Upload
yang-li
View
22
Download
0
Embed Size (px)
Citation preview
Superconductivity in Al-substituted Ba8Si46 clathrates
Yang Li,1,2,3,a) Jose Garcia,1 Ning Chen,3 Lihua Liu,2 Feng Li,2 Yuping Wei,2
Shanli Bi,2 Guohui Cao,2 and Z. S. Feng4
1Department of Engineering Science and Materials, University of Puerto Rico at Mayaguez, Mayaguez,Puerto Rico 00681-9044, USA2Department of Physics, University of Science and Technology Beijing, Beijing 100083, China3School of Materials Science and Engineering, University of Science and Technology Beijing, Beijing 100083,China4Department of Mathematics, University of Texas-Pan American, Edinburg, Texas 78541, USA
(Received 30 January 2013; accepted 6 May 2013; published online 24 May 2013)
There is a great deal of interest vested in the superconductivity of Si clathrate compounds with
sp3 network, in which the structure is dominated by strong covalent bonds among silicon atoms,
rather than the metallic bonding that is more typical of traditional superconductors. A joint
experimental and theoretical investigation of superconductivity in Al-substituted type-I silicon
clathrates is reported. Samples of the general formula Ba8Si46�xAlx, with different values of xwere prepared. With an increase in the Al composition, the superconducting transition
temperature TC was observed to decrease systematically. The resistivity measurement revealed
that Ba8Si42Al4 is superconductive with transition temperature at TC¼ 5.5 K. The magnetic
measurements showed that the bulk superconducting Ba8Si42Al4 is a type II superconductor. For
x¼ 6 sample Ba8Si40Al6, the superconducting transition was observed down to TC¼ 4.7 K which
pointed to a strong suppression of superconductivity with increasing Al content as compared
with TC¼ 8 K for Ba8Si46. Suppression of superconductivity can be attributed primarily to a
decrease in the density of states at the Fermi level, caused by reduced integrity of the sp3
hybridized networks as well as the lowering of carrier concentration. These results corroborated
by first-principles calculations showed that Al substitution results in a large decrease of the
electronic density of states at the Fermi level, which also explains the decreased superconducting
critical temperature within the BCS framework. The work provided a comprehensive
understanding of the doping effect on superconductivity of clathrates. VC 2013 AIP Publishing LLC.
[http://dx.doi.org/10.1063/1.4807316]
I. INTRODUCTION
Group-IV clathrate materials are extended Si, Ge, and
Sn cage-like solids with sp3-hybridized networks that have
received increasing attention over the past years. These
materials have a semiconducting framework into which
metal atoms can be substituted, providing a number of possi-
bilities for electronic materials.1,2 Furthermore, within the
sp3 hybridized networks, K, Na, Rb, Cs, Sr, Ba, I, and Eu
atoms can be encapsulated in the cages.3,4 Clathrates exhibit
metallic, semiconducting, or insulating behavior depending
either upon the substitutional atoms or the occupation frac-
tion of the group IV atoms in the cage framework. The study
of clathrates opens a field of new materials with the metals
arranged in a nanoscale array, and with a wide variety of
properties ranging from insulators to metals.5 New thermo-
electric applications have driven a great deal of this inter-
est.6,7 The cage structures can be filled with atoms that have
a strong propensity to scatter phonons. These factors greatly
influence the thermoelectric efficiency. Recently, Nuclear
Magnetic Resonance (NMR) and Mossbauer measurements
have directly demonstrated atomic hopping within the cages
of Sr8Ge30Ga16 (Ref. 8) and Eu8Ge30Ga16,9 respectively.
