High temperature thermoelectric properties of dy doped camno3 ceramics

The energy conversion efficiency is related to the materials intrinsic prop-erties which can be characterized by the dimensionlessfigure of merit ZT¼ S2sT/k, where T is the absolute temper

High Temperature Thermoelectric Properties of Dy-doped

Bin Zhan1), Jinle Lan1), Yaochun Liu2), Yuanhua Lin1)*, Yang Shen1), Cewen Nan1)

1) State Key Laboratory of New Ceramics and Fine Processing, School of Materials Science and Engineering,

Tsinghua University, Beijing 100084, China

2) School of Materials Science and Engineering, University of Science and Technology Beijing,

Beijing 100083, China

[Manuscript received May 22, 2013, in revised form June 26, 2013, Available online 25 January 2014]

Dy-doped CaMnO3ceramics have been synthesized by co-precipitation method combined with the solid-state reaction Phase composition and microstructure analysis indicate that high density and pure CaMnO3 phase can be achieved The electric conductivity can be enhanced by Dy doping, and result in a slight increase of the thermal conductivity The highest dimensionless figure of meritZT of 0.15 has been obtained at 973 K for

x ¼ 0.02 sample, which is about 4 times larger than that of the pure CaMnO3, which indicate that CaMnO3 can be a promising candidate forn-type thermoelectric material at high temperature

KEY WORDS: CaMnO 3 ; Thermoelectric properties; Thermal conductivity

1 Introduction

Environment friendly thermoelectric materials have attracted

widespread interests for potential applications in space

explora-tion, exhaust recycling, and clean cooling[1e3] The energy

conversion efficiency is related to the materials intrinsic

prop-erties which can be characterized by the dimensionlessfigure of

merit ZT¼ S2sT/k, where T is the absolute temperature, S is the

Seebeck coefficient,sis the electrical conductivity, andkis the

thermal conductivity The three parameters S, s, and k are

correlated to each other Therefore, it is difficult to optimize all

parameters simultaneously The good thermoelectric

perfor-mance needs high power factor (PF, S2s), and low thermal

conductivityk Normally, alloy semiconductors, such as Bi2Te3,

PbTe and SiGe[4e7], their ZT values can exceed 1.0, and show a

good practical prospects For oxides-based thermoelectrics, their

high chemical and thermal stability makes them to be promising

candidates at high temperature application Recently, various

oxides such as Ca3Co4O9[8,9], SrTiO3[10], ZnO[11] and

BiCu-SeO[12,13] have been investigated in detail to enhance the ZT

value

CaMnO3 (CMO) is a typically n-type oxide thermoelectric

materials[14e18] However, low electric conductivity leads to

poor performance in CMO system Thesof pure CMO is just w10 S/cm and ZT is less than 0.04 at 900 K In the previous work, the polycrystalline ceramics of La-doped CaMnO3were synthesized by conventional solid-state reaction (SSR)[14], and showed a porous structure Lan et al.[19]reported that the ZT value can been improved to 0.24 at 973 K though reducing the thermal conductivity infine grain size (200e400 nm) at a low sintering temperature Wang et al.[20]studied the electron-doped CaMnO3by rare earth, which shows that the Seebeck coefficient

is determined by the carrier concentration, while the electric conductivity and thermal conductivity can be tuned by the ion radius and ion mass, respectively

In this work, we attempted to control the microstructure and optimize thermoelectric properties The cold isostatic pressing (CIP) was used to achieve high density sample and Dy as a dopant to optimize the electrical properties Our results indicate CIP is available to enhance the density and Dy dopant is effectively to improve the ZT value

2 Experimental

Polycrystalline ceramic samples of Ca1 xDyxMnO3 (x ¼ 0, 0.02, 0.04, 0.06, 0.08, 0.10) were synthesized via a chemical co-precipitation method and solid-state reaction CIP was used to control the porosity For the co-precipitation method, Dy2O3

(99.90%) was carefully dissolved in nitric acid as a kind of starting material, and then mixed together with Ca(NO3)2$4H2O, Mn(NO3)2aqueous solution in deionized water to make the ni-trate stock solution NH4HCO3and ammonia solution were used

to control the reaction pH value in the range of 7.5e9.0 to make

* Corresponding author Prof., Ph.D.; Tel.: þ86 10 62773741; Fax: þ86

10 62771160; E-mail address: linyh@tsinghua.edu.cn (Y Lin).

