You are on page 1of 5

Journal of ELECTRONIC MATERIALS, Vol. 40, No.

5, 2011

DOI: 10.1007/s11664-011-1584-2
 2011 TMS

Effect of Sb Doping on the Thermoelectric Properties


of Mg2Si0.7Sn0.3 Solid Solutions
H.L. GAO,1 X.X. LIU,1 T.J. ZHU,1 S.H. YANG,1 and X.B. ZHAO1,2
1.State Key Laboratory of Silicon Materials, Department of Materials Science and Engineering,
Zhejiang University, Hangzhou 310027, Peoples Republic of China. 2.e-mail: zhaoxb@zju.edu.cn

This study focuses on Sb-doped Mg2(Si,Sn) thermoelectric material. Samples


were successfully fabricated using a hybrid synthesis method consisting of
three different processes: induction melting, solid-state reaction, and a hotpress sintering technique. We found that the carrier concentration increased
with Sb content, while the Seebeck coefficient exhibited a decreasing trend. Sb
doping was shown to improve the power factor and thermoelectric figure of
merit compared with the undoped material, yielding a peak figure of merit
(ZT) of 0.55 at 620 K, while leaving the band gap of Mg2Si0.7Sn0.3 almost
unchanged.
Key words: Thermoelectric material, Sb doping, fabrication method, solid
solution

INTRODUCTION
In recent decades, technology growth has
demanded more energy as a response to growing
human requirements. Eventually, we will have
consumed all the fossil fuels available on Earth. On
the other hand, use of fossil fuels is polluting the
environment. These considerations force human
beings to look for alternative energy sources called
green energy, i.e., energy sources which do not leave
a pollution footprint. Green energy sources such as
thermoelectric, photovoltaic, wind, biogas, and tidal
are important in and of themselves. In addition,
thermoelectric (TE) power is one of the most quickly
growing fields due to its ability for direct conversion
of heat into electricity and vice versa. It is unique
because of its features of solid-state assembly, low
noise, and long life. TE power is the best and most
efficient way to recover waste heat for use in a
heating or cooling system.
The conversion efficiency of a TE device is determined by the materials dimensionless figure of
merit ZT = a2rT/k, where a is the Seebeck coefficient, r is the electrical conductivity, j is the thermal conductivity, and T is the temperature in
(Received May 29, 2010; accepted February 5, 2011;
published online March 12, 2011)

830

Kelvin. Recent demand for high thermoelectric


conversion efficiency has prompted renewed
research efforts on novel and high-performance
thermoelectric materials, such as skutterudites,1
clathrates,2 AgPbmSbTem+2 alloys,3 and Zintl phases.4 However, an ideal TE material should not only
have higher ZT, but also should be nontoxic, as well
as being made up of abundantly available elements.
Mg2BIV (BIV = Si, Ge, Sn) based materials are
excellent candidates to meet these requirements.
According to the classical theory of thermoelectric
materials, the dimensionless figure of merit ZT is
proportional to the parameter A = (T/300)(m*/me)3/2
l/kph),5 where l is the carrier mobility in units of cm2/
V s, m* is the carrier effective mass, and jph is the
phonon thermal conductivity in mW/cm K. For simplicity, the units for A will be omitted throughout
this paragraph. Mg2BIV-based compounds have a
large A value of about 14, which is much larger than
for other thermoelectric silicides, such as p-type
FeSi2 (A = 0.8) and n-type SiGe (A = 2.6). Furthermore, the abundance of the constituent elements of
Mg2BIV is high. These are the main reasons why
Mg2BIV-based materials have attracted more attention. Moreover, a ZT value of 1.1 at 800 K has been
recently reported in Mg2Si0.4Sn0.6 solid solution,6
which is comparable to values of state-of-the-art
intermediate-temperature TE materials such as

Effect of Sb Doping on the Thermoelectric Properties of Mg2Si0.7Sn0.3 Solid Solutions

