鐵電氧化鉿鋯記憶體特性及其應用
No Thumbnail Available
Date
2018
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
近幾年來,鐵電材料已經被廣泛的使用,其應用在鐵電材料記憶體(Ferroelectric Random-Access Memory, FeRAM)上最多,因此被當成新興技術之非揮發性記憶體之一。鐵電閘極場效式電晶體(Ferroelectric Field Effect Transistor, FeFET)具有許多優點,如讀寫速度快、操作電壓低…等特性。雖然有著眾多優點但在過去,傳統鈣鈦礦鐵電材料,如Pb (ZrxTi1-x) O3 (PZT)和BaTiO3 (BTO)等有機材料一直無法直接整合於現今CMOS(Complementary Metal-Oxide-Semiconductor)技術上。經過多年研究發現若是使用二氧化鉿為基底,既能展現出鐵電材料之特性,又是能使用在CMOS技術上的高介電係數材料。
本研究將主題分為三個部分,第一部分為探討鐵電Hf1-xZrxO2場效應電晶體與鰭式電晶體,兩者在高數據保持率(Retention)和讀取耐久性(Endurance)的結果與差異;第二部分為在具有鐵電層的元件上,使用各種方式施加偏壓使鐵電極化翻轉,量測出穿隧現象;第三部分為利用具有鐵電特性元件,透過不同波型模組設定做深度學習觀察其差異及變化。
In recent years, ferroelectric material has been extensively investigated. Hf-based oxide materials with ferroelectricity have the potential for application on FeRAM. Therefore, it is considered as one of the non-volatile memory candidates of emerging technologies. The FeFET (Ferroelectric Field Effect Transistor) has many advantages, such as fast read and write speeds, low operation voltage, and high endurance operation. Conventional Perovskite-type ferroelectric material, such as BaTiO3 (BTO) and Pb (ZrxTi1-x) O3 (PZT), has the issues of the process compatibility and scaling down for FeRAM applications. The HfO2-based with ferroelectricity may for CMOS applications due to process compability. The subject of this study will be divided into three parts. The first part is the results of ferroelectric Hf1-xZrxO2 planar MOSFET (Metal-Oxide-Semiconductor Field-Effect Transistor) and FinFET (Fin Field-Effect Transistor) for data retention and endurance. The second part is FTJ (Ferroelectric Tunneling Junction) including the dependence on voltages, HZO thickness, and interface treatment in chapter 3. The third part is deep learning (training) of FeFET and MFM in chapter 4.
In recent years, ferroelectric material has been extensively investigated. Hf-based oxide materials with ferroelectricity have the potential for application on FeRAM. Therefore, it is considered as one of the non-volatile memory candidates of emerging technologies. The FeFET (Ferroelectric Field Effect Transistor) has many advantages, such as fast read and write speeds, low operation voltage, and high endurance operation. Conventional Perovskite-type ferroelectric material, such as BaTiO3 (BTO) and Pb (ZrxTi1-x) O3 (PZT), has the issues of the process compatibility and scaling down for FeRAM applications. The HfO2-based with ferroelectricity may for CMOS applications due to process compability. The subject of this study will be divided into three parts. The first part is the results of ferroelectric Hf1-xZrxO2 planar MOSFET (Metal-Oxide-Semiconductor Field-Effect Transistor) and FinFET (Fin Field-Effect Transistor) for data retention and endurance. The second part is FTJ (Ferroelectric Tunneling Junction) including the dependence on voltages, HZO thickness, and interface treatment in chapter 3. The third part is deep learning (training) of FeFET and MFM in chapter 4.
Description
Keywords
鉿基氧化物, 鐵電電晶體, 深度學習訓練, Hafnium-based Oxides, FeFET, deep learning training