Positional Stability of Skyrmions in a Racetrack Memory with Notched Geometry

The field of storage and memory is one of the most vibrant and important areas in information and technology landscape. Recent advances in technology and computer architecture have resulted in significant improvements in memory technology. Today, it is possible to pack more and more memory components in computers, servers, laptops and smartphones. Despite their remarkably small sizes, current memory devices have larger memory spaces than conventional devices.

Several steps have been taken to improve the performance of modern memory devices by addressing the limitations of conventional memory technology. One of these challenges is the inherent delay in fetching between memory elements and logic cores. Among the existing strategies for addressing these challenges, the application of magnetic skyrmions is the most promising. Magnetic skyrmions are potential candidates for applications as information-carrying bits in logic as well as low-power high-density nonvolatile memory due to their attractive features.

Racetrack memory is the most commonly studied skyrmionic-based platform. The positional stability of skyrmions is a crucial issue in the application of racetrack memory. However, maintaining the skyrmion positional stability for a certain duration of time remains a big challenge. Moreover, skyrmions are susceptible to undesirable motion behaviors, which are controlled by pinning skyrmions by employing confinement barriers, such as notches and defects. Although the dynamics and interactions of skyrmions with pinning sites like defects and notches have been extensively studied, a systematic analysis of the notched racetrack, energy barrier mechanics and their impacts on skyrmion stability, mobility, annihilation and unintended nucleation is missing.

To bridge this research gap, Dr. Md Golam Morshed, Dr. Hamed Vakili and Professor Avik Ghosh from the University of Virginia systematically analyzed the positional stability and the possibility of producing and tunning energy barrier in skyrmion-based racetrack memory with notched geometry. Micromagnetic simulations were used to compute energy barriers associated with the notch and positional stability along the racetrack. Their work is currently published in the journal, Physical Review Applied.

The authors showed that the occurrence of energy barrier was highly dependent on the reduction in the skyrmion sizes induced by the notch created in the racetrack. A 5-nm thick racetrack provided a range of energy barriers up to ~45 K_BT. This energy barrier could endow skyrmions with lifetime positional stability for long-term memory applications. It only required a moderate current of ~ 10₁₀ A/M²to move the skyrmions. Importantly, an optimal combination of the notch radius, skyrmion size and racetrack thickness was necessary to obtain a large enough energy barrier to ensure lifetime positional stability.

Based on the simulation results, quasi-analytical equations were derived to estimate the energy barrier. In addition, other pinning mechanisms, like the local material parameter variation in the region, were explored and compared with the notched geometry. Results showed that the notched geometry provided the highest energy barrier, which also depended on the shape of the notch. Furthermore, it was worth noting that an optimal barrier height and long-term positional stability were achieved by varying the material parameters: notch geometry, chiral Dzyaloshinskii-Moriya interactions and racetrack thickness.

In summary, the ability to produce skyrmion positional stability by creating notches along the racetrack was demonstrated. The results of the analytical energy equations were compared to the simulation data and a good agreement was established. The possible quantitative differences were attributed to the differences in the geometric boundary conditions. In a statement to Advances in Engineering, Professor Avik Ghosh stated that their findings would open up possibilities toward reliable and long-term skyrmion-based racetrack memory applications.