近年來電子產品的日新月異以及網際網路的盛行使得半導體IC產業的發展極為迅速,而技術與品質是我國電子產業得以在國際市場上生存的利器。國內有強實的研究發展體系支持且高科技廠商競相投入大量資金與第一流人才來改善IC製程的良率(Yield)與技術的革新,大幅提高整體電子工業的生產力與國際競爭力。 所謂的半導體產業,泛指所有生產製造和IC相關產品的產業。IC則是積體電路(Integrated Circuit )的簡稱,積體電路顧名思義就是將電晶體、電阻器、電容器、二極體等相關電路元件,整合聚集在一片矽晶片上,形成一個完整的邏輯電路,邏輯電路的功能可以用來控制電腦運作、電腦運算與記憶,這片矽晶片稱為IC可以說是最基本的電子元件;因此在半導體產業中每一項產品的關鍵製程之良率非常重要,只要關鍵製程良率提高,進而就能改善產品品質,已獲取競爭之利基。 本研究主要目的在於改善目前晶圓廠內蝕刻製程晶圓接觸窗阻值電性參數之研究,應用實驗設計2k因子設計法作為本研究之實驗方法,當有很多因子要研究時此設計方法特別有效,它提供了在一次完整因子設計裡可以研究k個因子的最小實驗次數;應用變異數分析( ANOVA )與迴歸分析之手法分析實驗結果,並透過擬定之實驗步驟規劃流程逐步完成資料驗證,求出本研究之最佳參數組合。
In the past few years, the glorious developing of semi-conductor IC industries were leading into a tremendous step, and this was result from the fast glowing of electronic products and the world-wide internet. What Taiwan can stand out in the world of electronic industries is the good product quality and great technology. This is because Taiwan has strong technology research support from government and related semi-conductor companies race to invest the techniques and manpower, which improve the IC production yield and techniques. And this promotes extensively the production ability of the whole electronic industries and the competition to the worldwide. What semi-conductor industry is the industry of producing IC related products. IC is short for “Integrated Circuit.” Namely, the “Integrated Circuit” is a complete logical circuit in a silicon wafer, which combines the transistors elements, resistances elements, capacitors elements and bi-polar elements, etc. The function of logical circuit in the wafer, IC, can control the basic operation and memory in the computer, which is the most basic electronic element. That’s why the key process yield is so important to every electronic product in semi-conductor industries that if the key process yields could be improved, the products quality could be increased. Naturally, the profit and the competition will come. My paper is to apply the 2k factorial experiment designs to improve the etching process of RC value of wafer contact window in the current wafer factory. This method is very suitable if there are too many factors in the experiment. Because this method provides the least experimental times of K factors in a complete factorial design. Not only I apply the ANOVA table and regression analysis to analyze my experimental data but I also verify the data completely by experiment step design before hand to study the important process factors in my research.