Superconducting transition at 12.5 K in RbxC60

21 July 1995 L" ?

CHEMICAL PHYSICS LETTERS

.;- c. :"

ELSEVIER

Chemical PhysicsLetters 241 (1995) 154-160

Superconducting transition at 12.5 K in

RbxC60

Hideyuki Funasaka, Koji Sakurai, Kenji Sugiyama, Kazunori Yamamoto, Takeshi Takahashi Nuclear Technology Development Division, Tokai Works, Power Reactor and Nuclear Fuel Development Corporation, Tokai-mura, lbaraki 319-11, Japan

Received 20 March 1995

Abstract

The effect of heat treatment of RbxCro (x = 3, 4, 5, and 6) on the superconducting transition has been studied with shielding diamagnetism measurements. For RbxC6o (x = 4, 5, and 6), a superconducting transition at 12.5 K appeared in the course of annealing treatment at 400°C and disappeared on further prolonged annealing, while a clear superconducting transition at 30 K due to the Rb3Ceo phase still remained. We speculate that this superconducting transition at 12.5 K is due to a certain phase which appears in the course of the annealing treatment.

1. Introduction

The most important aspect of the fullerenes as a new material is derived from the enormous potential for yielding derivatives [1], which can be roughly divided into two types, exohedral and endohedral. The simplest type of exohedral fullerene derivatives is based on the interstitial doping with alkali metals of crystalline fuUerene solids. Because those alkali metal fullerides exhibited superconductivity [2] at unprecedentedly high temperatures for organic conductors which were surpassed only by the copper-oxide-based ceramics, much attention has been paid to a variety of studies of their chemical and physical properties [3-6]. Particularly, efforts to increase the superconducting transition temperature (Tc) resulted in Rb3C60 (Tc = 28-30 K) [7,8] and RbCs2C60 (Tc = 33-34 K) [9,10]. At the same time, it has been shown that the T~ values for the A3C60 superconductors are involved both in their lattice parameters and in the density of

states at the Fermi level N ( E F) [11,12]. Therefore, clarification of the relationship between structure and superconducting property is essential for a detailed understanding of the mechanism of the superconductivity in these systems. So far, structural data have been reported for the non-conducting, rock-salt type A1C60 (where A denotes the alkali metals) [13], an intercalated bodycentered tetragonal (bct) structure for A4C60 [14], body-centered cubic (bcc) structures A6C60 [15], and the superconducting, intercalated face-centered cubic (fcc) material A3C60 [16]. In addition to the above, Zhu et al. [17] have found a dilute fcc doped phase, 0 < x < 1, and a substoichiometric bcc phase, x ,-, 5. However, an investigation of the growth process of the superconducting phase or determing the existence of other superconducting phases in RbxC60 with heat treatment has not been reported yet, because the kinetics of the reaction of C60 with rubidium metal is more complex. In this Letter, we have studied the properties of

0009-2614/95/$09.50 © 1995 Elsevier Science B.V. All fights reserved SSD1 0009-2614(95)00628-1

155

H. Funasaka et al. / Chemical Physics Letters 241 (1995) 154-160

alkali-metal doped C6o superconductors prepared by the reaction of C6o with r u b i d i u m metal. Particularly, the effect of heat treatment o f RbxC6o ( x = 3, 4, 5,

and 6) on the s u p e r c o n d u c t i n g transition was determ i n e d from shielding d i a m a g n e t i s m m e a s u r e m e n t s with successive a n n e a l i n g treatments. In addition,

(d)

(a)

i

O

i

,oo,..°,~.°°°.o.o,..I.

--o-.- •

1 •

°•

J~

°o

-I .2

~. -2

°

°°•°

°°°

u

i

FC

°o." -°-'°°'

FC

-0.6

¢

0

,. ° o,° ,: ,°°- .-,-,,.°~°----,,o.,,,,.°-°" °°o°~°..°.~.IN,°,

e

l

-1.8

.~

ZFC......... -'""'" ,o.,.,.o,,,,o....,,-,

;

-2.40

:'o

ZFC ...'"

-3

""

:'5

o, ,~°°°°°°,""'°'°°°

2'o z'5

X 310

315

40

-4 i 0

;

:'o

Temperature (K)

.°.,°°o:°O°* ,o°.~°~,°

~,o..°.°°.°°~..

