標題: PECVD護層薄膜應力/形變對MOS起始電壓偏移之影響
Effect of PECVD Passivation Layer Stress/Deformation on The Threshold Voltage Shift of MOS
作者: 鄭惇元
Cheng, Tun-Yuan
吳耀銓
Wu, YewChung Sermon
工學院半導體材料與製程設備學程
關鍵字: 護層;磷矽玻璃;形變;應力;起始電壓;金氧半;Passivation;PSG;Deformation;Stress;Threshold Voltage;MOS
公開日期: 2008
摘要: 對於0.5μm線寬的半導體製程而言,一般使用電漿增強化學氣相沈積 (PECVD) 技術來沈積元件線路保護層 (Passivation Layer) 的薄膜。在護層的薄膜中,經常使用雙層薄膜 (Multi-Layer) 的結構,第一層薄膜沈積磷矽玻璃 (PSG),第二層薄膜則沈積氮化矽 (SiN),即可將元件線路結構與外界污染物隔離。 許多的文獻顯示,磷矽玻璃在大氣下是具有吸水特性的薄膜材料,且磷矽玻璃薄膜吸水後,會導致薄膜應力 (Stress) 由伸張應力 (Tensile) 轉變為壓縮應力 (Compressive) 的形變 (Deformation)。此現象可能造成元件結構的缺陷 (Defect),進而嚴重降低元件的效能,這是半導體工廠不願意發生的情況。 在本文中,筆者為解決半導體工廠中時常發生的產品低良率 (Low Yield) 問題,實驗將使用電漿增強化學氣相沈積器沈積磷矽玻璃的薄膜材料。起初在大氣下利用監測晶圓 (Bare Wafer) 量測磷矽玻璃薄膜吸水後應力隨大氣中時間 (Q-Time) 的改變量。結果顯示,磷矽玻璃會隨著停留在Q-Time越長吸水越多而產生形變。若使用真空腔 (Vacuum Chamber) 將水氣隔絕,則磷矽玻璃的薄膜不會發生形變。更進一步,藉由修改沈積程式的配方 (Recipe) 來嘗試控制磷矽玻璃的應力避免形變,筆者稱之為黃金配方 (Golden Recipe)。 最後,使用量產產品 (Device Wafer) 成功驗證黃金配方,並針對異常發生的主因作一解釋。
For semiconductor fabrication process of 0.5μm critical dimension, in general, for protecting the device circuits issue, the foundry uses the Plasma Enhanced Chemical Vapor Deposition (PECVD) to deposit the passivation layer with multi structure. The first layer is phosphosilicate glass (PSG) and the second layer is silicon nitride (SiN). Now the passivation layer can isolate the device circuits from external contamination damage. Most of technical literatures show that the PSG have the characteristic of water absorption in air. After the PSG film absorbs water, it causes film stress deformation from tensile stress to compressive stress. This change can not only induce the defects which are in the device circuit structure, but also can reduce the device performance. Foundry dose not hope to see this abnormal issue occur. In order to solve this low yield problem, author uses the Plasma Enhancement Chemical Vapor Deposition (PECVD) to deposit PSG film layer for experiment. At first, depositing the PSG film on bare wafer, author measures the stress correlate with Q-time. The result shows that stress deformation is positive with the Q-time. If the water absorption can be isolated by vacuum chamber, PSG film will not be deformed. In Further step, author fines tune the deposition recipe and tries to control the PSG stress, it will avoid the occurrence of deformation. Author defines the non-deformation recipe as golden recipe. Eventually, the golden recipe which deposits PSG with non-deformation has been successfully verified by device wafer and author will explain the root cause of the abnormal issue.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT079575516
http://hdl.handle.net/11536/41617
顯示於類別:畢業論文