Figure 1. Schematic illustration of the real space (upper panel) and momentum space (lower panel) for ferromagnetism (left), antiferromagnetism (center) and altermagnetism (right).
Ferromagnetism (FM) and antiferromagnetism (AFM) are the two major classifications of magnetism over the past century. For FM and AFM, the neighboring spins are aligned parallelly and antiparallelly with each other, resulting in sizable (M >> 0) and zero (M ≈ 0) magnetization, respectively. In the momentum space, FM and AFM has spin-splitting and spin-degenerated bands, respectively, where the former is able to generate a spin polarized current upon charge current injection, but the later cannot.
Figure 2. Schematic illustration on the spin to charge conversion through the inverse altermagnetic spin-splitting effect (IASSE) and the inverse spin Hall effect (ISHE) from the spin current injection using spin Seebeck effect (SSE).
In the past few years, an emerging classification of magnetism has been identified as altermagnetism (AM), which combines the advantages of both ferromagnetism (FM) and antiferromagnetism (AFM). The AM has no magnetization (M ≈ 0) like an AFM, but has spin-splitting Fermi surfaces in the momentum space like an FM. It is an ideal material for spintronic applications, carrying no magnetization but efficiently delivers angular momentum. Therefore, comparing to FM, the AM based device is expected to have better performances, such as higher device density, higher stability and faster writing speed.
What material has altermagnetism? Theorists predict that when the symmetry of spin and lattice are different, AM could occur. The prototype material is the rutile RuO2 with C4 andC2 rotational symmetry for lattice and spin spaces, respectively. Depending on the relative orientation to the Néel vectors, the injected charge current is able to generate a longitudinal spin polarized current or a transverse pure spin current in RuO2 through the altermagnetic spin-splitting effect (ASSE). Reciprocally, if the spin current is injected along a specific direction relative to the Néel vectors, RuO2 converts the spin current into a charge flow through the inverse altermagnetic spin splitting effect (IASSE). Since the ASSE or IASSE depends on the orientation of the Néel vectors, a highly anisotropic behavior is a key signature for the detection of ASSE or IASSE.
Figure 3. A picture of the team from the Quantum Spintronics Lab at National Taiwan University. C.-T. Liao (first author, upper right), Y.-C. Tien (upper left), Y.-C. Wang (lower right), S. -Y. Huang (corresponding author, lower left) and D. Qu (corresponding author, lower center)
In order to capture the IASSE in RuO2 experimentally, a few conditions have to be satisfied. Firstly, we need a high-quality epitaxial thin film to ensure the alignment of Néel vectors in a single direction to capture the anisotropy; Secondly, an insulating spin current source is preferred to avoid complications from parasitic effects, such as anomalous Hall and Nernst effects. Thirdly, since RuO2 has spin-orbit coupling (SOT), the charge current generated from IASSE has to be separated from the conventional inverse spin Hall effect (ISHE) that is caused by the SOT. The fulfillment of each requirement is challenging, but excitingly we have achieved them all.
In our experiments, by injecting spin current from the spin Seebeck effect in YIG, we successfully observed the anisotropic spin to charge conversion in the high-quality YIG/RuO2/TiO2 (substrate) samples with single Néel vector orientation. By separating the results from ISHE, we confirmed the existence of the IASSE in RuO2. Moreover, we observed an anisotropy not only in the size but also in the shape of the detected voltage. The anisotropy in shape is caused by the pining of the YIG magnetization by the adjacent RuO2 Néel vectors along the [001] directions. Our results bring important insights in the study of altermagnetism.
Corresponding author’s email address: [email protected]
