Investigation of Energy Efficient Magnetic Switching in Novel Materials and Device Structures for Spintronics Application

Jason Cheng-Hsiang Hsu

EECS Department
University of California, Berkeley
Technical Report No. UCB/EECS-2024-199
December 1, 2024

http://www2.eecs.berkeley.edu/Pubs/TechRpts/2024/EECS-2024-199.pdf

Amidst the digital transformation era, data-intensive computing applications such as artificial intelligence and big data analytics are rapidly evolving. As a result, computing hardware development is under pressure to deliver higher performance while minimizing energy consumption. This presents a significant opportunity for hardware research, moving alongside and beyond traditional CMOS technology. Spintronic devices have emerged as a promising candidate to address the gaps in the computing memory hierarchy and serve as an emerging device for beyond von Neumann computing schemes. In the core of a spintronic device, the spin-charge conversion phenomenon plays a crucial role in determining its performance and energy efficiency. Spin-charge conversion enables the generation of a spin current, which can be utilized to manipulate magnetization and magnetic textures - spin-orbit torque. However, several significant challenges hinder the widespread adoption of spintronic devices in future computing hardware. These challenges encompass high operating power in comparison to traditional silicon counterparts, ensuring material compatibility with the silicon CMOS platform, developing highly scalable solutions for future process nodes, and addressing specific fundamental design hurdles. In this dissertation, my primary objective is to tackle various challenges by innovatively designing materials, fabrication processes, and device structures. Firstly, I discover significant spin-charge conversion efficiency in commercially available silicides, which exhibit a scalable underlying mechanism that could offer new insights into spin physics. Leveraging these silicides, I achieve highly energy-efficient magnetic switching, outperforming state-of-the-art heavy metal systems for magnetic memory applications. Moreover, I tackle the fundamental design challenge of achieving field-free spin-orbit torque switching by engineering a device fabrication process that allows for easy tunability of magnetic anisotropy. Lastly, I propose and experimentally explore a potential pathway towards the long-awaited pure voltage-driven magnetic bi-directional switching in commercially available magnetic tunnel junction devices.

Advisor: Sayeef Salahuddin

"; ?>

BibTeX citation:

@phdthesis{Hsu:EECS-2024-199,
    Author = {Hsu, Jason Cheng-Hsiang},
    Editor = {Salahuddin, Sayeef and Yablonovitch, Eli and Analytis, James and Hellman, Frances},
    Title = {Investigation of Energy Efficient Magnetic Switching in Novel Materials and Device Structures for Spintronics Application},
    School = {EECS Department, University of California, Berkeley},
    Year = {2024},
    Month = {Dec},
    URL = {http://www2.eecs.berkeley.edu/Pubs/TechRpts/2024/EECS-2024-199.html},
    Number = {UCB/EECS-2024-199},
    Abstract = {Amidst the digital transformation era, data-intensive computing applications such as artificial intelligence and big data analytics are rapidly evolving. As a result, computing hardware development is under pressure to deliver higher performance while minimizing energy consumption. This presents a significant opportunity for hardware research, moving alongside and beyond traditional CMOS technology. Spintronic devices have emerged as a promising candidate to address the gaps in the computing memory hierarchy and serve as an emerging device for beyond von Neumann computing schemes. In the core of a spintronic device, the spin-charge conversion phenomenon plays a crucial role in determining its performance and energy efficiency. Spin-charge conversion enables the generation of a spin current, which can be utilized to manipulate magnetization and magnetic textures - spin-orbit torque. However, several significant challenges hinder the widespread adoption of spintronic devices in future computing hardware. These challenges encompass high operating power in comparison to traditional silicon counterparts, ensuring material compatibility with the silicon CMOS platform, developing highly scalable solutions for future process nodes, and addressing specific fundamental design hurdles. In this dissertation, my primary objective is to tackle various challenges by innovatively designing materials, fabrication processes, and device structures. Firstly, I discover significant spin-charge conversion efficiency in commercially available silicides, which exhibit a scalable underlying mechanism that could offer new insights into spin physics. Leveraging these silicides, I achieve highly energy-efficient magnetic switching, outperforming state-of-the-art heavy metal systems for magnetic memory applications. Moreover, I tackle the fundamental design challenge of achieving field-free spin-orbit torque switching by engineering a device fabrication process that allows for easy tunability of magnetic anisotropy. Lastly, I propose and experimentally explore a potential pathway towards the long-awaited pure voltage-driven magnetic bi-directional switching in commercially available magnetic tunnel junction devices.}
}

EndNote citation:

%0 Thesis
%A Hsu, Jason Cheng-Hsiang
%E Salahuddin, Sayeef
%E Yablonovitch, Eli
%E Analytis, James
%E Hellman, Frances
%T Investigation of Energy Efficient Magnetic Switching in Novel Materials and Device Structures for Spintronics Application
%I EECS Department, University of California, Berkeley
%D 2024
%8 December 1
%@ UCB/EECS-2024-199
%U http://www2.eecs.berkeley.edu/Pubs/TechRpts/2024/EECS-2024-199.html
%F Hsu:EECS-2024-199