I

I

2'o zs

310

3i5

40

Temperature (K)

(b) •

(e)

•

' *'".~.................. FC' '..,...,....:,... ' . . .'. . . . . . '

i

°~°°°°°'°°°'-•''

FC

O

°°

°

~, -2.2

..,v.... °.-°°

ZFC

,,.%.,

-0.5

,,°,""

,,,,"

° •

~-4.4

~. -1

-~ ~ -6.6

°°

I

o°

..~ -1.5

ZFC .... """ " .,.,.,,~.o°O-" I

-8.8 0

I

5

10

I

i

I

I

I

15 20 25 Temperature (K) i

i

i

1

I

30

35

i

i

X

40

-2o

;

lJO

1~5 2~0 25 Temperature (K)

3~0

3J5

40

(c)

•b°E

,.°.....°,.°'• ..... .t............................ :? . . . . . .

FC

_0.9

i

°°, o.

-1.8 •

._.q-2.7 m

°

-3.6

°°

ZFC °.,'""

~..~,°'""°*°

°°°°.....~ ° ;

110

1#S

210

Z'5

310

315

40

Temperature (K)

Fig. 1. Temperature dependence of the magnetic susceptibility for a sample of nominal composition Rb6C60 reacted at 400°C for various reaction times. (a) For 50 h of annealing; (b) for 150 h of annealing; (c) for 200 h of annealing; (d) for 250 h of annealing; (e) for 300 h of annealing. The curves were obtained by cooling in a zero field, subsequent warming in a 10 Oe field (ZFC) and by cooling in a 10 Oe field (FC). Solid arrows show the 12.5 K anomaly.

156

H. Funasaka et al. / Chemical Physics Letters 241 (1995) 154-160

X-ray powder diffraction measurements of these samples were taken in order to elucidate the relationship between structure and superconducting property.

1

I

L

I

E 0 ~,

***** o **

o

.~-1 #

-2

2. E x p e r i m e n t a l

Carbon soot was prepared by an arc heating discharge in 100 Torr of helium. Details of the production of this arc heating discharge method have been described previously [18,19]. C60 was obtained by extraction with toluene from the carbon soot followed by high performance liquid chromatography (HPLC) with toluene/hexane eluants on an octadecylsilica bonded column. The C60 was then washed using C2HsOH and heated at 200°C for 6 h under vacuum to remove the solvent. The purity of C60 was confirmed to be at least 99.9% by HPLC and mass spectroscopy. The samples of nominal composition RbxC60 (x -- 3, 4, 5 and 6) were prepared by reaction of C6o with rubidium. Typically 30 mg of C6o were placed in a pyrex tube (6 mm outer diameter) along with the alkali dopant under argon atmosphere. The quantity of rubidium was controlled by cutting Rb-filled pyrex capillary tubing to the required length. The reagents were sealed in evacuated pyrex tubes and heated for a given time at 400°C. The dc magnetization M(T) of the RbxC60 sample was measured for every 50 h in a SQUID magnetometer (Quantum Design MPMS) in an applied field of 10 Oe. The sample was first cooled to 5 K in a zero field and data were taken in a 10 Oe field on warming the sample to 35 K (zero-fieldcooled (ZFC) data). The sample was then cooled in the same field down to the base temperature (fieldcooled (FC) data). In addition, the superconducting volume fraction exhibits shielding diamagnetism expressed as a percentage of a Nb standard measured at 5K. The diffraction patterns were collected with a Rigaku (RINT 1000) diffract,meter using Cu Kct radiation at room temperature.

3. Results and discussion

The temperature dependence of magnetic susceptibility for the nominal composition Rb6C60 reacted

_ ~

-3 ~'~

~. 0 £;

.

-4

~°

.

J

J

** o

°°

FC

o

*

._. ,~1.4

o

4-a

.

~

: o

b_2

e

e

-2.6

-3.2

~o~ZFC

0

i

10

i

20

J

30

40

Temperature (K) Fig. 2. Expanded view of a portion of Fig. lc. It is clear that the anomaly at 12.5 K in temperature dependence of the magnetic susceptibility is undoubtedly a superconducting transition.

at 400°C with various reaction times is shown in Fig. 1. At the 400°C reaction, the superconductivity at 30 K has already appeared after 50 h reaction. After 150 h reaction the superconductivity volume fraction had attained the maximum value, and the superconductivity finally disappeared at 300 h reaction. The onset superconducting transition temperature, Tc, was constant at 30 K throughout the annealing process. A superconducting transition-like anomaly at about 12.5 K appeared at the 400°C reaction for 200 h (as shown in Fig. lc) and disappeared at a reaction time of more than 200 h (as shown in Fig. ld). Fig. 2 shows the expanded view of Fig. lc. Considering the significant kinks in both the FC and the ZFC data, it is distinct that this anomaly exhibits the superconducting transition. The same measurement was performed on RbxC60 ( x = 3, 4, and 5 in the feed) for every 50 h of reaction time up to 400 h. This superconducting transition at 12.5 K was also observed on x = 4 and 5 for 250 and 200 h, respec-

