Automated inspection and classification of flip-chip-contacts using scanning acoustic microscopy

Industrial applications often require failure analysis methods working non-destructively, enabling either a rapid quality control or fault isolation and defect localization prior to a detailed defect investigation requiring target preparation. Scanning acoustic microscopy in the frequency range above 100 MHz provides high axial and lateral resolution, a moderate penetration depth and the required non-destructivity. In this study a method for an automated detection of defects in flip-chip-contacts was developed. Chip samples were manufactured in flip-chip technology containing a 750 mu m thick die with solder balls (80 mu m diameter) and underfill attached to an organic-layer substrate. For acoustic inspection a scanning acoustic microscope in combination with a 175 MHz transducer was used. Recorded echo signals were analyzed off-line applying custom-made MATLAB software. For differentiation between the flip-chip-contacts and the underfill, the recorded echo signals were pre-analyzed. Signals obtained from the contacts were then inspected by wavelet-, pulse separation- and backscatter amplitude integral analysis. Complementary X-ray- and SEM-inspection was performed for defect verification. The separation of pulses obtained from the interfaces of the contacts, the absolute values and the distribution of wavelet coefficients corresponded to the interconnecting condition. The success rate of detecting voids was 96.8% as verified by SEM-imaging, while manual X-ray inspection showed success only in 64% of the analysed cases.