MAGNETIC-FIELD DEPENDENCE OF THE NEUTRON-SCATTERING FROM ERRH4B4

Magnetism in the reentrant superconductor ErRh4B4 has been studied by neutron scatteringas a function of an applied magnetic field. For a temperature of 1.69 K long.rangeferromagnetism is found in fields higher than 1 kOe. Considerable hysteresis is found in the neutron scattering intensity vs magnetic field curve and long-range order with a small Er momentremains when the field is reduced to small values. conducting high magnetic a a ferromagnet the


1
Institute for Pure and Applied Physical Sciences, University of California, San Diego, La Jolla, CA 92093, U.S.A. (Recieved 28 November 1979 by H. Suhi) Magnetism in the reentrant superconductor ErRh 4B4 has been studied by neutron scattering as a function of an applied magnetic field. For a temperature of 1.69 K long.range ferromagnetism is found in fields higher than 1 kOe. Considerable hysteresis is found in the neutron scattering intensity vs magnetic field curve and long-range order with a small Er moment remains when the field is reduced to small values.
BECAUSE of its unusual magnetic and superconducting used to remove high order contamination from the properties [11 there has been a lot of interest in the monochromator. The monochromator was pyrolytic ternary compound ErRh4B4. The material becomes a graphite and collimation after the sample of 20' was superconductor at 8.7 K and superconductivity is used. Two sets of measurements were made. In the first destroyed and long-range magnetic order is established set of measurements the sample was placed in a superat about 0.9 K. Neutron diffraction experiments by conducting magnet capable of producing high magnetic Moncton et al.
[21 have shown that ErRh4B4 is a fields. However, the superconducting magnet had a ferromagnet below 0.9 K with the moment direction in small remanent field and thus it was necessary to check the basal plane. The moment value was found to be the low field results with a different magnet system. 5.6p~which is well below the free ion value of~Additional low field measurements were thus made with Fertig et a!. [11showed that the superconducting state a conventional pumped 4He cryostat placed in a is destroyed with the application of a magnetic field.
Helmholtz pair of magnet coils. This gave a uniform This paper reports neutron diffraction measurements magnetic field and a negligible remanent field. The on ErRh 4B4 taken as a function of an applied static experiments were performed at the lowest temperatures magnetic fIeld, achievable by the two cryostat systems which were The sample was prepared by arc melting the rare 1.69 K for the superconducting magnet and 1.79 K for earth tetraboride with Rh followed by annealing. The the cryostat with the Helmholz coils. aIB isotope was used to decrease the absorption cross We will first discuss the measurements made in the section for slow neutrons. The experiments were per-superconducting magnet. Magnetic diffraction peaks formed at the High Flux Isotope Reactor using con-could easily be observed in fields larger than about ventional techniques. An incident neutron wavelength 1 kOe. in zero field. Figure 1 shows the (101), (110) Fig. 2. Field dependence of the magnetic contribution applied field (open circles) and in an applied field or to the (101) peak intensity measured in a supercon. 10 kOe (closed circles).
ducting magnet for a sample temperature of 1.69K.
When making powder diffraction measurements in about 1 kOe was applied reflecting the superconductivity a magnetic field considerable care must be taken that in this field range. The scattering intensity then increases the field does not orient the powder particles with increasing field, the ordered moment values correspreferentially. This commonly occurs with fuse particles ponding to 5.06 ± 0.5 fB at 10 kOe and 6.9 ±0.5~at that have sizeable magnetic moments. In the present 20 kOe. This is a similar value to that obtained in the case, the sample consisted of rather coarse polycrystal-magnetization measurements. The intensity found upon line particles held in a flat sample holder so that the lowering the applied field appears to be quite different sample size was about 2 x 2 x 0.1 cm 3. The field was from the magnetization measurements although the applied vertically so that it was perpendicular to the reversible magnetization data are only published for low scattering plane and along one of the long dimensions field values. We see considerable hysteresis in the intenof the plate. The thin plate was used so that neutrons sity vs field curve and at zero applied field we still see could be transmitted through the sample which is quite some long-range ordered moment. The field value is not absorbing to slow neutrons because of the high absorp-brought identically to zero since the magnet assembly tion cross-section of Rh. The sample particles were and spectrometer have some remanent field amounting packed tightly in the holder and it seems unlikely that to about 100 Oe. they would move in the presence of the applied field.
Good statistics were obtained after decreasing the Nevertheless, to check that the field was not producing field to zero applied field (point 1 on the graph) showpreferred orientations that could influence the interpret. ing that the long-range order found is a real effect. If ation of our results, powder diffraction patterns were one starts from point 1 and increases the field, the intentaken before and after the fields were applied and at sity remains on the decreasing field curve and follows it several values of the applied field. Analysis of the dif-back to higher values. If one starts at point I and warms fraction patterns showed no evidence that the field was the sample to a temperature above the superconducting producing preferential alignment of a particular crystal-transition temperature (8.7 K) and then recools to lographic axis.
1.69 K one returns to the zero intensity value corres-The field dependence of the magnetic scattering ponding to no long-range moment. The scattering intendetermined from the intensity of the (101) reflection sity at point 1 corresponds to a long-range ordered is shown in Fig. 2. The nuclear component to the reflec-moment of 0.7 ±0~2PBFreeman and Jarlborg [5] have tion has been subtracted. The field was produced by the in fact previously suggested that upon lowering an superconducting magnet operating in the persistent external field from a value greater than the critical field mode and thus the field value was very stable during it may be possible to form a mixed state in which normal the measurements at each point. The sample tempera. and ferromagnetic regions of the compound coexist with ture was held at 1.69 ±0.02 K during the course of the superconductivity. experiment. The curve for increasing field is in some The results of the second set of measurements are respects similar to that established by Fertig et a!. [I] shown in Fig. 3. These measurements are confused to and Ott eta!. [3,4]