1005-0302/$ e see front matter Copyright Ó 2014, The editorial office of

Journal of Materials Science & Technology Published by Elsevier

Limited All rights reserved.

http://dx.doi.org/10.1016/j.jmst.2014.01.002

ScienceDirect

J Mater Sci Technol., 2014, 30(8), 821e825

the metal ions precipitate completely The resultant suspension

was aged to remove supernatantfluid and then was subjected to

suctionfiltration The precursor powder was calcined at 1073 K

for 4 h in air, and then pressed to pellets The bulks sealed in

glove and compacted by CIP at 200 MPa in oil to enhance the

density (LDJ-100/320-300 Sichuan Airlines Industry Chuanxi

Machinery Factory, Yaan, China) The samples were sintered at

1473 K for 12 h

X-ray diffraction (XRD) with a Rigaku D/MAX-2550V

diffractometer (Rigaku, Tokyo, Japan; CuKa radiation) and

scanning electron microscopy (SEM, JSM-6460LV, JEOL,

Tokyo, Japan) were used to investigate the phase composition

and microstructure of CMO bulks, respectively The temperature

dependence of electric conductivity was measured from room

temperature to 973 K by a four-probe method Seebeck coef

fi-cient was obtained from the slope of the linear relation between

DV andDT, whereDV is the thermoelectromotive force produced

by the temperature gradientDT The thermal conductivitykwas

determined by the following parameters: the thermal diffusivity

(a), the heat capacity (Cp), and the density (r), using the

rela-tionshipk ¼aCpr The relative bulk density was measured by

the Archimedes method, and a Netzsch LFA 457 (Selb,

Ger-many) laser flash apparatus was used to measure the thermal

diffusivity and the specific heat

3 Results and Discussion

3.1 Morphology and structure

Fig 1(a) shows the XRD patterns for Dy-doped CMO

sam-ples All the samples are corresponded to CaMnO3phase with an

orthorhombic perovskite-type structure in Pnma space group

The lattice parameters have been calculated by the XRD data, as

shown inFig 1(b) The lattice constant a and cell volume are

increased as dopant concentration increases This behavior is

independent of the ionic radius of Dy[21,22], in spite of the radius

of Dy3þions (1.08 nm, C.N IX) is slightly smaller than that of

Ca2þions (1.18 nm, C.N IX) The electron doping will induce the presence of Mn3þwithin the Mn4þmatrix for charge bal-ance The ionic radius of Mn3þis larger than that of Mn4þ(r [Mn3þ]¼ 0.64 nm, C.N VI and r[Mn4þ]¼ 0.53 nm, C.N VI), which leads to the MnO6 octahedra to be distorted in CMO structure

Fig 2 shows the SEM images of microstructure of CMO samples The precursor powder is typical spherical and diameter changes from 1 to 10 mm as seen in Fig 2(a), which are composed of many small grains ofw100 nm in size as shown in the insert of Fig 2(a) Fig 2(b) shows the surface of sample without CIP, and some pores with diameter of 1e2mm appeared

As shown in Fig 2(c), the macroporous pores disappeared by CIP technology and the density of samples can increase to 95% after CIP as compared with the density 80% of samples without CIP It can be observed that the high density will deteriorate the thermal conductivity, which can reduce thermoelectric proper-ties ComparedFig 2(c) and (d), pores will further decrease with more dopant And the binding between grains are more closely, which indicates that the addition of Dy can act as the sintering aids and contribute to the formation of grains, which is helpful to make CMO ceramic be densified

3.2 Electric properties

Fig 3(a) shows the temperature dependence of electric con-ductivitysof samples from 300 to 973 K With increasing doping concentration,sgradually increases and reaches the largest value (184 S/cm) at x¼ 0.10 at 973 K As compared with undoped CMO (sw10 S/cm), a significant increase of electric conductivity can be observed by Dy doping, which is mainly derived from the varia-tion of carrier concentravaria-tion The substituvaria-tion of Dy3þfor Ca2þ will import a large number of electron carriers and induce Mn3þ appeared in Mn4þmatrix The increased carrier concentration is directly affect the s, and the presence of Mn3þis beneficial to electron hopping in perovskite manganites Therefore, Dy doping can facilitate the transport of carriers by hopping mechanism and then enhance the electric conductivity

The slope ofseT curve is positive relationship (ds/dT> 0) at low temperature, which indicates semiconducting behavior And then it shows a metallic behavior (ds/dT< 0) as the temperature further increases The insulator-metal (IM) transition temperature

TIMincreases from 400 to 550 K with the Dy content increasing For semiconductor part, the temperature dependence of the conductivity is generally described using the small polaron model given by Mott as the following equation[23]

