| 摘要: | 微影製程是半導體製造技術最關鍵技術之一，微影製程主要含塗佈光阻、曝光、顯影三步驟。近年來因環保減碳意識抬頭及企業的成本管控等考量，減少光阻使用量正成為各家半導體公司及相關科技行業著墨的重點之一，本研究主要就是針對微影製程之旋轉塗佈部分的正型光阻用量，進行減量實驗，藉由改善旋轉塗佈製程實驗參數的調整，尋求較佳的製程參數組合，並驗證所得之優化參數結果是否符合現行規格與量產條件。
本實驗選用12吋雙面拋光矽晶圓，以正型光阻減量優化為目標導向，探討旋轉塗佈製程參數於正型光阻減量的過程中，對光阻薄膜厚度平均值、厚度間距差值與晶圓表面光阻覆蓋品質的影響，其中包含旋轉塗佈模組變異、光阻消耗減少溶劑預塗轉速、初轉速、主轉速，並研究與分析量產的可行性。由實驗結果得知，在正型光阻減量並改變增加初轉速度時，其光阻薄膜厚度平均值結果不變，但厚度間距差值結果逐漸變差。後續藉由光阻消耗減少溶劑預塗轉速與初轉速的配置變化，對於正型光阻減量的過程中，所造成厚度間距差值變差的結果而有所改善。本文最後正型光阻減量優化結果，藉由旋轉塗佈製程的優化可將原正型光阻使用量由0.9g減量至0.6g，再經由正型光阻減量旋轉塗佈穩定性與產品驗證相關項目結果分析，均與正型光阻減量前無所差異，預計可減少約33.3%的光阻使用成本。
由本實驗研究最後所獲得之成果與結論，用意在於建立一套快速有效的模式與方法，爾後在評估不同種類、不同黏滯係數的光阻於微影製程應用中，有利於縮短光阻塗佈製程條件之調整時程，能較快找出適合量產的製程參數及優化整體生產效益，同時期許能貢獻提供給半導體及其相關科技業界，在光阻塗佈製程領域之參考。;The photolithography process is one of the most critical technologies in semiconductor manufacturing technology, the photolithography process mainly includes three steps: coating photoresist, exposure, and development. In recent years, due to the rising awareness of environmental protection and carbon reduction and the cost control of enterprises, reducing the amount of photoresist is becoming one of the emphases of various semiconductor companies and related technology industries. This research is mainly aimed at the spin coating part of the photolithography process. The amount of positive photoresist used in the experiment was reduced, and by improving the adjustment of the experimental parameters of the spin coating process, a better combination of process parameters was sought, and the results of the optimized parameters are verified to meet the current specifications and mass production conditions.
In this experiment, a 12-inch double polished silicon wafer is selected, and the optimization of positive photoresist reduction was the goal. We study the influence to the difference between film thickness mean value, film thickness range value, and the quality of photoresist coverage on the wafer surface during the process of reducing the positive photoresist. The conducted factors used in the spin coating process includes the variation of spin coating module, RRC (Reduced Resist Consumption) pre-wetting speed, spin-up speed and spin-off speed, we also study and analyze the feasibility of mass production. The experimental results show that when the positive photoresist is reduced and the spin-up speed is changed, the film thickness mean value remains unchanged, but the film thickness range value gradually become worse. Furthermore, by conducting the configuration changes of the RRC pre-wetting speed and spin-up speed, the worsening phenomenon of the film thickness range value during the process of positive photoresist reduction has been improved. The study demonstrates the optimization result of the positive photoresist weight reduction, through the optimization of the spin coating process, the original positive photoresist usage can be reduced from 0.9g to 0.6g. Moreover, through the analysis result of the positive photoresist weight reduction spin coating stability and product verification test items, it shows no difference of the stability and test results before and after the positive photoresist reduction. It is expected to reduce the cost of photoresist by about 33.3%.
The results and conclusions obtain in this experimental study are intended to establish a set of fast and effective models and methods to further shorten the time to adjust photoresist coating manufacturing parameter while evaluating different types of photoresist with different viscosity coefficients in the application of the photolithography process. This method can help to target on the manufacturing parameter good for mass production and optimize the overall production efficiency. Moreover, it is expected that this method can be contributed to the semiconductor and related technology industries as a reference in the field of photoresist coating process. |
