Enhancement in the thermoelectric performance by y substitution on SrSi2

Article abstract of DOI:10.1063/1.3136847

We report the results of the Y substitution in Sr_1-xY _xSi₂ with x≤0.15 via measuring the temperature- dependent electrical resistivity, thermal conductivity, as well as Seebeck coefficient. Upon substituting Y onto the Sr

Full text of DOI:10.1063/1.3136847

Enhancement in the thermoelectric performance by Y substitution on SrSi 2
C. S. Lue, M. D. Chou, N. Kaurav, Y. T. Chung, and Y. K. Kuo
Citation: Applied Physics Letters 94, 192105 (2009); doi: 10.1063/1.3136847
View online: http://dx.doi.org/10.1063/1.3136847
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/94/19?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Hot deformation induced defects and performance enhancement in FeSb2 thermoelectric materials
J. Appl. Phys. 114, 184904 (2013); 10.1063/1.4831668
Intrinsically modified thermoelectric performance of alkaline-earth isovalently substituted [Bi2 AE
2O4][CoO2]y single crystals
J. Appl. Phys. 114, 043705 (2013); 10.1063/1.4816315
Enhancement of thermoelectric performance in strontium titanate by praseodymium substitution
J. Appl. Phys. 113, 053704 (2013); 10.1063/1.4790307
High-temperature thermoelectric properties of the double-perovskite ruthenium oxide (Sr1−x La x )2ErRuO6
J. Appl. Phys. 112, 073714 (2012); 10.1063/1.4757632
Enhanced thermoelectric figure-of-merit ZT for hole-doped Bi2Sr2Co2O y through Pb substitution
J. Appl. Phys. 111, 103709 (2012); 10.1063/1.4720075
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
84.88.136.149 On: Thu, 18 Dec 2014 09:30:05