The variety of electronic behavior attained by chemical sub-
stitution and doping suggests that significant new features
may be produced in this system. In a search for better pho-
non scattering efficiency, Ge clathrates filled with rare earth
Eu have been synthesized, indicating that clathrates of this
type containing local magnetic moments are possible.10
Further studies have identified ferromagnetic behavior in
magnetically substituted Ba8Mn4Ge42 (type I clathrate)11
and Ba6Fe3Ge22 (chiral type clathrate).12 The potential of
such magnetic clathrates is quite significant, since the clath-
rate structure can be tailored for the desired magnetic proper-
ties by substitution and alloying. Therefore, clathrates also
have potential applications in magnetic sensors and new
magnetic semiconductors.12–14
Inspired by the discovery of superconductivity in alkali
metal-doped C60 fullerene, efforts have been made to explore
the superconductivity of Group IV clathrates with particular
attentions to the sp3 hybridized networks. In contrast to carbon,
silicon and germanium do not form sp2-like networks.
Therefore, superconductivity of Si clathrate superconductors
with sp3 network should be unique. In a precursor study, the
Caplin group investigated the conductivity and magnetic sus-
ceptibility of silicon clathrates containing Na atoms as guests
but found no superconductivity.5 Then Ba-encapsulated silicon
clathrates were found to exhibit superconductivity, with
TC¼ 8 K for the best samples with pure Ba encapsulation.15,16
a)Author to whom correspondence should be addressed. Electronic mail:
0021-8979/2013/113(20)/203908/6/$30.00 VC 2013 AIP Publishing LLC113, 203908-1
JOURNAL OF APPLIED PHYSICS 113, 203908 (2013)
The structure of clathrate superconductor is dominated by
strong covalent bonds between silicon atoms, rather than the
metallic bonding that is more typical of traditional supercon-
ductors. Isotope effect measurements have revealed that super-
conductivity in Ba8Si46 is of the classic BCS kind, arising from
the electron-phonon interaction.17 Study of the band structure
for Ba8Si46 has shown a strong hybridization between the Si46
band and Ba orbitals, resulting in a very high density of states
(DOS) N(EF) at the Fermi level.18–20 Both the strong hybridiza-
tion of Ba with the conduction band and the high N(EF) are
believed to play a key role in the superconductivity of these
compounds, and further studies of Si and Ge clathrates indicate
the superconductivity to be an intrinsic property of the sp3 net-
work.21 The combined experimental and theoretical studies of
the effect of Cu- and Ga-doping on the superconductivity of
clathrates were performed, repectively.20,22 The doped atoms
are found to be strongly hybridized with the cage conduction-
band state. The substitution results in a lowering of the carrier
concentration and decreasing of density of states at Fermi level.
These play key roles in the suppression of superconductivity.
We are interested in the effect of Al-doping on the supercon-
ductivity of Ba8Si46, as well as the change of electronic struc-
ture in the clathrates. The Ba8(Si,Al)46 system exhibits a wide
variety of physical properties; with the Al content increasing
from x¼ 0, the clathrate behavior changes from superconduct-
ing Ba8Si46 (TC¼ 8 K)16 to the heavily doped semiconductor
Ba8Si30Al16. Investigation of Al doping can also increase our
understanding of the electronic structure and superconducting
mechanism in clathrate materials.
This paper reports a joint experimental and theoretical
study of Al substitution in Ba8Si46�xAlx clathrates. The
Ba8Si42Al4 and Ba8Si40Al6 are shown to be superconductors,
however, with Al content increasing, the superconducting TC
decreases. The first-principles calculations are used to build
a detailed picture of the atomic and electronic structure of
Al-substituted clathrates. By comparing the electronic struc-
tures of different Al-substituted silicon clathrates, the theo-
retical results show that Al-doping gives rise to a lower
density of states at the Fermi-level [N(EF)], which is
explained as one of the reasons of the destructive effect of
Al-doping on superconductivity in Si-clathrates. For dilute
levels, the changes induced by substitution of Al for Si are
approximately rigid band in character; therefore, it is possi-
ble to change the electron concentration by framework sub-
stitution while leaving the superconducting character of the
sp3 network intact.
