Dear Chen DF,
I have encountered situations like the one you described in the past. It
would be very surprising if the high resistance you are measuring is, in
fact, due to a change in resistivity in the coonductive epoxy. My guess is
that the actual location of the high rtesistance has not been determined
yet.
My first suggestion is to check for proper adhesion of the epoxy to the pad.
Try pressing down on the epoxy a little with your multimeter probe and see
if that reduces resistance. A high resistance layer can also be formed due
to segregation of resin at the pad interface, impeding the conductive
particles to form intimate contact with the surface of the pad.
Depending on what your bonding surface is, there could also be metal
oxidation at the pad surface occuring during the curing cycle, this would
increase resistance as well. Even if the surface is supposed to be gold,
there may be a thin layer of contaminants or metal residues on the surface,
undetectable to all but surface sensitive analysis techniques (XPS, Auger,
SIMS).
Hope this helps,
Gabriel Rojano
Senior Engineer
Reliability Analysis Laboratory