Superconductivity and magnetic order in CeCo2

Superconductivity and magnetic order in CeCo2

Journal of Magnetism and Magnetic Materials 140-144 (1995) 2057-2058 ~ Journalof magnetism magnetic ,~ m~erlals ELSEVIER Superconductivity and m...

163KB Sizes 0 Downloads 25 Views

Journal of Magnetism and Magnetic Materials 140-144 (1995) 2057-2058

~

Journalof magnetism magnetic

,~

m~erlals

ELSEVIER

Superconductivity and magnetic order in CeCo 2 J.-G. Park *, M. Ellerby, K.A. McEwen, M. de Podesta Department of Physics, Birkbeck College, University of London, Malet Street, London WCIE 7HX, UK

Abstract We report the study of the magnetisation of CeCo 2 at temperature below 1 K, in which a superconducting transition (Tc = 0.82 K) is observed. Field cooling the sample from above Tc in a 5 mT field suppressed the superconducting transition for temperatures down to 0.4 K. Magnetisation measurements below Tc show behaviour typical of a type II superconductor, except for an unusual feature below Hoe, with Hcl = 7.5 mT and Hc2 = 0.275 T at 0.4 K.

Recent progress in both high-Tc compounds and heavy fermion compounds has led to a re-examination of superconductivity in CeRu 2 [1-4]. Yagasaki et al. [2] found that CeRu 2 exhibits behaviour that may be described as a re-entrant superconductor. Whilst this is not the first rareearth compound to show such behaviour [5], it is amongst the first binary compounds to be described as such. Band calculations [6] and X-ray photoemission studies [7] for CeRu 2 both suggest that the Ce 4f electrons are to some extent delocalised and form a narrow 4f band near to the Fermi level. Coles [8] has suggested that the development of a gap at the Fermi surface on becoming superconducting would reduce the Kondo screening below Hc2. An early study of CeCo 2 made by Luo et al. [9] considered the effect of changing composition on Tc. They found Tc to be strongly peaked (Tc = 1.5 K) for a lattice parameter of 7.165 A. However a pressure of 20 kbar produced a negligible change in the value of Tc. From this they concluded that the maximum in Tc is not associated with a critical value of the volume, but that Tc reflects the degree of order in the sample. In this study of CeCo 2 we present magnetisation data which will be compared with the earlier results [9] for CeCo 2. We also compare the results with those of CeRu 2. The sample was made by arc melting the constituents under an argon atmosphere and then annealed at 900°C for 1 week. X-ray powder diffraction (XRD) measurements reveal that the sample contains a second phase which might be accounted for by the presence of CeCo 3 or Ce2Co 7. However neither of these patterns provided a close fit to the second phase. The lattice parameter for this sample of CeCo 2 is calculated from XRD to be 7.161 A.

The magnetisation was measured using a vibrating sample magnetometer (VSM) built by Oxford Instruments. The VSM is equipped with a 12 T vertical field magnet and a cryostat which operates between T = 0.3-80 K. The base temperature is achieved through the use of liquid 3He which is pumped using a sorption pump. Fig. 1 shows the magnetisation as a function of temperature. The sample was first cooled to 0.3 K in zero field and then heated in an applied field of 5 mT. The superconducting ground state gradually collapses in a broad transition which is completed at T = 0.82 K. On subsequently reducing the temperature from T = 0.9 K in a 5 mT field, the sample did not undergo a superconducting transition. Instead the moment remained positive down to T = 0.4 K. All magnetisation measurements were made on the sample after zero-field cooling. Measurements made at T = 0.4 K show that CeCo 2 exhibits type II superconducting behaviour with He1 = 7.5 mT and He2 = 0.275 T. Further experiments indicate that He1 decreases rapidly from 7.5 mT at 0.4 K to 1 mT at 0.8 K. Fig. 2a shows the low field detail of the hysteresis loop 0.1 o ~

-0.1

FC

~

41.2

'~

41.3

"~

41.4

+44-171-631

6220; email:

/""

/

f

CeCoj

~ -o.5 -0.6 .,I

41.7 0.3

* Corresponding author. Fax: [email protected]

/

, 0.4

i,,,i,,,ll,,i,,,t,, 0.5

0.6 0.7 0.8 Temperature (K)

0.9

Fig. 1. The magnetisation as a function of temperature for CeCo2, both heating (ZFC) and cooling (FC) in a 5 mT field. Tc is taken as the temperature where the broad transition is observed to be complete.