APPLIED PHYSICS LETTERS 94, 192105 ͑2009͒
Enhancement in the thermoelectric performance by Y substitution on SrSi2
C. S. Lue,1 M. D. Chou,1 N. Kaurav,2 Y. T. Chung,2 and Y. K. Kuo2,a͒
1Department of Physics, National Cheng Kung University, Tainan 70101, Taiwan
2Department of Physics, National Dong Hwa University, Hualien 97401, Taiwan
͑Received 23 February 2009; accepted 23 April 2009; published online 14 May 2009͒
We report the results of the Y substitution in Sr1−xYxSi2 with xՅ0.15 via measuring the
temperature-dependent electrical resistivity, thermal conductivity, as well as Seebeck coefficient.
Upon substituting Y onto the Sr sites, the electrical resistivity exhibit semiconducting behavior and
the room-temperature electrical resistivity tends to reduce for xՅ0.08. The thermal conductivity
also decreases with increasing the Y content. Moreover, the Seebeck coefficient has a substantial
increase and a maximum of about 220 ␮V/K at around 80 K has been found for x=0.08. These
promising effects lead to a significant enhancement in the thermoelectric performance characterized
by the figure-of-merit, ZT. A room-temperature ZT value of approximately 0.4 is thus achieved for
Sr0.92Y0.08Si2, about one order of magnitude larger than that of stoichiometric SrSi2. © 2009
American Institute of Physics. ͓DOI: 10.1063/1.3136847͔
Silicides with semiconducting or semimetallic character-
istics have attracted considerable attention because of their
practical applications in electronics and thermoelectrics.1–3 It
is of particular importance that those silicides are composed
of nontoxic and naturally abundant elements in the earth’s
crust. Recently, SrSi2 was reported to be a narrow-gap semi-
conductor with a band gap of 35 meV estimated from the
Hall coefficient between 180 and 300 K.4 Band structure
calculations,5–8 on the other hand, indicated the presence of a
sharp pseudogap in the Fermi-level density of states ͑DOS͒,
suggesting a semimetallic character for SrSi2. As proposed
by Mahan and Sofo,9 materials with sharp electronic band
features of a few tens of meV from the Fermi level would be
promising candidates for developing highly efficient thermo-
electrics. As a matter of fact, SrSi2 was reported to have a
large Seebeck coefficient ͑S͒ of approximately 130 ␮V/K
at 300 K.10 However, the thermoelectric performance is
hampered by its relative high electrical resistivity ͑␳
Ӎ1 m⍀ cm͒ and thermal conductivity ͑␬Ӎ5 W/m K͒ at
room temperature,10 with respect to the dimensionless figure-
of-merit ZT=S2T/␳␬.
In general, the difficulty in achieving good thermoelec-
tric performance is characterized by the need to minimize the
electrical resistivity and the thermal conductivity of materi-
als, while enhancing their Seebeck coefficient. The strategy
to optimize these parameters usually involves the substitu-
tion of different elements to alert the electronic band struc-
tures in the vicinity of the Fermi level and to enhance the
phonon scattering by introducing crystallographic disorder.11
For the present case of SrSi2, the application of an external
pressure has an effect to reduce the electrical resistivity, as
observed by Imai et al.12 This encouraging finding suggested
that further improvements of the thermoelectric performance
on the SrSi2-based alloys may be made by substituting ele-
ments with a small atomic radius which would induce a posi-
tively chemical pressure in the system, similar to the pres-
sure effect. With this motivation, we investigate the effects of
the chemical substitution on the thermoelectric properties of
Sr1−xYxSi2͑0ՅxՅ0.15͒ as Y has a smaller atomic size than
Sr. In addition, the replacement of Sr by Y would cause a
band structure modification since Y has one more electron in
its valence shell than Sr, providing an opportunity for the
proximity of the Fermi level to a sharp peak in the DOS.
Polycrystalline Sr1−xYxSi2 samples were prepared by
mixing appropriate amounts of elemental metals. Mixture of
high-purity elements were placed in a water-cooled copper
crucible and then melted several times in an argon arc-
melting furnace. An x-ray analysis taken with Cu K␣ radia-
tion on powder specimens was consistent with the expected
P4332 structure.13,14 The variation in lattice constant as a
function of the Y concentration is shown in Fig. 1. It is
clearly seen that the lattice constant gradually decreases with
x, indicating that the Sr sites are successfully replaced by Y
atoms, according to Vegard’s law.
The temperature variation in the electrical resistivity for
the Sr1−xYxSi2 alloys is illustrated in Fig. 2͑a͒, obtained by a
standard dc four-terminal method during warming process.
The electrical resistivity of SrSi2 exhibits a semimetallic
character, showing weak temperature dependence, in agree-
ment with the prediction by the band structure calculation.
On the other hand, ␳͑T͒ exhibits a negative temperature co-
efficient of resistivity ͑TCR͒ for each Y-substituted sample,
finding reminiscent of semiconducting behavior. However,
other mechanisms such as a variable range hopping ͑VRH͒
process could also induce the negative TCR response. To
6.535
Sr1-xYxSi2
6.534
6.533
6.532
0.00
0.05
0.10
0.15
Y concentration, x
FIG. 1. ͑Color online͒ Lattice constant vs Y concentration in Sr1−xYxSi2 as
obtained from x-ray diffraction.
a͒
Electronic mail: ykkuo@mail.ndhu.edu.tw.
0003-6951/2009/94͑19͒/192105/3/$25.00
94, 192105-1
© 2009 American Institute of Physics
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
84.88.136.149 On: Thu, 18 Dec 2014 09:30:05

