Abstract
Tin whisker growth has become a reliability concern during the course of Pb-free solder transition. A common mitigation strategy is to use conformal coatings to prevent electrical shorts by tin whiskers. However, recent research indicates that under elevated temperature and humidity, or in areas of thin covering, whiskers can grow and penetrate conformal coatings. Additionally, for long environmental exposure, the effectiveness of conformal coatings may be compromised.
The main objective of this work is thus to develop a test procedure (1) to determine the initial adhesive and cohesive strengths of an applied coating and (2) to subsequently assess the rate of degradation under accelerated testing conditions for long-term evaluation.
We propose a blister-type test method to assess both the adhesive and cohesive strengths of conformal coatings. The proposed approach offers unique advantages. The testing method mimics actual tin whisker growth and allows for a quantitative comparison of the adhesive and cohesive strengths of the coating. In addition, blister test specimens can be subjected to harsh environments, which make this test perfect for accelerated testing.
In the typical blister test of adhesive strength, pressure is applied through a hole in a substrate. As the pressure is increased, the coating layer on top of the substrate will begin to bulge out like a blister. The blister will continue to grow until it reaches a maximum blister height, at which point delamination will occur [1]. In a modified blister test for cohesive strength, a thin film coating will be fixed between two copper substrates. A ramp pressure will be applied to the hole in the lower substrate until the blister reaches a critical height, at which point the film will fail due to rupture.
The typical blister tests of coating layers are marred by the presence of large scatter in their results, making the quantitative assessment of adhesion difficult, and the ability to distinguish the effects of environmental exposure even more challenging. In the proposed study a novel concept to create a pre-defined initial crack diameter is proposed and implemented to reduce deviation in test results.
Two common conformal coatings, silicone rubber and urethane, were selected for testing. Preliminary experimental results indicate that silicone coating samples fail due to the limited cohesive strength of the coating, while urethane coating samples fail either cohesively or adhesively due to weak adhesion between the coating and the tin-coated substrate. Each conformal coating was tested before and after subjecting it to two accelerated testing conditions. The first accelerated testing condition is a high temperature and humidity chamber maintaining the specimens at 55°C/85%RH for a 2000 hour period. Adhesion and cohesion testing was then performed every 1000 hours. The second accelerated testing condition is a temperature cycling chamber which cycled the temperature from -55°C to 125°C for 1000 cycles. Testing was performed for every 500 cycles in the T/C chamber. From the blister test for adhesive strength, the critical pressure is determined, and an analytical equation developed by Gent and Lewandowski [2] is applied to determine the adhesion strength. From the cohesive blister test, the maximum blister height is determined, and a finite element model is employed to determine the cohesive strength.
The effectiveness of the proposed test procedure is reviewed by using the experimental data of initial properties. The results from the accelerated testing are followed to evaluate the key parameters of two conformal coatings for long-term reliability.