157

H. Funasaka et al. / Chemical Physics Letters 241 (1995) 154-160

tively, and disappeared on further prolonged annealing. With increasing of the rubidium composition from x = 4 to x = 6, the volume of the superconducting transition at 12.5 K was increased, while the transition temperature was constant at 12.5 K. However, for the x---3 specimen, this transition at 12.5 K was not observed. The reaction time dependence of the superconducting phase volume fractions in RbxC60 (x = 3, 4, 5, and 6) is shown in Fig. 3. In the sample prepared with the nominal composition Rb3C60 ( x - - 3 ) , the superconducting phase volume fraction reached a maximum of greater than 40%, a high value for a powder sample, while the volume fraction of RbxC60 (x = 4, 5, and 6) increased with the reaction time, reached a maximum, and then decreased, respectively. This reaction time dependence of the superconducting phase volume is related to the growth and disappearance between the superconducting phase (fcc for x = 3) and non-superconducting phases (bct for x = 4 and bcc for x = 6). Now this superconducting transition at 12.5 K appeared shortly after the superconducting phase volume fraction began to decrease. It is assumed that this behavior occurs not

50

in the mixing stage but in the homogenization process [8]. In the case of a fullerene superconductor, the transition temperature (Tc) in accordance with the BCS model is given by the well-known expression Tc = toph e x p [ - 1 / V N ( E F ) ] ,

(1)

where OJph is the frequency of intramolecular vibrations determining the pairing of electrons, N(EF) is the energy density of electron states at the Fermi level, and V is the electron-phonon coupling strength. The increased character of the empirical dependence Tc (ao: lattice constant) might be explained on the basis of expression (1) taking into account the well-known dependence N(E F) ~ a~ and presuming the magnitude of V, which corresponds to the interaction of an electron with intramolecular vibrations to be independent on a o. Indeed, in the superconducting fcc A3C60 alkali metal compounds, the ionic radius of the alkali metal cations determines the lattice constant (ao) , and this in turn correlates with the superconducting temperature through changes in the density of states at the Fermi level [20]. In other words, alkali atoms with larger

I

I

I

I oo

200 Reaction time (h)

3oo

RbxC60

40 a~ e-

.o 30

E 20 _m 0 > 1o

o~ o

400

Fig. 3. Reaction time dependence of the superconducting phase volume fraction for RbxC 60 ( x = ( O ) 3, ( [] ) 4, ( ~ ) 5, and ( A ) 6) reacted at 400°C. The superconductingtransition at 12.5 K appeared shortly after the volume fraction began to decrease for x = 4, x = 5, and x=6.

H. Funasaka et al. / Chemical Physics Letters 241 (1995) 154-160

158

x=3"L °

~

A~.

~

~

o ~

~

--

¢-

~

,~o

o o

v

o

o

o

u'J e .-

e

¢-

5

40

20

60

2 0 (deg.) Fig. 4. X-ray powder diffraction profiles for RbxC6o reacted at 400°C (nominal x values of 3, 4, 5, and 6). The annealing time is 350 h for x = 3, 250 h for x = 4, 200 h for x = 5 and 200 h for x = 6, respectively. Open circles, filled circles and solid triangles show peaks from the fcc-(I), the bcc and the bet phases, respectively. The peaks indicated by an asterisk are due to the Mylar film of the sample holder.

ionic radii produce a larger lattice constant, which should increase the density of states at the Fermi level and thus yield a higher To. In addition, there are few reports about a superconducting transitionlike anomaly except for A 3 C 6 0 . Kobayashi et al. [21] reported that this superconducting transition-like anomaly at about 9 K appeared on KxC60, and this 40

i

,

(a) 35 3O

was due to the dilute bcc phase. It has also been reported that for pressures 0.2 < p < 1.09 GPa a small anomaly appears in the temperature range 7 < T < 10 K in single-phase K3C6o [22]. Therefore, it is considered that an increase in Tc from 9 to 12.5 K is also due to the ionic radii of alkali metals (potassium and rubidium). With increasing the annealing tem40

-/ i

I

I

35 0

30

oo~ 25

25 v

v o

I

(b)

o I-

20

ge

15

10

III

21

IQ

1! 11

.. s D 7 - + ~ 2 c 8

•

•

5 TU~b 0

13.8

di

U:'~ i/,

I

w

14.0

14.2

14.4

Lattice Constant a 0A

14.6

13.6

14.0

I 142

I 14.4

14.6

Lattice Constant a0A

Fig. 5. (a) The relationship between Tc and a o (lattice constant) for A3C60 superconductors over a wide range of a 0 [6]. (b) The correlation between Tc and the 300 K foe lattice constant, with hydrostatic pressure as the implicit variable [24]. ([2]) Rb2CSC6o; (11) Rb3C6o; ( 0 ) K3C60; ( • ) Na2RbC6o at one atom.

