透過您的圖書館登入
IP:3.139.85.170
  • 學位論文

二硫化鎢電阻式記憶體開發及研究

Development and Investigation of the Tungsten Disulfide Resistive Random-Access Memory

指導教授 : 吳肇欣
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


因應人工智慧及物聯網技術的蓬勃發展,半導體產業對於運算效能提升及儲存級記憶體的需求與日俱增,打破傳統范紐曼計算機架構的運算瓶頸與新興記憶體技術的開發及應用在後摩爾定律時代可謂勢在必行。而其中,二維材料具有可層狀堆疊的特性,並且相較於塊材而言,在原子級薄尺度下仍維持高電子遷移率,並具備符合熱預算之後段製程可行性。 首先,透過文獻回顧,吾人發現現行二維材料電阻式記憶體的研究在元件結構上的問題,忽略尺寸效應及過厚的中間層。吾人利用機械剝離法製備交叉點結構之二硫化鎢電阻式記憶體,且經由統計驗證在此製程結構下元件具有一定程度的均勻性。 其次,透過導入化學氣象沉積的大面積轉移技術,成功將元件的二硫化鎢中間層微縮至數層,在電性表現上獲得較小的操作電壓,並在製程方面展現可擴充性。 再者,吾人探討元件的傳導機制。初步先透過載子傳導的公式進行電性擬合,並輔以電性變溫量測的結果作為驗證。而後更進一步以GineatraTM軟體分析平台作為輔助,試著更完整地了解阻絲形貌及缺陷的動態改變。 最後,吾人製備兩種分別以氧電漿轟擊及紫外光臭氧處理的二硫化鎢及氧化鎢異質結構元件,經由電性量測及材料分析之結果,發現經紫外光臭氧處理的元件電流下降,並探討兩種處理方式的結構差異。

並列摘要


Due to the blooming development of AI and IoT, the demand for boosting high-performance computing and the amount of storage class memory is sky-rocking in the semiconductor industry. Therefore, in the post-Moore's law era, how to break through the bottleneck of the traditional Von Neuman computer architecture and the development as well as application of emerging nonvolatile memory are urgent matters to be solved. Under this circumstance, Two-dimensional materials could be one of the promising candidates since 2D materials own property of layered-structure and 2D Transition metal dichalcogenide still hold high mobility under atomic-scale compared with traditional bulk material in Silicon technology. To start with, through the literature review of Two-dimensional material Random Resistive-Access Memory, the main issue is found that many pieces of research are conducted with interlayer thicker than 100 nm and device area larger than thousands of μm2 in common bottom electrode structure, which leads to size effect in real application considering the nowadays technology node. Therefore, in this study, a cross-point device structure is applied and focuses on Tungsten disulfide-based RRAM under the scale of several μm2. For the second part, WS2 RRAM is acquired by mechanical exfoliation titanium electrode, which is studied in electrical characteristic including reliability and proved to have uniformity to a certain degree through statistical analysis. Continuously, by implying CVD grown large area transfer WS2, the switching layer shrinks to about 2 nm, which results in a forming-free phenomenon and a lower operating voltage. In the third part, the modeling and characterization of exfoliated RRAM are performed by analytical formula and multilevel platform GinestraTM, assisted by the temperature-dependent IV of both HRS and LRS for further evidence. The traditional carrier transport mechanism and the morphology of conduction filament are studied and characterized. Eventually, in the fourth part, the treatment of O2 plasma RIE and UV ozone exposure are conducted to form WOX/WS2 heterostructure RRAM. Through the electrical characteristic analysis, the device with UV ozone treatment shows about 2 order decrease in the current while keeping the on/off ratio. In conclusion, this work demonstrates WS2 RRAM on a reasonable scale and tried to characterize it. For the power consumption consideration, the ultrathin film WS2 by CVD transfer is realized and leads to a lower voltage, as well as the UV ozone treatment is applied to suppress the current.

參考文獻


[1] J, Von Neumann. (2013, March 14). First Draft of a Report on EDVAC [Online]. Available: https://web.archive.org/web/20130314123032/http:/qss.stanford.edu/~godfrey/vonNeumann/vnedvac.pdf
[2] F. T. Hady, “The Role of Intel® Optane™ Persistent Memory in High Performance Computing.” 2020 IEEE International Electron Devices Meeting (IEDM)., Short Course. CA., USA.
[3] D. Reinsel, J. Gantz, J. Rydning. (2018 November). The Digitization of the World From Edge to Core. [Online]. Available: https://www.seagate.com/files/www-content/our-story/trends/files/idc-seagate-dataage-whitepaper.pdf
[4] H. Li, “In Alternative Technologies for SRAM.” 2020 IEEE International Electron Devices Meeting (IEDM)., Short Course. CA., USA.
[5] Z. Wang, H. Wu, G. W. Burr, C. S. Hwang, K. L. Wang, Q. Xia, and J. J. Yang, “Resistive switching materials for information processing,” Nature Reviews Materials, vol. 5, no. 3, pp. 173–195, 2020.

延伸閱讀