In this paper we propose a new multistep characterisation method to be able to map out the dependency of moisture diffusion parameters of a polymeric material over a range of temperature and humidity conditions in a limited amount of time. We do that by (1) using a moisture sorption analyser which can continuously monitor weight changes with microgram accuracy, (2) using thin samples which speeds up the diffusion process and (3) already switch to the next humidity level at 90 or 95% completion of a diffusion step. A multistep diffusion model was developed to account for the overlapping diffusion steps. This model showed to be extremely accurate for fitting experiments consisting of five absorption and one desorption steps. We show that for temperatures up to 85 °C and humidity level between 0 and 85% RH the diffusion of our material was essentially Fickian with a diffusion coefficient ranging from 3.8 × 10⁻⁷ mm²/s at 20 °C to 3.6 × 10⁻⁶ mm²/s at 85 °C. The moisture saturation concentrations showed a slightly non-linear variation with the applied humidity level.

In microelectronic packages, generally the chip is encapsulated by a molding compound (MC). The MC provides a mechanical support for the chip and isolates it from the environment and as a result protects the encapsulated chip. It is well known that MC's are polymer-based materials. When packages are exposed to a harsh environment such as to high-temperature storage or to thermal cycling, the mechanical properties of the MC's can change significantly. Consequently this could result into reliability issues of these packages. For a long time, there was no simple and efficient model method available to simulate the mechanical behavior of these packages under thermal aging conditions. As a result, it was hard to forecast the package reliability after a period of thermal aging. Since in our previous work [1,2] the thermomechanical properties of MC's before and after thermal aging were systematically characterized, the above problem was merely solved. A simple and efficient modeling method was proposed to simulate the thermal aging effects on MC's [2]. In this paper, a bi-material sample consisting of a MC layer on a Copper substrate is prepared and used to verify the proposed modeling method at two different thermal conditions: High-temperature storage (HTS) and Temperature cycling (TC). Based on the proposed modeling method the mechanical behavior of the bi-material sample after aging under these (different) thermal conditions are established throug FEM simulation. The simulation results match the experiment results quite well.