H. Funasaka et al. / Chemical Physics Letters 241 (1995) 154-160

perature from 400 to 450°C, the appearance time of the superconducting transition at 12.5 K was reduced from 200 to 20 h. In view of the results described above, it can be considered that this superconducting transition at 12.5 K is due to a certain structure in RbxC60. In order to elucidate the structure of this superconducting transition at 12.5 K, X-ray diffraction patterns were obtained for the samples with x values of 4, 5, and 6 for 250, 200, and 200 h, respectively (as shown in Fig. 4) in which the superconducting transition at 12.5 K appeared. In addition to these samples, the diffraction pattern for a sample of nominal composition Rb3C60 for 350 h is appended as a reference. All of the lines can be indexed to a fcc-(I) structure with a lattice constant of a 0 = 14.38 ± 0.02 ,~. This observed lattice constant and almost all of the relative integrated intensities except for the (422) peak are in fair agreement with those previously reported [17,23]. On the other hand, the samples with x values of 4, 5, and 6 for 250, 200, and 200 h, respectively, as shown in Fig. 4, consist of mixtures of three or more phases. For the sample of nominal composition Rb6C60 , three phases were observed: the majority of the lines can be indexed too a bcc phase with a lattice constant of 11.54 ± 0.01 A which is isostructual with Rb6C60 , described by Zhou et al. [15], the others lines can be indexed to a fcc-(I) phase which exhibits the superconducting transition at 30 K, and to a bct o phase with lattice constant a 0 = 11.97 + 0.03 A, c = 11.02 + 0.05 ~, which has been already reported by Fleming et al. [14]. For the sample with x = 4 and 5, three phases were also observed, that is, a fcc-(I), a bct and a bcc phase. Additionally, in those samples (x = 4, 5, and 6) extra diffraction peaks appeared, indicating the formation of other phase than fcc-(I), bct and bcc. Special attention was given to the presence of a new phase with the lattice constant 14.1 A. The correlation of Tc with the unit cell size of A3C6o [20] and data on the pressure dependence of Tc [22,24] and the lattice constant (a 0) [25] suggest that the lattice parameter of the phase for this superconducting transition at 12.5 K should be near 14.1 A for the transition temperature (as shown in Fig. 5). Some of the observed diffraction peaks were assigned to a cubic phase with the lattice constant 14.1 A. How-

159

ever, this new phase with the lattice constant a 0 = 14.1 ,A was less clarified, because the diffraction peaks for this new phase partly overlap those for other coexisting phases, and the superconducting volume fraction for 12.5 K is not large enough to be clearly observed in X-ray diffraction measurements. On further prolonged annealing, the superconducting transition at 12.5 K vanished, subsequently the superconducting transition at 30 K also vanished with the disappearance of the fcc-(I) (as shown in Fig. le), and finally the sample with x = 6 because only a single phase with a bcc structure.

4. Conclusion The effect of heat treatment of RbxC60 (x = 3, 4, 5, and 6) on the superconducting transition was determined from shielding diamagnetism measurements. For RbxC60 (x = 4, 5, and 6), a superconducting transition at 12.5 K appeared on prolonged annealing at 400°C and disappeared with further annealing, while a clear superconducting transition at 30 K due to the Rb3C60 phase still remained. We speculate that this superconducting transition at 12.5 K is due to a certain phase which appears in the course of the annealing treatment.

Acknowledgement We express thanks to Toshiaki Ishiguro (Genshiryoku Gijyutu Co.), and Yoshiharu Kano (Genshiryoku Gijyutu Co.) for their experimental help.

References [1] D.E. Koshland Jr., Science 254 (1991) 1706. [2] A.F. Hebard, M.J. Rosseinsky, R.C. Haddon, D.W. Murphy, S.H. Glarum, T.M. Palstra, A.P. Ramirez and A.R. Kortan, Nature 350 (1991) 600. [3] J.E. Fischer and P.A. Heiney, J. Phys. Chem. Solids 54 (1993) 1725. [4] D.W. Murphy, M.J. Rosseinsky, R.M. Fleming, R. Tycko, A.P. Ramirez, R.C. Haddon, T, Siergrist, G. Dabbagh, J.C. Tully and R.E. Walstedt, J. Phys. Chem. Solids 53 (1992) 1373. [5] W.E. Pickett, in: Solid state physics, Vol. 48, eds. H. Ehrenreich and F. Spaepen (Academic Press, New York, 1994) p. 225;

