February_EDFA_Digital

edfas.org ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 23 NO. 1 6 initiated to compare the performance of two sets of bill-of- materials (BOM) for a given device, a device that has bond over active circuit (BOAC) metal structures with exposed Cu along the sidewalls and nickel/palladium layers on top was evaluated. Corrosive constituents canoften react with the exposed Cu sidewalls. For this particular investiga- tion, autoclave testing, which is an environmental test that measures device resistance to moisture penetration and the resultant effects of galvanic corrosion, was inclu- ded to find out if the change in BOM would affect Cu sidewall corrosion. Decapsulation was chosen as the preferential sample preparation to allow top down inspection of the BOAC structures for evidence of Cu corrosion. As acid decapsu- lation would have resulted in Cu sidewall corrosion and would have removed any evidence of Cu migration, the samples were decapsulated using atmospheric pressure MIP. For each BOM group, some samples were subjected to 96 hr autoclave exposure, and other samples were subjected to 168 hr autoclave. Four samples for each auto- clave condition fromeach group were decapsulated. One sample did show evidence of Cu corroded and migrated from one BOAC line and deposited onto the other BOAC line (see Figs. 5 and 6). Only one corroded area was found on the entire decapsulated die surface for this one unit. Fig. 5 Optical images of reference sample and failed sample, close-up image showing corrosion and discoloration on failed sample. Fig. 6 Oblique viewSEM image showing corrosion (cavities in the Cu sidewall of the adjacent BOAC line). Fig. 7 (a) After acid decapsulation, carbonized mold compound residue remains. (b) Further MIP de- capsulation removed the carbonizedmold compound and exposed the original EOS site in the die circuitry without artifact.