Flux quantization in a high-T_c superconductor

C. E. Gough^*, M. S. Colclough^*, E. M. Forgan^*, R. G. Jordan^*, M. Keene^*, C. M. Muirhead^*, A. I. M. Rae^*, N. Thomas^*, J. S. Abell^† & S. Sutton^†

^*Department of Physics, University of Birmingham, Birmingham B15 2TT, UK ^†Department of Metallurgy and Materials, University of Birmingham, Birmingham B15 2TT, UK

We have observed the quantization of magnetic flux in a high-T_c yttrium-based ceramic superconductor¹ and obtain a value for the flux quantum in this material. The value of the flux quantum, h/2e where h is Planck's constant and e is the electron charge, implies that the charge carriers of superconductivity are electron pairs.

The measurements were made on a ring (outer-diameter 10.0 mm, inner-diameter 4.5 mm and thickness 4.0 mm) formed into this shape by sintering Y_1.2Ba_0.8CuO₄ prepared from powders of Ba CO₃Y₂O₃ and CuO heated in air at 950°C. The sample had a broad superconducting transition between 50 and 85 K.

The ring was immersed in liquid helium at 4.2 K inside a superconductivity shield. The flux inside the ring was monitored by a radio frequency superconducting quantum interference device (r.f.-SQUID) magnetometer weakly coupled to the ring via a flux transformer and flux was applied to the ring by a solenoid passing through it. As the current in the solenoid was changed, supercurrents were induced in the ring until some critical current was exceeded when flux passed in or out of it. A typical magnetization curve obtained in this way is shown in Fig. 1. Such behaviour is characteristic of measurements made on a superconducting circuit containing a weak link².

The use of the long solenoid (diameter 1.86 mm with a turn spacing of 0.142 mm) bent into a nearly complete loop passing through the centre of the ring allows us to make an accurate determination of the applied flux. Along the sloping sections of the magnetization curve the total flux remains constant and the induced magnetometer signal is therefore determined entirely by the flux in the solenoid. We note that there is a slight overall slope to the magnetization curve due to a small amount of direct coupling between the solenoid and the flux transformer. After allowing for this we obtain an accurate calibration of the magnetometer signal in terms of the flux produced by the circulating supercurrent whatever its actual path through the ring. In this way we deduce that the magnitude of the flux jumps shown in Fig. 1 is typically 100 (h/2e).

Fig. 1 Typical magnetization curve showing the flux output of the r.f.-SQUID magnetometer as the flux applied through the long solenoid is swept through a single cycle. | high-resolution version |

Fig. 2 Output of the r.f.-SQUID magnetometer showing small integral numbers of flux quanta jumping in and out of the ring. | high-resolution version |

To obtain a value for the flux quantum, we set the solenoid current to zero and periodically expose the ring to a local source of electromagnetic noise causing the ring to make small jumps between quantized flux-states. This is illustrated in Fig. 2, in which a number of equally-spaced lines have been superimposed to emphasize the quantum nature of the flux transitions. Using these results and the earlier calibration we obtain a value for the flux quantum of 0.97 ± 0.04 (h/2e) thus demonstrating that superconductivity in this high-T_c material involves the pairing of electrons as in conventional BCS (Bardeen–Cooper–Schrieffer) superconductivity. The observed flux jumps may be between the outside of the ring and the hole, or may be between the hole and the bulk of the material. We are not able to distinguish between these two possibilities at present.

We have performed experiments similar to those shown in Fig. 2 in which we observed no detectable change in the magnetometer output for 1,000 s. On this basis the resistance of our sample is less than 10^–13 ohms.

The above measurements are of interest not only because they demonstrate the existence of coherent superconducting states in the newly discovered class of ceramic superconductors but also because they suggest that SQUID-type devices could be made from simple fabricated shapes using the intrinsic properties of the weakly superconducting ceramic material to provide its own weak link. Indeed, we have already used our ring to observe noisy r.f.-SQUID characteristics at 20 MHz typical of a weak-link loop with too large a critical current. We are currently extending all these measurements to other rings and to higher temperatures to investigate the compositional and temperature dependence of the magnetic and SQUID behaviour.

We thank G. R. Walsh for technical assistance and Professor W. F. Vinen for support and encouragement.

Received 8 April; accepted 10 April 1987.

1. Wu, M. K. et al. Phys. Rev. Lett. 58, 908–911 (1987).
2. Zimmerman, J. E. & Silver, A. H. Phys. Rev. 157, 317–341 (1967).

| return to looking back index page |