192105-2
Lue et al.
Appl. Phys. Lett. 94, 192105 ͑2009͒
1000
30
25
20
15
10
5
x=0
Sr1-xYxSi2
(a)
x=0.03
x=0.05
x=0.08
x=0.10
x=0.15
100
10
1
x=0
x=0.03
x=0.05
x=0.08
x=0.10
x=0.15
200
150
100
50
(b)
x=0
x=0.03
x=0.05
x=0.08
x=0.10
x=0.15
0
0
10
100
T
(K)
-50
-100
-150
FIG. 3. ͑Color online͒ Temperature dependence of the total thermal conduc-
tivity in Sr1−xYxSi2.
0
50 100 150 200 250 300
increase and consequently decreases after passing through a
highest one for x=0.08. The change in S with higher Y
amounts indicates the modification of the electronic band
structure. In these alloys, the Seebeck coefficient develops a
broad maximum, a common feature due to the contribution
of the thermally excited electrons across their pseudogaps.
For x=0.03, 0.05, and 0.10, the sign of the low-
temperature S data is negative accompanied by negative
peaks at around 50 K. Such a feature can be qualitatively
interpreted in terms of two-carrier conduction mechanism.
Accordingly, the total S can be expressed as
T
(K)
FIG. 2. ͑Color online͒ ͑a͒ Electrical resistivity as a function of temperature
for Sr1−xYxSi2. ͑b͒ Seebeck coefficient vs temperature in Sr1−xYxSi2.
examine this scenario, the data of ␳͑T͒ were fitted to the
Greaves15 VRH conduction, where the dc resistivity ␳
dc
1/4
ͱ
obeys the relationship ␳dcϰ T exp͓͑To/T͒ ͔, with the char-
acteristic constant To inversely propositional to the Fermi-
level DOS. The satisfactory fitting suggests that the presence
of microstructural inhomogeneity or antisite disorder in
Sr1−xYxSi2 could be the main source for the observed nega-
tive TCR. Moreover, the deduced value of To initially de-
creases with increasing the Y concentration and then in-
creases with further adding of more Y content. Such a trend
is consistent with the variation in the electrical resistivity,
indicative of the reliable interpretation for the electrical
transport of Sr1−xYxSi2. With these respects, it is not appro-
priate to classify these Y-doped samples as semiconductors
with real band gaps in their Fermi-level DOS. Rather, they
should be realized as semimetals with pseudogaps in the vi-
cinity of the Fermi levels. Note that the evolution of the
electrical resistivity caused by the chemical substitution is
not quite consistent with the case of the pressure effect. It is
likely that the Fermi-level band structure of SrSi2 has been
severely changed with varying the Y content, especially for
the higher Y concentration. As a matter of fact, employing Y
as an electron donor in SrSi2 would modify its electronic
band feature.
␴
␴
p
n
S =
Sn +
Sp,
͑1͒
␴n + ␴
␴n + ␴
p
p
where Sn,p and ␴ represent the Seebeck coefficients and
n,p
electrical conductivities for the n- and p-type carriers, re-
spectively. Since the signs of Sn and Sp are opposite, tuning
these quantities could result in a sign change in S. For these
Y-doped samples, the low-temperature Seebeck coefficients
are mainly dominated by the n-type carriers. However, with
increasing temperature, the intrinsic electrons and holes are
excited across the pseudogaps. If the holes have a slightly
higher mobility than the electrons, the thermal transport is
increasingly governed by the p-type carriers, as we observed
in these materials.
In Fig. 3, we display the observed thermal conductivity
for all studied samples. At low temperatures, ␬ increases
with temperature and a maximum appears between 20 and
40 K. This is a typical feature for the reduction in thermal
scattering in solids at low temperatures. A remarkable trend
found in ␬ is that the height of the low-temperature peak
decreases drastically with increasing the substitution level,
indicative of a strong enhancement in the phonon scattering
by lattice imperfections. On the other hand, the thermal con-
ductivity exhibits a marginal reduction with the substitution
level for TՆ200 K, showing a room-temperature ␬ of about
5 W/mK. This is attributed to the dominant lattice thermal
Figure 2͑b͒ shows the Seebeck coefficient as a function
of temperature for the Sr1−xYxSi2 alloys. The positive sign of
S for the stoichiometric compound of SrSi2 implies that the
hole-type carriers dominate the heat transport. The room-
temperature S value of about 135 ␮V/K is consistent with
the previously reported result.10 For the slightly substituted
sample ͑x=0.03͒, the magnitude of S tends to reduce, pre-
sumably attributed to the band filling effect, as Y has one
more electron in its valence shell than Sr. Upon further sub-
stituting Y for Sr ͑xՆ0.05͒, the magnitude of S becomes to
conductivity ͑␬ ͒ to the total ␬, as analyzed by Hashimoto
L
et al.10 for SrSi . The absence of an effective decrease in ␬
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at:2http://scitation.aip.org/termsconditions. DownloadeLd to IP:
84.88.136.149 On: Thu, 18 Dec 2014 09:30:05

