摘要
金属互连电迁移有断路失效和短路失效两种常规失效模式,其中短路失效是由于发生了析出效应。目前对电迁移断路失效的研究较多,但是对于析出效应（短路失效）的研究较少。研究发现在金属电迁移析出效应监测过程中易产生两种电介质击穿效应,分别为在实验刚开始发生的瞬时电介质击穿（TZDB）效应和测试过程中产生的时间依赖性电介质击穿（TDDB）效应。此外,电介质层材料的介电常数值越高,其耐电介质击穿的能力越高。析出效应的监测电场强度的设定值应该同时考虑电介质层材料与测试结构的特性,监测电场强度的设定范围建议为0.15~0.24 MV/cm,以防止在析出效应监测过程中发生电介质击穿,混淆两种不同的失效机理,造成误判。
Metal interconnection open and short are the two normal failure models observed in electromigration test. The reason of the short circuit is due to extrusion effect. The researches in electromigration were mainly focused on the open failure model,but less on the extrusion（ short） failure model. The studies show that two kinds of dielectric breakdown may be found during monitor extrusion in electromigration test,one of which occurs at the beginning of the experiment called time-zero dielectric breakdown（ TZDB）,and the other one occurs during the process of the experiment called time dependent dielectric breakdown（ TDDB）. Moreover,the resistance to dielectric breakdown is positively correlated with the value of dielectric constant. The characteristic of dielectric material and test structure should be considered when setting the electric field intensity that used to monitor extrusion. The recommended range of the electric field intensity used to monitor extrusion is 0. 15- 0. 24 MV / cm,in order to avoid dielectric breakdown,which can prevent confusing the two different failure mechanisms to ensure the accuracy of electromigration test.
出处
《半导体技术》
CAS
CSCD
北大核心
2015年第4期314-318,共5页
Semiconductor Technology
