Session: 10-01 Interactive Presentations

Paper Number: 100316

100316 - A Novel Approach to Reliability Assessment of 2nd Level Underfill in Bga Packages 

 Underfill materials are firstly used to improve the thermomechanical reliability of flip chips due to the CTE mismatch between the die and substrate. Nowadays it has also been widely used to improve the 2nd level interconnection of BGA packages in harsh environments. The general notion is the thermomechanical reliability of interconnections can always benefit from underfilling.   However, improper underfill can also reduce the reliability during thermal cycling environments. Due to the variation of the stiffness rigidity, efficient CTE, and physical structure of the package itself, the application of 2nd underfilling should be studied specifically.  There is no universal underfilling solution for each package.

Instead of the costly thermal cycling test, this study proposed a novel approach to thermomechanical reliability assessment of 2nd level underfill in BGA packages. The BGA packages with two major underfill reinforcement methods (corner bonding and full bottom surface bonding) and no-underfill were studied. The material properties of two underfill materials are different. To quantitatively measure the deformation of solder balls, the BGA packages were cross-sectioned before thermal cycling. The two-dimensional digital image correlation (DIC) technique was used to capture the in-plane deformation of the critical solder ball in thermal cycling intervals. The accumulated plastic strain of the BGA solder was collected every 10 thermal cycles. After 50 thermal cycles, the accumulated plastic strain per cycle of each package can be obtained and this value denotes the fatigue life of solder balls in each package. The FEA model was used to correlate and verify the performance of each package during thermal cycling. 1/8 symmetric model was used to relieve the computation cost. Based on the experiment and simulation results, we observed that in comparison with the no-underfilled package,  the package with fully underfilled alleviated the averaged plastic strain on the critical solder ball, meanwhile, the package with corners underfilled had a larger plastic strain. This observation also proves that there is no golden rule for underfill applications. The performance of underfill is associated with its material properties, bonding methods, and also packages’ material properties and geometry.

Instead of the conventional thermal cycling testing, which is costly in both time and energy, by using this approach, we can assess the thermomechanical reliability of 2nd level underfill quickly and smartly.

Presenting Author: Yangyang Lai Binghamton University