Evaluation of polymeric coatings for MCM applications

Abstract

An effort was initiated under the Reliability without Hermeticity (RwoH) project to evaluate the performance of nonhermetic coatings for multichip module (MCM) applications. As a means of selecting coating materials, Sandia ATCO1 test chips, with and without chip passivation (SiN), were coated with three nonhermetic polymer materials: a silicone gel, a filled epoxy, and a polyimide. Some of the coated parts were preconditioned (500 temperature cycles and 96 h of salt spray), and some were not. The coated parts were subjected to either pressure cooker test (PCT-121/spl deg/C, 99.6% RH, no bias) or temperature-humidity-bias (THB-85/spl deg/C, 85% RH, 40 V bias) tests, for 7000 and 10 000 h, respectively. In this paper, the moisture protection performance of the coatings on nonpassivated test chips is presented in detail. In addition, some overall comparisons of the failures in passivated and nonpassivated chips are given. For the preconditioned parts, THB median lives (times to 50% failures) in excess of 10 000 h were obtained for the epoxy and the gel coatings. Polyimides exhibited a THB median life of 6950 h. In PCT, the three coatings exhibited similar median lives in the range of 350-400 h, but, much different failure kinetics. Based on earlier published studies, these median lives are representative of good coating performance. The failures in the nonpreconditioned coated parts were much lower (20-50%) for the gel and the polyimide coatings, at the end of THB tests. Epoxies on the other hand did not exhibit such influence of preconditioning steps on failures. The comparison of passivated and nonpassivated coated ATC01's revealed that the presence of passivation enhances the moisture protection performance of all three coatings. This effect was especially pronounced for the silicone gel.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>