IN THE PRESENT PHASE of development, microelectronics has reached a state in which, because of the diminutive size of structures, quantum mechanical effects on the one hand side, and thermal and mechanical effects on the other hand side, gain increasing significance. As a result of the ever-shrinking size of the semiconductor devices, more and more components per given chip area can be integrated for the purpose of handling highly complex tasks more efficiently. This high integration density of semiconductor components entails new challenges for their design, operation, and reliability.

While -- on the average -- over a period of eighteen months the density of integration of components per chip doubles, over the same span of time the resulting power-loss density keeps growing exponentially -- and, concurrently, resulting a dramatic increase in the thermal load on the components. Consequently, it becomes essential to provide considerably better cooling for these high-performance semiconductor devices than for traditional components of the same type.

In order to deal with these effects, a deepened insight into thermal effects and developments is required, while in addition, these effects and their consequences should be rigorously considered by an effective simulation software tool already during the development phase. This way, by the use of suitable optimizing strategies, the components involved can be optimally designed to meet specific operating conditions. As a result, effectively combining the simulation of electrical and thermal effects will be a task of ever increasing importance. Furthermore, for components of such diminutive size, it will become essential to also include mechanical aspects into relevant investigations and research, since particularly these effects are of considerable relevance for reliability.

Since both thermal and mechanical loads have significant bearing on the electrical properties as well as on the reliability of semiconductor components, it becomes a must to optimize them for their intended specific use. An absolute requirement for this is determining material properties with utmost precision, so that characteristics of critical components can be optimized for their specific purpose.

Furthermore, this dissertation is to demonstrate how -- by means of simple transient electrical measurements and sophisticated optimizing strategies -- important electrical and thermal material parameters can be identified without taking recourse to costly and time-consuming caloric measuring procedures. Subsequently, material parameters thus identified find use in complex compound structures in order to provide the even more precise results that are required for describing the transient temperature developments, and thus being able to calculate the resulting mechanical loads.

Stefan Holzer 2007-11-19