration dependent Seebeck coefficient (Fig. 4b) and Hall mobility (Fig. 4c) can be well understood by the SPB model. A state effective mass (m*) of ~5me and a deformation potential coefficient (Edef) of ~8 eV at room temperature are estimated for Mn1-xAgxTe2. A rigid band behavior for Mn1-xAgxTe2 can be seen from the independence of both m* and Edef on carrier concentration. Fig. 4d shows temperature dependents deformation potential coefficient (Edef) and density of state effective mass (m*) for the Mn1-xAgxTe2 (x≤0.04). Fig. 5a and 5b show temperature versus electrical
be seen that the lattice parameter remains nearly unchanged due to Ag-doping (Fig. S2), which is presumably due to the similar ion radii between Mn and Ag and the low doping concentration. The precipitation of Ag-rich phases at x>0.03 would affect the electronic transport, but the low concentration leads their effect on carrier concentration to be negligible. It is found that the measured carrier concentration is in good agreement with the expected hole concentration, supposing each substitutional Ag atom releases one hole. This 2
a
ρ (mohm*cm)
100
x=0 x=0.005 x=0.01 x=0.02 x=0.03 x=0.04
b400 350
S (µ µ V/K)
Mn1-xAgxTe2
300 250
150 300
400
500
600
T (K)
700
800
900
T
-1
1.2
0.8
300
400
500
600
700
800
， ，
1.2
zT
0.5 0.4 0.3 0.2 0.1
T (K)
0.00
0.01
d 0.8
zT
0.6
T=300K T=500K T=850K
0.4
SC
0.7
Model Expt. data
1.6
300 400 500 600 700 800 900
c0.8
， ，
0.8
min κL
0.4
x=0 x=0.005 x=0.01 x=0.02 x=0.03 x=0.04
κ L (W/mK)
1.6
κ κL
RI PT
κ ,κ L (W/mK)
Mn1-xAgxTe2
0.0
300 400 500 600 700 800 900
0.7
T=850K Edef~5eV
0.6
m*~5me
0.02
0.03
0.04
0.05
x κLmin κL~0.5 κL~0.6
0.5
κL~0.7
0.4 0.3 0.2 1E+20
1E+21 nH (cm ) Fig. 6. Temperature dependent total thermal conductivity and lattice thermal conductivity (a), the reduction in lattice thermal conductivity due to Ag-doping along with the point defect scattering model predictions at different temperatures (b), temperature dependent zT (c) for Mn1-xAgxTe2 (x≤0.04) and Hall carrier concentration versus zT predicted by SPB model with different lattice thermal conductivity at 850 K (d).
T (K)
TE D
200 10
introduced. As a result, the lowest κL of ~0.5 W/m-K is achieved, which is closed to the κLmin of MnTe2. The MANUSCRIPT Debye-Callaway model57, 58 is used to understand the κL-reduction due to Ag-doping. Because of the large contrast on mass and strain between the host Mn and the substitutional Ag, the experimentally observed κL-reduction can be well expected according to the point defect scattering model, as shown in Fig. 6b at three different temperatures for Mn1-xAgxTe2 (x≤0.04). a2.0 b2.0
M AN U
resistivity and Seebeck coefficient for Mn1-xAgxTe2 (x≤0.04), respectively. Both of them decrease with increasing Ag-doping ACCEPTED concentration, because of the increased nH. The solubility of ~3% for Ag in MnTe2 limits the further decrease with an even higher Ag concentration (x=4%). All the samples are p-type conduction due to the positive value of Seebeck coefficient. Pristine MnTe2 shows an intrinsic p-type conduction with a hole concentration of 2×1019 cm-3, which suggests the existence of Mn vacancies as the major type of defects in this compound. Since the binary phase diagram of Mn-Te system (Fig. 1) shows no phase transition for MnTe2. The slight fluctuation in Seebeck coefficient (Fig. 5b) is believed to be due to measurement uncertainties. Due to the minority carriers exited thermally, electrical resistivity and Seebeck coefficient decrease with increasing temperature for low-nH sample. Majority of the samples obtained in this work shows a degenerated semiconducting behavior. Temperature versus thermoelectric power factor (PF=S2/ρ) for Mn1-xAgxTe2 (x≤0.04) is shown in Fig. S4. The PF for the Ag-doped samples is significantly enhanced in the entire temperature range, comparing to that of pristine MnTe2. Using the average Edef and m* as shown in Fig. 4d, the SPB model leads to a further prediction on power factor dependent carrier concentration at different temperatures. The model prediction indicates the obtained nH by Ag-doping is very close to the optimum, leading to a maximal PF of ∼5 µW/cm-K2 at 850 K.
