Full metadata record
DC FieldValueLanguage
dc.contributor.author黃子譯en_US
dc.contributor.authorHuang,Tzu-Ien_US
dc.contributor.author汪大暉en_US
dc.contributor.authorWang,Tahuien_US
dc.date.accessioned2014-12-12T02:33:38Z-
dc.date.available2014-12-12T02:33:38Z-
dc.date.issued2012en_US
dc.identifier.urihttp://140.113.39.130/cdrfb3/record/nctu/#GT070050101en_US
dc.identifier.urihttp://hdl.handle.net/11536/71877-
dc.description.abstract在本篇論文中,我們採用輸入脈衝式電壓的操作方式來控制電阻式記憶體,接著觀測在SET/RESET 過程中出現的隨機電報雜訊,並且針對在不同阻態下的隨機電報雜訊進行量測以及統計分析。實驗結果顯示當電阻式記憶體操作在高阻態時,其隨機電報雜訊的振幅統計分布會有一偏離指數分布的tail存在,並且隨著電阻式記憶體的阻值越大時,其偏離指數分布的tail現象會越明顯,在tail分布上的隨機電報雜訊造成的最大振幅變化量也會跟著越益嚴重。而造成上述現象的物理原因為:當電阻式記憶體儲存的阻態越高時,對應之導電燈絲斷開的區間便會越寬,使得在導電燈絲斷開區間內的單一缺陷捕捉以及釋放電子造成的隨機電報雜訊對電流變動的影響程度會越劇烈。 除此之外,我們比較不同介電層厚度的元件內所量測到的隨機電報雜訊,實驗結果顯示隨著介電層的厚度越厚,其元件操作在高阻態下所量測到的隨機電報雜訊之振幅統計分布出現tail的現象便會越顯著。 最後,我們利用trap-assisted-tunneling model為物理基礎建立一蒙地卡羅模型,此模型能成功的模擬出在電阻式記憶體中隨機電報雜訊之振幅統計分布的tail現象,驗證在導電燈絲斷開區間內的缺陷是造成隨機電報雜訊之振幅統計分布出現tail現象的主要原因。zh_TW
dc.description.abstractIn this thesis, we use pulse operation method for RRAM devices, and then we measure the RTN signal under different SET/RESET cycles. We investigate the RTN amplitude distributions under different resistance states and find that there is a tail which is deviating from the RTN amplitude exponential distribution when the resistance state of RRAM is at high resistance state (HRS). The tail phenomenon in the RTN amplitude distribution under HRS becomes severe when the resistance of RRAM becomes large. And the maximum value of the RTN amplitude in the distribution tail becomes large with the increasing of the resistance state. The reason is that the ruptured region thickness will increase as the resistance of RRAM becomes large, and a single oxide trap induced RTN in the ruptured region would cause a large variation in the read current owing to the trap-assisted-tunneling conduction mechanism. Besides, RRAM devices with different oxide thickness are measured for the RTN signal in order to provide an insight into the electron conduction mechanism in RRAM devices. The experimental result shows that the tail phenomenon in the RTN amplitude distribution under HRS becomes severe when the thickness of the dielectric layer increases. In order to verify the physical model we discussed above, a quasi-2D Monte Carlo simulation model including trap-assisted-tunneling mechanism is developed. This primary model can successfully simulate the tail phenomenon of the RTN amplitude distribution in RRAM devices under HRS which confirms that the tail phenomenon in the RTN amplitude distribution under HRS is attributing to the trap located in the ruptured region of the dielectric layer of RRAM.en_US
dc.language.isoen_USen_US
dc.subject電阻式記憶體zh_TW
dc.subject隨機電報雜訊zh_TW
dc.subjectRRAMen_US
dc.subjectRTNen_US
dc.title氧化鉿電阻式記憶體之隨機電報雜訊分析與蒙地卡羅模擬zh_TW
dc.titleRandom Telegraph Noise (RTN) Analysis and Monte Carlo Simulation in HfOx-based Resistive Random Access Memoryen_US
dc.typeThesisen_US
dc.contributor.department電子工程學系 電子研究所zh_TW
Appears in Collections:Thesis