Laser-based Package Singulation and Trenching for SiP

Abstract

Pulsed lasers are solving some packaging challenges for SiP applications. SiP devices designed for mobile devices with communication capability often have a need for EMI shielding. Nanosecond UV and picosecond green lasers are finding use in cutting these multi-compound packages as well as trench for exposing board contacts used for grounding. Mechanical cutting may not be an option if the perimeter of the package is not a simple rectangle. Water jet cutting may bear the risk of the package to fail to to water seeping in between the package layers during cutting. We will demonstrate how laser-based SiP package cutting and trenching can ameliorate both of those concerns and point toward limitations of laser-based processes. Today's high power nanosecond UV lasers are capable of cutting 1mm thick packages at an effective cutting speed of more than 10mm/s. The resulting cut surface shows no charring or melting. The surface roughness allows for excellent adhesion of EMI shielding paint on the cut conformal compound. Since the laser beam can be controlled with a programmable scanner, arbitrary shapes can be cut just as easily as simple rectangles. Difference ablation rates between polymer and copper allow the laser to trench the conformal coating and stop automatically when a metal layer is exposed. This self-terminating process leads to a precise exposure even when the conformal coating thickness differs from part to part. For the utmost in surface quality picosecond lasers can be employed. Their shorter pulse duration allows for an even lower surface roughness and less thermal impact on the package. Using an ultrafast laser, such as a 10ps laser, has the additional advantage that the process scales in speed with increased laser power. The excellent edge quality and possibility to cut random shapes allow laser-based processes to complement mechanical processes in high-end applications. We will present a full cost of ownership model that allows the designer to predict the exact singulation and trenching cost for a given package size.