Unambiguous separation of the inverse spin Hall and anomalous Nernst Effects within a ferromagnetic metal using the spin Seebeck effect

TL;DR

This study demonstrates a method to distinctly measure and separate the inverse spin Hall effect and anomalous Nernst effect in a ferromagnetic metal using the spin Seebeck effect, clarifying their individual contributions.

Contribution

The paper introduces a technique to unambiguously distinguish ISHE and ANE in a ferromagnetic metal by using a non-magnetic spacer and magnetic coercivity differences.

Findings

ISHE and ANE can occur simultaneously with opposite polarity in a single ferromagnetic metal.

Using a spacer material allows direct observation of pure spin flow.

The contributions of ISHE and ANE to the thermal response can be separated unambiguously.

Abstract

The longitudinal spin Seebeck effect is measured on the ferromagnetic insulator Fe$_3$O$_4$ with the ferromagnetic metal Co$_{0.2}$Fe$_{0.6}$B$_{0.2}$ (CoFeB) as the spin detector. By using a non-magnetic spacer material between the two materials (Ti), it is possible to decouple the two ferromagnetic materials and directly observe pure spin flow from Fe$_3$O$_4$ into CoFeB. It is shown, that in a single ferromagnetic metal the inverse spin Hall effect (ISHE) and anomalous Nernst effect (ANE) can occur simultaneously with opposite polarity. Using this and the large difference in the coercive fields between the two magnets, it is possible to unambiguously separate the contributions of the spin Seebeck effect from the ANE and observe the degree to which each effect contributes to the total response. These experiments show conclusively that the ISHE and ANE in CoFeB are separate phenomena with different origins and can coexist in the same material with opposite response to a thermal gradient.