PbTe and filled skutterudites. Thus, Mg2BIV-based


compounds are believed to be cost-effective and ecofriendly candidates for thermoelectric power generation in the future in large-scale commercial
applications.
To improve the thermoelectric figure of merit of
Mg2BIV compounds, tremendous effort has been
invested in research into Mg2Si-Mg2Sn and Mg2SiMg2Ge solid materials,7 which are expected to have
an optimized band structure and reduced phonon
thermal conductivity. Mg2(Si,Sn)-based materials
with ZT exceeding unity have been reported at
composition near Mg2Si0.4Sn0.6.6,8 The quasibinary
phase diagram9,10 of Mg2Si-Mg2Sn shows the presence of an immiscibility gap between Mg2Si0.4Sn0.6
and Mg2Si0.6Sn0.4. These two compositions may
have a maximal solid solubility, yielding low phonon
thermal conductivity. Zhang et al. demonstrated
that a material composition near the gap boundaries of Mg2(Si,Sn) contains a number of natural
nanostructures, which can effectively enhance phonon scattering and decrease thermal conductivity.8
In this paper, the composition Mg2Si0.7Sn0.3 was
selected for study of its structure and properties.
Moreover, fabrication of Mg2(Si,Sn) solid solutions is difficult due to the large difference in
melting points among its constituents, and the
higher saturated vapor pressure of magnesium.
Various fabrication methods have been reported,
such as direct melting,8,1115 mechanical alloying,9
solid-state reactions,16 and reactive spark plasma
sintering (SPS).17 In our work, Sb-doped
Mg2Si0.7Sn0.3 samples were produced by a novel,
hybrid fabrication method. In this method, we separately fabricated binary Mg2Sn and Mg2Si, by
direct melting and solid-state reaction, respectively.
Then, we hot-pressed them together to obtain solid
solutions of Mg2(Si,Sn). The effect of Sb doping on
the thermoelectric properties of Mg2Si0.7Sn0.3 was
systematically investigated.

831

between the x-ray peaks of the corresponding binary


Mg2Si and Mg2Sn, indicating the formation of a
Mg2(Si,Sn) solid solution. A trace amount of MgO
impurity phase was also detected.
The thermal diffusivity and the specific heat were
measured using laser flash apparatus (Netzsch LFA
457) and a thermal analyzer (Netzsch DSC 404),
respectively. The thermal conductivity j was calculated from the relationship j = DaCp, where a is
the thermal diffusivity, Cp is the specific heat
capacity, and D is the density of the material. The
electrical conductivity r and the Seebeck coefficient
a were measured simultaneously using computerassisted apparatus.11 Hall coefficient (RH) measurements were conducted using a Quantum Design
PPMS-9T system with the van de Pauw method at
room temperature. The carrier concentration n and
the Hall mobility lH were calculated using the
relations n = 1/eRH, and lH = rRH, respectively.
RESULTS AND DISCUSSION
The temperature-dependent electrical conductivity r and Seebeck coefficient a for all the
Mg2Si0.7xSn0.3Sbx samples are plotted in Fig. 1a

EXPERIMENTAL PROCEDURES
Binary Mg2Sn was first prepared by direct melting of Mg (99.9% purity) and Sn (99.9% purity)
using induction heating under argon atmosphere,
whereas Sb-doped Mg2Si was produced by solidstate reaction. Mg was added in excess to compensate Mg loss due to evaporation in both processes.
The obtained Mg2Sn and Sb-doped Mg2Si were
crushed into powders, before being mixed to form
the compositions Mg2Si0.7xSn0.3Sbx, where x = 0,
0.0055, 0.0065, 0.0075, 0.0085, and 0.0095. The
mixed powders were hot-pressed under pressure of
80 MPa at 983 K for 2 h. The relative densities of all
the sintered samples were above 98%. The phase
structures of the sintered samples were analyzed by
x-ray powder diffraction (XRD) on a Rigaku
D/MAX-2550PC diffractometer using Cu Ka radiation. The XRD patterns of the samples showed
that almost all the diffraction peaks were located

Fig. 1. Temperature dependence of (a) electrical conductivity, and


(b) Seebeck coefficient for the Mg2Si0.7xSn0.3Sbx samples.

832

Gao, Liu, Zhu, Yang, and Zhao

Table I. Basic parameters for Sb-doped Mg2Si0.7Sn0.3


Doping
Content (ppm)
0
5500
6500
7500
8500
9500

n (/cm3)
1.89
6.94
5.99
9.29
7.65
7.16

9
9
9
9
9
9

1019
1019
1019
1019
1019
1019

lH (cm2/V/s)

Eg (eV)

