目前砷化鎵(GaAs Wafer)晶圓之切割、鍵合、覆晶黏晶與檢驗等製程設備之定位,由於無法直接穿透晶圓表面而辨別內部線路位置,故鍵合精度較差。所以,本研究將導入近紅外線定位法(NIR Alignment)之技術,可用於改善現行非貫穿孔以及非透明材質之視覺定位上的缺失,利用此一原理可執行砷化鎵晶片定位動作,達到超高之鍵合精度,及縮短定位時間。 實驗使用波長大於886nm之視覺設備,包含一般CCD具紅外線功能(波長介於400~1000nm)之鏡組,搭配NIR-Lens取像(波長介於400~1200nm),照射近紅外線光源(波長介於800~1300nm)的波長區域於上層砷化鎵晶片,透過近紅外線鏡組,穿透晶圓的光滑砷化鎵晶片表面,辨識到晶圓內層線路,將辨識的結果,搭配自行撰寫的軟體,提供使用者人性化介面的操作,達到定位的動作。 研究在軟體方面,利用Borland C++ Builder 6軟體及搭配Matrox Imaging Library 7.0 函式庫輔助,自行撰寫程式,使程式具有一般套裝軟體的水準,達到近紅外線視覺系統定位之需求。研究在硬體方面,使用現有視覺設備取像,提供影像供撰寫之軟體應用,並於程式中模擬在定位時之動作;近紅外線視覺設備方面,首先整理具有近紅外線波長之CCD與LEN資料與砷化鎵晶圓材料,運用近紅外線視覺設備做測試,得到最佳之近紅外線穿透設備所需之規格。最後,運用所得之最佳系統分別作定位驗證。其結果證實利用影像處理直接擷取晶圓內部電路影像於砷化鎵晶圓定位設備是可行的且絕對更快更準確。特別是於需高精度複雜IC線路之定位應用,更是最佳選擇。
For the current GaAs wafer processing such as dicing, die bonding, Filp-Chip bonding and inspection after assembly, all need the precision positioning of the wafer. Because it is unable to see through the wafer surface to detect the inside circuit for the alignment directly, it is relatively poor and indirect in the precision when using outer alignment marks. The study will apply the near infrared ray (NIR Alignment) alignment technique to improve the current positioning method in those non-through holes and non-transparent materials such as the GaAs wafer. By utilizing this technique can have much better accuracy and shorten the alignment time. For the experiment, it uses a near infrared with the wavelength greater than 886 nm and a general CCD (wavelength is 400~1000nm), together with a NIR-Lens (wavelength is 400~1200nm) and the infrared light source (wavelength is 800~1300nm) to see through the smooth side of GaAs wafer surface. It is able to detect the circuit inside the wafer. The resulting images are then fitted into the self-developed software for the processing. A user -friendly interface is designed in the software so that the system is easy to operate in the wafer position alignment. The software is developed with the Borland C ++ 6 tool and combined with the application of Matrox Imaging Library 7.0. The program is designed to have similar functions and interfaces as what the general commerce software has. For the hardware setups, the study surveys many of the CCD and Lenses capable of handling the infrared. And do trial tests with the GaAs wafer to obtain the best-fit system. It is then finalized to use in the alignment test. The alignment technique described in the study uses the circuit image inside the wafer directly for the positioning, it is considered as a very efficient and better method in the application of GaAs wafer equipments that need precision positioning.