A publication of the National Electronics Manufacturing Center of Excellence September 2005

EMPF Director

Michael D. Frederickson
mfrederickson@aciusa.org


Sign up to receive email notifications of the newests issues of the EMPFasis!

A customer contacted the EMPF Helpline after observing adhesion failure of chip components that were bonded with conductive epoxy.

EMPF staff evaluated the adhesion failures. The components were bonded to a gold-plated ceramic substrate. Adhesion failure was observed after the device was hermetically sealed and operating in the field. Failures were also observed on devices that were not in service. Two assemblies were provided for analysis. Both experienced adhesion failure where the chip component separated from the board, as seen in Figures 3-1 and 3-2.

Top-down scanning electron microscopy (SEM) with energy dispersive spectroscopy (EDS) was performed on the pad surface and the conductive epoxy surface. Scanning Auger analysis was performed on two areas of the failed gold surface. Scans were performed in the as-received condition as well as after a 30-second ion etch.

Top-down SEM analysis of the pad surface in the areas of adhesion failure did not show evidence of contamination. There were small areas of conductive epoxy present and a number of pin holes in the gold surface. Elemental analysis of the pad surface showed the presence of gold and trace amounts of carbon on the surface. The carbon levels were low (less than 10 wt. %) and were not considered to be a source of contamination. There was no contamination found in the pin holes. The conductive epoxy that remained on the pad after separation was composed mainly of carbon and silver. Figures 3-3 and 3-4 show high magnification images of the gold surface in the area where the epoxy failed.

SEM surface examination of other gold-plated areas on the substrate revealed pin holes similar to what was observed in the pad areas. No evidence of surface contamination was present. Like the pad areas, small amounts of carbon were also detected. An SEM image of the gold surface away from the pad is shown in Figure 3-5 below.

No contamination was detected on the surface of the conductive epoxy from the separated chip component. Elemental analysis detected only carbon and silver.

Auger analysis revealed the presence of carbon, gold, oxygen, silver, nitrogen, and chlorine on the gold pad surface before ion etching. Gold, oxygen, and trace amounts of silver were detected after etching. All the elements detected are common components of conductive epoxy. No abnormal elements were found.

Conclusions
The epoxy failure was likely caused by improper epoxy application or poor epoxy performance. No evidence of surface contamination was detected on the failed gold pad surface or in other areas of the gold plating.

Surface analysis of pads also showed traces of the conductive epoxy materials remaining on the pad after failure. This may indicate epoxy that was not fully cured.

Many of the other components on the board, including wire bonded components, did not display adhesion failures. The other components were smaller and possessed a lower thermal mass. The higher thermal mass of the chip components may contribute to an insufficient epoxy cure. In addition, if contamination was present on the board, wire bond adhesion failure would also be likely.


[site map]