- Email: [email protected]

Temperature dependent Hell effect in Nd2.xCexCuO4 end "60 K" YBa2CU3Oy single crystals. Z.Z. Wang', D.A. Brawner 1, T.R. Chien, N.P. Ong 1, J.M. Tarascon =, and E. Wang =, Department of Physics ~, Princeton University, Princeton, N.J. 08544 Bell Communications Research 2, Redbank, NJ 07701 The Hall coefficient is studied in single crystals of the two superconductors Nd=.=Ce,CuO4 and "60 K" YBa=CusOy. The strong temperature dependence is discussed. We also report the temperature dependence of the out-of-plane resistivity in the latter. 1. Introduction We have studied the transport properties of single crystals of Nd2.xCexCuO4 ('NCCO") 1 and the "60 K" phase of YBa2Cu3Oy (123) 2. Superconducting crystals of NCCO (Tc ~20 K) are obtained by reducing as-grown crystals. Crystals of the "60 K" phase of 123 are made by reducing as-grown "90 K" crystals in an Ar atmosphere (8 % oxygen) for 7-13 days.

0 -2 -4

NdCeCuO O

-6 O

~ -8 LJ ~ -10

O

'o -12

2. Hall effect of Nd2.xCexCuO4 From Hall studies on several crystals we find that the Hall coefficient RH in NCCO changes significantly at low temperatures T. (Earlier reports 3 Stopped at 50 K.) Between 200 K and 4 K, RH changes from -2x10 -9 to -14 x 10-9 m3/C (Fig. 1). [Below 20 K, the applied field H ( - 15 T) exceeds Hc2 (= 6 T) for H normal to the CuO 2 planes.] In a second crystal we see a similar divergence of R H as T~0. However, the divergent term is positive. ( RHCrosses zero at 80 K.) In all the superconducting crystals, Tc lies between 18 to 20 K. Unlike ceramic samples, the in-plane resisitvity Pac is metallic and equal to -450 IJ-Qcm at 290 K. Interestingly, Pac saturates to a Tindependent value below ~30 K. The strong T-dependence of RHWaS previously observed in 1234,5. In NCCO, RH appears to diverge at low T, where Pac is independent of T. Since the mobility (or mobilities in a two-band model) must be T-independent in this range of T, the conventional Lorentz Hall resistivity cannot 0921-4534/89/$03.50 © Elsevier Science Publishers B.V. (North-Holland)

0 0 O O

0 -14

oo o

T¢

4'0 6'o 85

I

I

I

;2014o

200

T(K) FIG 1 The temperature dependenceof the Hall coefficient RHIn slngle-cwstal NCCOwtth Tc= 20 K. The field of 15 T is applied normal to the CuO2 planes. change with T. Hence, Fig. 1 presents a strong argument that the divergent Hall signal arises from an extra anomalous term which varies approx, as 1/T. (This term is unlikely to be magnetic skew scattering. The effect of impurities on the asymmetric scattering in older high-Tc systems was recently reported6.) The results here are the strongest evidence to date for the existence of the anomalous Hall term in the normal state of the high-Tc oxides.

3. Hell effect of "60 K" 123. In the 60 K phase of 123, we find that RH is also T-dependent (Fig. 2). The Hall density nH

Z.Z. Wang et al. I Temperature dependent Hall effect in Nd2_ ~Ce~CuO~

1014

10 60K YBCO

2.5

0.3

'E 2.0 u £ 1.5 T ,==

8

1.0 w

~

,#

.

/

o o

2 0

o

o

i

i

T

,

f

i

i

,

r

,

i

i

~

,

Sample

oi

~n

• 2

A0.2 E

o

3

o

u

<, 0.I

0.5c~ 0

i

4'o'8'o

i

;io'

lio T(K)

i

0 -r

, , 280 0 240

FIG 2 RH vs. T(open circles) in single-crystal "60 K" YBa2Cu3Oy. The variation of the Hall density nH (solid circles) is also shown. Field is normal to the CuO2 planes. falls on a straight line which has a finite intercept at T =0. (The magnitude of R H at 95 K is ~40% higher than the "plateau" value at y=6.6 in a study with ceramic samples4.) In comparison with the 90 K system, we find that the T-dependence of R H is slightly suppressed, but still quite pronounced, and n H is approximately halved. Using the ceramic results 4, we infer that y is quite close to 6.5 in these crystals. We have also measured the in-plane and out-of-plane resistMties (Pab and Pc resp.) in the 60 K crystals. In four crystals studied, Pab (~340 I~Qcm at 290 K) decreases approx, linearly with T, showing no evidence of an upturn near Tc (60 K). However, Pc shows a monotonic increase with decreasing T, reaching ~0.2 ~cm at 80 K (Fig. 3). Because of our sample preparation conditions, the oxygen content- if inhomogeneous - is higher in the core than on the surface. From this, we rule out the possibility that the increase in Pc is caused by an oxygen deficient core. We conclude that the different signs of the derivatives dpc/dTand dPab/dT is an intrinsic property of the 60 K system. Earlier studies 7 on the 90 K phase also showed a similar sign

0

0

4'0

. . . . . .

80

120

160 T(K)

, , , , , I

200

240

280

FIG 3 Variation of the out-of-plane resistivityPcwith T in four crystals of "60 K" 123.Pc increases monotonically w~thdecreasing T. The increased data scatter below 120 K in Samples 1 and 4 is caused by the exponential suppression of the voltage drop when the current is applied in-plane2,6. disparity below 150 K. This difference in sign cannot be accounted for by Anderson localization along ¢. It remains a major problem in understanding the normal state properties. Thus, the divergent behavior of the resitivities in the two directions and the Hall effect, present serious challenges to theories which attempt to explain transport in these oxides using conventional theories. Work at Princeton is supported by the Office of Naval Research (Contract N00014-88 -K-0283). The Hall experiments were performed at the National Magnet Lab. (M.I.T., Cambridge) which is supported by the National Science Foundation. References

1. J.M. Tarascon et al, Phys. Rev. B, submitted. 2. D.A. Brawner, Z.Z. Wang and N.P. Ong, submitted. 3. H. Takagi, private communication; Y. lye, unpublished. 4. Z.Z. Wang et al, Phys. Rev. B 36, 7222 (1987); 5. P. Chaudari et al, Phys. Rev. B 36, 8903 (1987); T. Penney et al, Phys. Rev. B 38, 2918 (1988). 6. J. Clayhold et al, Phys. Rev. B 39, 7324 (1989). 7. S. W. Tozer et al, Phys. Rev. Left. 59, 1768 (1987); S. Hagen et al, Phys. Rev. B 37, 7928 (1988).