II. EXPERIMENTAL RESULTS
The synthesis of Ba8Si46�xAlx was based on the multi-
step melting of Ba, Al, and Si under argon atmosphere and
subsequent solid-state reaction.12 The powder samples were
characterized and analyzed by X-ray diffraction (XRD). The
bulk samples were analyzed for resistivity and susceptibility
by a cryogen-free physical properties measurement system.
Analysis by powder X-ray diffraction showed character-
istic type-I clathrate reflections. Structural refinement of the
powder X-ray diffraction data was carried out using the GSAS
software package.23,24 As shown in Fig. 1, the sample of
Ba8Si42Al4 with dilute Al-doping, exhibited the main phase
to be type I clathrate with a little impure phases such as sili-
con when analyzed by X-ray diffraction at room temperature.
Ba8Si42Al4 crystallizes into the type-I clathrate structure
[cubic space group Pm�3n (No.223) as shown in Fig. 2]. The
experimental pattern is in agreement with the simulated one
for the entire 2h region. R values for the fit are Rwp¼ 0.09
and Rp¼ 0.08. The measured structural parameters were
selected as the input data for the model simulations. As a
result of the refinement, Al was found to preferentially
occupy the 6c framework sites for dilute doping, however,
for heavy substitution, Al inclined towards a random distri-
bution of the other 16i and 24k sites. As shown in the inset
of Fig. 1, the refined lattice parameters of Ba8Si46�xAlx(x¼ 4, 6, and 16) are 10.389, 10.436, and 10.517 A, respec-
tively, which exhibit an increasing trend with x due to the
larger atomic size of Al than that of Si.
The four-probe transport measurement using a Cryogen-
free Physical Properties Measurement System confirms that
FIG. 1. X-ray refinement for Ba8Si42Al4. Upper curve: data and fit, with dif-
ference plot below. Blue, black, and red ticks, respectively, show peaks
indexed according to the type-I clathrate, BaSi2 (orthorhombic phase) and Si
structure. Inset: lattice parameters of Ba8Si46�xAlx (x¼ 4, 6, and 16), show-
ing an increasing trend with x. The lattice parameter of Al-free Ba8Si46
clathrate was taken from Ref. 16.
FIG. 2. The type-I clathrate structure, built from a regular arrangement of a
combination of Si20 (Ih) and Si24 (D6d) cages. The Ba atoms are on the center
of the Si cage.
203908-2 Li et al. J. Appl. Phys. 113, 203908 (2013)
Ba8Si42Al4 has a metallic behavior from room temperature
to 6 K. This sample resistance starts to drop dramatically at
6.0 K, thus, indicating the initiation of the superconducting
state. Such temperature is defined as the superconducting
onset temperature (TC,onset). As shown in Fig. 3, the resist-
ance of sample sharply decreases from its high temperature
value of R¼ 60 mX to zero resistance at 5.3 K (TC,zeroR). The
superconducting transition width was observed to be about
0.7 K. The superconducting critical temperature TC is defined
as the temperature at which the dR/dT-T curve reaches a
maximum. The sample Ba8Si42Al4 has a superconducting
critical temperature TC¼ 5.5 K.
The superconducting transition temperatures for
Ba8Si42Al4 were systematically measured at various mag-
netic fields from H¼ 0 to 90 kOe with an interval 5 kOe as
shown in Fig. 4. A decrease in the superconducting transition
temperatures was observed with an increasing magnetic
field. In the inset of Fig. 4, the magnetic field dependent
superconducting transition temperatures are plotted. As the
field reached 4.5 T, the sample resistance temperature
TC,zeroR dropped below 2 K. However, a clear superconduct-
ing signal until the field up to 9 T could still be observed,
which implies that the critical field is fairly high for the sam-
ple. Within the standard BCS approach to superconductivity,
as has been applied for other superconducting Si clathrates,17
the value of the superconducting gap at 0 K from the critical
temperature (TC¼ 5.5 K) was deduced using the well-known
relation,25 2DT!0K ¼ 3:52 kBTC. In this way, the supercon-
ducting gap was determined to be about 0.8 meV for
Ba8Si42Al4.
