Lead-Free Solder Material and Chip Thickness Impact on Board-Level Reliability for Low-K WLCSP

Abstract

This paper evaluates the effect of the solder ball material and chip thickness on board-level reliability using low-K wafer-level chip-scale packaging (WLCSP). The composition of the three evaluated solder ball materials includes SnAg1.0Cu0.5, SnAg2.0Cu0.5, and SnAg2.6Cu0.6. The chip thickness is ranging from 200, 300, and 775 ¿m. Initially, the high temperature storage and ten-time multiple reflow tests were used as the wafer-level reliability test items. The work uses both the low speed (500 ¿m/s) and the high speed (200 mm/s and 1 m/s) solder ball shear test to compare the shear force, displacement, and fracture energy of the three solder compositions. The high speed shear test result shows the high silver content correlates to small displacement and the low fracture energy. Next, the finite element model was employed to compare the board level chip and solder ball stress for these three different chip thicknesses. The finite element method simulation results indicate the thinner chip implies the preferred thermal fatigue cycles. In addition, the drop test results show the chip thickness of 200 and 300 ¿m improve the drop test performance (81 drops and 88 drops) by 20.9%-31.34% than that of chip thickness of 775 ¿m. Moreover, the chip thickness of 200 and 300 ¿m substantially enhance the thermal fatigue life (665 cycles and 655 cycles) by 81.44%-84.21% than that of chip thickness of 775 ¿m using the SnAg1.0 Cu0.5. Notably, the experimental results correlate well with the simulation trend. This work provides a design guideline for selecting the favorable solder materials and the chip thickness to obtain the satisfactory board-level reliability of the low-K WLCSP packaging.