Texp

Ea

kBT



;

where C, kB, and Ea are the pre-exponential terms, Boltzmann constant, and activation energy, respectively Fig 3(b) shows the activation energy increases with increasing content of Dy

at low temperature interval, which raising from 0.048 to 0.070 eV With more dopant, the densifying of ceramic may be beneficial to electronic transport The carrier concentration increases and more Mn3þ ions can be formed, and then a higher s and TIM can be obtained This indicates that the increase in Mn3þ concentration is favorable for the formation

of polaron in this temperature range[19]

Fig 3(c) displays the temperature dependence of Seebeck coefficient All samples exhibit negative Seebeck coefficient,

Fig 1 XRD patterns (a) and the lattice constant a and cell volume (b) of

which indicates that electrons are the predominant charge

car-riers (n-type conduction) The x¼ 0.02 sample has a very large S

value, being about370mV K1at 300 K The absolute value

of S decreased obviously as Dy content increasing, which arises

from the increase of carrier concentration For x¼ 0.02, absolute

value of Seebeck coefficient decreases with increasing

temperature, which shows a typical characteristic of nonmetal-like temperature dependence With more Dy doping, the abso-lute value of S decreases with increasing temperature and exhibit metallic behavior This difference should be attributed to the contribution of the oxygen deficiency[18,24] The turning point came in the x¼ 0.04 sample, which did not appear ins

Fig 3 Temperature dependence of electric conductivity (a), Seebeck coefficient (c), and power factor (d); and (b) activation energy plotted vs Dy content.

Fig 2 Typical SEM images of CaMnO 3 samples: (a) x ¼ 0.06 powder by co-precipitation route; (b) CMO ceramic without CIP; (c) x ¼ 0.02 ceramic by CIP; (d) x ¼ 0.06 ceramic by CIP.

For materials with more than one type of charge carrier, the

diffusion Seebeck coefficient can be expressed as

i

si

s



Si

where siand Si are the partial electrical conductivity and the

partial Seebeck coefficient associated with the ith group of

carriers, respectively We can rewrite S of CMO as

sinþsex;defactSinþ sex;defact

sinþsex;defactSex;defact

where sin and Sin are the contribution from intrinsic carriers;

sex,defect and Sex,defect are the contribution from extrinsic

carriers due to the oxygen defects Since the increase of

electrical conductivityðweE a =ðK B TÞÞ is faster than the decrease

of S (wEa/(KBT )) for semiconductors, one could expect the

second term in above equation would increase and therefore

the absolute value of Seebeck coefficient for CMO would

increase, which should be responsible for the simultaneous

increase of the electrical conductivity and absolute value of S

with increasing temperature at low temperature interval It

indicates that the existence of oxygen deficiency is important

for Seebeck coefficient

3.3 Thermoelectric properties

The temperature dependence of the power factor (PF) is

shown inFig 3(d) At 973 K, the maximum of PF can reach

3.82mW cm1K2when x¼ 0.02, which is a high value for a

kind of n-type oxide thermoelectric material.Fig 4(a) shows the

temperature dependence of thermal conductivity of CMO

sam-ples Thermal conductivitykcan be expressed generally by the

sum of lattice thermal conductivity (kl) and electronic thermal conductivity (ke) ask¼klþke Thekecan be calculated by the WiedemanneFranz lawke¼ LTs, where L is the Lorentz con-stant In order to simplify the calculation, the Lorentz constant Lo

(Lo ¼ p2

kB2/(3e2) ¼ 2.44  108 WUK2) was used The calculated ke in the entire temperature range is quite small as compared to kl (less than 15%), which ranges from 0.168 to 0.436 W m1K1at 973 Kklis the predominant component in thermal conductivity, and the variation ofkis mainly caused by the change ofklas shown in the insert ofFig 4(a) The lowest value ofk is 2.48 W m1K1at 973 K for x¼ 0.02 As we mentioned before, CIP process can increase the density, which is

an important reason for highk The variation of thermal diffusion coefficient is the dominant factor for this result, which indicates that CIP not only makes the ceramic be densified, but also changes some intrinsic properties of materials

As shown in Fig 4(b), the dimensionless ZT of

Ca1 xDyxMnO3was calculated The optimal ZT is 0.15 at 973 K when x¼ 0.02, which is about 4 times larger than that of the pure CaMnO3(w0.038) More detailed further work is desirable

to further enhance the thermoelectric performance of CaMnO3

4 Conclusion

In summary, high density Ca1 xDyxMnO3 (0  x  0.10) ceramic has been prepared and the microstructure and thermo-electric properties have been investigated Nanostructured pre-cursor powders were obtained by co-precipitation method and the bulks can reach a high density with CIP The electrical conductivity can be obviously improved by Dy doping The maximum power factor can reach 3.82mW cm1K2at 973 K

in sample x¼ 0.02 The highest dimensionless figure of merit ZT

of 0.15 has been obtained at 973 K in the air for

Ca0.98Dy0.02MnO3 Acknowledgments This work wasfinancially supported by the Ministry of Sci

