CHARACTERIZATION OF FLIP CHIP BUMP FAILURE MODES USING HIGH FREQUENCY ACOUSTIC MICRO IMAGING

Flip chip technology is being used increasingly as a method of interconnecting a die to a substrate. There are reliability concerns associated with the technique such as pre-existing voids in the solder joints or non-wet of the solder to the bond pads. Defects in the interconnections can also occur during use of the device or during life simulation tests such as thermal cycling. For reliability determination a nondestructive method is needed to examine the solder joints in the “as assembled ” units as well as in the “post stress ” condition.Acoustic micro imaging has been used as a nondestructive method to evaluate flip chip attach with promising results. Acoustic microscopes utilize high frequency ultrasound to examine the internal features in materials and components. The instrument used in this study was a C-SAM (C-Mode Scanning Acoustic Microscope) operating in the reflection mode at frequencies of 100 to 180 MHz. A previous paper presented analyses done on flip chip test samples These samples contained known induced defects at the chip/bump and bump/substrate interfaces. Using these samples the optimum techniques for inspection of the flip chip attach were developed. This information was used as baseline for the work presented in this paper. This presentation describes the analyses on a set of samples which showed failures after thermal cycling. The failures according to destructive physical analysis were thought to be cracks in the solder joints. However, a one to one correlation could not be seen between acoustic data and initial destructive physical analysis. With further investigation unexpected laminar cracks were discovered under the bond pad and in the surrounding glassivation layers on the silicon die. These were difficult to see in the cross sections unless specifically looked for by referring to the acoustic analysis. The data presented in this paper shows the acoustic images of flip chip samples before and after thermal cycling and describe the acoustic signatures of the defect types along with the supporting correlative analysis