The temperature dependences of the resistivity of a
Ba8Si40Al6 sample under various magnetic fields are shown
in Fig. 5. The sample showed a very sharp superconducting
transition at 4.7 K when without applied magnetic field.
Superconducting critical temperature TC decreases with field
increasing, as shown in the inset of Fig. 5. We notice that the
sample Ba8Si40Al6 has no zero-resistance down to 2 K,
which was attributed to the weak linkage on the interface of
grains. The Al substitution for Si suppresses the supercon-
ducting critical temperature TC. From the inset of Fig. 3, we
can clearly see the trendy of TC decreasing with aluminum
doping x; the superconducting critical temperature TC
of Ba8Si46�xAlx decreases from TC¼ 8 K for x¼ 0 to TC
¼ 4.7 K for x¼ 6.
A temperature dependent susceptibility measurement
was performed for the Ba8Si42Al4 sample in 100 Oe static
field, as shown in Fig. 6. The susceptibility had virtually no
temperature dependence for 100> T> 6 K. At about 6 K, the
sample started to exhibit superconducting characteristics;
the susceptibility v suddenly dropped. The large change in
the susceptibility was accompanied by a distinct drop in elec-
trical resistivity of the Ba8Si42Al4 sample confirming that
Ba8Si42Al4 enters into a superconducting state at 5.7 K.
Figure 6 presents the susceptibility of Ba8Si42Al4 as a func-
tion of temperature, under conditions of zero field cooling
(ZFC) and field cooling (FC) at 100 Oe. The ZFC magnetiza-
tion data were taken while heating after the sample cooling
in zero applied field was performed, whereas the FC magnet-
ization was measured as a function of both decreasing and
increasing temperature in the applied field. The enhancement
FIG. 3. The resistance of Ba8Si42Al4 versus temperature at the field H¼ 0.
Inset: superconducting critical temperature TC of Ba8Si46�xAlx versus Al-
substitution content x. The TC of Al-free Ba8Si46 clathrate was taken from
Ref. 16.
FIG. 4. The resistance of Ba8Si42Al4 versus temperature at various fields Hfrom 0 to 90 kOe with an interval 5 kOe. Inset: the superconducting onset
transition temperature TC,onset, the zero resistance temperature TC,zeroR, and
superconducting critical temperature TC with magnetic field.
FIG. 5. The resistance of Ba8Si40Al6 versus temperature at various magnetic
fields (H¼ 0, 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, and 9 T). Inset: the
superconducting critical temperature TC of Ba8Si40Al6 versus magnetic field.
203908-3 Li et al. J. Appl. Phys. 113, 203908 (2013)
of the diamagnetism below TC originated from the screening
supercurrents (ZFC regime) and the Meissner effect of mag-
netic flux expulsion (FC regime). Figure 6 demonstrates that
there is a difference between vg in ZFC and FC. As can be
seen from the plot for both ZFC and FC, vg exhibit a super-
conducting drop for T< TC. The TC found for Ba8Si42Al4was lower than that of pure Ba8Si46 (TC¼ 8 K depending on
sample preparation).15,16 However, for this case, where the
Al substitution for Si was 9 at. %, TC is not very heavily sup-
pressed. Such result is quite different from Cu- and Ga-
doping in Ba8Si46.20 For example, in Ba8Si42Cu4, the onset
TC is reduced to 2.9 K while for Ba8Si40Cu6 no superconduc-
tivity was observed down to 1.8 K.20 For Ba8Si40Ga6, the
onset of the superconducting transition occurs at
TC¼ 3.3 K.22 Al-doping has a relatively weaker suppression
on superconductivity. It is clear that the similarity of Al and
Si in electronic structure helps to maintain the superconduct-
ing sp3 network. The theoretical simulations discussed in the
subsequent section confirm this result. Also as shown in
Fig. 6, the existence of the hysteresis between the two mag-
netization curves for the ZFC and the FC modes indicates
that the compound is a type-II superconductor.
