Crystal field and exchange effects in CeZn

(1)

HAL Id: jpa-00218970

https://hal.archives-ouvertes.fr/jpa-00218970

Submitted on 1 Jan 1979

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Crystal field and exchange effects in CeZn

A. Murasik, H. Ptasiewicz-Bik, A. Zygmunt

To cite this version:

A. Murasik, H. Ptasiewicz-Bik, A. Zygmunt. Crystal field and exchange effects in CeZn. Journal de Physique Colloques, 1979, 40 (C5), pp.C5-143-C5-144. �10.1051/jphyscol:1979553�. �jpa-00218970�

(2)

JOURNAL DE PHYSIQUE Colloque C5, supplément au n° 5, Tome 40, Mai 1979, page C5-143

Crystal field and exchange effects in CeZn

A. Murasik (*), H. Ptasiewicz-Bak (*) and A. Zygmunt (**)

(*) Institute of Nuclear Research, Swierk Research Establishment, 05-400 Otwock, Poland (**) Institute for Low Temperature and Structural Research 50-950 Wroclaw, Poland

Résumé. — Nous avons réalisé à 80 et 5 K des expériences de diffusion inélastique des neutrons sur le composé intermétallique CeZn.

Abstract. — Inelastic neutron scattering experiments at 80 and 5 K have been performed on the intermetailic compound CeZn.

1. Introduction. — We report neutron inelastic scat- tering experiments on the intermetallic compound CeZn, performed with the aim to investigate the CEF and MF splitting, the knowledge of which is useful for a proper understanding of its various physical behaviour. CeZn crystallizes in the CsCl structure with a small tetragonal distortion below the ordering temperature. Magnetic susceptibility and neutron-diffraction measurements revealed that below 30 K this compound orders antiferromagnetically [1, 2]. Its magnetic structure consists of ferro- magnetic (001) sheets coupled antiferromagnetically, with magnetic moments of 1.93 /iB oriented along

< 001 > direction. From a preliminary estimation based on neutron spectroscopy measurements in the paramagnetic phase, the crystal-field splitting of 5.4 meV with the T8 quartet lying lowest was reported [2]. In the present study, neutron scattering measurements are extended to the low temperature region.

2. Experimental. — Approximately 30 g of poly- crystalline CeZn was obtained by melting of stoi- chiometric quantities of Ce and Z. Details of sample preparation are given elsewhere [2]. Neutron inelastic scattering experiments have been performed on the triple-axis spectrometer at the EWA reactor, Swierk.

The incident neutrons had a constant energy of 33.96 meV. Measurements were carried out in the energy loss configuration i.e. the Ce ion is excited from lower to higher state. The scattering angle i//

was held constant at 8°, 14° and 20°. The data were collected at 80 and 5 K.

3. Results and analysis. — The energy spectra of neutrons scattered from CeZn at 80 and 5 K are shown in figure 1. The spectrum taken at 80 K shows a large peak at zero energy transfer corresponding to nuclear incoherent scattering. Its width at the bottom is apparently enlarged due to a complete

500f ~ - / \ Scattering angle

400- ; \ y=%°

I \ I

^5K

300 - / ! J

'^200- / A l l / *

/«A\ / \

§ 500-

8 Scattering angle

t⁴° ° ' I A *80K°

| 300- I l \ \ 200 - J I I \

100 - ' / f ' A \ .

/y| "125 \ \^ ^ t ^

5 0 - 5 -10 Energy transfer [meV]

Fig. 1. — Energy spectra of neutrons scattered from polycrystalline CeZn in the paramagnetic and ordered states. The solid lines correspond to the calculated spectra for the parameters : y = 5.4 meV, A = 5.4 meV at 80 K and A = 5.4 meV, y = 2meV, Hm = 950 kG for the spectrum taken at 5 K.

overlap of the inelastic and quasielastic crystal-field contributions. When the temperature is lowered to 5 K then, apart from nuclear incoherent peak, a well defined maximum at about 8 meV appears.

For both temperatures the magnetic origin of observed intensities was confirmed by the variation of scattering angle i.e. by changing the range of momentum transfers %Q.

