Here is the abstract you requested from the IMAPS_2013 technical program page. This is the original abstract submitted by the author. Any changes to the technical content of the final manuscript published by IMAPS or the presentation that is given during the event is done by the author, not IMAPS.
|Stitch Bond Process of Pd-Coated Cu Wire: Experimental and Numerical Studies of Process Parameters and Materials|
|Keywords: stitch bond process, capillary surface finish, finite element analysis|
|Cu wire has been replacing conventional Au wire due to its lower price and better mechanical and electrical properties. The main disadvantage of Cu wire is its tendency to oxidize, which decreases its shelf life and reduces the stitch bondability. Pd-coated Cu wire was introduced recently to overcome the oxidation issue of Cu wire. However, the process window is still narrow compared to that of bonding processes with Au wire. More fundamental studies are required to understand the influences of bonding parameters and bonding tools to improve stitch bondability. The stitch bond process of 0.7 mil diameter PCC wire on a Au/Ni/Pd-plated QFN PPF substrate is investigated in this study. Three capillaries with the same geometry, but different surface finishes, are used to investigate the effect of capillary finish on the stitch bond process. The capillaries have different surface roughness values: the roughest is called granular finish; the next in roughness is the light-granular capillary; and the smoothest is the polished surface capillary. To compare capillaries, process windows are investigated with all of the capillary finishes. The process parameters included bond force and low frequency table motion during bonding. Two different types of table motion are used in this study. First, a single in plane motion along the wire direction, called XY – Distance (Skid). Second, a vibratory motion perpendicular to the wire (Table Scrub). The process window test results revealed that the more granular the surface, the larger the process window and the lower the chance of short tail occurrence. The largest process window is observed with the granular capillary. It has been shown that generally a higher scrub amplitude increases the chance of a short tail; however the granular capillary is the least influenced. To further compare the capillary surface finishes, a DOE with five cells including different bond force, skid, and scrub is designed. The cells are selected based on the earlier process window observation (overlapping the process window with different surface finishes and finding parameters that result in successful bonding for all three capillaries). Among the five DOE cells tested, the granular capillary again shows the best quality. It has the highest stitch pull strength and an acceptable tail pull strength. A finite element model (FEM) of the process is developed to better understand the experimental observations. Friction coefficients needed for the model are found by experiment. The model is calibrated with a previously done in-situ measurement of wire deformability. The amount of surface expansion (plastic deformation) of the wire at the wire/substrate interface is extracted from the model and attributed to the degree of adhesion (bonding). The model is able to reproduce the experimental observation of adhesion (bonding) with higher/lower granularity. The model is well suited to be extended to the more complex process parameters.|
|Alireza Rezvani, PhD Candidate
University of Waterloo