ACT · Advanced gate stacks for CMOS technology

Final Activity Report Summary - ACT (Advanced gate stacks for CMOS technology)

Marie Curie project ACT targeted the heart of advanced CMOS technology, which is the dielectric at the gate of the transistors. Until very recently, micro and nanoelectronics relied on the well-known silicon dioxide (SiO2) insulator in order to regulate the flow of charge in transistors. As cost and performance drive the integrated circuits to smaller dimensions, all critical device parameters including the SiO2 thickness shrunk to the point where they became leaky and inadequate to function as gate dielectrics. Scientists in ACT and in other European initiatives tackled the difficult problem of finding alternatives for SiO2. The work in ACT focused on Hafnium (Hf)-based gate dielectric materials. It is worth noting that Hf-based dielectrics are in production as part of the new 45 nm CMOS microprocessors since the end of 2007, marking the biggest and most radical change in micro-and nano-electronics in 40 years. The ACT results in combination with the results of another European project INVEST bring us closer to the next CMOS generation. One of the biggest challenges in future technology nodes (e.g. 32 nm CMOS) is to adjust the threshold voltage of transistors at the required value. Too large a value could make the transistors run slower, to small could result in power loss. While threshold voltage adjustment was not so difficult with the SiO2 dielectric, now it becomes a big challenge as we move to Hf-based oxides replacements. The work in ACT and INVEST showed that by adding Lanthanum (La) in Hafnium oxide (HfO2) it is possible to tune the threshold voltage of nMOSFETs to the required value, opening a whole new area of research which focuses on doping HfO2 with rare earth elements. ACT also showed that La improves the reliability of devices with HfO2 k dielectric. ACT showed that thickness, preparation methodology and nitrogen treatment all influence the reliability of devices with Hf-based (e.g. HfSiO(N)) dielectrics. There has been a large debate among scientists, engineers and technologists in 'top-notched' conferences such as IEDM and VLSI Technology as to what could be the mechanism which makes rare earth elements suitable dopants for HfO2. No matter what the outcome of this debate will be, it appears that key ICs manufacturers keep rare-earth doped Hf-based oxides in their research agenda as an option for the 32 nm CMOS products expected in about one to two years from now. The scientific achievements within ACT created the best possible conditions for the training of the young scientists involved giving them the opportunity to start and later establish a successful carrier. ACT gave the scientists the means to broaden their knowledge from basic materials science to processing and device testing and further develop their skills in engineering and performance evaluation. On the other hand, the European semiconductor industry benefited a lot from a highly qualified expertise that was trained in one of the most advanced and competitive technological fields. This is expected to add to the effort to maintain both intellectual property and semiconductor manufacturing in Europe.

Data: CORDIS, © European Union