利用阻抗分析探討雙層電阻式記憶體之物理機制

Investigating Resistive Switching Mechanism of Bilayer RRAM Using Impedance Analysis

張育愷(Yu-Kai Chang)

指導教授 ： 侯拓宏

國立交通大學/電機學院/電子工程系所/碩士(2020年)

摘要

近年來，類神經網路在許多方面展現出色的成果及潛力，相較於傳統馮·諾依曼的計算架構，其數據在讀取與儲存時需不斷往返處理器與記憶體間，因此消耗大量的時間及功率；而類神經網路透過將計算及儲存整合來縮短溝通距離。因此，許多研究則正致力於尋找適合的突觸元件來達成硬體實踐。
　　非阻絲形態之電阻式記憶體由於具備突觸特性而受到廣泛的注目，像是在相同的電壓刺激之下可以達到連續的電阻變化。然而，對於高密度陣列之應用，元件可靠度仍然是一個嚴重的問題，特別是在有限的操作次數以及阻態維持的穩定度上；儘管許多物理模型相繼被提出，像是在氧化層中氧缺陷的移動以及載子的儲存等，但缺乏完整明確的物理機制阻礙了我們對元件的了解及開發，其根本的原因來自於微觀世界的變化，像是離子移動或電荷變化，是較困難藉由直流量測的方式來直接觀察，這也驅使我們使用交流訊號之阻抗分析來了解其背後的物理機制。
　　在此篇論文中，我們提出了阻抗分析的方法並應用於分析其物理切換機制。而就我們所知，這是第一個將阻抗分析方法應用於非阻絲形態之電阻式記憶體上。在二氧化鈦(TiO2)的元件當中，我們藉由阻抗分析了解其電容及電阻有強烈的相關性，將雙層氧化層之電容率變化詳細地在此篇論文加以討論，並且提出了等效電路模型來描述元件電容及電阻之動態切換的過程，進而應用於可靠度之分析上。我們也發現了在氧化鉭(TaOx)的元件，其不同於二氧化鈦元件之切換機制，我們也對元件特性及阻抗分析進行詳細地分析，並推測其可能之物理機制。在論文的最後，我們整理了近幾年關於非阻絲形態之電阻式記憶體的相關研究，將其元件之電流密度與阻態維持之可靠度加以整理及分析，提供我們在這個領域更進一步的了解。

關鍵字

電阻式記憶體 ； 阻抗分析 ； 可靠度分析及改善

並列摘要

Recently, neuromorphic computing has shown attractive potential as a next-generation computing paradigm to improve conventional computing algorithms based on the von Neumann architecture, which faces challenges in transporting data between processor and memory. Brain-inspired architecture shortens the communication by distributing the computing in neurons and storage in synapses. As a result, many researchers are looking for promising solid-state synaptic devices for hardware implementation.
Non-filamentary resistive random-access memory (RRAM) has been attracting great attention due to its promising synaptic features, such as continuous resistance change (analog weight update) under the repeated voltage stimuli. However, device reliability, especially retention and endurance, remains an issue for the implementation in a large-scale array. Different switching mechanisms, such as charge-trapping and oxygen vacancies migration, have been proposed to explain the switching behavior. The ambiguous physical mechanism hinders device engineering from improving reliability further. The more detailed understanding of microscopic-level phenomena, such as dynamic evolutions of ionic movement and charge trapping, cannot be achieved through the DC electrical measurement, which only gives the resistance information in any electrical state. This motivates us to investigate non-filamentary switching by using AC impedance analysis.
In this thesis, we developed a methodology of impedance analysis that unravels the underlying resistive switching mechanism. To the best of our knowledge, this is the first detailed impedance analysis on the non-filamentary RRAM. A strong correlation between capacitance and resistance based on TiO2-based devices is demonstrated. The systematic extraction of the permittivity of the bilayer stack is also discussed in detail. Moreover, an impedance model is proposed to describe the switching dynamics of the capacitance and resistance change, and its application on reliability analysis is also shown by engineering different device designs. A different non-filamentary RRAM system based on TaOx-based devices is also investigated. In the end, we benchmark various non-filamentary RRAM devices reported in the literature and discuss the correlation between the device current density and reliability characteristics. This might provide an engineering pathway towards better reliability characteristics.

並列關鍵字

RRAM ； Impedance Analysis ； Reliability Analysis and Engineering

參考文獻

[1] D. Silver, A. Huang, C. J. Maddison, A. Guez, L. Sifre, G. Van Den Driessche, J. Schrittwieser, I. Antonoglou, V. Panneershelvam, and M. Lanctot, "Mastering the game of Go with deep neural networks and tree search," nature, vol. 529, no. 7587, p. 484, 2016.

[2] S. Yu, "Neuro-inspired computing with emerging nonvolatile memorys," Proceedings of the IEEE, vol. 106, no. 2, pp. 260-285, 2018.

[3] S. Yu, Neuro-inspired computing using resistive synaptic devices. Springer, 2017.

[4] B. Hudec, C.-W. Hsu, I.-T. Wang, W.-L. Lai, C.-C. Chang, T. Wang, K. Fröhlich, C.-H. Ho, C.-H. Lin, and T.-H. Hou, "3D resistive RAM cell design for high-density storage class memory—a review," Science China Information Sciences, vol. 59, no. 6, p. 061403, 2016.

[5] R. Waser and M. Aono, "Nanoionics-based resistive switching memories," in Nanoscience And Technology: A Collection of Reviews from Nature Journals: World Scientific, 2010, pp. 158-165.