0304-8853/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0 3 0 4 - 8 8 5 3 ( 9 4 ) 0 1 4 9 1 - 4

J.-G. Park et al. /Journal of Magnetism and Magnetic Materials 140-144 (1995) 2057-2058

2058

"'IT '''''l

0.63

KCeC°l =0,4 ¢

0.02

~ ' mI ' ' ' I ' ' ' ~ ' ' '

0.01

:j" 4).02

-0.03 1-0.4

-0.3

-0.2

0.l

-0.1

0.2

0.3

0.4

0

(b)

CeCo "~

0.05

-~g

o

/

T=OAK

-0.05

-0.1

, , ,

-12

i , , ,

-8

i .

-4

, ,

4

8

Magneticfield ( T e s l a )

Fig. 2. (a) The low field detail of the hysteresis loop showing an anomaly just below He2. The arrows indicate the direction in which the field was ramped. In (b) the full hysteresis loop indicates that there were no high field anomalies and there was no apparent saturation of the moment.

for T = 0.4 K. There is a small anomaly at B = 0.27 T which appears on both sides of the hysteresis loop. In addition we find a ferromagnetic contribution of 6 × 10-3/xB/f.u. at B = 0.0 T. Magnetisation measurements made above Tc suggest that the ferromagnetic component is saturated above 10 mT. Fig. 2b shows the magnetisation at T = 0 . 4 K for fields up to 12 T. In this figure we observe no additional anomalies at higher fields and the moment at B = 12 T is 0.055/~a/f.u. The moment shows no sign of saturation at 12 T. The units have been quoted in formula units (f.u.) since there is a possibility that the cobalt may make a contribution to the observed moment. The presence of ferromagnetic behaviour requires some consideration. As outlined earlier, the XRD study shows the existence of a second phase with the extra peaks being probably explained by CeCo 3 or Ce2Co 7. In a review by Buschow [10] the moments quoted for CeCo 3 and Ce2Co 7 were 0.2 and 0.9/zB/Ce atom respectively. There is no quoted value [10] for a saturation moment in CeCo 2. If we assume that the ferromagnetic moment derives from one or other of the possible impurities then we arrive at a value for the impurity concentration of ~ 3%. The observation of a ferromagnetic signal in CeCo 2 has also been made in EPR measurements on CeCo 2 [11,12].

According to Ref. [9], with the lattice parameter determined from XRD, we should expect a value for Tc of ~ 0.5 K. This is clearly not the case. The observation [9] that the value of Tc depends in some way on the lattice parameter may be associated with a change in the magnetic impurity concentration. Unlike a conventional type II superconductor, the magnetisation curve (Fig. 2a) shows an unusual feature below He2 at B = 0.27 T. We find the moment below He2 to be greater than above H~2. Similar observations have been reported for another Laves phase mixed valence superconductor, CeRu 2. The re-entrant behaviour observed for CeRu 2 varies from one sample to another, and although CeCo 2 shows some similarity to CeRu2, we have not seen the so called re-entrant behaviour in the magnetisation of CeCo 2 . We note that the values of H~I, He2 and Tc may be influenced to some extent by the presence of ferromagnetic impurities. However we believe our observations are qualitatively sound and figures quoted for Hal, He2 and Tc constitute lower limits for CeCo 2. CeCo 2 has an unusual feature below H~2 which suggests behaviour similar to that observed in CeRu 2. The ferromagnetic signal, observed in the M ( H ) measurements, may be responsible for the disappearance of the superconducting transition when the sample is cooled in a 5 mT field. The origin of the ferromagnetic signal remains, as yet, unresolved. Acknowledgements: We acknowledge helpful discussions with Professor B.R. Coles. We also thank Dr. H. Stone for his help with sample preparation. This work has been supported by the UK Science and Engineering Research Council. References

[1] S.B. Roy, Philos. Mag. B 65 (1992) 1435. [2] K. Yagasaki et al., J. Phys. Soc. Jpn. 62 (1993) 3825. [3] A.D. Huxley et al., J. Phys. Condensed Matter 5 (1993) 7709. [4] J.-G. Park (unpublished, 1992). [5] O. Fischer, Ferromagnetic Materials, Vol. 5, eds. K.H.J. Buschow and E.P. Wohlfarth, (North-Holland, 1990). [6] A. Yanase, J. Magn. Magn. Mater. 52 (1985) 403. [7] J.W. Allen et al., Phys. Rev. B 26 (1982) 445. [8] B.R. Coles (private communications). [9] H.L Luo et al., Phys. Lett. A 27 (1968) 519. [10] K.H.J. Buschow, Ferromagnetic Materials, Vol. 1, ed. E.P Wohlfarth (Noah-Holland, 1980). [11] R.C. Barnes et al., Phys. Rev. Lett. 16 (1966) 233. [12] A.C. Gossard and J.H. Wernick, Phys. Rev. Lett. 16 (1966) 995.