LaPtSb: a half-Heusler compound with high thermoelectric

performance
Q. Y. Xue, H. J. Liu*, D. D. Fan, L. Cheng, B. Y. Zhao, J. Shi

Key Laboratory of Artificial Micro- and Nano-Structures of Ministry of Education and School of

Physics and Technology, Wuhan University, Wuhan 430072, China

The electronic and transport properties of the half-Heusler compound LaPtSb are investigated

by performing first-principles calculations combined with semi-classical Boltzmann theory and

deformation potential theory. Compared with many typical half-Heusler compounds, the LaPtSb

exhibits obviously larger power factor at room temperature, especially for the n-type system.

Together with the very low lattice thermal conductivity, the thermoelectric figure of merit (ZT) of

LaPtSb can be optimized to a record high value of 2.2 by fine tuning the carrier concentration.

1. Introduction
Owing to serious energy crisis and environment pollution, the demand for
sustainable and clean energy source becomes more and more important.
Thermoelectric materials can directly convert waste heat into electricity, which have
attracted intensive attention from the science and industry communities. The
performance of thermoelectric materials is usually determined by the dimensionless
figure of merit

S 2σ
ZT = T, (1)
κl + κe

where S , σ , T , κ l and κ e are the Seebeck coefficient, the electrical

conductivity, the absolute temperature, the lattice and electronic thermal conductivity,
respectively. An ideal thermoelectric material behaves as phonon-glass

electron-crystal (PGEC), which requires maximize the power factor ( S 2σ ) and/or

*
Author to whom correspondence should be addressed. Electronic mail: [email protected].
1
minimize the thermal conductivity ( κ l + κ e ). However, these transport coefficients ( S ,

σ , κ e ) are usually coupled with each other, and it is still a great challenge to obtain a

high ZT value, especially for the bulk materials.

Ternary half-Heusler (HH) system is a class of intermetallic compounds, some of
which are potential thermoelectric materials because of their narrow band gaps, high
temperature stability, large Seebeck coefficients, and moderate electrical
conductivities [1-5]. However, most HH compounds were found to exhibit higher
thermal conductivity in the order of magnitude of 10 W/mK, which is undesirable for
high performance thermoelectric materials [6]. Over the past two decades, the
thermoelectric properties of a few HH compounds have been extensively investigated,
especially for the (Ti, Zr, and Hf)NiSn-based and (Ti, Zr, and Hf)CoSb-based
compounds [1-17]. Among them, Chen et al. reported that a ZT of ~1.2 can be
achieved around 900 K for n-type Hf0.6Zr0.4NiSn0.995Sb0.005 alloys without the need of
nanostructures [16]. They further found that a ZT of ~1.3 can be achieved for
n-type Hf0.65Zr0.25Ti0.15NiSn0.995Sb0.005/nano-ZrO2 near 850 K [17]. Fu et al. found that
the p-type FeNb0.88Hf0.12Sb and FeNb0.86Hf0.14Sb alloys exhibit a ZT of about 1.5 at
1200 K [15]. However, these ZT values of HH compounds remain low compared
with the target value of about 3.0, which can compete with the efficiency of
conventional refrigerators or power generators. To achieve higher ZT value, an
efficient strategy is to search thermoelectric materials with intrinsically low thermal
conductivities. Recently, Carrete et al. [18] theoretically investigated the thermal
conductivities of a large number of HH compounds and found that the compound
LaPtSb has a particularly low lattice thermal conductivity of 1.72 W/mK at room
temperature. It is therefore interesting to ask whether such compound can exhibit
better thermoelectric performance. In this work, the thermoelectric properties of
LaPtSb are predicted by multiscale calculations which combine the first-principles
calculations, Boltzmann theory, and deformation potential theory. We demonstrate
that by fine tuning the carrier concentration of LaPtSb, a maximal ZT value of 2.2
can be achieved for n-type system at room temperature, which makes it a very
2
promising candidate for high performance thermoelectric materials.

