The selection of die backside metallization for solder
In support of the need to achieve processor speeds above 2.5 GHz, the feasibility of metallurgical heat transfer between chip and heat sink is under study. The selection of a suitable metallization on the chip backside was undertaken. Candidate stackss were evaluated based on reliability and thermal...
محفوظ في:
| المؤلفون الرئيسيون: | , |
|---|---|
| التنسيق: | text |
| منشور في: |
Animo Repository
2001
|
| الموضوعات: | |
| الوصول للمادة أونلاين: | https://animorepository.dlsu.edu.ph/faculty_research/13346 |
| الوسوم: |
إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
|
| المؤسسة: | De La Salle University |
| الملخص: | In support of the need to achieve processor speeds above 2.5 GHz, the feasibility of metallurgical heat transfer between chip and heat sink is under study. The selection of a suitable metallization on the chip backside was undertaken. Candidate stackss were evaluated based on reliability and thermal resistance measurements. Temperature cycle (250x, -55 to 125°C), biased HAST (100h,3.3V, 85%RH, 135°C)and extended bake tests (200h, 200°C) were done. These were to assess fatigue, electromigration and arrhenius diffusion failure mechanisms. After stress units were subjected to thermal resistance testing, x-ray and accoustic microscopy (transmission and C-scan). Four stacks were short listed based on literature and prior tests: Ti/NiV/Au, Ti/NiV/Ag, Ti/Au, Ti/Ag. Ag stacks had poor thermal performance because of Ag leaching by the In solder matrix. Ti/Au had poor fatigue performance as detected by transmission accoustic microscopy. X-sections reveal failures due to the absence of Au-Ni intermetallic compounds. Ti/NiV/Au was chosen despite a significant impact in manufacturing time and capacity. |
|---|