900
T (K)
Fig. 5. Temperature versus electrical resistivity (a) and Seebeck coefficient (b).
AC C
EP
Fig. 6a shows temperature dependent total (κ) and lattice (κL) thermal conductivities for Mn1-xAgxTe2 (x≤0.04). The total thermal conductivity mainly consists of the two parts: lattice thermal conductivity (κL) and electronic thermal conductivity (κE) calculated by the Wiedemann-Franz law, κE=LT/ρ, where L is the Lorenz factor, which is estimated by the SPB model with acoustic phonon scattering. And κL can be estimated by subtracting the κE from κ. κL decreases with increasing temperature by T-1, showing a dominant phonon scattering in MnTe2 by Umklapp process. In order to understand the low κL in MnTe2, longitudinal (vl) and transverse (vt) sound velocities are measured and listed in Table S1. The corresponding physical parameters, such as Debye temperature (θD), Poisson ratio (ε), bulk modulus (B) and Grüeneisen parameter (γ), are estimated based on the measured vl and vt and listed in Table S1. It is shown that the change in sound velocities due to Ag-doping does not exceed 5%. The amorphous limit (κLmin) (dashed curve in Fig. 6a) is estimated by the Cahill model56. It is also found that κL reduces with increasing Ag-doping level, which mainly results from the additional scattering of phonon by the point defects
3
-3
Fig. 6c shows the temperature dependent zT for Mn1-xAgxTe2 (x≤0.04). zT increases with increasing Ag concentration and temperature. A peak zT of 0.7 is achieved, resulting from the synergistic effects of optimization-nH and reduction-κL by Ag-substitution. Using the measured κL and the minimum predicted by Cahill model, the SPB model makes a further a prediction on zT versus nH for each composition at 850K, as shown in Fig. 6d. The experimental results are in good agreement with the prediction. It is found that the hole concentration obtained in this work is close to the optimum for maximizing zT at 850 K for Mn1-xAgxTe2 with x≥0.04). It should be noted that the maximum PF of ~5 µW/cm-K2 here is much lower than those of conventional thermoelectrics59-61 due to their high band degeneracy (Nv of 4 or higher32, 60, 62, 63). The low PF of MnTe2 in p-type is presumably due to the high m* and low Nv of 1 (Fig. 3a). According to the calculated band structure (Fig. 3a), it is found that the band effective mass for conduction band at Γ point is lower than of the valence band, indicating higher carrier mobility for n-type MnTe2. Moreover, the Γ and R conduction bands show a small band energy offset, therefore, a heavily doped MnTe2 would show a multiple-band transport behavior with a higher effective Nv in n-type. Both low band effective mass and high band degeneracy suggest a great potential for a higher thermoelectric performance in n-type MnTe2. Unfortunately, attempts on doping quite a few
SC
