Electrolytic Deflash & High-pressure Water Jet Simulation Driven Delamination Risk Assessment in Plastic Packages

Plastic molded lead frame products are widely used in the semiconductor industry. The Epoxy Mold Compound (EMC) selected for these products is designed to have excellent adhesion and integrity to the die and other materials in contact. In manufacturing, mold flash is often unavoidable, and cleaning is required to enhance mechanical and thermal functionality as well as cosmetic appearance. Deflash, the commonly used cleaning method, is typically integrated into post-plating operations that may also finish the package with Sn and other lead-free plating. Deflash typically consists of passing the molded package through an alkaline bath (such as KOH) that softens and degrades the adhesion of the EMC flash. This process can be energized using Electrolytic Deflash (ED) i.e., by applying DC current between the product and the alkaline bath. ED cleaning effectiveness increases with higher current density. Next, High-Pressure Water Jet (HPWJ) removes the loosened flash in line with the deflash process. An issue/risk is that ED and HPWJ processes can induce unwanted interface delamination between the leads and the EMC at the boundary of the package. In this paper, we propose a computational framework to study the mechanism of delamination during flashing process. During the two-step analysis, electrostatics and mechanical simulations are performed to assess the impact of ED and HPWJ processes and identify regions of the package that are susceptibility for delamination. Current density was predicted to vary widely across the lead frame in-plane features and with bath concentration. Regions of the lead frame at the package’s boundary with high current density were identified as a risk for delamination. The electrolytic reaction results in crack initiation at these regions and furthermore the crack propagates under HPWJ. This is understood from HPWJ simulation where the high stress region coincides with high current density region in the ED simulation. Overall, the experimental observed delamination regions agreed with the deflash simulation predicted delam risk locations. This developed computational framework thus enables pathways to improve the strip designs and deflash process control maintenance.