The compound ErFe4Ge2 undergoes on cooling a first-order transition from a paramagnetic to a magnetically-ordered state at TN = 44 K. The high-resolution powder X-ray diffraction patterns collected at BM16 revealed a double symmetry breaking at the magnetic transition (Figure 77). The X-ray data enabled us to distinguish between the structural and the magnetic satellite peaks present in the neutron-diffraction patterns below TN [1]. The satellites display a highly complex peak topology, and were very difficult to interpret. The interdependence of the structural and magnetic transitions is now unveiled.

The powder-diffraction measurements show that the high temperature paramagnetic phase (space group P42/mnm) disproportionates at TN = 44 K into two symmetrically distinct phases with orthorhombic symmetry. The respective space groups are Cmmm and Pnnm. Both phases coexist in the intermediate temperature range (20 K < T < TN) in proportions varying with temperature (Figure 78). The Pnnm phase reaches its highest concentration of ~31% near 30 K. Below 20 K, the Cmmm phase dominates, up to 95%.

This is the first time such a phenomenon has been observed. We suggest that it originates from competing magneto-elastic mechanisms, involving the Er crystal field anisotropy, the Er-Er, Er-Fe and the Fe-Fe exchange interactions and their coupling with the lattice strains. The transition is accompanied by strong microstrain effects, which lead to a noticeable (hkl) dependence of the line broadening. The shear strain accompanying the P42/mnm Cmmm transition was attributed to the magneto-elastic coupling between the Er-Er dominant interaction and the lattice strains [1]. The observed ao/bo 1 deformation (referring to the 2at, 2at, ct cell) is due to the fact that the Er-Er exchange interaction is negative along bo and positive along ao. The P42/mnm Pnnm transition is most probably driven by the tensile strain resulting from the Fe displacement field (Figure 78). Given that the shortest Fe-Fe distances are in the strongly deformed tetrahedron, the antiferromagnetic Fe-Fe interaction probably plays a dominant role in this transition as a result of geometric frustration. We assume that below TN the Er-Er and Fe-Fe interactions are of comparable strength but involve different order parameters, leading to the double symmetry breaking as a consequence of the coupling between the magnetic order parameter and the lattice strains.

[1] P. Schobinger-Papamantellos, J. Rodríguez-Carvajal, G. André, C.H. de Groot, K.H.J. Buschow, J. Magn. Magn. Mat., 191, 261 (1999).

Principal Publication and Authors
P. Schobinger-Papamantellos (a), J. Rodríguez-Carvajal (b), K.H.J. Buschow (c), E. Dooryhee (d), A.N. Fitch (d), J. Magn. Magn. Mat., 210, 121-137 (2000).

(a) Lab. für Kristallographie, ETHZ Zürich (Switzerland)
(b) Lab. Leon Brillouin, (CEA - CNRS) Saclay (France)
(c) Van der Waals-Zeeman Institute, University of Amsterdam, (The Netherlands)
(d) ESRF