楊宏智Young, Hong-Tsu臺灣大學:機械工程學研究所陳鯤仁Chen, Kuen-RenKuen-RenChen2010-06-302018-06-282010-06-302018-06-282008U0001-2307200814123300http://ntur.lib.ntu.edu.tw//handle/246246/187367化學機械研磨為半導體製程當中最常被採用的平坦化技術，研磨墊表面需要適度的以鑽石修整器修整，以維持化學機械研磨製程之穩定性及產能。本研究計算鑽石修整器上所有鑽石刮劃研磨墊之次數分佈，以模擬預測修整過後研磨墊表面之外型。鑽石與研磨墊之相對速度經實驗證實對修整率並無影響，因此統計刮劃次數時並不需要考慮速度的影響。本研究首先模擬不同鑽石分布方式的修整器之修整結果，如環形與全面分布、不同鑽石間距、整齊與不規則排列等。接著探討各種設備上對應之修整參數，如轉速、修整器掃掠之幅度及頻率等對研磨墊外形之影響。另外，本研究定義出一個修整過程的均勻度指標，可用來最佳化掃掠頻率以確保研磨墊各處隨時都被均勻地修整。模擬預測之研磨墊外型亦與實際使用過的研磨墊做比對，兩者之趨勢一致，證實刮劃次數可有效的預測研磨墊外型。根據本研究所建立的模擬程式，使用者可以直接針對設備上可調的參數進行模擬，進而找出最佳的修整參數。Chemical mechanical polishing (CMP) is the planarization technology most often used for semiconductor processes. The polishing pad needs to be dressed by a dresser to maintain the stability and the throughput of the planarization process. This study simulated and predicted the pad profile by calculating the distribution of scratch numbers of diamonds against the pad. The effect of the relative velocity between a dresser and a pad has been experimentally verified to be insignificant on dressing rate, so the simulation did not take the effect of relative velocity into account. First, different types of dressers, including ring-types and full-type, different pitches and arrangements, were simulated. Then, the effects of different dressing conditions, rotational speeds, sweep amplitudes and frequencies were examined. A uniformity index was defined to optimize the uniformity of dressing process. The simulated pad profile was then compared to the actual pad profiles and results were very similar. According to the simulation model, users can directly modify the parameters that are adjustable in the equipment and find the optimum dressing condition.口試委員會審定書 I謝 III要 IVbstract Vontents VIist of Figures IXist of Tables XIhapter 1. Introduction 1.1 Background 1.1.1 Introduction to Semiconductor Process 1.1.2 Applications of CMP to Semiconductor Process 3.1.3 Why Dressing? 6.1.4 Challenges in CMP 7.1.5 Modeling of Dressing 7.2 Motivation 10.3 Objectives 11.4 Organization 12hapter 2. Chemical Mechanical Polishing 13.1 Fundamental of Planarization Technology 13.2 Process of Chemical Mechanical Polishing 16.3 Equipment in CMP 19.3.1 Overview of CMP Equipment 19.3.2 Kinematics Systems 22.3.3 Carrier of Wafer 25.3.4 Endpoint Detection 27.4 Consumables in CMP 28.4.1 Slurry 29.4.2 Pad 31.4.3 Dresser 34hapter 3. Methods 37.1 Overview of the Model 37.2 Diamond Pattern 39.2.1 Diamond Pitch 39.2.2 Arrangement of Diamond 40.2.3 Distribution Area of Diamond 41.3 Trajectory Equation 43.4 Solving the Trajectory Equation 47.4.1 Bisection method 48.4.2 Fixed-point method 48.4.3 Newton-Raphson method 49.4.4 Secant Method 50.4.5 Selection of Numerical Method 50.4.6 Key Points of Solving Equations 51.5 Effect of Relative Velocity 56.5.1 Experimental Equipments 56.5.2 Experimental Design 60.5.3 Measurement 61.5.4 Results 63hapter 4. Simulation Results and Discussion 64.1 Simulation Conditions 64.2 Diamond Distribution of Dresser 65.2.1 Diamond Pitch 65.2.2 Grid and Random Arrangement 66.2.3 Distribution Area of Diamonds 67.2.4 Summary of Diamond Distribution 68.3 Dressing Conditions 69.3.1 Start Position and End Position 69.3.2 Ratio of Rotational Speed 72.3.3 Sweep Frequency 74.4 Uniformity of Dressing Process 76.5 Examination of Pad Profile 82hapter 5. Conclusions and Future Works 84.1 Conclusions 84.2 Future Works 86eferences 873399671 bytesapplication/pdfen-US化學機械研磨鑽石修整器研磨墊軌跡分析Chemical Mechanical PolishingDiamond DresserPadTrajectory Analysisthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/187367/1/ntu-97-R95522729-1.pdf