RI PT
29, 1606768. 12. Jiawen Shen, Xinyue Zhang, Zhiwei Chen, Siqi Lin, Juan Lin, MANUSCRIPT Wen Li, Shasha Lin, Yue Chen and Y. Pei, Journal of Materials Chemistry A, 2017, 5, 5314-5320. 13. S. D. Mahanti and D. Bilc, J Phys-Condens Mat, 2004, 16, S5277-S5288. 14. Y. Pei, L. Zheng, W. Li, S. Lin, Z. Chen, Y. Wang, X. Xu, H. Yu, Y. Chen and B. Ge, Advanced Electronic Materials, 2016, 2, 1600019. 15. Wen Li, Siqi Lin, Xinyue Zhang, Zhiwei Chen, Xiangfan Xu and Y. Pei, Chem Mater, 2016, 28, 6227-6232. 16. W. Li, S. Lin, B. Ge, J. Yang, W. Zhang and Y. Pei, Advanced Science, 2016, 3, 1600196. 17. S. Lin, W. Li, S. Li, X. Zhang, Z. Chen, Y. Xu, Y. Chen and Y. Pei, Joule, 2017, 1, 816-830. 18. Z. Chen, X. Zhang and Y. Pei, Advanced materials, 2018, 1705617. 19. G. CHEN, M. S. DRESSELHAUS, J.-P. FLEURIAL and T. CAILLAT, Anglais, 2003, 48, 45-66 20. S. Lin, W. Li, Z. Chen, J. Shen, B. Ge and Y. Pei, Nature communications, 2016, 7, 10287. 21.Y. Pei, X. Shi, A. LaLonde, H. Wang, L. Chen and G. J. Snyder, Nature, 2011, 473, 66-69. 22.Y. Pei, H. Wang, Z. M. Gibbs, A. D. LaLonde and G. J. Snyder, NPG Asia Materials 2012, 4, e28. 23. Y. Pei, A. D. LaLonde, N. A. Heinz, X. Shi, S. Iwanaga, H. Wang, L. Chen and G. J. Snyder, Advanced materials, 2011, 23, 5674-5678. 24. J. Shen, Z. Chen, L. Zheng, W. Li and Y. Pei, Journal of Materials Chemistry C, 2016, 4, 209-214. 25. W. Li, Z. Chen, S. Lin, Y. Chang, B. Ge, Y. Chen and Y. Pei, Journal of Materiomics, 2015, 1, 307-315. 26. W. Li, Y. Wu, S. Lin, Z. Chen, J. Li, X. Zhang, L. Zheng and Y. Pei, ACS Energy Letters, 2017, 2, 2349-2355. 27.Linglang Zheng, Wen Li, Siqi Lin, Juan Li, Zhiwei Chen and Y. Pei, ACS Energy Letters, 2017, 2, 563-568. 28. Wen Li, Linglang Zheng, Binghui Ge, Siqi Lin, Xinyue Zhang, Zhiwei Chen, Yunjie Chang and Y. Pei, Adv. Mater., 2017, 29, 1605887. 29. Juan Li, Zhiwei Chen, Xinyue Zhang, Yongxing Sun, Jiong Yang and Y. Pei, NPG Asia Mater, 2017, 9, e353. 30. W. Liu, X. J. Tan, K. Yin, H. J. Liu, X. F. Tang, J. Shi, Q. J. Zhang and C. Uher, Phys Rev Lett, 2012, 108, 166601. 31. M. Inoue, Y. Takai, K. Fukui, H. Yagi and T. Tatsukawa, J Magn Magn Mater, 1983, 36, 255-258. 32. C. G. Fu, T. J. Zhu, Y. Z. Pei, H. H. Xie, H. Wang, G. J. Snyder, Y. Liu, Y. T. Liu and X. B. Zhao, Adv. Energy Mater., 2014, 4, 1400600. 33.Y. Xu, W. Li, C. Wang, J. Li, Z. Chen, S. Lin, Y. Chen and Y. Pei, Journal of Materials Chemistry A, 2017, 5, 19143-19150. 34. Y. Ren, J. Yang, Q. Jiang, D. Zhang, Z. Zhou, X. Li, J. Xin and X. He, Journal of Materials Chemistry, 2017, 00, 1-9. 35. G. S. Mann and L. H. V. Vlack, Physical Chemistry, 1977, 8, 53-57. 36. M. N. Mamonntov, Inorganic materials, 1996, 32, 1190-1194. 37. Y. Mita, Y. Ishida, M. Kobayashi and S. Endo, Acta Physica Polonica A, 2008, 113, 617-620. 38. M. T. Heinilä and A. S. Oja, Physical Review B, 1994, 49, 11995-12002.