2. Computational Methods
The structure optimization and electronic structure calculations of LaPtSb are
performed by using the projector-augmented wave (PAW) method [ 19 ], as
implemented in the Vienna ab initio simulation package (VASP) [20, 21, 22]. The
treatment of exchange and correlation energy is adopted in the form of
Perdew-Burke-Ernzerhof (PBE) [23] with the generalized gradient approximation
(GGA). It’s necessary to add a Hubbard-type term U [24] to act on the La 5d and Pt
5d states, which can greatly improve the quality of PBE and give the correct
electronic structure [25]. We use an effective Hubbard U of 8.1 eV for La and 3.0 eV
for Pt, as obtained from the AFLOWLIB.org consortium repository [26]. The plane
wave cut-off energy is set to be 400 eV, and the Brillouin zone is sampled with a
Monkhorst-Pack k-mesh of 15×15×15. Optimal atomic positions are determined until
the magnitude of the forces acting on every atom becomes less than 0.01 eV/Å.
The electronic transport coefficients are evaluated by using the semi-classical
Boltzmann transport theory [27] with relaxation time approximation. In terms of the
KK
so-called transport distribution function Ξ(ε ) = ∑
K
vkK vkKτ kK , the Seebeck coefficient S
k

and the electrical conductivity σ are given by

ek B ⎛ ∂f 0 ⎞ ε −µ
S=
σ ∫ d ε ⎜⎝ − ∂ε ⎟⎠Ξ (ε ) k BT
, (2)

⎛ ∂f 0 ⎞
σ = e2 ∫ d ε ⎜ − ⎟Ξ ( ε ) , (3)
⎝ ∂ε ⎠

Here f 0 is the equilibrium Fermi-Dirac distribution function, µ is the chemical

K K
potential, k B is the Boltzmann’s constant, and vkK is the group velocity at state k .

The calculations of transport coefficients are performed in the so-called BoltzTraP

code [27]. The relaxation time τ kK is estimated by the deformation potential (DP)

theory as proposed by Bardeen and Shockley [28]. In addition, the electronic thermal

3
conductivity κ e is calculated via the Wiedemann-Franz law [29]

κ e = Lσ T , (4)

Here the Lorenz number L can be obtained by [30, 31]

ke ⎛ k B ⎞ ⎡ 3F2 ⎛ 2 F1 ⎞ ⎤
2 2

L= = ⎢ −⎜ ⎟ ⎥, (5)
σ T ⎜⎝ e ⎟⎠ ⎢ F0 ⎝ F0 ⎠ ⎥
⎣ ⎦
∞ i
where Fi is a function of the reduced Fermi energy ς given by Fi (ς ) = ∫ ( xx−ςdx .
0
)
e +1

3. Results and Discussions

HH compounds crystallize into the cubic MgAgAs-type structure [32] with the

space group F 43m [33]. Figure 1 shows the unit cell of the HH compound LaPtSb,

where La atoms are located at the position of 4a(0,0,0), Pt atoms at the 4c(1/4,1/4,1/4),
and Sb atoms at the 4b(1/2,1/2,1/2) in Wyckoff coordinates. The structure can also be
regarded as four interpenetrating fcc lattices, containing a lattice of La atoms, a lattice
of Pt atoms, a lattice of Sb atoms, and a lattice of vacancies. The La and Sb atoms
form a rock salt structure. The optimized lattice constant of LaPtSb is 6.87 Å. The
total valence electron count (VEC) in the primitive cell satisfies the criterion of
VEC=18 [34], as generally found in HH compounds.
We first focus on the electronic band structure of LaPtSb, which is shown in Figure
2 along the high-symmetry lines of the irreducible Brillouin zone (IBZ). We can see
that both the valence band maximum (VBM) and the conduction band minimum
(CBM) are located at Γ point, resulting in a direct band gap of 0.23 eV. Moreover, we
find that the energy band near CBM is rather sharp, which indicates a quite small
effective mass. In contrast, the energy band near VBM is relatively flat, and has a
threefold degeneracy. All these observations will play a significant role in the
transport properties of LaPtSb, as discussed in the following.
To evaluate the electronic transport coefficients within the framework of
Boltzmann theory, the relaxation time has to be determined which is usually very
complicated since it depends on the detailed scatting mechanisms. As the wave length
4
of thermally activated carriers at room temperature is much larger compared with the
lattice constant, the scattering of carriers is dominated by the electron-acoustic
phonon coupling [35]. In this respect, we evaluate the relaxation time with the help of
deformation potential theory, which can effectively describe the electron-acoustic

phonon interactions. For three-dimensional system, the relaxation time along the β

direction at temperature T can be expressed as:

2 2π Cβ = 4
τβ = , (6)
3(k BTm dos* )3/2 Eβ2

1 ∂2 E
In this formula, Cβ is the elastic constant defined as Cβ = , where
V0 ∂ (∆l / l0 ) 2 l = l0

