Kondo effect induced suppression of superconductivity in Y1−xPrxBa2Cu3O7−δ

PHYSICA@ ELSEVIER

Physica C 282-287 (1997) 1395-1396

Kondo Effect Induced Suppression of Superconductivity in Yl-=PrxBa2Cu307_~ Duo Jin a, Jianlin Luo a, Yupeng Wang% Hao Jin b, Guanghui Rao b, Shusheng Yan c * aCryogenic Laboratory, CAS, P.O.Box 2711, Beijing 100080,China q n s t i t u t e of Physics, CAS, P.O.Box 603, Beijing 100080, China CDepartment of Physics, Peiking University, Beijing 100871, China Low temperature specific heat measurements on YI-=Pr=Ba2Cu307 compouds with x=0.45, 0.6 in temperatures ranging from 0.1 to 30K, and resistance measurements with x=0.45, 0.5 for both polycrystalline and film sample in temperatures ranging from 4.2 to 100K are reported. The Kondo effect has been observed from both the specific heat and resistance mearements. The Tc depression and metal-insulator transition in this system are attributed to localization due to the Kondo interaction of Pr 4f-electrons with hole carrier in Cu02 plane.

Among R~Ba:Cu307 compounds with R = Y and rare earth elements, only PrBa2Cu307 does not exhibit superconductivity. The superconducting t e m p e r a t u r e Tc of Yl-xPr~Ba2Cu307 decreases monotonically with x increase, and finally the system undergoes a metal-insulator transition at the critical doping density x,-~0.55[1]. Several models were proposed to account for the effect of Pr-doping on superconductivity, such as hole filling, pair breaking and hole localization. Here we report an investigation of low temperature specific heat and resistance measurements on YI-xPr~Ba2Cu307 materials. Kondo interaction between P r magnetic moments and conduction holes has been observed and the results support the mixed local state model proposed by Y.Wang et al[2]. We measured the specific heat C in temperatures ranging from 0.1 to 30K. C decreases as the t e m p e r a t u r e decreases from high temperatures. When T drops to a b o u t 10K, broad peaks are found for b o t h samples. The upturns below 1K can be observed clearly, which is due to the contribution of a nuclear Schottky anomaly. The measured specific heat consists of four main contributions: the nuclear Schottky anomaly contribution CN, the lattice contribution Cph, The conduction electron contribution Ce and the P r magnetic contribution CK. The nuclear specific heat CN can be discerned from very low tem*This work is supported by the National Center of Research on Superconductivity 0921-4534/97/$17.00 © Elsevier Science B.V. All rights reserved. Pll S0921-4534(97)00792-2

peratures since it is T -2 t e m p e r a t u r e dependent. The lattice contribution Cph and electron contribution Ce can be obtained from C(T) vs T d a t a at high temperatures, because the t e m p e r a t u r e dependence of specific heat above 18K exhibits a linear t e r m plus T 3 term.

,

4000

,

3200

2 4 0 0 ~ 1600 BOO

3200~ 2400 1600 0

, ""

""

""

I

4000

BOO

O

3

6

9

1~

T(K) Figure 1. Pr magnetic contribution to the specific heat, solid lines are Kondo fits.

CK can be calculated by subtracting the nuclear, the phonon and conduction electron contributions from the total specific heat. Fig.1 shows CK versus T. CK fits well with a Kondo effect

D. Jin et al./Physica C 282-287 (1997) 1395-1396

1396

expression[3]:

1.0

C g = (1 -- a)(2 -- a)RHT~ln(2S + 1) T2(1 + TK/T) ~

(1)

i

0.8

0

0.6 where TK is the K o n d o t e m p e r a t u r e , S is the P r spin and RH is the universal gas constant. The best fitted p a r a m e t e r s are a = 3.8, S = I , and TK=14, 15 for x=0.45 and 0.6 respectively. The spin S = I indicates t h a t the configuration of the ground state of P r ions is 4 f 2, so the valence of the P r ion is +3.

0

i

0

0

!

0

0

I

0

0

-

oo 0.4

_

0.2

.;"

0.0 0

0.2

,

0.4

0.6

0.8

X

26 Figure 3. 5 vs. x for Yl-xPrxBa2Cu307-~, the solid circles indicate the sample is superconducting, and the open circles indicate the sample is nonsuperconducting. The solid line represents x + 5=0.5.

25 24

~" 2

,

~:

~"

,

"

'

40

•

"

'

60

'

80

24

20 16 12 0

i

I

~

i

20

40

60

80

100

T(K)

Figure 2. Resistance (R) versus T e m p e r a t u r e (T). Fig.2 shows the resistance versus t e m p e r a t u r e for the polycrystalline Yo.55Pro.45Ba2Cu307 and the film Yo.sPro.5Ba2Cu307 between 4.2K and 100K. For b o t h samples, with decreasing the temperature, R decreases at first to a minium value, then goes up and a peak can be observed before entering the superconducting state. We found t h a t the normal state resistance of the two samples can be expressed by the following equation:

R = Ro + AT + CIn(TK/T)

(2)

T h e third t e r m on the right hand side of eq.(2) is from the K o n d o effect. T h e solid lines in fig.2 are the best fits using eq.(2). The results coincides with the specific heat behaviour carried out on the same sample with x=0.45.

In YI-xPr~BauCu307, the Pr ion has a larger radius compared to Y and other rare earth elements (except for Ce), the carrier holes in the Cu02 plane can be localized by f-electrons of the Pr ion and form a mixed local state[2] because of a strong Kondo interaction between them. When P r concentration increase from 0 to a b o u t 0.5, the mobile holes in Cu02 plane decrease gradually and so lead to the depression of Tc, and eventually the metal-insulator transition occurs at xN0.55 when almost all the holes are localized. Such a result is similar to the depression of Tc with decreasing 5 in the YBa2Cu307_~ system. The recent work[4] on YI-~Pr~Ba2Cu307-~ with different Pr-concentration x and oxygen content also supports the above view. Fig.3 shows versus x for YI-~PrxBa2Cu307-~ samples. We found approximately all the superconducting samples have x + 6