Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Utilizing acoustic emission signals to evaluate BEoL stack damages of two different sample systems caused by Cu-pillar shear-off

: Silomon, Jendrik; Clausner, André; Zschech, Ehrenfried


Institute of Electrical and Electronics Engineers -IEEE-; IEEE Electron Devices Society; IEEE Reliability Society:
IEEE International Integrated Reliability Workshop, IIRW 2020 : Virtual Conference, October 4 - November 1, 2020
Piscataway, NJ: IEEE, 2020
ISBN: 978-1-7281-7059-6
ISBN: 978-1-7281-7058-9
5 S.
International Integrated Reliability Workshop (IIRW) <2020, Online>
Fraunhofer IKTS ()
advanced packaging; acoustic emission; BEol stack stability testing; Cu-Pillar shear-off; chip-package interaction (CPI)

In this work, a novel approach is introduced to identify and categorize damages occurring in the back end of line (BEoL) stack utilizing acoustic emission (AE) measurements. The damages in this experimental approach are induced by a tribo-indenter system shearing off copper (Cu-) pillars with a customized tip. The research focuses on the comparison of the acoustic signals of two sample systems manufactured by GLOBALFOUNDRIES. One test vehicle (TV) is a 28 nm technology chip bumped with round Cu-pillars and the other one a 22 nm technology chip bumped with oblong pillars. Several Cu-pillar shear-off experiments with different parameters were performed on single Cu-pillars of both sample systems. To trigger different damage modes, the shear height on the round pillars was varied as well as the shear angle on the oblong pillars. The force-progression over time and the AE signal during the shear process were recorded and aligned with the resulting damage. The acoustic emission data was analyzed further by comparing the shape, the amplitude, and the occurring frequencies of the respective signals. For the latter, a Fast Fourier Transformation (FFT) was applied. The different AE signals showed specific characteristics depending on the damage mode and the sample system used. These results suggest that the AE technique is a suitable tool to provide valuable additional information regarding the BEoL stack stability evaluation.