Competing magnetic states, disorder, and the magnetic character of Fe3Ga4

TL;DR

This study investigates the magnetic and electronic properties of Fe3Ga4 single crystals, revealing how disorder influences magnetic states, with evidence of complex magnetic behavior and electronic structure changes across transitions.

Contribution

It provides new insights into how disorder and sample history affect the competition between ferromagnetic and antiferromagnetic states in Fe3Ga4, supported by experimental and electronic structure calculations.

Findings

Magnetic transitions at 68 K and 360 K are sensitive to annealing conditions.

Hall effect shows a significant anomalous component and sign change with temperature.

Evidence of a topological Hall effect in the antiferromagnetic phase.

Abstract

The physical properties of metamagnetic Fe$_3$Ga$_4$ single crystals are investigated to explore the sensitivity of the magnetic states to temperature, magnetic field, and sample history. The data reveal a moderate anisotropy in the magnetization and the metamagnetic critical field along with features in the specific heat at the magnetic transitions $T_1=68$ K and $T_2=360$ K. Both $T_1$ and $T_2$ are found to be sensitive to the annealing conditions of the crystals suggesting that disorder affects the competition between the ferromagnetic (FM) and antiferromagnetic (AFM) states. Resistivity measurements reveal metallic transport with a sharp anomaly associated with the transition at $T_2$. The Hall effect is dominated by the anomalous contribution which rivals that of magnetic semiconductors in magnitude ($-5 \mu \Omega$ cm at 2 T and 350 K) and undergoes a change of sign upon cooling into the low temperature FM state. The temperature and field dependence of the Hall effect indicate that the magnetism is likely to be highly itinerant in character and that a significant change in the electronic structure accompanies the magnetic transitions. We observe a contribution from the topological Hall effect in the AFM phase suggesting a non-coplanar contribution to the magnetism. Electronic structure calculations predict an AFM ground state with a wavevector parallel to the crystallographic $c$-axis preferred over the experimentally measured FM state by $\approx$ 50 meV per unit cell. However, supercell calculations with a small density of Fe-antisite defects introduced tend to stabilize the FM over the AFM state indicating that antisite defects may be the cause of the sensitivity to sample synthesis conditions.