192105-3
Lue et al.
Appl. Phys. Lett. 94, 192105 ͑2009͒
tion in ␬ should be further improved to enlarge the ZT
value.
L
0.4
Sr1-xYxSi2
An extensive investigation of the effect of Y doping in
Sr1−xYxSi2 is presented via the thermoelectric measurements
as a function of temperature. We clearly demonstrate that the
Y substitution represents a good opportunity for improving
its ZT, making this system attractive for possible candidates
for low-and intermediate-temperature thermoelectric applica-
tions. Small amounts of Y dramatically affect the electronic
band structure, and thus the electrical transport properties of
these materials. However, one of the challenges which still
remain is to reduce the lattice thermal conductivity. Efforts to
utilize the mass fluctuation scattering to lower ␬ are being
employed.
x=0
0.3
0.2
0.1
0.0
x=0.03
x=0.05
x=0.08
x=0.10
x=0.15
This work was supported by National Science Council,
Taiwan under Grant Nos. 97-2628-M-259-001-MY3 ͑YKK͒
and NSC-95-2112-M-006-021 ͑CSL͒.
0
50 100 150 200 250 300
1T. M. Tritt, M. Kanatzidis, H. B. Lyon, Jr., and G. Mahan, Symposium
Proceedings of the Materials Research Society ͑Materials Research Soci-
ety, Pittsburgh, 1997͒, 478.
T (K)
FIG. 4. ͑Color online͒ ZT value as a function of temperature for Sr1−xYxSi2.
2C. S. Lue, Y.-K. Kuo, C. L. Huang, and W. J. Lai, Phys. Rev. B 69,
125111 ͑2004͒.
3Y. K. Kuo, K. M. Sivakumar, S. J. Huang, and C. S. Lue, J. Appl. Phys.
98, 123510 ͑2005͒.
is associated with the weak mass fluctuation between Sr and
Y, since their mass difference is only about 1%. The results
imply that a further reduction in ␬ is possible if large atomic
mass fluctuation is present by alloying heavier elements in
SrSi2.
4M. Imai, T. Naka, T. Furubayashi, H. Abe, T. Nakama, and K. Yagasaki,
Appl. Phys. Lett. 86, 032102 ͑2005͒.
5Y. Imai and A. Watanabe, Intermetallics 10, 333 ͑2002͒.
6Y. Imai and A. Watanabe, Intermetallics 14, 666 ͑2006͒.
7S. Brutti, D. Nguyen Manh, and D. G. Pettifor, Intermetallics 14, 1472
͑2006͒.
Figure 4 shows ZT as a function of temperature for
Sr1−xYxSi2. Remarkably, the Sr0.92Y0.08Si2 specimen has the
highest ZT=0.41 at room temperature. This is mainly due to
the fact that both ␳ and S are optimized with this concentra-
tion, although no significant reduction in ␬ is achieved. This
result can be ascribed by the simultaneous presence of the
sharp Fermi-level band edge as well as the high mobility of
the carriers for this composition. While the highest ZT value
among the studied compositions is a bit smaller than that of
the optimized Bi2Te3͑ZTӍ0.87͒,16 it is comparable to the
nonoptimized Bi2Te3 having ZTӍ0.55. Furthermore,
Sr0.92Y0.08Si2 shows SӍ135 V/K and ␳Ӎ0.28 m⍀ cm at
room temperature, resulting in a high thermoelectric power
factor S2/␳Ӎ6.5ϫ10−3 W/m K2. This value is larger than
those of the thermoelectric materials such as Bi2Te3,
K2Bi8Se13, CoSi1−xGex,16–18 and even comparable to that
achieved by pressurizing Bi2Te3.19 It is apparent that the
thermoelectric power factor of Sr1−xYxSi2 is indeed enhanced
with a suitable dopant level. However, the necessary reduc-
8M. Imai, T. Naka, H. Abe, and T. Furubayashi, Intermetallics 15, 956
͑2007͒.
9G. D. Mahan and J. O. Sofo, Proc. Natl. Acad. Sci. U.S.A. 93, 7436
͑1996͒.
10K. Hashimoto, K. Kurosaki, Y. Imamura, H. Muta, and S. Yamanaka, J.
Appl. Phys. 102, 063703 ͑2007͒.
11F. J. DiSalvo, Science 285, 703 ͑1999͒.
12M. Imai, T. Naka, H. Abe, and T. Furubayashi, Intermetallics 15, 956
͑2007͒.
13J. Evers, J. Solid State Chem. 24, 199 ͑1978͒.
14M. Imai and T. Kikegawa, Chem. Mater. 15, 2543 ͑2003͒.
15G. N. Greaves, J. Non-Cryst. Solids 11, 427 ͑1973͒.
16H. Scherrer and S. Scherrer, in CRC Handbook of Thermoelectrics, edited
by D. M. Rowe ͑CRC, Boca Raton, FL, 1995͒, p. 211.
17D.-Y. Chung, K.-S. Choi, L. Iordanidis, J. L. Schindler, P. W. Brazis, C. R.
Kannewurf, B. Chen, S. Hu, C. Uher, and M. G. Kanatzidis, Chem. Mater.
9, 3060 ͑1997͒.
18E. Skoug, C. Zhou, Y. Pei, and D. T. Morelli, Appl. Phys. Lett. 94, 022115
͑2009͒.
19S. V. Ovsyannikov, V. V. Shchennikov, G. V. Vorontsov, A. Y. Manakov,
A. Y. Likhacheva, and V. A. Kulbachinskii, J. Appl. Phys. 104, 053713
͑2008͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
84.88.136.149 On: Thu, 18 Dec 2014 09:30:05