Abstract
Achieving a reliable package requires ensuring the reliability of each of its components and their interconnections. In practice, die strength is adversely affected when dies undergo various manufacturing processes. High stresses are also induced when dies are subject to conditions such as testing, handling, and operation. Die reliability depends on processed die strength under possible stress conditions. Existing information on the die strength is limited to what is known about perfect, unprocessed silicon. The accurate measure of the strength of processed dies is achieved through evaluation of the real component. This work focuses on the comparative analysis of test methodologies and the effects of processing. Characterization of actual die defects dies, primarily due to processing, is followed by experimentation utilizing a system designed at Motorola’s AISL specifically for the testing of small dies. Failure analysis of fractured dies reveals the effects of different test methods and the roots of die fracture. Coupling the experimental data from this work with predicted stress results produced through finite element analysis, a measure of the adequacy of a die’s strength for a given package can be made, providing valuable information for purposes of package reliability and enhancement.
