Spintronics & MAGNEPIC

MAGNEPIC aims to tackle some materials and physics challenges spintronics faces nowadays by using radically different approaches and new material combinations.

Information technologies are rapidly evolving. We need more powerful computers, larger data storage capabilities, and faster reading, writing, and digital data processing. Therefore, we are in a continuous quest to shrink the size of memory elements and find more efficient ways to process the data without compromising an increasing energy and material cost. Spintronics offers one of the most viable solutions to these endeavors, and consequently, has become a prominent field of research in the past 35 years. 

Spintronics employs a fundamental property of a fundamental particle: the spin of an electron. Spin is an intrinsic angular momentum, a purely quantum property that does not have a classical counterpart. In a simplified picture, you can think of an electron as a tiny magnet with north and south poles where the orientation can represent ‘0’ and ‘1’ in a binary data system. Additionally, spins can be manipulated within a material using external stimuli (magnetic field, electric field, temperature gradient, etc.), thereby providing a powerful platform to execute all of the tasks required in digital data management.

The discovery of giant magnetoresistance (GMR) in the late 1980s has initiated the spintronics research officially. The GMR and later on tunnel magnetoresistance (TMR) effects have provided relevant electrical tools to read the data in hard disk drives and revolutionized the magnetic recording industry (see the Nobel Prize in Physics 2007). Since then, spintronics research has evolved into an interdisciplinary field encompassing physics, materials science, electrical engineering, and nanotechnology to deal with a wide spectrum of activities from the fundamental understanding of spin physics to real-world applications.

MAGNEPIC aims to tackle some materials and physics challenges spintronics faces by using radically different approaches and new material combinations. More specifically, we aim to study a family of magnetic insulators called garnets and place them in a favorable position for spintronics research and applications by exploiting emerging physics and materials/device engineering.

Magnetic Insulators

We are on the verge of a paradigm shift in spintronics where insulating magnetic materials are becoming a major player.

Early breakthrough discoveries in spintronics, namely the giant and tunnel magnetoresistance effects, magnetic interlayer coupling, interfacial perpendicular magnetic anisotropy (PMA), and spin-transfer-torques (STT), exclusively relied on conducting magnetic materials such as Co, Fe, and Ni. Due to the lack of knowledge and methods to electrically probe and control magnetization, magnetic insulators remained less explored in the spintronics context up until recently. Discoveries such as the spin Seebeck effect, spin pumping, and spin hall magnetoresistance have granted us the tools to generate and probe spin currents by actively using magnetic insulators in devices. More recently, we have achieved current-induced magnetization switching and domain wall motion in magnetic insulators by spin-orbit torques, just how it is done in conducting magnetic materials. With these recent advances, we are now on the verge of a paradigm shift in spintronics where insulating magnetic materials are becoming a major player. In MAGNEPIC, we investigate ferrimagnetic insulators. A ferrimagnet is in-between a ferromagnet and an antiferromagnet. The adjacent spins are aligned antiparallel to each other, like in an antiferromagnet, but these spins possess unequal magnetic moments, so the net magnetization is non-zero like in a ferromagnet. Because one can greatly vary the forming elements and their relative compositions in a ferrimagnetic insulator, they offer multiple tuning knobs to adjust their magnetic properties. Some ferrimagnetic insulators of interest to this project are thulium iron garnet (Tm3Fe5O12, TmIG), terbium iron garnet (Tb3Fe5O12, TbIG), and yttrium iron garnet (Y3Fe5O12, YIG).

Perpendicular magnetic anisotropy

Perpendicular magnetic anisotropy offers great advantages for spintronics experiments and future device applications.

Magnetic anisotropy is the tendency for the magnetization to point along certain preferential axes in a magnetic material. It is an important property for spintronic devices as the magnetic data can be encoded in the magnetization orientation, taking two equally stable values (e.g., up -> 1, down -> 0). There is an energy barrier associated with reversing the magnetization between these two stable states. A large barrier makes the anisotropy strong and the memory device robust against external disturbances (temperature and magnetic field); however, more energy is required to change its state (write the data). With the same reasoning, low anisotropy devices would be more “energy-efficient” but may be prone to inadvertent memory loss. Therefore, anisotropy engineering is a crucial task for spintronic devices to fulfill the required operational conditions. Magnetic anisotropy pointing perpendicular to the film surface (perpendicular magnetic anisotropy, PMA) is the most suitable option for downsizing magnetic memory bits while retaining large (and tunable) energy barriers.

In MAGNEPIC, we develop ferrimagnetic garnets with PMA. Thanks to the compositional tunability, ferrimagnetic garnets offer an excellent platform for anisotropy engineering for future spintronic applications. The mechanism to obtain PMA in our garnet films relies on the mismatch of their lattice parameters relative to the substrate. We grow our films on deliberately chosen substrates in optimized conditions (temperature, pressure, etc.) to induce tensile strain. As a consequence, the PMA emerges in order to minimize the overall magnetic energy. Since the PMA in our ferrimagnetic garnet films is a bulk property, we can change the thickness of our films to a great extent from a few to tens of nanometers while maintaining the PMA. This aspect offers great flexibility for experiments and future device applications.

Current-induced Spin-Orbit Torques

Spin-orbit torques are presently the state-of-the-art current-induced magnetic manipulation method in spintronic devices.

Spin-orbit torques (SOTs) describe the current-induced magnetic torques originating from the spin-orbit coupling of conduction electrons in the bulk and interfaces of materials. First demonstrated experimentally in semiconductors and then in metals, SOTs have attracted enormous attention from the spintronics and community due to their immense potential and flexibility in manipulating magnetization by electricity. Over the past decade, a handful of  SOT generating mechanisms have been discovered, including the spin Hall effect (SHE), interfacial Rashba–Edelstein effect, topological surface and bulk states, etc. Among these, the SHE is primarily considered as a SOT source for current-induced magnetization manipulation in ferrimagnetic garnets, of interest to our project. The SHE emerges in the bulk of the materials characterized by strong spin-orbit coupling (e.g., Pt, W, and Ta) and describes the spin-dependent scattering of conduction electrons, generating a pure spin current transverse to the charge current injection direction. A magnetic layer in interfacial contact with a SHE material can then absorb the resulting spin current, which acts as a spin-torque on the magnetization and enables its reversal. While the spin current generation mechanism differs in the other listed SOT sources, the applied torque has comparable symmetry. 

Chiral magnetism at interfaces