E is the total energy of the system, l0 is the lattice constant along the direction of

β , ∆l = l − l0 is the corresponding lattice distortion, and V0 is the equilibrium

volume of the unit cell. The deformation potential constant Eβ is calculated as

∂Eedge
Eβ = , which represents the shift of band edges (VBM or CBM) per unit
∂ (∆l ) / l0
1
strain. The density-of-states effective mass m *
dos is given by m *
dos = (m m m
*
x
*
y z)
* 3
,

where m*x , m*y , and m*z are the effective mass along the x, y, and z directions,

respectively. These three quantities ( Cβ , Eβ , m dos* ) can be readily obtained from

first-principles calculations. It should be noted that we must consider the band

degeneracy mentioned above when calculating the m dos* [36]. The derived relaxation

time of LaPtSb compound is summarized in Table I for both p- and n-type carriers. It
is obvious to find that the relaxation time of electrons is much larger than that of holes,
which mainly originates from a very small density-of-states effective mass of the
former, as can be also seen from the dispersion relations around CBM and VBM in
Fig. 2. Note that such significant difference in the relaxation time between electrons
and holes has been also found in other systems such as few-layer black phosphorus
and phosphorene [37, 38]. On the other hand, the relaxation time of both electrons
5
and holes are larger than those of typical thermoelectric materials such as Bi2Te3 [39]
and MoS2 [35], which suggests the favorable thermoelectric performance of LaPtSb
compound.
In Figure 3, we plot the calculated Seebeck coefficient S , the electrical

conductivity σ , and the power factor S 2σ of LaPtSb as a function of carrier

concentration at room temperature. We can see that the absolute values of Seebeck
coefficient decreases with increasing carrier concentration and becomes vanishing
when the concentration is larger than 1021 cm−3 for both p-type and n-type systems.
On the other hand, we find that the electrical conductivity of n-type system is much
larger than that of p-type system at the same carrier concentration, which is caused by
the obvious difference between the relaxation time of electrons and holes. Comparing
Fig. 3(a) with 3(b), we note there is a sharp increase of the electrical conductivity at
the carrier concentration where the Seebeck coefficient has very small absolute value.
It is thus natural to find a balance between them so that the power factor can be
maximized by fine tuning the carrier concentration. Indeed, we see from Fig. 3(c) that
the power factor of LaPtSb can reach peak values at particular concentration for both
electrons and hole. For example, at a concentration of 2.7×1020 cm−3 , the power
factor of n-type system can be optimized to as high as 0.43 W/mK2, which is much
larger than those of many traditional thermoelectric materials such as Bi2Te3 [39].
Such characteristic suggests again that the HH compound LaPtSb could have
favorable thermoelectric performance.
To evaluate the ZT value of LaPtSb compound, one also needs to know the

thermal conductivity, which includes the contributions from both phonons ( κ l ) and

charge carriers ( κ e ). Using the Wiedemann-Franz Law given by Equation (4), the

electronic thermal conductivity can be readily obtained and is shown in Fig. 3(d) as a

function of carrier concentration at room temperature. We see that the behavior of κ e

is basically the same as that of electrical conductivity shown in Fig. 3(b). This is
reasonable since the Lorenz number calculated from Equation (5) has little change

6
with the carrier concentration. For the phonon part, we take the result of Carrete et al.
where a room temperature thermal conductivity of 1.72 W/mK is predicted by solving
phonon Boltzmann transport equation [18]. Such a low value of thermal conductivity
is very desirable for thermoelectric applications when compared with other HH
compounds.
We can now evaluate the thermoelectric performance of LaPtSb compound by
inserting all the transport coefficients into Equation (1). Figure 4 plots the room
temperature ZT values as a function of carrier concentration. For the p-type system,
a maximal ZT of 1.3 appears at the concentration of 1.2×1019 cm−3. The ZT value
can be further enhanced to 2.2 for the n-type system when the carrier concentration is
tune to 3.3×1019 cm−3. It should be mentioned that such a ZT value exceeds those of
many good thermoelectric materials, and in particular the typical (Ti, Zr, and
Hf)NiSn-based and (Ti, Zr, and Hf)CoSb-based HH compounds. In the case of
rare-earth based HH compounds, although there is currently no experimental study on
LaPtSb, we note that the thermoelectric properties of several other rare-earth based
HH systems, such as LnPdSb (Ln=Ho, Er, Dy), ScMSb (M=Ni, Pd, Pt), YMSb
(M=Ni, Pd, Pt), ErPdX (X=Sb, Bi), PtYSb, and YNiBi have been experimentally
reported [40-45]. Compared with the LaPtSb predicted in the present work, these HH
compounds exhibit relatively smaller power factors and larger lattice thermal
conductivities. As a result, the thermoelectric performances are less favorable than
that of LaPtSb.

