Frustrated itinerant ferromagnets, with non-collinear static spin structures, are an exciting class of material as their spin chirality can introduce a Berry phase in the electronic scattering and lead to exotic electronic phenomena such as the anomalous Hall effect (AHE). This study presents a reexamination of the magnetic properties of Fe(3)Sn(2), a metallic ferromagnet, based on the two-dimensional kagome bilayer structure. Previously thought of as a conventional ferromagnet, we show using a combination of SQUID (superconducting quantum interference device) measurements, symmetry analysis and powder neutron diffraction that Fe(3)Sn(2) is a frustrated ferromagnet with a temperature-dependent non-collinear spin structure. The complexity of the magnetic interactions is further evidenced by a re-entrant spin glass transition ([Formula: see text] K) at temperatures far below the main ferromagnetic transition (T(C) = 640 K). Fe(3)Sn(2) therefore provides a rare example of a frustrated itinerant ferromagnet. Further, as well as being of great fundamental interest our studies highlight the potential of Fe(3)Sn(2) for practical application in spintronics technology, as the AHE arising from the ferromagnetism in this material is expected to be enhanced by the coupling between the conduction electrons and the non-trivial magnetic structure over an exceptionally wide temperature range.

The kagome-bilayer material Fe(3)Sn(2) has recently been shown to be an example of a rare class of magnet-a frustrated ferromagnetic metal. While the magnetism of Fe(3)Sn(2) appears to be relatively simple at high temperature, with localized moments parallel to the c-axis (T(C) = 640 K), upon cooling the competing exchange interactions and spin frustration become apparent as they cause the moments to become non-collinear and to rotate towards the kagome plane, forming firstly a canted ferromagnetic structure and then a re-entrant spin glass (T(f) approximately equal 80 K). In this work we show that Fe(3)Sn(2) possesses an unusual anomalous Hall effect. The saturated Hall resistivity of Fe(3)Sn(2) is 3.2 µΩ cm at 300 K, almost 20 times higher than that of typical itinerant ferromagnets such as Fe and Ni. The anomalous Hall coefficient R(s) is 6.7 × 10(-9) Ω cm G(-1) at 300 K, which is three orders of magnitude larger than that of pure Fe, and obeys an unconventional scaling with the longitudinal resistivity, ρ(xx), of R(s) is proportional to ρ(xx)(3.15). Such a relationship cannot be explained by either the conventional skew or side-jump mechanisms, indicating that the anomalous Hall effect in Fe(3)Sn(2) has an extraordinary origin that is presumed to be related to the underlying frustration of the magnetism. These findings demonstrate that frustrated ferromagnets, whether based on bulk materials or on artificial nanoscale structures, can provide new routes to room temperature spin-dependent electron transport properties suited to application in spintronics.

First‐order phase transitions, where one phase replaces another by virtue of a simple crossing of free energies, are best known between solids, liquids, and vapors, but they also occur in a wide range of other contexts, including even elemental magnets. The key challenges are to establish whether a phase transition is indeed first order, and then to determine how the new phase emerges because this will determine thermodynamic and electronic properties. Here it is shown that both challenges are met for the spin reorientation transition in the topological metallic ferromagnet Fe3Sn2. The magnetometry and variable temperature magnetic force microscopy experiments reveal that, analogous to the liquid–gas transition in the temperature–pressure plane, this transition is centered on a first‐order line terminating in a critical end point in the field‐temperature plane. The nucleation and growth associated with the transition is directly imaged, indicating that the new phase emerges at the most convoluted magnetic domain walls for the high temperature phase and then moves to self‐organize at the domain centers of the high temperature phase. The dense domain patterns and phase coexistence imply a complex inhomogenous electronic structure, which can yield anomalous contributions to the electrical conductivity.

Abstract. The zero-point entropy of a spin glass is a difficult property to experimentally determine and interpret. Spin glass theory provides various predictions, including unphysical ones, for the value of the zero-point entropy, however experimental results have been lacking. We have investigated the magnetic properties and zero-point entropy of two spinel Cu 2+ based spin glasses, CuGa2O4 and CuAl2O4. Dc-and ac-susceptibility and specific heat measurements show many characteristic spin glass features for both materials. The spin glass freezing temperature is determined to be T f = 2.89 ± 0.05 K for CuGa2O4 and T f = 2.30 ± 0.05 K for CuAl2O4. By integrating the specific heat data we have found that CuGa2O4 and CuAl2O4 have zero-point entropies of S0 = 4.96 JK −1 mol −1 and S0 = 4.76 JK −1 mol −1 respectively. These values are closest to the prediction for a Sherrington-Kirkpatrick XY spin glass, however they are notably higher than all of the theoretical predictions. This indicates that CuGa2O4 and CuAl2O4 have a greater degeneracy in their ground states than any of the spin glass models.