Hyperfine Interactions 126 (2000) 299–303

Magnetism of ErAgSn studied by M¨ossbauer spectroscopy

299

119 Sn

K. Ła¸tka a , E.A. G¨orlich a , J. Gurgul a and R. Kmie´c b a

M. Smoluchowski Institute of Physics, Jagiellonian University, PL-30-059 Krak´ow, Poland b H. Niewodnicza´nski Institute of Nuclear Physics, PL-31-342 Krak´ow, Poland

Antiferromagnetic ErAgSn compound was investigated in detail by 119 Sn M¨ossbauer spectroscopy in a temperature range between 2.2 and 300 K. The 119 Sn spectra recorded below 4.2 K can be well fitted with a single main magnetic component in agreement with recent neutron diffraction studies [1]. A broad distribution of magnetic hyperfine fields observed above 4.2 K and enhanced spin correlations among Er3+ ions at T > TN = 5.6 K are the remarkable features of the investigated system.

1.

Introduction

Recently, magnetic susceptibility and neutron diffraction studies of the compound ErAgSn have been reported [1]. The crystal structure of this compound was described either in terms of the hexagonal CaIn2 -type structure (space group P63 /mmc) [1–3] or in its ordered derivative the GaGeLi-type structure (space group P63 mc) [1]. In the GaGeLi-type of structure all ions have unique crystallographic positions. Rather broad maximum of magnetic susceptibility observed around TN = 6.2 K indicates that ErAgSn compound orders antiferromagnetically but neutron investigations show that TN is a little lower and equal to 5.6 K [1]. Below TN , magnetic moments of Er are aligned along the hexagonal c-axis and at low temperatures this compound exhibits simple collinear antiferromagnetic structure described by the wave vector k = [1/2, 0, 0] while incommensurate sine modulated magnetic structure characterised by the wave vector k = [0.426(0), 0.035(1), 0] has been found near TN [1]. In the present work, X-ray diffraction and 119 Sn M¨ossbauer investigations were undertaken and critically compared with the results of neutron diffraction measurements made on the same sample [1]. 2.

Results and discussion

2.1. Crystal structure The preparation of the ErAgSn sample was described previously in [1,2]. The X-ray powder diffraction pattern recorded at room temperature with the Siemens D-501  J.C. Baltzer AG, Science Publishers

300

K. Ła¸tka et al. / Magnetism of ErAgSn studied by M¨ossbauer spectroscopy

Figure 1. The X-ray diffraction pattern of ErAgSn with Rietveld fitting (solid line). The bottom line shows the difference between observed and calculated intensities.

diffractometer using the Ni-filtered CuKα radiation is displayed in figure 1, and can be indexed in accord with neutron diffraction measurements [1] on a hexagonal unit cell with the GaGeLi-type structure. The obtained values of lattice constants: a = b = 0.46623(11) nm, c = 0.7291(2) nm are in perfect agreement with neutron studies [1]. The additional peaks (figure 1) which were excluded from the refinement procedure, performed with the FULLPROF computer program [4], are due to an unknown impurity phase. 2.2.

119 Sn

M¨ossbauer spectroscopy

The 119 Sn M¨ossbauer investigations were carried out in the temperature range 2.2–300 K using a standard equipment in transmission geometry and a Ba 119m SnO3 source kept at temperatures close to 4.2 K. Absorber was made of the fine powdered compound with optimal thickness of 35 mg cm−2 and a palladium foil of 0.05 mm thickness was applied as a critical absorber for tin X-rays. The spectra were fitted within the Lorentz approximation and with full hyperfine interaction Hamiltonian Hhf . The room temperature spectrum (figure 2) can be well fitted by superposition of two quadrupole-split subspectra. The main component attributed to ErAgSn takes about 88% of the total intensity whereas the minor component (the isomer shift δisimp = 2.50(1) mm/s, the quadrupole splitting |∆EQimp | = |eQVzz | = 1.63(2) mm/s, and the experimental half-width Γimp = 0.80 mm/s) clearly seen at the right wing of the spectrum originates from an unknown Sn-containing phase which was also detected in the X-ray diffraction pattern as mentioned above. Table 1 summarizes the results for the main component ascribed to ErAgSn. During the numerical analysis of all magnetically split spectra it was assumed that the neighbourhood of the Sn atom possesses three-fold symmetry axis being parallel to crystallographic c-axis and thus the asymmetry parameter η = 0.

K. Ła¸tka et al. / Magnetism of ErAgSn studied by M¨ossbauer spectroscopy

Figure 2.

119

301

Sn resonance spectra for ErAgSn sample recorded at 300 and 2.2 K. The continuous lines represent the least-squares fit to the experimental points.