160

[6] [7]

[8] [9] [10]

[11]

[12] [13]

[14]

H. Funasaka et al. / Chemical Physics Letters 241 (1995) 154-160

C.M. Lieber and Z. Zhang, in: Solid state physics, Vol. 48, eds. H. Ehrenreich and F. Spaepen (Academic Press, New York, 1994) p. 349. K. Tanigaki, I. Hirosawa, J. Mizuki and T,W. Ebbesen, Chem. Phys. Letters 213 (1993) 395. M.J. Rosseinsky, A.P. Ramirez, S.H. Glarum, D.W. Murphy, R.C. Haddon, A.F. Hebard, T.T,M. Palstra, A.R. Kortan, S.M. Zahurak and A.V. Makhija, Phys. Rev. Letters 66 (1991) 2830. K. Holczer, O. Klein, S. Huang, R.B. Kaner, K. Fu, R.L. Whetten and F. Diederich, Science 252 (1991) 1154. K. Tanigaki, T.W. Ebbesen, S. Saito, J. Mizuki, J.S. Tsai, Y. Kubo and S. Kuroshima, Nature 352 (1991) 222. K. Sakurai, M. Kato, H. Funasaka, K. Yamamoto, Y. Wada and Y. Klshimoto, The Japan. Soc. of Appl. Phys. and Related Soc. The 39th Spring Meeting (1992). K. Tanigaki, I. Hirosawa, T.W. Ebbesen, J. Mizuki, Y. Shimakawa, Y. Kudo, J.S. Tsai and S. Kuroshima, Nature 356 (1992) 419. M.A. Schluter, M. Lannoo, M. Needels and G.A. Baraff, Phys. Rev. Letters 68 (1992) 526. D.M. Poirier, T.R. Ohno, G.H. Kroll, P.J. Benning, F. Stepniak, J.H. Weaver, L.P.F. Chibante and R.E. Smalley, Phys. Rev. B 47 (1993) 9870. R.M. Fleming, M.J. Rosseinsky, A.P. Ramirez, D.W. Murphy, J.C. Tully, R.C. Haddon, T. Siegrist, R. Tycko, S.H. Glarum, P. Marsh, G. Dabbagh, S.M. Zahurak, A.V. Makhija and C. Hampton, Nature 352 (1991) 701.

[15] O. Zhou, J.E. Fischer, N. Coustel, S. Kycia, Q. Zhu, A.R. McGhie, W.J. Romanow, J.P. McCauley Jr., A.B. Smith III and D.E. Cox, Nature 352 (1991) 462. [16] P.W. Stephens, L. Mihaly, P. Lee, R.L. Whetten, S.-M. Huang, R.B. Kaner, F. Diederich and K. Holczer, Nature 351 (1991) 632. [17] Q. Zhu, O. Zhou, N. Coustel, G. Vaughan, J.P. Mccauley Jr, W.J. Romanow, J.E. Fischer and A.B. Smith III, Science 254 (1991) 545. [18] K. Yamamoto, H. Funasaka, T. Takahashi and T. Akasaka, J. Phys. Chem. 98 (1994) 2008. [19] H. Funasaka, K. Sugiyama, K. Yamamoto and T. Takahashi, J. Phys. Chem. 99 (1995) 1826. [20] R.M. Fleming, A.P. Ramirez, M.J. Rosseinsky, D.W. Murphy, R.C. Haddon, S.M. Zahurak and A.V. Makhija, Nature 352 (1991) 787. [21] M. Kobayashi, Y. Akahaka, H. Kawamura, H. Shinohara, H. Sato and Y. Saito, Japan. J. Appl. Phys. 31 (1992) L1051. [22] G. Sparn, J.D. Thompson, S.-M. Huang, R.B. Kaner, F. Diederich, R.L. Whetten, G. Gruner and K. Holczer, Science 252 (1991) 1829. [23] P.W. Stephens and L. Mihaly, Phys. Rev. B 45 (1992) 543. [24] J.E. Schirber, D.L. Overmyer, W.R. Bayless, M.J. Rosseinsky, D.W. Murphy, Q. Zhu, O. Zhou, K. Kniaz and J.E. Fischer, J. Phys. Chem. Solids 54 (1993) 1427. [25] O. Zhou, G.B.M. Vaughan, Q. Zhu, J.E. Fischer, P.A. Heiney, N. Coustel, J.P. McCauley Jr. and A.B. Smith III, Science 255 (1992) 833.