A New Approach for Testing and Modeling Silicon Dioxide Reliability

[9420585]罗森鲍姆将调查三个问题，其答案将为二氧化硅主题的未来工作提供实际的重点。首先，物理实验将用于研究极薄氧化物的可靠性，并发现目前认识中的不足之处。其次，通过对实验数据的分析和可靠性工程师的广泛讨论，将对氧化物可靠性的现有模型进行评估，并注意那些被认为不足的模型。最后，将确定氧化物可靠性的哪些方面将在实践中限制未来集成电路的可靠性。我们将考虑各种电路类型，如逻辑电路、非易失性存储器和DRAM，因为不同的电路可能受到不同氧化物可靠性问题的限制。这项工作的最终目标是将氧化物可靠性模型整合到可靠性设计CAD工具中；然而，这些活动的重点是可靠性物理和建模，而不是算法开发。由于这项研究的动机是开发有用的设计工具，因此在资助期间将作出重大努力，与这些工具的预期用户，即工业可靠性和设计工程师进行互动。＊＊＊

9420585 Rosenbaum Three questions will be investigated and the answers will provide a practical focus for future work on the subject of silicon dioxide. First, physical experimentation will be used to study reliability of very thin oxides and to discover deficiencies in the current understanding. Second, through analysis of experimental data and extensive discussions will reliability engineers from industry, present models for oxide reliability will be evaluated and note will be made of those which are considered inadequate. Finally, a determination will be made as to which aspects of oxide reliability will, in practice, limit the reliability o future integrated circuits. A variety of circuit types will be considered, such as, logic circuits, non-volatile memories, and DRAM, as different circuits may be limited by different oxide reliability problems. The final goal of this work is incorporation of models for oxide reliability in a design-for-reliability CAD tool; however, the focus of the activities on reliability physics and modeling rather than on algorithm development. Since development of a useful design tool provides the motivation for this research, significant effort will be made during the grant period to interact with the intended users of such tools, namely, industrial reliability and design engineers. ***