Table 1 The hyperfine interaction parameters inferred from the 119 Sn resonance spectra of the ErAgSn intermetallic compound. Temp. T (K)

main Heff (kOe)

∆EQ a (mm/s)

δis b (mm/s)

θ (deg)

Γc (mm/s)

imp Heff (kOe)

θimp (deg)

χ2

2.2 3.0 4.2 78 300

12.5(1) 12.5(1) 12.8(1) – –

−0.79d −0.79d −0.79d ±0.81(1) ±0.79(1)

1.71(1) 1.71(1) 1.72(1) 1.67(1) 1.69(1)

25(1) 25(1) 26(1) – –

0.94(2) 0.95(2) 0.94(2) 0.99(1) 0.85(1)

21.8(7) 21.7(7) 21.8e – –

38(4) 34(5) 38e – –

1.02 1.30 1.95 0.93 1.01

∆EQ = eQVzz . δis is relative to the Ba 119m SnO3 source. c Half-width of the resonance line. d Parameters kept constant during the fit equal to the respective values obtained above TN . e Parameter was constrained to a constant value derived at 2.2 K. For Eγ = 23.875 keV γ-transition in 119 Sn: 1 mm/s corresponds to 7.963(2) × 10−8 eV or 19.253(6) MHz. a

b

This assumption is strictly fulfilled for the GaGeLi-type of structure but it seems to be a good approximation also for the CaIn2 -type of structure [6,7]. The low temperature spectra recorded below TN (between 2.2 and 4.2 K) reveal magnetic hyperfine splitting. Since for collinear structure observed in this temperature range all tin positions in ErAgSn are magnetically equivalent so the respective spectra could be well analysed using one unique set of hyperfine parameters for the main component (see table 1). The half-width Γ treated as a free parameter remains unaltered which indicates that simple collinear magnetic structure persists up to 4.2 K. On the other hand, for the impurity component the magnetic hyperfine splitting is much larger and changes imp imp only slightly if at all between 2 and 4.2 K. The values of δis , Γimp and |∆EQ |

302

K. Ła¸tka et al. / Magnetism of ErAgSn studied by M¨ossbauer spectroscopy

Figure 3. The temperature evolution of the 119 Sn M¨ossbauer spectra and the corresponding field distribution functions F (Heff ), obtained from computer fits at various temperatures above 4.2 K. For explanation, see the main text.

for the impurity component were constraint to those obtained at room temperature. imp Fits with the negative quadrupole interaction ∆EQ = −1.63(2) mm/s were much better. In contrast to the results obtained in the temperature range 2.2–4.2 K, the spectra recorded between 4.2 and 9 K reveal some features characteristic for continuous distribution of magnetic hyperfine fields and an approach developed by Wivel and Mørup [5] was used. This method has an advantage of making no assumptions concerning the shape of the field distribution function, F (Heff ). Figure 3 illustrates the temperature evolution of field distribution function, F (Heff ), derived from the experimental spectra obtained above 4.2 K. The magnetic hyperfine splitting survives above TN pointing to the existence of a certain spin correlations among Er3+ ions. These correlations have certainly short-range character since no spatial coherence of magnetic moments over the sample volume was observed by neutron diffraction technique.

Acknowledgements We thank W. Chajec for experimental assistance at the early stage of this work and Prof. A. Szytula (Jagiellonian University) for providing us with ErAgSn sample. The authors are grateful to the State Committee for Scientific Research in Poland for financial support within Grants 2 P03B116 12 and 2 P03B114 16.

K. Ła¸tka et al. / Magnetism of ErAgSn studied by M¨ossbauer spectroscopy

303

References [1] S. Baran, J. Leciejewicz, N. St¨usser, A. Szytuła, A. Zygmunt and Y. Ding, J. Magn. Magn. Mater. 170 (1997) 143. [2] D. Mazzone, D. Rossi, R. Marazza and R. Ferro, J. Less-Common. Met. 80 (1981) P47. [3] J. Sakurai, S. Nakatani, A. Adam and H. Fujiwara, J. Magn. Magn. Mater. 108 (1992) 143. [4] J. Rodriguez Carvajal, in: XVth Congr. Int. Union of Crystallography, Proc. of the Meeting on Powder Diffraction, Toulouse, France (1990) p. 127. [5] C. Wivel and S. Mørup, J. Phys. E 14 (1981) 605. [6] R. Kruk, R. Kmie´c, K. Ła¸tka, K. Tomala, R. Tro´c and V.H. Tran, J. Alloys Compounds 232 (1996) L8. [7] K. Ła¸tka, E.A. G¨orlich, R. Kmie´c, R. Kruk, A.W.J. Pacyna and W. Chajec, Mol. Phys. Rep. 22 (1998) 87.