以實驗設計法探討高透射率減光型相移圖罩最適化

Abstract

在不更動曝光波長的情況下，使用解像度增進技術，亦可縮小元件尺寸。應用實驗設計法可在不增加額外製程成本下，快速尋找最適化參數與製程設定。使用光學鄰近效應修正則可有效改善阻劑圖案失真。 本論文可分為三個部分。第一部份為改良透射率控制圖罩。隨著線寬愈來愈細，要在圖罩接觸孔內製作減光區，以避免兩孔間阻劑架橋，製作難度甚高，如將整個接觸孔改用高透射率材料製作，可降低製作困難度，模擬證明可獲得良好改善效果。第二部分以田口玄一實驗設計法，針對0.07微米線幅孤立線微影製程，利用L9直交表研究相關阻劑參數與光學參數對微影製程之影響。模擬結果發現，對0.07微米孤立線，使用UV-6阻劑，阻劑參數最適化設定為預烤溫度120℃、預烤時間60秒、照後烤溫度130℃、照後烤時間55秒、顯影時間50秒；光學參數最適化設定為四孔偏軸發光孔心相擾度當量值0.6、孔徑相擾度當量值0.22、數值孔徑0.7、減光型相移圖罩透射率29﹪。最適化結果使焦深自0.31微米增為0.46微米；照射寬容度自3.81﹪增為7.62﹪。在光學鄰近效應修正方面，使用散條法可有效改善圖案失真。第三部份亦使用田口玄一實驗設計法，針對0.13微米線幅密集線隙微影製程，利用L9直交表研究相關阻劑參數與光學參數對微影製程之影響。模擬結果發現，對0.13微米密集線（線/隙=1），使用UV-6阻劑，阻劑參數最適化設定為預烤溫度125℃、預烤時間90秒、照後烤溫度130℃、照後烤時間35秒、顯影時間45秒；光學參數最適化設定為環形偏軸發光外環相擾度當量值0.80、內外環相擾度當量值比2/3、數值孔徑0.7、減光型相移圖罩透射率25﹪。最適化結果使焦深自0.43微米增為0.77微米；照射寬容度自4.59% 增為10.59%。在光學鄰近效應修正方面，使用圖案偏差法可有效改善。 未來12吋晶圓搭配90奈米線幅，細線化製程之難度甚高，使用電腦模擬配合田口實驗設計法可快速有效且可系統性的分析與研究眾多參數。

Without changing exposure source, the size of device can be reduced by resolution enhancement techniques (RETs). The application of design of experiment (DOE) can increase the process latitude by optimizing parameters and process settings with no extra cost. The using of optical proximity correction (OPC) can correct the pattern distortion of resist effectively. This thesis can be divided into three parts. The first part is modification of transmittance control mask (TCM). As the line widths get narrower, in order to avoid the photoresist bridging between two contact holes by fabricating an attenuated area in contact hole on mask, the fabrication is quite difficult. If high transmittance material were used in the entire contact hole area, the difficulty of fabrication will be reduced. Simulation has proved that this modification has a good improvement. The second part, the Taguchi Genichi DOE has been applied by using L9 orthogonal array to study the effects of resist parameters and optical parameters on the 248 nm DUV lithographic process for the 0.07μm isolated lines. By simulation, the optimized resist parameters were found out to be prebake temperature 120℃, prebake time 60 sec, post exposure bake temperature 130℃, post exposure bake time 55 sec and development time 50 sec. The optimized optical parameters were found out to be Quadrupole Off-Axis Illumination (OAI) center sigma equivalent 0.6, radius sigma equivalent 0.22, numerical aperture 0.7, transmittance of attenuated phase-shifting mask (APSM) 29%. The optimized results indicated that depth of focus increased from 0.31μm to 0.46μm；exposure latitude increased from 3.81% to 7.62%. Regarding OPC, the using of scattering bar can correct pattern distortion effectively. The third part, the Taguchi method has been applied by using L9 orthogonal array to study the effects of resist parameters and optical parameters on the 248 nm DUV lithographic process for the 0.13μm dense lines. The optimized resist parameters were found out to be prebake temperature 125℃, prebake time 90 sec, post exposure bake temperature 130℃, post exposure bake time 35 sec and development time 45 sec. The optimized optical parameters were found out to be Annular OAI outer ring sigma equivalent 0.8, ratio of inner ring over outer ring sigma equivalent 2/3, numerical aperture 0.7, transmittance of APSM 25%. The optimized results indicated that depth of focus increased from 0.43μm to 0.77μm；exposure latitude increased from 4.59% to 10.59%. Regarding OPC, the using of feature bias can correct pattern distortion effectively. 12 inches wafer will be coupled with 90 nm linewidth in future, the difficulty of fabrication of narrow linewidth is high, using computer simulation combined with Taguchi DOE could analyze and study various process parameters efficiently and systematically.