Full metadata record
DC FieldValueLanguage
dc.contributor.author吳炤華en_US
dc.contributor.authorWu Chao-Hwaen_US
dc.contributor.author唐麗英en_US
dc.contributor.authorProf. Lee-Ing Tongen_US
dc.date.accessioned2014-12-12T02:22:14Z-
dc.date.available2014-12-12T02:22:14Z-
dc.date.issued1999en_US
dc.identifier.urihttp://140.113.39.130/cdrfb3/record/nctu/#NT880031008en_US
dc.identifier.urihttp://hdl.handle.net/11536/65166-
dc.description.abstract由於全球資訊業蓬勃的發展,積體電路(Integrated Circuit, IC)製造業的國際競爭力成為眾所矚目的焦點。對一般IC的半導體工業廠商而言,產品的良率(yield)為評估公司生產能力的重要指標之一。而在複雜的晶圓製程中,缺陷(defect)的產生是不可避免的,而有些缺陷會導致晶片(chip)的失效(fault)而降低產品的良率,因此為有效地提升IC產品的良率,在IC生產線上,針對在製品實行統計製程管制(statistical process control, SPC)以監控製程缺陷的分佈及掌握製程狀況是非常重要的。 傳統對晶圓上以缺陷數為品質特性所製作的缺陷數管制圖(c-chart)是根據卜瓦松(Poisson)分配而繪製,但是在超大型積體電路(very large scale integration,VLSI)的製造技術下,隨著晶圓面積的增加,晶圓上缺陷的分佈出現群聚(clustering)的現象,因而違反了卜瓦松分配的假設,導致缺陷數管制圖不再適用,而目前所提出的一些相關之修正缺陷數管制圖尚有不完善之處。 本研究的主要目的是考慮在缺陷出現群聚現象的情況下,提出一套簡便且合理的晶圓缺陷數管制圖繪製方法,以有效的管制晶圓表面的缺陷數。本研究所提出的方法為結合缺陷數及缺陷群聚嚴重程度對良率的影響,使用Hotelling T2多變量管制圖(Multivariate Quality Control Chart)來對缺陷數及缺陷群聚現象同時進行監控。本研究所提之管制圖經新竹科學園區某積體電路公司之實際資料應用後,結果不但與最終估計良率之趨勢吻合,且透過本研究之分解方法可以進一步找出造成異常的原因為何。zh_TW
dc.description.abstractAll industrial manufacturers ultimately strive to obtain maximum profits. For the integrated circuits (IC) manufacturer, the yield on each wafer is an important index to evaluate profit. During the complicated manufacturing process of a wafer, defects on the wafer surface are nearly impossible to avoid. Some defects permanently damage the chips, subsequently leading to yield loss. In addition, the extent of defect-clustering is also another important factor to affect yield loss. To enhance the yield of IC products, most manufacturers use control charts for their statistical process control (SPC) under controlling defects. Generally, the attribute control chart has been utilized to monitor the defect counts on a wafer. However, this control chart for defects may cause erroneous results without the consideration of the clustered defects. In this study, we present a novel means of constructing a Hotelling T2 control chart that can detect defects and defect-clustering simultaneously. A case study is also presented, demonstrating the effectiveness of the proposed control chart. 1.1 研究動機與目的 1 1.2 研究方法 2 1.3 研究範圍與假設 2 1.4 研究架構 3 第二章 文獻探討 4 2.1 缺陷數管制圖之探討 4 2.1.1 傳統積體電路製程缺陷數管制圖的探討 4 2.1.2 修正的缺陷數管制圖 5 2.2 缺陷群聚現象之探討 7 2.2.1 缺陷群聚現象 7 2.2.2 缺陷群聚嚴重程度的指標 8 2.3 HOTELLING T2多變量管制圖 8 2.3.1 Hotelling T2多變量管制圖之回顧 8 2.3.2 Hotelling T2多變量管制圖之探討 13 第三章 結合缺陷數與缺陷群聚的管制圖之構建程序 14 第四章 實例驗證 18 4.1 結合缺陷數與缺陷群聚的管制圖之構建 18 4.2 管制圖之驗證 22 4.3 分析與比較 24 4.3.1 傳統的缺陷數管制圖 24 4.3.2 Albin與Friedman的缺陷數管制圖 25 4.3.3 比較結果 26 第五章 結論 27 參考文獻 28en_US
dc.language.isozh_TWen_US
dc.subject積體電路zh_TW
dc.subject晶圓zh_TW
dc.subject缺陷zh_TW
dc.subject群聚現象zh_TW
dc.subject良率zh_TW
dc.subject統計製程管制zh_TW
dc.subjectIntegrate circuiten_US
dc.subjectHotelling T2 control charten_US
dc.subjectdefecten_US
dc.subjectclusteren_US
dc.subjectyielden_US
dc.subjectstatistical process control (SPC)en_US
dc.title積體電路生產線上結合缺陷數與缺陷群聚之Hotelling T2多變量管制圖zh_TW
dc.titleUsing Hotelling T2 Control Chart for Clustered Defects in IC Fabricationen_US
dc.typeThesisen_US
dc.contributor.department工業工程與管理學系zh_TW
Appears in Collections:Thesis