Research history

After graduating in physics in 2013 at ETH Zurich, I joined the Institute of Geophysics as a research assistant. Applying fundamental physical concepts for studying the magnetism of Earth materials has always fascinated me, which is why in turn I started a PhD in April 2014 in collaboration with the Laboratory of Metal Physics and Technology (LMPT). My research was on monoclinic pyrrhotite (an iron monosulfide with the ideal formula Fe₇S₈) that has an ordered defect structure (Fig. 1a) which accounts for a rich variety of physical properties. After my PhD defense in late 2018, I stayed for one more year as a post-doc at the LMPT.

Fig. 1. a) Monoclinic pyrrhotite structure with the distribution of Fe-sites (spheres) and Fe vacancies (open circles). The sulfur atoms are omitted for clarity reason b) Magnetization curves measured in 100 and 3000 mT fields within the ab-plane. The vertical dashed line at T_B ≈ 31 K indicates the low-temperature transition.

Thesis description

My PhD work was a combined experimental and numerical approach to assess the magneto-structural properties of ferrimangnetic monoclinic pyrrhotite down to low temperature. Particular focus was on the physical mechanism of a low-temperature transition around T_B ≈ 31 K (Fig. 1b). Since its discovery in 1964, there has been an ongoing controversy as to whether this transition is caused by a change in the crystal structure (in analogy to the T_V ≈ 120 K Verwey transition of Magnetite), or whether it is associated with a purely magnetic phenomenon. The outcome of my PhD provided a substantial contribution in solving this controversy, and the key findings can be summarized in three parts:

1. In order to understand the magnetism of a material, it is crucial to also understand its underlying structural properties. For monoclinic pyrrhotite, the alternating stacking sequence of vacancy-free and vacancy-bearing iron layers (Fig. 1a) is key for the resulting magnetic and anisotropy properties, i.e., the directional dependence of the Fe²⁺ spins. Theoretical considerations postulated that the Fe²⁺ spins of those sites that have an inter-layer vacancy as nearest neighbour, exhibit different anisotropy properties compared to those with either no, or an intra-layer vacancy as nearest neighbour. Given this, the monoclinic pyrrhotite is a system with two magnetic components that differ in their anisotropy properties.

2. Magnetic torque measurements, which are key for assessing magnetic anisotropy, were conducted down to low temperature to uncover the magnetic thermodynamics. The experimental data was concisely described by implementing a prototypical phenomenological model based on statement 1) (Fig. 2a). The model was able to infer the gradual out-of plane Fe²⁺ spin rotation that occurs at T < 200 K, and the individual components of the two groups of Fe-sites (Fig. 2b). At the T_B ≈ 31 K transition, a reversed rotation sets in, which within this model has been explained by an interaction between the two anisotropy components.

Fig. 2. a) Magnetic torque measured between 200 and 2 K by rotating a B=500 mT field in the bc-plane (open circles) and their fit (solid line). b) Temperature dependence of the total (\( \theta_{\textnormal{tot}} \)), and the individual out-of-plane Fe²⁺ spin rotation angles of the two groups (\( \theta_{I} \), \( \theta_{II} \)) as inferred from the fit in a).

3. The findings from statement 2) were indicative of a magnetic phenomenon associated with the transition but they could not unambiguously exclude a crystallographic change. For this purpose, powder neutron diffraction measurements were accomplished in order to simultaneously study the crystal structure and magnetic long-range order. The diffraction data between 298 and 5 K were well described using a monoclinic structural model from the literature (Fig. 3a). Moreover, the refinement of the diffraction data revealed a convergence of the Fe-Fe bond lengths across the transition (Fig. 3b), which induces a change in the spin-orbit coupling, i.e., magnetic anisotropy. Given this, the anisotropy modification across the transition as postulated in statement 2) is triggered by structural changes on an atomic level within a crystallographically stable system.

Fig. 3. a) Powder neutron diffraction patters (spheres) and their fit (solid line) with a monoclinic structural model b) Temperature evolution of Fe-Fe bond lengths inferred from the fit in a)

The results of my PhD work were published in peer-reviewed physical and geophysical journals. You can access the links to my publications here. Moreover, you can also download a copy of my PhD thesis from the ETH Research collection here.