AC C
EP
TE D
M AN U
elements, such as Fe, Al, Ga, In, La, Co, Ni, Sn, Pb at Mn site and I at Te site, turn out to be unsuccessful for a sufficient ACCEPTED high electron concentration in this work. This opens a question to the field for a realization of possibly high thermoelectric performance in n-type MnTe2. In summary, this work reveals a new promissing thermoelectric material MnTe2. The broad carrier concentration ranging from ~2×1019 cm-3 to ~4×1020 cm-3, enabled by Ag-doping, not only optimizes the hole carrier concentration but also reduces the lattice thermal conductivity, both of which contribute to the peak zT as high as 0.7. The broad carrier concentration further enables a systematical analysis of elemental material properties determining thermoelectric performance by the SPB model with acoustic scattering. The band structure calculations indicate an even higher thermoelectric performance in n-type MnTe2 due to its high band degeneracy and low band mass, which deserves further investigation. This work demonstrates MnTe2 as a promising matrix material for high-efficient thermoelectric applications. Acknowledgements This work is supported by the National Natural Science Foundation of China (Grant No. 11474219 and 51772215), and the national Recruitment Program of Global Youth Experts (1000 Plan) and the Fok Ying Tung Education Foundation (Grant No. 20170072210001). CW and YC acknowledge the financial support from the Early Career Scheme of RGC under Project Number 27202516 and the research computing facilities offered by ITS, HKU. References 1. K. F. Hsu, S. Loo, F. Guo, W. Chen, J. S. Dyck, C. Uher, T. Hogan, E. K. Polychroniadis and M. G. Kanatzidis, Science, 2004, 303, 818-821. 2. K. Biswas, J. He, I. D. Blum, C.-I. Wu, T. P. Hogan, D. N. Seidman, V. P. Dravid and M. G. Kanatzidis, Nature, 2012, 489, 414-418. 3. B. Poudel, Q. Hao, Y. Ma, Y. C. Lan, A. Minnich, B. Yu, X. A. Yan, D. Z. Wang, A. Muto, D. Vashaee, X. Y. Chen, J. M. Liu, M. S. Dresselhaus, G. Chen and Z. F. Ren, Science, 2008, 320, 634-638. 4. Y. Pei, J. Lensch-Falk, E. S. Toberer, D. L. Medlin and G. J. Snyder, Advanced Functional Materials, 2011, 21, 241-249. 5. L.-D. Zhao, S.-H. Lo, Y. Zhang, H. Sun, G. Tan, C. Uher, C. Wolverton, V. P. Dravid and M. G. Kanatzidis, Nature, 2014, 508, 373-377. 6. D. T. Morelli, V. Jovovic and J. P. Heremans, Phys Rev Lett, 2008, 101, 035901. 7. H. Liu, X. Shi, F. Xu, L. Zhang and W. Zhang, Nature materials, 2012, 11, 422-425. 8. H. Liu, X. Yuan, P. Lu, X. Shi, F. Xu, Y. He, Y. Tang, S. Bai, W. Zhang, L. Chen, Y. Lin, L. Shi, H. Lin, X. Gao, X. Zhang, H. Chi and C. Uher, Advanced materials, 2013, 25, 6607-6612. 9. S. I. Kim, K. H. Lee, H. A. Mun, H. S. Kim, S. W. Hwang, J. W. Roh, D. J. Yang, W. H. Shin, X. S. Li and Y. H. Lee, Science, 2015, 348, 109-114. 10. Zhiwei Chen, Binghui Ge, Wen Li, Siqi Lin, Jiawen Shen, Yunjie Chang, Riley Hanus, G. Jeffrey Snyder and Y. Pei, Nature Communications, 2017, 8, 13828. 11. Zhiwei Chen, Zhengzhong Jian, Wen Li, Yunjie Chang, Binghui Ge, Riley Hanus, Jiong Yang, Yue Chen, Mingxin Huang, Gerald Jeffrey Snyder and Y. Pei, Adv. Mater., 2017,
4
SC
RI PT
