MRS Meetings and Events

 

EL20.11.04 2023 MRS Fall Meeting

Hf0.5Zr0.5O2 Based Ferroelectric Memories for Applications Down to Deep Cryogenic Temperatures

When and Where

Nov 30, 2023
4:45pm - 5:00pm

Hynes, Level 3, Room 301

Presenter

Co-Author(s)

Sayani Majumdar2,Heorhii Bohuslavskyi1,Kestutis Gregorios1,Mika Prunnila1

VTT Technical Research Centre of Finland1,Tampere University2

Abstract

Sayani Majumdar2,Heorhii Bohuslavskyi1,Kestutis Gregorios1,Mika Prunnila1

VTT Technical Research Centre of Finland1,Tampere University2
Since the last decade, extensive studies on HfO<sub>2</sub>-based ferroelectric devices have been carried out to unlock their applications in high-density, low-power, and high-speed electronics, and most importantly in non-volatile memories and logic circuits. Emergence of HfO<sub>2</sub>-based ferroelectric memories has its origin in the CMOS compatibility and scalability potential, therefore bringing the prospect of easy integration with the existing foundry processes [1]. Among ferroelectric hafnia, Zr-doped HfO<sub>2</sub> (HZO) has received most attention due to its low crystallization temperature, making it a suitable candidate for CMOS back-end-of-line (BEOL) integration. More recently, memories operating at cryogenic temperatures became a topic of high interest due to their potential applications in high-performance computing, quantum computing and space technologies [2]. To realize the memory devices functional down to deep cryogenic temperatures with reasonably low power dissipation, understanding of fundamental device physics and proper material engineering are of vital importance. Previously, perovskite oxide based ferroelectric tunnel junctions, when cooled down to low temperatures, were demonstrated to have several key parameters improved as compared to room temperature operation, featuring higher on/off ratio, faster switching and improved endurance [3]. In terms of HZO-based capacitors or junctions, a few recent papers reported operation down to low temperatures [4-6]. However, a comprehensive understanding of the cryogenic physics and properties of HZO devices within the wide temperature window from 300 K down to 4 K temperatures as well as the optimization of interfaces and material stack for cryogenic operation are currently missing. In the present work, we report high remnant polarization from quasi wake-up free Hf<sub>0.5</sub>Zr<sub>0.5</sub>O<sub>2</sub> thin film devices together with state-of-the-art endurance down to deep cryogenic temperatures [7]. The HZO films, grown and annealed under CMOS BEOL-compatible conditions, were characterized within the 300 – 4 K temperature range with a particular focus on the analog switching properties of the films. Finally, we evaluate the overall performance of our HZO devices as cryogenic memories aiming at optimizing the domain dynamics control for achieving high classification accuracy of in-memory computing circuits [7,8].<br/><br/>The authors gratefully acknowledge the financial support from the Academy of Finland projects (Grant no. 350325, 345068, and 350667). This work was realized using experimental facilities of Micronova National Research Infrastructure for Micro- and Nanotechnology.<br/><br/>[1] S. Majumdar, Advanced Intelligent Systems 4, 2100175 (2022).<br/>[2] S. Alam et al., Nature Electronics 6, 185–198 (2023).<br/>[3] Q. H. Qin et al., Advanced Materials 28, 6852-6859 (2016).<br/>[4] D. Wang et al., Japanese Journal of Applied Physics 58, 090910 (2019).<br/>[5] J. Hur et al., IEEE Journal on Exploratory Solid-State Computational Devices and Circuits, vol. 7, no. 2, pp. 168-174 (2021).<br/>[6] J. Hur et al., IEEE Transactions on Electron Devices, vol. 69, no. 10, pp. 5948-5951, (2022).<br/>[7] H. Bohuslavskyi et al., <i>in preparation </i>(2023).<br/>[8] S. Majumdar, Neuromorphic Computing and Engineering 2 (4), 041001 (2022).

Symposium Organizers

Gina Adam, George Washington University
Sayani Majumdar, Tampere University
Radu Sporea, University of Surrey
Yiyang Li, University of Michigan

Symposium Support

Bronze
APL Machine Learning | AIP Publishing

Publishing Alliance

MRS publishes with Springer Nature