Possibilities and limitations of no-flow fluxing underfill

The performance of three developmental no-flow fluxing underfills was examined in a variety of reflow profiles. The underfills differed in regards to their resin system, fluxing agents and cure kinetics. The reflow profiles differed in regards to the ramp rate, soak time, soak temperature, time above liquidous and peak temperature. For each underfill and reflow profile combination, the performance attributes of solder wetting, underfill voiding and completeness of underfill cure were determined. Experimental results for the nine cases were then compared to calculated underfill cure and volatilization data, which were obtained using a kinetic model. A comparison of experimental and theoretical results showed that solder wetting correlates well with the calculated process-window (i.e. the time between when the solder melts and when the underfill gels or the solder resolidifies). Similarly, the post-reflow underfill hardness correlates with the calculated % resin cure. For a given underfill, the data also suggests a direct relationship between voiding and volatile outgassing. However, this simple relationship fails when comparing materials of diverse compositions. In general, voiding appears to result from a complex interaction of underfill cure kinetics, volatile outgassing, fluxing chemistry, and the reflow profile employed.