According to Fig. 6, the superconducting onset tempera-
ture is 5.7 K, while the bulk transition occurs at around
T¼ 5.1 K. The superconducting volume fraction was esti-
mated to be 29% of the theoretical Meissner value according
to the ZFC susceptibility under 100 Oe. For these estimates,
the theoretical density of 3.56 g/cm3 was used, as estimated
from the XRD data for Ba8Si42Al4. Furthermore, the sample
was roughly a half-disc, with measurements made parallel to
the long axis, so the demagnetization factor (N) for this
direction was estimated to be about 2 in CGS units. In this
case, the value �1/(4p-N) corresponds to the bulk ideal
Meissner effect. These results, therefore, indicate that
Ba8Si42Al4 is indeed a bulk superconductor. Temperature de-
pendent FC magnetization values under different magnetic
fields H¼ 100, 500, 1 k, 5 k, 10 k, 15 k, 20 k, 25 k, 30 k, 35 k,
49 k, 45 k, and 50 k Oe are shown in the inset of Fig. 6. Flux
expulsion (Meissner effect) decreases with increasing exter-
nal field; the magnetic field easily suppresses the magnitude
of superconducting response. As shown in the inset of Fig. 6,
an increasing applied field leads to only a small suppression
in TC but causes a strong reduction in superconducting vol-
ume, a similar behavior has been observed in the Ga-doped
Si clathrates.22
III. THEORETICAL RESULTS
In order to explain the effect of Al doping on supercon-
ductivity, first-principles calculations for Ba8Si46�xAlx(x¼ 0, 6, and 16) systems were carried out using the
CASTEP code with the generalized gradient approximation.
CASTEP are first-principles ab initio calculation packages,
using plane-wave basis sets and suited for periodic sys-
tems.26 The input structural parameters were obtained from
experimental parameters by the geometry optimization func-
tion (see Table I). As discussed above, X-ray diffraction
refinement has shown that for dilute doping, Al is preferen-
tially placed at the 6c site in clathrates. In order to simplify
the Al-substitution model and to give prominence to Al sub-
stitution on 6c sites which bridge the Si20 and Si24 cages, Al
was assumed to be located on all 6c sites for Ba8Si40Al6, and
to occupy 16i and 24k sites with random distribution for
Ba8Si30Al16.
The density of states N(E) for Ba8Si46, Ba8Si40Al6 and
Ba8Si30Al16 are shown in Fig. 7. With the Al-doping increas-
ing, the bands of Si(Al)-3s and -3p have a dispersion and the
fundamental gap effectively shrinks. We attribute this to a
consequence of the enhancement of hybridization by Al-
doping. For Ba8Si46, the Fermi level is just positioned on the
peak of density of states, which is similar to what has previ-
ously been calculated for Ba8Si46.19,20 This peak value is
�33 states/eV at Fermi level.
For Ba8Si40Al6, the density of states N(E) exhibits me-
tallic character which is agreement with resistivity
FIG. 6. The magnetic susceptibility of Ba8Si42Al4 vs. temperature. Main
plot: the susceptibility for conditions of ZFC and FC in the measurement
field of 100 Oe. Inset: The FC susceptibility under different fields from 100
to 50 k Oe for Ba8Si42Al4.
TABLE I. Calculated structures and inequivalent atomic positions for clathrate phases Ba8Si46, Ba8Si40Al6, and Ba8Si30Al16. The notation for atomic positions
follows that of the International Tables for Crystallography.