the Less Common Metals, 1965, 8, 272-287. 52. S. L. Dudarev, G. A. Botton, S. Y. Savrasov, C. J. Humphreys MANUSCRIPT and A. P. Sutton, Physical Review B, 1998, 57, 1505-1509. 53. Z. M. Gibbs;, A. Lalonde; and G. J. Snyder, New Journal of Physics, 2013, 15, 075020. 54. Y. I. Ravich, B. A. Efimova and I. A. Smirnov, Semiconducting Lead Chalcogenides, Plenum Press, New York, 1970. 55.W. Xie, S. Populoh, K. Gałązka, X. Xiao, L. Sagarna, Y. Liu, M. Trottmann, J. He and A. Weidenkaff, Journal of Applied Physics, 2014, 115, 103707. 56. D. G. Cahill and R. O. Pohl, Annu. Rev. Phys. Chem., 1988, 39, 93-121. 57. J. Callaway and H. C. Vonbaeyer, Physical Review, 1960, 120, 1149-1154. 58. P. G. Klemens, Physical Review, 1960, 119, 507-509. 59. C. G. Fu, S. Q. Bai, Y. T. Liu, Y. S. Tang, L. D. Chen, X. B. Zhao and T. J. Zhu, Nature Communications, 2015, 6, 7. 60. A. D. LaLonde, Y. Pei and G. J. Snyder, Energy & Environmental Science, 2011, 4, 2090–2096. 61. L.-D. Zhao, G. Tan, S. Hao, J. He, Y. Pei, H. Chi, H. Wang, S. Gong, H. Xu and V. P. Dravid, Science, 2016, 351, 141-144. 62.Y. Pei, X. Shi, A. LaLonde, H. Wang, L. Chen and G. J. Snyder, Nature, 2011, 473, 66-69. 63. Y. Tang, Z. M. Gibbs, L. A. Agapito, G. Li, H.-S. Kim, M. B. Nardelli, S. Curtarolo and J. Snyder, Nat Mater, 2015, 14, 1223-1228.
AC C
EP
TE D
M AN U
39. O. O. and M. T., Journal of the Physical Society of Japan, 1977, 43, 343-344. 40. J. M. Hastings, L. M. Corliss, M. Blume andACCEPTED M. Pasternak, Physical Review B, 1970, 1, 3209-3210. 41. M. Pasternak and A. L. Spijkervet, Physical Review, 1969, 181, 574-579. 42. P. Burlet, E. Ressouche, B. Malaman, R. Welter, J. P. Sanchez and P. Vulliet, Physical Review B, 1997, 56, 14013-14018. 43. T. Chattopadhyay and H. Fjellvag, Physics Letters A, 1987, 120, 44-46. 44. J. M. Hastings, L. M. Corliss, W. Kunnmann and D. Mukamel, Physical Review B, 1986, 33, 6326-6330. 45. K. V. Kaznacheyev, T. Muro, T. Matsushita, T. Iwasaki, Y. Kuwata, H. Harada, S. Suga, H. Ishii, T. Miyahara and T. Mizokawa, Physical Review B, 1998, 58, 13491-13497. 46. A. Sawaoka and S. Miyahara, Journal of the Physical Society of Japan, 1965, 20, 2087. 47.B. Muller and H. D. Lutz, Solid State Communications, 1991, 78, 469-471. 48. P. Vulliet, J. P. Sanchez, D. Braithwaite, M. Amanowicz and B. Malaman, Physical Review B, 2001, 63, 184403. 49. H. Fjellvag, W. A. Grosshans, W. Honle and A. Kjekshus, Journal of magnetism and Magnetic Materials, 1995, 145, 118-124. 50. H. Fjellvag and A. Kjekshus, Physics Letters A, 1985, 112, 411-413. 51. W. D. Johnston, R. C. Miller and D. H. Hamon, Journal of
5
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Yanzhong Pei has been working on advanced thermoelectric semiconductors for more than a decade, from synthesizing the materials to understanding the underlying physics and chemistry. He holds a B. E. from Central South University in China, a Ph. D. from the Shanghai Institute of Ceramics, CAS, and postdoctoral research experience of around 5 years from Michigan State University and the California
AC C
EP
TE D
M AN U
SC
RI PT
Institute of Technology. He became a National 1000 Plan junior professor in 2012 at the School of Materials Science and Engineering, Tongji University, Shanghai, China. His interests are focused on materials physics and chemistry for energy applications.