Interface Characterization and Failure Modeling for Semiconductor Application

Abstract

Due to the fact that microelectronics is made up from various materials with highly dissimilar thermo-mechanical properties, generally the interface between two adjacent materials is the place where delamination related failure most likely would occur. Failure of these interfaces results in decreased reliability and loss of performance of microelectronic components. Therefore, adequate knowledge of delamination prediction is desirable. To be able to predict delamination growth the incremental energy being released when a crack propagates over an incremental crack area growth, the “Energy Release Rate’’ G, should be firstly calculated. The value can be compared with the “Critical Energy Release Rate” or so-called interface toughness Gc. This interface toughness can be obtained through combined experimental and numerical approaches. The measurements are complicated due to the fact that the critical energy release Rate is temperature, moisture and stress state (mode mixity) dependent. Moreover, the prediction of delamination depends very much on the accuracy of the material model used in the simulations. In particular an accurate material model is necessary to calculate the stresses and strains due to cooling down from the molding temperature and due to moisture absorption. The success of analyzing and predicting the interfacial delamination problems in microelectronic components strongly depends on the accuracy of the established critical interfacial energy release rate. The goal of this research is to build a test method for establishing the delamination strength of interfaces between epoxy molding compound (EMC) and copper lead frame. Among various test method possibilities, the so-called mixed mode bending (MMB) test method was selected as a base of the method development. It should be realized that for an interface delamination, due to the mismatch in thermal-mechanical properties of the materials adjacent to the interface and also because of the residual stress state the crack will always propagate under mixed mode conditions. The test samples are cut from electronic components being fabricated with the regular production process. Consequently, the specimen dimensions are relatively small and therefore, a dedicated small-size test set-up was realized. The test setup allows the possibility to transfer two separate loadings (mode I and mode II) on a single specimen. The test setup is flexible and adjustable for measuring specimens with various dimensions. For measurements under various temperatures and moisture conditions, a special climate chamber is designed. The “current crack length” is required for the interpretation of measurement results through FEM-fracture mechanics simulations. Therefore, during testing the “current crack length” is captured using a CCD camera. In order to be able to establish the interface fracture toughness accurately, the viscoelastic material properties of the applied molding compound are first established and subsequently considered in the FEM simulations. A special loading procedure is used to investigate the fracture properties in the viscoelastic transition (temperature) region of the EMC. The critical fracture properties are obtained by interpreting the experimental results through dedicated finite element modeling. The (stress) mode mixity is defined as a ratio of mode I (tensile-) to mode II (shear-) loading on the interface near the crack tip. In the present study the mode mixity is obtained through an alternative manner, using the crack opening displacements ahead of the crack tip. The FEM model used to simulate the viscoelastic material behavior will be discussed. The delamination toughness (critical energy release rate) as a function of mode mixity at different temperatures will be given in the result section.