Ba8Si46 Ba8Si40Al6 Ba8Si30Al16
Symmetry Pm�3n (No.223) Pm�3n (No.223) P1 (No.1)
Lattice constant a (A) 10.328 10.436 10.517
6c (Si,Al) x¼ 0.25, y¼ 0, z¼ 0.5 x¼ 0.25, y¼ 0, z¼ 0.5 x¼ 0.25, y¼ 0, z¼ 0.5
16i (Si, Al) x, y, z¼ 0.6877 x, y, z¼ 0.6836 x, y, z¼ 0.6777
24k (Si,Al) x¼ 0, y¼ 0.6986, z¼ 0.8775 x¼ 0, y¼ 0.6959, z¼ 0.8806 x¼ 0, y¼ 0.6956, z¼ 0.8875
2a (Ba) x, y, z¼ 0 x, y, z¼ 0 x, y, z¼ 0
6d (Ba) x¼ 0.25, y¼ 0.5, z¼ 0 x¼ 0.25, y¼ 0.5, z¼ 0 x¼ 0.25, y¼ 0.5, z¼ 0
203908-4 Li et al. J. Appl. Phys. 113, 203908 (2013)
measurement; the fundamental gap, with width about 0.7 eV
is located well below EF. The Fermi level for Ba8Si40Al6 is
close to a N(E) peak. Our calculations show the N(E) value
is about 30 states/eV at Fermi level. The small reduction in
N(EF) due to dilute Al-doping (the lower valence count) is
consistent with the no big change in TC observed, in the BCS
model for superconductivity. Thus, for dilute substitution of
Al, we find a nearly rigid-band displacement of the Fermi level,
leaving the sp3-connected electronic structure of the framework
relatively unchanged and still conducive to superconductivity.
However, for more heavily substituted Ba8Si30Al16, larger
changes in electronic structure are observed. In Ba8Si30Al16,
the Fermi level is located on the border of fundamental gap,
according to the expected semiconducting behavior, since the
temperature dependence of electrical resistivity of Ba8Si30Al16
is typical for heavily doped semiconductor.
A comparison of the N(E) of Ba8Si46, Ba8Si40Al6, and
Ba8Si30Al16 shows that there are significant changes brought
about by the additional substitution of Al. The valence and
conduction bands are broadened, and the fundamental gap
correspondingly narrowed. The Al-doping effect is similar to
the effect of pressure on Ba8Si46, which also reduces TC as
the lattice constant is reduced.27 Ba8Si30Al16 is a Zintl com-
pound, as there are nominally 184 valence electrons in
Ba8Si30Al16, contributing an average of four electrons for ev-
ery framework atom, enough for a completely filled four-
bonded network. Indeed, the simulation shows a lower total
energy for Ba8Si40Al6 than Ba8Si46 implying that the Al sub-
stituted phase is more stable. An additional stabilization,
besides the Zintl mechanism may come from the increased
polarity of the Al-Si bonds, as experimentally we find that
the Al-substituted materials can be formed more easily as
single-phase materials, as compared to Ba8Si46 which gener-
ally requires high pressure techniques, and this is true even
with relatively dilute Al substitution.
IV. DISCUSSIONS
Isotope effect measurements have revealed that super-
conductivity in Ba8Si46 is of the classic type, arising
from the electron-phonon interaction.17 In the conventional
BCS theory for phonon-mediated superconductivity,28 TC
can be estimated in terms of the Debye temperature HD,
the effective electron-phonon repulsive interaction l*, and
the electron-phonon coupling constant kep: TC ¼ HD
1:45
exp�1:04ð1þkepÞ
kep�l�ð1þ0:62kepÞ
� �. Furthermore, kep can be expressed as
the product of N(EF) and the average electron pairing inter-
action Vep.
The Debye temperature HD¼ 370 K has been evaluated
by specific heat measurement in Ba8Si46.17 We make the rea-
sonable assumption that HD should have the same magnitude
in Ba8Si40Al6. The estimation of kep from TC using the
McMillan formula is not very sensitive to the value of HD.
