Search

FCCSP (Flip Chip Chip Scale Package) is widely used in current electronic products, such as: mobile devices. Although it's a mature technology package type so far, it still need to overcome strip warpage issue after die bonded and molding procedures. The most challenge is strip warpage behavior always performed convex on short edge after die bonded process. This behavior is odd with normal situation since CTE mismatch might induce deformation on long edge. It induced multi-issues and need more man power instead of automation on next procedure. Hence, this study focused on investigating constraint condition during flip chip bond process to affect strip deformation. Moreover, we adopted aMA (Advanced Metrology Analyzer) to figure out the strip deformation behavior and EMC (Epoxy Molding Compound) shrinkage volume and then realized in mechanical simulation model. Consequently, this study would suggest the optimal design to improve the overall strip warpage and advise the appropriate EMC properties to pass manufacture limitation.

Carrier is considered a good tool for process warpage control in wafer/panel level packaging technology. Glass is the most common carrier material. Its rigid, strong property makes it have the high potential for recycle in assembly. And its adjustable CTE (coefficient of thermal expansion) also have the advantage for warpage control. Thus glass strength and warpage must be investigated for the developing advanced technology, such as panel level fan-out package. Glass is tested by 4PB (4-point bending) which follows ASTM C158-02 to come out strength. Two kinds of standard glass sample (around 50×50 mm2) are chosen to compare with glass supplier’s datasheet for benchmark. The standard sample has the same material, dimension and test condition with supplier. 3 types of standard glass sample with different CTE are tested for strength. Those samples have similar average strength and strength range. Two kinds of panel glass (300×300 mm2) with different CTE are tested by 4PB to benchmark with standard sample using the same material. The panel glass strength is within the database of standard sample. For 12″ wafer glass, there is no equation to transfer breaking load into glass strength. A standard glass is chosen as the base to build modeling, so that the wafer load can be converted into strength by simulation. And the simulation by finite element method result shows a liner relationship between load and strength which is the same with strength equation from ASTM. Three dimensional digital image correlation (3D-DIC) theory developed equipment named aMA (advanced metrology analyzer) is used for warpage measurement. For standard sample, its warpage is very small, approximately 6 um. But warpage of wafer, panel glass are much bigger than standard sample due to those large sizes. 12″ wafer glass samples, CTE ranges from 3.0 to 8.7, have similar warpage behavior, and the CTE can also be calculated by aMA. The CTE from aMA is similar with that from datasheet. For A-supplier 12″ wafer glass, it shows a saddle warpage shape if the glass is attached a PI film. For B-supplier 12″ wafer glass, the warpage becomes much large if the glass CTE is large than 9.6. And the warpage shape behaves a concentric circle. To control process warpage, glass is added some material such as PI film and jig. The glass with PI film or jig also shows a concentric circle warpage shape at 30°C.