4. Summary
In summary, our theoretical work demonstrates that the HH compound LaPtSb
could be optimized to exhibit very good thermoelectric performance, which is
believed to be benefited from the relatively large power factor and low lattice thermal
conductivity. At a moderate carrier concentration of 3.3×1019 cm−3, the room
temperature ZT value can be enhanced to as high as 2.2 for n-type system, which
exceeds that of the HH compounds reported so far. Our calculations suggest that the
n-type LaPtSb is a very promising thermoelectric material and major efforts should be
7
directed to developing comparatively effective p-type LaPtSb so that efficient
thermoelectric modules based on this HH compound could be realized.

Acknowledgements
This work was supported by the National Natural Science Foundation (Grant No.
11574236 and 51172167) and the “973 Program” of China (Grant No.
2013CB632502).

8
Table I The deformation potential constant E1 , the elastic constant C , the

*
density-of-states effective mass mdos , and the relaxation time τ of LaPtSb at room

temperature.

Carrier type E1 (eV) C *

mdos ( me ) τ (fs)

electron 21.73 1.45 0.03 2063

Hole 14.43 1.45 0.25 194

9
Figure 1 (Color online) The unit cell of the half-Heusler compound LaPtSb.

10
Figure 2 The calculated energy band structure of LaPtSb.

11
Figure 3 The calculated electronic transport coefficients of LaPtSb: (a) the Seebeck

coefficient S , (b) the electrical conductivity σ , (c) the power factor S 2σ , and (d)

the electronic thermal conductivity κ e as a function of carrier concentration at room

temperature. Negative and positive carrier concentrations represent n- and p- type

carriers, respectively.

12
Figure 4 The calculated ZT values of LaPtSb as a function of carrier concentration
at room temperature.

13
References

[1] H. Hohl, A. P. Ramirez, C. Goldmann, G. Ernst, B. Wlfing, and E. Bucher, J. Phys.