21.34
21.26
29.02
19.71
27.88
28.46

0.37
0.38
0.37
0.37
0.33
0.37

and b, respectively. The room-temperature values of


electrical conductivity ranged from 24,000 S/m to
34,000 S/m for Sb-doped samples, which is five to
seven times higher than for undoped Mg2Si0.7Sn0.3
samples. Here, each added Sb atom substitutes a Si
atom, releasing one electron into the conduction
band. All the Sb-doped samples exhibited typical
degenerate semiconductor behavior. The measured
carrier concentrations are listed in Table I. The
electrical conductivity increased with increasing Sb
doping amount, corresponding to increased carrier
concentration. The electrical conductivity for the
undoped Mg2Si0.7Sn0.3 sample was 6460 S/m with
carrier concentration of 1.89 9 1025/m3, whereas in
the Mg2Si0.7Sn0.3 sample doped with 9500 ppm Sb
this value was boosted to 32,615 S/m with carrier
concentration of 7.16 9 1025/m3.
In Fig. 1b, the negative sign of the Seebeck coefficient confirms the existence of n-type conduction in
the sample, representing the effect of n-type doping.
The absolute value of the Seebeck coefficient shows
an increasing dependency on temperature, which
declines after Sb doping. The absolute value of a is
as large as 246 lV/K for the undoped Mg2Si0.7Sn0.3
sample, whereas this value reduces to about 104 lV/K
for the sample doped with 9500 ppm Sb. This can be
interpreted based on the following equation:
a /  kB r  ln n=e, where kB is Boltzmanns constant, e is the elementary charge, r is the scattering
factor, and n is the carrier concentration.14 The
scattering factor r represents the exponent of the
relaxation time versus energy and is assumed to be
constant. Carrier concentration decreases with
increasing temperature, and as a result, a also
decreases. With addition of Sb, the carrier concentration n increased but a decreased.
The energy band gap Eg could be estimated by the
relationship Eg = 2eamaxTmax, where e is the elementary charge, amax is the maximum value of the
Seebeck coefficient in the studied temperature
range, and Tmax is the temperature at which amax is
reached. The calculated results are shown in
Table I. The band gap, Eg, of the Sb-doped samples
varied little with the amount of Sb dopant as compared with the undoped Mg2Si0.7Sn0.3 sample. The
trivial changes in Eg suggest that doping with Sb
atoms has only a minor effect on the band gap of
Mg2Si0.7Sn0.3.

Fig. 2. Power factor for Mg2Si0.7xSn0.3Sbx samples as a function of


temperature.

Figure 2 shows the temperature dependence of


the power factor a2r for the Mg2Si0.7xSn0.3 Sbx
samples. The maximum power factor doubled with
Sb doping, which is a consequence of the increasing
electrical conductivity. Furthermore, with increasing Sb doping, the maximum power factor shifted
towards higher temperature.
The thermal conductivity j and phonon thermal
conductivity jph of all samples are shown in Fig. 3.
To determine the respective contributions from
charge carriers and phonons, we estimated the
carrier thermal conductivity je via the Wiedemann
Franz relation: je = L0Tr, where the degenerate
Lorenz number L0 is given a value of 2.4 9 108
V2/K2. The phonon thermal conductivity jph can
then be deduced from the simple relation
j = je + jph. Since the electrical conductivity of all
these samples is not very high, the electrical contribution to the total thermal conductivity is much
smaller than the lattice contribution. This means
that the total thermal conductivity is mostly determined by jph rather than by je. For the lattice
thermal conductivity, phononphonon interaction is
the primary source of thermal resistance. j and jph
exhibit a decreasing behavior with temperature, as
phononphonon scattering is predominant at higher
temperatures. An interesting point to be noted is
that the phonon thermal conductivity decreases
dramatically from 7.9 W/m/K for Mg2Si and 5.9 W/
m/K for Mg2Sn to the range 2.4 W/m/K to 3.6 W/m/
K for the solid alloys Mg2Si0.7xSn0.3Sbx at 300 K.
Substitution of Sb for Si introduces mass fluctuation scattering and strain field fluctuation scattering for phonons due to the mass and size
differences between substituting atoms and host
atoms. In general, the high-temperature phonon
thermal conductivity is sensitive to point-defect
scattering. Figure 3 clearly shows the dependency
of jph on the amount of Sb doping. jph increases
with increasing Sb doping, until the amount of Sb

Effect of Sb Doping on the Thermoelectric Properties of Mg2Si0.7Sn0.3 Solid Solutions

833

Fig. 4. Figure of merit as a function of temperature for the


Mg2Si0.7xSn0.3Sbx samples.

higher power factor and low thermal conductivity. It


is evident that the Sb doping has enhanced ZT. The
highest value of ZT obtained in this work is 0.55 at
620 K.
CONCLUSIONS

Fig. 3. Temperature dependence of (a) thermal conductivity, and (b)


phonon thermal conductivity for the Mg2Si0.7xSn0.3Sbx samples.

doping reaches x = 0.0075, which is equivalent to


7500 ppm. The Si atoms have been substituted by
Sb atoms, which has attributed to a change in mass
and size difference between the Si and Sn sites(each
of them has different atomic size). The atomic radii
for Si, Sn, and Sb are 110 pm, 145 pm, and 145 pm,
respectively. The decline of the value of jph for
x > 7500 ppm may be a result of increasing point
defects induced by the entry of Sb atoms into the
lattice. Higher point defects lead to higher mass
fluctuation scattering and strain field fluctuation
scattering.
We now turn to the figure of merit of the thermoelectric effect. Figure 4 shows the figure of merit
ZT as a function of temperature with and without
Sb doping for Mg2Si0.7Sn0.3. All samples have similar ZT values at room temperature, but they exhibit
a temperature dependency at higher temperature.
After the temperature reaches 570 K, Sb-doped
samples have higher ZT values than the undoped
sample. The x = 0.0055 sample exhibited the highest ZT value of 0.54 at 650 K, corresponding to

