An inverse cassette tape: racetrack memory may offer a more efficient means of digital data storage
Many of the devices we use daily rely on the incredible density of digital data storage available to us. Improving on this technology is a high priority for many businesses and researchers, as it unlocks the possibility for miniaturising digital technologies and improving computational power. These possibilities have motivated researchers in the NYU Center for Quantum Phenomena (CQP) to help develop a new format of digital memory.
This technology, generally referred to as “racetrack memory,” works by moving configurations of magnetic fields around the surface of a conductive magnetic material. Imagine a cassette tape; they read data by moving a material (the tape) across a reader, which decodes the information written on the material to reproduce sound. With racetrack memory, it is the inverse; the material stays in place, and the information itself is moved across the reader. There are numerous methods for creating and manipulating information to be applicable for digital data storage. Researchers and students in Professor Andrew Kent’s lab are working on two options: domain walls and skyrmions. Properly developed and implemented, racetrack memory could supplant current methods of data storage (namely flash memory and disc drivers) due to its improved density of information storage, faster computations and lower energy use.
The first candidate, domain walls, are boundaries on a magnet where a region of magnetic polarization in one direction meets a region polarized in the opposite direction (see Figure 1). These boundaries are not fixed; their location can be moved by applying an electric current to the magnet they are on. These states are only stable in very specific material environments, so identifying the ideal materials for domain walls is a first priority for making the technology applicable.
Post-doctorate fellow Yassine Quessab has been characterizing how these microscopic magnetic charge configurations behave on different materials. Quessab’s work involves fabricating these candidate materials and testing their suitability for maintaining stable domain walls. These materials take the form of incredibly thin layers of metallic alloys, sandwiched between other conductive and non-conductive materials. This combination of different materials is essential for creating an environment in which domain walls can be used. At present, Quessab is working on an alloy of cobalt and gadolinium, surrounded by platinum and tungsten, with each layer only a few nanometers thick (see Figure 2). Given their incredible thinness, researchers must “grow” them in an ultra-high vacuum, by slowly building up layers a single atom thick. In Figure 2, these distinct layers of atoms can be seen directly.
With the material ready, it can then be analysed. It is set on a small cell and wired up so an electric current can be passed through. To observe the domain walls on this magnet, the material is examined under a Magneto-Optic Kerr Effect (MOKE) microscope. The resulting image can be seen in Figure 3; the current flows from the central grey rectangle, across the magnet (the thin vertical segment) to the outer grey region. On the right image, the white regions show the parts of the magnet which have a magnetic polarisation pointing down, while the grey regions are pointing up. The borders between these regions are domain walls.
Beyond visualising domain walls, researchers in Professor Kent’s lab work to identify the ideal temperature ranges for these materials to support stable domain walls. In April 2020, the team will travel to the Lawrence-Berkeley National Laboratory in California with their selected materials to perform further trials with the instruments available there.
Ultimately, the team aims to use their knowledge of domain walls to shift to constructing and manipulating skyrmions in their place. Skyrmions, like domain walls, are configurations of magnetic polarisations which can be moved around a suitable conductor. They offer a technological advantage over domain walls as they may be grouped together much more densely on the conductor, allowing for more dense memory storage. However, skyrmions are more challenging to create and manipulate. Rather than a simple boundary between positive and negative magnetic polarisation, skyrmions resemble discs with magnetic fields which continuously rotate across their face (see Figure 4). Professor Kent explains them as a “vortex of magnetic field”; their complex topology (their configuration of magnetic polarization) allows them to be stable on the conductor as a vortex is stable in a fluid.
More of the team’s work on skyrmions can be read in their forthcoming paper. This work falls within the larger focus of Professor Kent’s lab on spintronics—the investigation of how the “spin” of electrons particles interact with magnetization. An understanding of these interactions can lead to new capacities to manipulate magnetic and electric fields. Further work remains to be done to develop these capacities such that they may be implemented in consumer electronics, but skyrmion racetrack memory may soon become a critical mode of data storage, enabling a new wave of miniaturisation.