Condens. Matter 11, 1697 (1999).
[2] Q. Shen, L. Chen, T. Goto, T. Hirai, J. Yang, G. P. Meisner, and C. Uher, Appl.
Phys. Lett. 79, 4165 (2001).
[3] G. S. Nolas, J. Poon, M. Kanatzidis, and MRS Bull. 31, 199 (2006).
[4] J. Yang , H. Li , T. Wu , W. Zhang , L. Chen , and J. Yang , Adv. Funct. Mater. 18,
2880 (2008).
[5] C. Yu, T. J. Zhu, R. Z. Shi, Y. Zhang, X. B. Zhao, and J. He, Acta Mater. 57, 2757
(2009).
[6] W. Xie, A. Weidenkaff, X. Tang, Q. Zhang, S. J. Poon, and T. M. Tritt,
Nanomaterials 2, 379 (2012).
[7] S. Bhattacharya, A. L. Pope, R. T. Littleton IV, T. M. Tritt, V. Ponnambalam, Y.
Xia, and S. J. Poon, Appl. Phys. Lett. 77, 2476 (2000).
[8] S. Sakurada and N. Shutoh, Appl. Phys. Lett. 86, 082105 (2005).
[9] S. R. Culp , S. J. Poon , N. Hickman , T. M. Tritt , and J. Blumm , Appl. Phys.
Lett. 88, 042106 (2006).
[10] S. R. Culp, J. W. Simonson, S. J. Poon, V. Ponnambalam, J. Edwards, and T. M.
Tritt, Appl. Phys. Lett. 93, 022105 (2008).
[11] X. Yan, G. Joshi, W. S. Liu, Y. C. Lan, H. Wang, S. Lee, J. W. Simonson, S. J.
Poon, T. M. Tritt, G. Chen, and Z. F. Ren, Nano Lett. 11, 556 (2011).
[12] G. Joshi, X. Yan, H. Wang, W. Liu, G. Chen, and Z. F. Ren, Adv. Energy Mater. 1,
643 (2011).
[13] S. Chen , K. C. Lukas , W. Liu , C. P. Opeil , G. Chen , Z. F. Ren , Adv. Energy
Mater. 3, 1210 (2013).
[14] C. G. Fu , T. J. Zhu , Y. T. Liu , H. H. Xie , and X. B. Zhao , Energy Environ. Sci.
8, 216 (2015).
[15] C. G. Fu, S. Q. Bai, Y. T. Liu, Y. S. Tang, L. D. Chen, X. B. Zhao and T. J. Zhu,
Nat Commun, 6, 8144 (2015).
[16] L. Chen, S. Gao, X. Zeng, A. Mehdizadeh Dehkordi, T. M. Tritt, and S. J. Poon
Appl. Phys. Lett. 107, 041902 (2015).
[17] L. Chen, X. Zeng, T. M. Tritt, and S. J. Poon, arXiv:1604.05397 (2016).
[18] J. Carrete, W. Li, and N. Mingo, Phys. Rev. X 4, 011019 (2014).
14
[19] P. E. Blochl, Phys. Rev. B 50, 17953 (1994).
[20] G. Kresse and J. Hafner, Phys. Rev. B 47, 558 (1993).
[21] G. Kresse and J. Hafner, Phys. Rev. B 49, 14251 (1994).
[22] G. Kresse and J. Furthmüller, Comput. Mater. Sci. 6, 15 (1996).
[23] J. P. Perdew, K. Burke, and M. Ernzerhof, Phys. Rev. Lett. 77, 3865 (1996).
[24] A. I. Liechtenstein, V. I. Anisimov, and J. Zaane, Phys. Rev. B 52, R5467 (1995).
[25] P. Larson , S. D. Mahanti , S. Sportouch , and M. G. Kanatzidis , Phys. Rev. B 59,
15660 (1999).
[26] S. Curtarolo, W. Setyawan, S. Wang, J. Xue, K. Yang, R. H. Taylor, L. J. Nelson,
G. L. W. Hart, S. Sanvito, M. Buongiorno Nardelli, N. Mingo, and O. Levy, Comput.
Mater. Sci. 58, 227 (2012).
[27] G. K. H. Madsen and D. J. Singh, Comput. Phys. Commun. 175, 67 (2006).
[28] J. Bardeen and W. Shockley, Phys. Rev. 80, 72 (1950).
[29] C. Kittel, Introduction to Solid State Physics (John Wiley and Sons, New York,
(2005).
[30] L. D. Hicks and M. S. Dresselhaus, Phys. Rev. B 47, 16631 (1993).
[31] L. D. Hicks and M. S. Dresselhaus, Phys. Rev. B 47, 12727 (1993).
[32] W. Jeitschko, Metall. Trans. 1, 3159 (1970).
[33] A. Neumann, L. Offernes, A. Kjekshus, B. Klewe, J. Alloys Compd. 274, 136
(1998).
[34] B. R. K. Nanda, and I. Dasgupta, J. Phys.: Condens. Matter. 15, 7307 (2003).
[35] Y. Q. Cai, G. Zhang, and Y. W. Zhang, J. Am. Chem. Soc. 136, 6269 (2014).
[36] M. A. Green, J. Appl. Phys. 67, 2944 (1990).
[37] J. Qiao, X. Kong, Z.-X. Hu, F. Yang, and W. Ji, Nat. Commun. 5, 4475 (2014).
[38] J. Zhang, private communication.
[39] T. J. Scheidemantel, C. Ambrosch-Draxl, T. Thonhauser, J. V. Badding, and J. O.
Sofo, Phys. Rev. B 68, 125210 (2003).
[40] K. Mastronardi, D. Young, C.-C. Wang, P. Khalifah, R. J. Cava, and A. P.
Ramirez, Appl. Phys. Lett. 74, 1415 (1999).
[41] J. Oestreich, U. Probst, F. Richardt, and E. Bucher, J. Phys.: Condens. Matter 15,
635 (2003).
[42] T. Sekimoto, K. Kurosaki, H. Muta, and S. Yamanaka, J. Appl. Phys. 99, 103701
(2006).
15
[43] S. Ouardi, G. H. Fecher, C. Felser, J. Hamrle, K. Postava, and J. Pištora, Appl.
Phys. Lett. 99, 211904 (2011).
[44] G. H. Li, K. Kurosaki, Y. Ohishi, H. Muta, and S. Yamanaka, Jpn. J. Appl. Phys.
52, 041804 (2013).
[45] S. M. Li, H. Z. Zhao, D. D. Li, S. F. Jin, and L. Gu, J. Appl. Phys. 117, 205101
(2015).

16