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:

ylibp@hotmail.com

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)