The spectrum at 80 K has been analysed on the basis of single-ion Hamiltonian and thermal neutron

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1979553

(3)

C5-144 A. MURASIK, H. PTASIEWIEZ-BAK AND A. ZYGMUNT

cross section derived by de Gennes [3]. Since it was not possible to separate the crystal-field transitions from the elastic line, a least squares fitting procedure has been applied. All crystal-field transition lines were assumed to be of equal width and have been approximated by Gaussian. In the fitting procedure the three parameters varied were : the crystal-field splitting A, the linewidth y, and the scale factor K.

The parameters corresponding to the best fit are : A = 5.4 meV, y = 5.4 meV with the quartet T, as ground state.

In the analysis of the energy spectrum at 5 K we restrict ourselves to the cubic crystal field i.e. we assume that by neglecting small tetragonal distortion the essential characteristics of the wave functions and energy level scheme are not much altered. At first, the overall CF splitting A was kept constant at the value found at 80 K and three adjustable parameters were the molecular field H,, the linewidth parameter y and the scale factor K. The fitting procedure was based on a single-ion Hamiltonian which describes exchange in a molecular field approximation. Such an approach does not take into account the dispersion of magnetic excitations, which for polycrystalline samples result in a weighted average over the Brillouin zone, because of random orientation of the crystallites. However, energy spectra measured at scattering angles : 70, 140 and 200 turned out to be (within the experimental error) independent of the scattering vector Q. The energy of the peak at 8.3 meV did not change by more than 0.2 meV for particular angle setting. Thus, we concluded that in a first approximation these effects might be neglected.

The best fit has been done for the values : Hm = 950 kG, y = 2 meV. With these parameters in mind, the peak at 8.2 meV could be recognized as a transverse J--transition from the ground to first excited state. Transitions to higher states have negligible probabilities, so that the observation of these transitions seems to be unlikely by neutron spectroscopy on polycrystalline sample. Nevertheless, in the range of energy transfers covered in the present experiment (20 meV) no other transitions were detect- ed.

The magnetic moment of 2.1 p, derived by using Hm = 950 kG and A = 5.4 meV compares well with experimental value of 1.93 p, [2], but the calculated NCel temperature (65 K) falls out of the range accep- table value of 30 K as determined from neutron- diffraction experiments. After this observation, an attempt was made to see whether a better consistency

could be achieved by the assumption that the crystal- field splitting is temperature dependent. Thus in the second stage of fitting, the overall crystal-field splitting A was changed step by step and fora given A , the parameters H,, y were varied until the best fit was obtained. Subsequently, for resulting parameters Hm and A, the NCel temperature and magnetic moment at 5 K were calculated. We noticed that the resulting linewidth parameter y and quality of the fit were almost independent of particular crystal-field splitting.

Results are shown in figure 2. Inspection of figure 2 reveals that for A lying in the range of 10-11 meV, H, is equal to

-

550 kG and TN is close to 30 K.

In this range the calculated magnetic moment of 2.0 p, agrees well with experimental value, but as a consequence of the molecular field reduction the overall crystal-field splitting drastically increases.

Summarizing we can admit that no unique solution which definitively supports particular set of values Hm and A can be inferred from the low temperature experiments. This ambiguity is partly due to the absence of reliable data on the magnitude of tetragonal distortion which may modify the level sequence below the ordering temperature. We think that studies on single crystal would be highly desirable to further elucidate these questions.

0 5 10 15 20 Crystal-field splitting

A = E ( y ) - E ( 5 ) in meV

Fig. 2. - Variation of the molecular field H,,, and the Ntel tempe- rature T, with the crystal-field splitting. The solid line represents the dependence of H,,, with A, for the best fit (left scale) whereas the dotted line shows the corresponding changes in the calculated Nee1 temperatures (right scale).

References

[I] KANEMATSU, K., ALFIERI, G. T. and BANKS, E., J. Phys. Soc.

Japan 26 (1969) 244.

[2] MURASIK, A., PTASIEWICZ-BAK, H. and ZYGMUNT, A., Phys.

Stat. Sol. (a) 46 (1978) K-75.

[3] DE GENNES, P. G., Magnetism, Vol. 3, p. 115 (Academic Press, New York and London).