Moreover, we set the effective electron-phonon repulsion l*to 0.24, which was estimated for Ba8Si46.17,21 From this, we
find that kep¼ 0.88, somewhat smaller than the value found
for Ba8Si46 (kep¼ 1.05, Ref. 21). This implies that Al-doped
Ba8Si40Al6 has a relatively weaker electron-phonon cou-
pling. Using N(EF)¼ 30 states/eV for Ba8Si40Al6, the aver-
age electron pairing interaction Vep was estimated as
29 meV. This value is also smaller than that obtained for Al-
free Ba8Si46, Vep¼ 32 meV based on N(EF)¼ 33 states/eV
and kep¼ 1.05. It thus appears that the TC decrease with Al-
doping can be partially assigned to the weakening of
electron-phonon coupling as well as a decrease of density of
FIG. 7. Density of states for (a) Ba8Si46, (b) Ba8Si40Al6, and (c)
Ba8Si30Al16. Density of states is calculated using 0.1 eV Gaussian broaden-
ing of the band structure.
203908-5 Li et al. J. Appl. Phys. 113, 203908 (2013)
states at Fermi level, however, given the range of TC
observed in various samples of Ba8Si46, it remains possible
that the observed reduction in Ba8Si40Al6 is due entirely to
the small drop in N(EF).
In the framework of BCS superconductivity model, the
superconductivity is strongly related to N(EF). Therefore, an
increase in N(EF) is associated with the occurrence of super-
conductivity. We have systematically investigated the
change in electronic density of states with increasing Al con-
tent, which is helpful for understanding the suppression of
superconductivity in the Al-substituted clathrates. The calcu-
lated N(E) is shown in Fig. 7 for Ba8Si46�xAlx (x¼ 0, 6, and
16). With increase in Al substitution content, the Fermi level
shifts toward lower energy from the peak of conduction-
band to the border of conduction-band due to decrease of
valence electrons of Al. Moreover, because of the Zintl
bond-filling mechanism and strong hybridization between Al
with Si network, the DOS features broaden, which results in
the fundamental gap narrowing. Since the band width is
related to the degree of electron overlap, it makes sense that
with Al content increasing the hybridization between Al and
Si framework become tighter and electron overlap tends to
enhance, so the Ba8Si30Al16 would have a larger bandwidth.
The most important is the density of states decrease with Al
increasing, for example, N(EF) decreases from 30 states/eV
in Ba8Si40Al6 to 8 states/eV in Ba8Si30Al16, which implies
that no superconductivity probably occurs. It is worth men-
tioning that the Al substitution causes a drop of the electron
number in the conduction-band, which directly associated
with carrier concentration. The concentration of electron is
estimated by the area integration of conduction-band occu-
pied is 9.08, 5.07, and 0.62 corresponding to Ba8Si46�xAlx(x¼ 0, 6, and 16), respectively. Therefore, the Al-doping
induced decrease of carrier concentration also plays an im-
portant role in the suppression of superconductivity.
In conclusion, a combined experimental and theoretical
study of the effect of Al substitution on the superconductiv-
ity of the type I clathrate Ba8Si46�xAlx is presented. In
Al-doped clathrates, the Al state was found to be strongly
hybridized with the cage conduction-band state. Al substitu-
tion resulted in a shift toward a lower energy, a decrease of
N(EF) and a lowering of the carrier concentration. These
play key roles in the suppression of superconductivity.
ACKNOWLEDGMENTS
This work was supported in part by the National Science
Foundation (DMR-0821284), NASA (NNX10AM80H and
NNX07AO30A), and the National Natural Science Foundation
of China (51072023).
1J. L. Cohn, G. S. Nolas, V. Fessatidis, T. H. Metcalf, and G. A. Slack,
Phys. Rev. Lett. 82, 779 (1999).2G. S. Nolas, T. J. R. Weakley, J. L. Cohn, and R. Sharma, Phys. Rev. B
61, 3845 (2000).3S. Bobev and S. Sevov, J. Solid State Chem. 153, 92 (2000).4H. Shimizu, T. Kume, T. Kuroda, S. Sasaki, H. Fukuoka, and S.