We successfully fabricated Mg2Si0.7xSn0.3Sbx


(x = 0, 0.0055, 0.0065, 0.0075, 0.0085, 0.0095)
samples via a hybrid method that incorporated
the induction melting, solid-state reaction, and
hot-press sintering techniques. Doping with Sb
increased the carrier concentration, which
enhanced the electrical conductivity from 6000 S/m
to 34,000 S/m. At the same time, it also decreased
the value of the Seebeck coefficient. The Sb doping
had a minor effect on the band gap of Mg2Si0.7Sn0.3.
Also, it shifted the maximum power factor towards
higher temperature. The thermal conductivity of
Mg2Si0.7xSn0.3Sbx was in the range from 2.4 W/m/K
to 3.6 W/m/K at room temperature. Thermal conductivity j was mainly controlled by the phonon
thermal conductivity jph, which was dependent on
the amount of Sb doping. The values of ZT for all Sbdoped Mg2Si0.7Sn0.3 samples were similar at room
temperature, but became larger than that of the
undoped Mg2Si0.7Sn0.3 sample after the temperature exceeded 570 K. The Sb doping improved ZT
and shifted the peak ZT to higher temperature.
ACKNOWLEDGEMENTS
The work was financially supported by the
National Natural Science Foundation of China
(50731006 and 50971115) and the Science and
Technology Program of Zhejiang Province (2009
C34007 and Z4090204).
REFERENCES
1. J.L. Mi, X.B. Zhao, T.J. Zhu, and J.P. Tu, Appl. Phys. Lett.
91, 172116 (2007).

834
2. G.S. Nolas, J.L. Cohn, G.A. Slack, and S.B. Schujman, Appl.
Phys. Lett. 73, 178 (1998).
3. K.F. Hsu, S. Loo, F. Guo, W. Chen, J.S. Dyck, C. Uher, T.
Hogan, E.K. Polychroniadis, and M.G. Kanatzidis, Science
303, 818 (2004).
4. C. Yu, T.J. Zhu, S.N. Zhang, X.B. Zhao, J. He, Z. Su, and
T.M. Tritt, J. Appl. Phys. 104, 013705 (2008).
5. C. Vining, Proc. 10th Int. Conf. on Thermoelectrics (Cardiff:
IEEE, 1991), p. 249.
6. V.K. Zaitsev, M.I. Fedorov, E.A. Gurieva, I.S. Eremin, P.P.
Konstantinov, A.Y. Samunin, and M.V. Vedernikov, Phys.
Rev. B 74, 045207 (2006).
7. T. Aizawa, R. Song, and A. Yamamoto, Mater. Trans. 46,
1490 (2005).
8. Q. Zhang, J. He, T.J. Zhu, S.N. Zhang, X.B. Zhao, and T.M.
Tritt, Appl. Phys. Lett. 93, 102109 (2008).
9. M. Riffel, J. Schilz, and Ieee. 1996. Pasadena, Ca.

Gao, Liu, Zhu, Yang, and Zhao


10. I. Jung, D. Kang, W. Park, N. Kim, and S. Ahn, Calphad 31,
192 (2007).
11. Q. Zhang, J. He, X.B. Zhao, S.N. Zhang, T.J. Zhu, H. Yin,
and T.M. Tritt, J. Phys. D Appl. Phys. 41, 185103
(2008).
12. Q. Zhang, T.J. Zhu, A.J. Zhou, H. Yin, and X.B. Zhao,
Physica Scripta T129, 123 (2007).
13. Q. Zhang, X.B. Zhao, T.J. Zhu, and J.P. Tu, Phys. Stat. Sol.
RRL 2, 56 (2008).
14. Q. Zhang, X.B. Zhao, H. Yin, and T.J. Zhu, J. Alloys Compd.
464, 9 (2008).
15. Q. Zhang, H. Yin, X.B. Zhao, J. He, X.H. Ji, T.J. Zhu, and
T.M. Tritt, Phys. Stat. Sol. A Appl. Mater. Sci. 205, 1657
(2008).
16. L.M. Zhang, C.B. Wang, H.Y. Jiang, Q. Shen, and Ieee 2003.
La Grande Motte, France.
17. J. Tani and H. Kido, J. Alloys Compd. 466, 335 (2008).

You might also like