Processing design rules for reliable reflowable underfill application

Recently there has been a growing interest in reflowable underfill materials for flip chip processes as a way to increase productivity - especially in consumer products or applications with a high degree of miniaturization. Former investigations had shown that the reliability obtained with commercially available products still needs improvement. Therefore it is especially important to observe the progress made in material development and to establish optimized processes. The scope of this work is the characterization and evaluation of different currently available reflowable underfill materials for optimized flip chip processes. A method is developed for the analysis of the underfillers that can be used for a standardized incoming inspection and preliminary process specification. It includes the determination of material properties (cure behavior, Tg, CTE, modulus) as well as the investigation of the flow behavior during placement. This is realized by a special test vehicle which carries structures that can simulate certain typical board characteristics (topography of solder mask, vias, conductor lines) under the die. The influence of these structures together with process parameters like deposition pattern, placement temperature and placement speed is studied. This analysis is a fast way to obtain the data required for optimizing the process flow for the materials chosen and for establishing basic design rules. With the processes defined from these results test samples with Daisy Chain structures are assembled. Solder joint formation and quality of the underfill layer are the criteria which are evaluated by X-ray, CSAM and cross sectioning. Together with the material properties these results allow the definition of design rules for flip chip processes using reflowable underfill. In order to verify the results and evaluate the underfillers selected the samples are then subjected to thermal cycling and humidity testing to determine the reliability obtained. The results allow a comparison with standard underfill materials and a statement on the applicability for production