Yamanaka, Phys. Rev. B 68, 212102 (2003).5S. B. Roy, K. E. Sim, and A. D. Caplin, Philos. Mag. B 65, 1445
(1992).6J. S. Tse, K. Uehara, R. Rousseau, A. Ker, C. I. Ratcliffe, M. A. White,
and G. MacKay, Phys. Rev. Lett. 85, 114 (2000).7B. B. Iversen, A. E. C. Palmqvist, D. E. Cox, G. S. Nolas, G. D. Stucky,
N. P. Blake, and H. Metiu, J. Solid State Chem. 149, 455 (2000).8W. P. Gou, Y. Li, J. Chi, J. H. Ross, Jr., M. Beekman, and G. S. Nolas,
Phys. Rev. B 71, 174307 (2005).9R. P. Hermann, V. Keppens, P. Bonville, G. S. Nolas, F. Grandjean, G. J.
Long, H. M. Christen, B. C. Chakoumakos, B. C. Sales, and D. Mandrus,
Phys. Rev. Lett. 97, 017401 (2006).10B. C. Sales, B. C. Chakoumakos, R. Jin, J. R. Thompson, and D. Mandrus,
Phys. Rev. B 63, 245113 (2001).11T. Kawaguchi, K. Tanigaki, and M. Yasukawa, Appl. Phys. Lett. 77, 3438
(2000).12Y. Li and J. H. Ross, Jr., Appl. Phys. Lett. 83, 2868 (2003).13Y. Li, J. Chi, W. P. Gou, S. Khandekar, and J. H. Ross, Jr., J. Phys.:
Condens. Matter 15, 5535 (2003).14Y. Li, W. P. Gou, J. Chi, V. Goruganti, and J. H. Ross, Jr., AIP Conf.
Proc. 772, 331 (2005).15H. Kawaji, H. O. Horie, S. Yamanaka, and M. Ishikawa, Phys. Rev. Lett.
74, 1427 (1995).16S. Yamanaka, E. Enishi, H. Fukuoka, and M. Yasukawa, Inorg. Chem. 39,
56 (2000).17K. Tanigaki, T. Shimizu, K. M. Itoh, J. Teraoka, Y. Moritomo and S.
Yamanaka, Nature Mater. 2, 653 (2003).18S. Saito and A. Oshiyama, Phys. Rev. B 51, 2628 (1995).19K. Moriguchi, M. Yonemura, A. Shintani, and S. Yamanaka, Phys. Rev. B
61, 9859 (2000).20Y. Li, Y. Liu, N. Chen, G. H. Cao, Z. S. Feng, and J. H. Ross, Jr., Phys.
Lett. A 345, 398 (2005).21D. Conn�etable, V. Timoshevskii, B. Masenelli, J. Beille, J. Marcus, B.
Barbara, A. M. Saitta, G.-M. Rignanese, P. M�elinon, S. Yamanaka, and X.
Blase, Phys. Rev. Lett. 91, 247001 (2003).22Y. Li, R. H. Zhang, Y. Liu, N. Chen, Z. P. Luo, X. Q. Ma, G. H. Cao,
Z. S. Feng, C.-r. Hu, and J. H. Ross, Jr., Phys. Rev. B 75, 054513
(2007).23A. C. Larson and R. B. von Dreele, Los Alamos National Laboratory
Report No. LAUR 86–748, 2000.24B. H. Toby, J. Appl. Cryst. 34, 210 (2001).25J. Bardeen, L. N. Cooper, and J. R. Schrieffer, Phys. Rev. 108, 1175
(1957).26M. C. Payne, M. P. Teter, D. C. Allan, T. A. Arias, and J. D.
Joannopoulos, Rev. Mod. Phys. 64, 1045 (1992).27P. Toulemonde, A. San Miguel, A. Merlen, R. Viennois, S. Le Floch, C.
Adessi, X. Blase, and J. L. Tholence, J. Phys. Chem. Solids 67, 1117
(2006).28W. L. McMillan, Phys. Rev. 167, 331 (1968).
203908-6 Li et al. J. Appl. Phys. 113, 203908 (2013)