In-situ identification of interface properties within 3D-microelectronic devices

The functionality of many modern devices, such as smartphones, LED-lighting systems, and flat displays, relies on microelectronics that are partly composed of thin material films stacked on top of each other. A persistent problem during processing and application of these devices is adhesive failure of the material stack, especially at the interfaces between different materials. This means that different material layers delaminate from one another, jeopardizing the reliability of the device. To understand the micromechanical behavior of the material stack, and eventually improve its production process, it is crucial to know the relevant interfacial properties. This calls for an advanced measurement technique, making use of a combined numerical-experimental approach, by which these properties can directly be identified within the complex 3D-microelectronic device itself.

PhD candidate: ir. Andre Ruybalid   
Supervisors: dr. ir. Johan Hoefnagels, dr. ir. Hans van Dommelen, dr. ir. Olaf van der Sluis, prof. dr. ir. Marc Geers
Institutions: Materials innovation institute (M2i), Philips Research

The aim of the project is to develop a combined approach of (1) interface testing (experimental), and (2) interface modeling (numerical), in order to realize direct, in-situ interface characterization of 3D-microelectronic devices. This allows for in-line testing of products right off the production line, obviating the need for time-consuming and expensive testing procedures that increase the time-to-market. Furthermore, not using simplified test-specimens, as is conventionally done when experiments are performed, but testing the actual 3D-device, allows the test results to be ultimately realistic. To characterize interface materials, and acquire material parameters, information is needed about the amount of loading of material and the resulting deformation. First, an experiment is conducted by which the complex 3D-microelectronic device is loaded, while being monitored by a microscopic technique. The resulting microscopic images contain information about the local deformation within the observed materials.

After the experimental data is acquired, a Finite Element Model (FEM) is used to simulate the deformation of the interface materials. This FEM-simulation requires material parameters as input. However, these parameters are not yet known, since they are what we are looking for. So, only initial guesses of these parameters can be given as input to the FEM-model. The FEM-simulated deformation results are then applied to the experimentally acquired images, and used in a Digital Image Correlating (DIC) procedure. How well the images can be correlated by DIC is determined by the correctness of the material parameters used in the FEM-simulation. By updating the material parameters used for FEM-simulation and repeating this process, FEM and DIC are iteratively alternated, converging towards a solution for the sought interfacial parameters. This identification technique is called FEM-based integrated DIC. The combination of numerical simulations with experimentation allows for direct identification of the interfacial properties of the complex material stacks.