Research

The field of magnetism has seen a surge of interest in a novel classification known as altermagnetism over the past few years. This interest is driven by the unique physical properties of altermagnetic materials, which takes advantages of both conventional antiferromagnets and ferromagnets through the altermagnetic spin-splitting effect (ASSE). The non-relativistic ASSE enables the generation of not only a longitudinal spin polarized current but also a transverse pure spin current upon injecting a charge current along certain crystal orientations. However, experimentally unequivocal observation of the ASSE is challenging. This difficulty arises from the inevitable mixing of ASSE with the spin Hall effect (SHE) caused by the material's relativistic spin-orbit coupling and the ASSE's dependence on the hard-to-probe and hard-to-control Néel vectors.

In our recent work, we address these challenges in the study of ASSE in an altermagnet candidate RuO2 [1]. Our experimental observations revealed a highly anisotropic spin-to-charge conversion in the epitaxial RuO2 thin film. We attribute the anisotropy to an altermagnetic origin by ruling out the complication of an anisotropic SHE. We found the IASSE exhibited an opposite sign compared to the inverse spin Hall effect (ISHE). Remarkably, the efficiency of the IASSE was found to be consistently 70% of that of the ISHE in RuO2 for thicknesses ranging from 5 nm to 32 nm. Furthermore, we demonstrated that the ASSE/IASSE effects are observable only when the Néel vectors are well-aligned by modifying the Néel vector domains via RuO2 crystallinity through different substrates. Interestingly, the shape of the thermal voltage hysteresis loop is anisotropic and is consistent with the anisotropic magnetic hysteresis loop of YIG, which verifies the [001] orientation of the RuO2 Néel vectors. Our study provides significant insights into the spin-splitting effect in altermagnetic materials, paving the way for future advancements in spintronic technologies.

In condensed matter physics, there are various states of materials, including insulators, metals, magnets, and superconductors. And in the last few decades, the topological version of them have been actively identified, such as topological insulators, where the bulk is insulating and the surface is conducting; the topological semimetals, where the bulk band has gapless band touching point; the topological magnets, such as skyrmions, where spins rotate 180 degrees when traverses from the center to the peripheral; and the topological superconductors where the Majorana fermions are expected to be excited. The topology of these materials can be described by their topological invariants.

Among these topological materials, the Weyl semimetal has attracted our interest. The Weyl semimetal is a type of topological semimetal with broken inversion or time-reversal symmetry. Unlike the topological insulator which is insulating in the bulk, the Weyl semimetal is a gapless conducting material, with conducting surface states known as Fermi arcs and conducting bulk with band touching points known as the Weyl nodes. They are robust against symmetry perturbations. The Weyl nodes behave like the magnetic monopoles in the momentum space and are serving as sources and sinks for the Berry curvature. A large berry curvature can have a large intrinsic contribution to the anomalous Hall and spin Hall effect. And the large transverse spin-dependent response is an ideal feature for spintronic applications.

An electron carries not only charge but also angular momentum (spin). This additional degree of freedom makes spintronics a potential solution to overcome the bottleneck of Moore's Law in the semiconductor industry.  Compared to electronic devices, spintronic devices are more energy-efficient, faster, and stable. They allow information to be stored at zero consumed energy, reach a computational speed up to the THz range, and are more stable against external perturbations.

There are a few methods one can use to generate a spin current, including the charge current, heat current, magnetic precessions, and others. Reciprocally, the spin current is also able to be converted back to the charge current, heat current, and magnetic precessions, and thus be detected.

Our study focuses on the generation, manipulation, and detection of spin current via the established spintronic effects as well as the discovery of novel spintronic phenomena. For example, in the spin/charge conversion scenario, the conventional ways utilize the spin Hall and inverse spin Hall effects to convert charge to spin and spin to charge, respectively. In these effects, the electrons are scattered to the transverse direction due to the spin-orbit coupling, thus the charge current, spin current, and spin polarization are mutually orthogonal. We have studied the spin/charge conversion in various materials and determined their conversion efficiencies (spin Hall angles). 

More recently, we also excitingly observed the magnetization-dependent spin Hall effect in a magnetic material. This effect allows arbitrary control of the spin polarization at the film surface by the magnetization direction of magnets, lifting the strict orthogonal relation in the conventional spin Hall effect. Thus, it opens a versatile and controllable route to explore the unconventional spin and charge conversions in magnetic materials and is beneficial for next-generation spintronic applications.

Mother nature has endorsed us with great blessings. One of such is the endless amount of energy we received from the sun. In return for nature's favor, green energy harvesting is of ultimate importance. 

More than two hundred years ago, Thomas Johann Seebeck discovered the phenomenon where heat can be directly converted into a charge current. This effect is named the Seebeck effect and is the guiding principle for thermopiles, thermocouples, thermoelectric generators, and more. And only until the last two decades, researchers have observed the spin version of the Seebeck effect, where a heat current generates a spin current, and this effect is called the spin Seebeck effect.

Experimentally, there are two geometries to observe the spin Seebeck effect, the transverse and longitudinal geometries. The transverse geometry encounters great difficulty in the unequivocal detection of the spin current, while the longitudinal geometry is well established in the magnetic insulators.

Our research has been mostly focused on utilizing the spin Seebeck effect in the ferromagnetic insulator YIG to generate the spin current. And we use the thermally generated spin current as a tool to study the spin to charge conversion in various materials of interest. The advantages of using the spin Seebeck method are that it utilizes the metal-insulator structure, thus is free from the complication of current distribution in the metal/metal structure; It has only one interface which greatly reduced the loss of spin current when traversing through multiple interfaces; And it is highly self-consistent when keeping the same experimental conditions. 

Spin waves, or magnons, are collective excitations of electron spins in magnetic systems. They can propagate in ferromagnetic and antiferromagnetic materials, including metals and insulators, and can cover a wide frequency band, from a few tenths of a GHz up to the THz range with wavelengths extending from the micrometer to the nanometer regime. The advantages of using magnonic devices are substantially low-energy dissipation, which can be achieved by reducing the Joule heating due to the absence of charge current, and device miniaturization owing to the orders of magnitude shorter wavelengths of spin waves compared to that of electromagnetic waves. 

Efficient generation and guidance of spin waves are ultimate goals for realizing nano-magnonic devices. In our work, we provide a compelling method to generate, propagate, and couple the spin-wave modes to a domain wall with micrometer decay length. The spin-wave is excited by an out-of-plane oscillating magnetic vortex core, which is induced by an out-of-plane radio-frequency (rf) magnetic field under zero dc bias field, in a nanodisk-film hybrid structure. We obtain tunable spin waves with frequencies up to 15 GHz and wavelengths down to 80 nm. The spin waves can further couple to and propagate along the nanochannel formed by the domain wall in the Py film with a decay length of about 1.36 μm.