& Tech of China through a 973 Project, under grant No 2013CB632506, the National Natural Science Foundation of China under Grant Nos 51025205 and 11234012, and the Specialized Research Fund for the Doctoral Program of Higher Education, under grant No 20120002110006

REFERENCES

[1] R Venkatasubramanian, E Siivola, T Colpitts, B O ’Quinn, Nature

413 (2001) 597 e602 [2] A.J Minnich, M.S Dresselhaus, Z.F Ren, G Chen, Energy En-viron Sci 2 (2009) 466 e479

[3] J.W Fergus, J Eur Ceram Soc 32 (2012) 525 e540 [4] J.R Sootsman, D.Y Chung, M.G Kanatzidis, Angew Chem Int Edit 48 (2009) 8616 e8639

[5] B Poudel, Q Hao, Y Ma, Y.C Lan, A Minnich, B Yu, X Yan, D.Z Wang, A Muto, D Vashaee, X.Y Chen, J.M Liu, M.S Dresselhaus, G Chen, Z.F Ren, Science 320 (2008)

634 e638 [6] J.P Heremans, V Jovovic, E.S Toberer, A Saramat, K Kurosaki,

A Charoenphakdee, S Yamanaka, G.J Snyder, Science 321 (2008)

554 e557 [7] X.W Wang, H Lee, Y.C Lan, G.H Zhu, G Joshi, D.Z Wang, J Yang, A.J Muto, M.Y Tang, J Klatsky, S Song, M.S Dresselhaus,

G Chen, Z.F Ren, Appl Phys Lett 93 (2008) 193121 [8] Y Wang, Y Shi, J.G Cheng, X.J Wang, W.H Su, J Alloy Compd.

477 (2009) 817 e821 Fig 4 Temperature dependence ofk(a) and ZT (b) value for Dy-doped

CMO.

[9] Y.H Lin, J Lan, Z.J Shen, Y.H Liu, C.W Nan, J.F Li, Appl Phys.

Lett 94 (2009) 072107.

[10] A Kikuchi, N Okinaka, T Akiyama, Scripta Mater 63 (2010)

407 e410

[11] M Ohtaki, K Araki, K Yamamoto, J Electron Mater 38 (2009)

1234 e1238

[12] J Li, J.H Sui, Y.L Pei, C Barreteau, D Berardan, N Dragoe,

W Cai, J.Q He, L.D Zhao, Energy Environ Sci 5 (2012)

8543 e8547

[13] Y Liu, L.D Zhao, Y.C Liu, J.L Lan, W Xu, F Li, B.P Zhang,

D Berardan, N Dragoe, Y.H Lin, C.W Nan, J.F Li, H.M Zhu,

J Am Chem Soc 133 (2011) 20112 e20115

[14] J.L Lan, Y.H Lin, A Mei, C.W Nan, Y Liu, B.P Zhang, J.F Li,

J Mater Sci Technol 25 (2009) 535 e538

[15] S Populoh, M Trottmann, M.H Aguire, A Weidenkaff, J Mater.

Res 26 (2011) 1947e1952.

[16] L Bocher, M.H Aguirre, D Logvinovich, A Shkabko, R Robert,

M Trottmann, A Weidenkaff, Inorg Chem 47 (2008) 8077 e8085 [17] C.L Wang, L Shi, X.M Xu, S.M Zhou, J.Y Zhao, Y.Q Guo, H.F Liu, L.F He, X Cai, G.J Xu, Appl Phys A 112 (2013)

1003 e1009 [18] A Bhaskar, C.J Liu, J.J Yuan, C.L Chang, J Alloy Compd 552 (2013) 236 e239

[19] J.L Lan, Y.H Lin, H Fang, A Mei, C.W Nan, Y Liu, S.L Xu,

M Peters, J Am Ceram Soc 93 (2010) 2121 e2124 [20] Y Wang, Y Sui, W Su, J Appl Phys 104 (2008) 093703 [21] H Muguerra, B Rivas-Murias, M Traianidis, C Marchal,

Ph Vanderbemden, B Vertruyen, C Henrist, R Cloots,

J Alloy Compd 509 (2011) 7710 e7716 [22] R.D Shannon, Acta Crystallogr A 32 (1976) 751 e767 [23] A.J Bosmana, H.J van Daala, Adv Phys 19 (1970) 1e117 [24] A Bhaskar, J.J Yuan, C.J Liu, J Electroceram 31 (2013) 124e128.