Germanium: The Past and Possibly a Future Material for Microelectronics

Abstract

Germanium possesses higher bulk mobilities than silicon and was used for the first transistors. However, by the 1960s its use was largely supplanted with Si due largely to Si's high quality thermal oxide. Today, with the 45 nm technology node in production, high k dielectrics are beginning to replace SiO2 in the gate, and as such, one of the key reasons for using Si is no longer as relevant. This, combined with performance concerns for Si based devices for and beyond the 22 nm node has made Ge a worthy area for research for high performance devices. In this paper, we give an overview of some of the major issues for Ge MOSFETs, illustrating recent progress using data from IMEC. Key results include a factor of 10 reduction in threading dislocations for epi Ge on Si by the use of an ~850{degree sign}C anneal, the first successful use of As halo implants, progress on optimization of activation anneals, the use of a thin epitaxial Si passivation and its impact on threshold voltage, a 2-step anneal NiGe process flow to drastically reduce defects, and the importance of a hydrogen anneal following metallization.