he quest for miniaturization, technology has come to a point where CMOS scaling beyond the 45nm node cannot be achieved by simply reducing transistor dimensions. Moreover, the need for small form factors coupled with the stringent requirement of low current leakage/low energy performance has become critical, especially in next-generation mobile devices. Within imec’s Germanium and III-V devices program, imec and its core partners are exploring the efficacy of high-mobility channel materials for CMOS devices for advanced nodes. Together with Riber, the bottleneck issue of gate stack passivation was tackled, resulting in effective passivation techniques for Ge and GaAs. Riber’s 200mm III-V and metal oxide MBE cluster offered the required extremely clean background and absence of any interfering gas phase components, enabling material and interface control on the atomic level. This resulted in the successful development of a passivation scheme for the MOS gate stack module. Amongst others, it was shown that controlling the GaAs surface reconstruction followed by a H2S passivation treatment and in-situ high-k deposition was crucial to create a well-passivated MOS structure with record-low interfacial state density. Moreover, the world’s first successful MOS capacitors on a new high-mobility candidate material, GeSn, were made in the 200mm Riber MBE cluster.
In the new project, the suitability of Riber’s 300 mm UHV chamber (ISA300), equipped with in-situ tools for surface analysis, and clustered with state-of-the-art 300mm Si CMOS production equipment, will be evaluated for the production of advanced CMOS devices based on high-mobility Ge and IIIV channels. The aim of the project is threefold
1) use Riber’s UHV chamber for study and control of surface structures;
2) bring knowledge on gate stack passivation from a 200mm research environment to a 300mm fab compatible platform; 3) demonstrate the technological viability of a 300mm MBE-module, clustered with ‘standard’ 300mm Si CMOS production equipment.
Frédérick Goutard, Riber CEO: “Participating in early stage research is intrinsic to Riber’s aim to strengthen our leading position in the compound semiconductor industry. Molecular Beam Epitaxy (MBE) is considered to be most efficient technology to leverage compound semiconductors towards high-end applications with a tremendous market potential, such as mobile phones, telecom, automotive, satellite, ....
Using our ISA300 chamber in a 300mm cluster configuration will allow for higher performance in terms of devices operating at high frequencies with reduced energy consumption, etc. We are therefore convinced that this project will clearly demonstrate the attractiveness of our technology in the production of next generations CMOS devices.”
Hans Lebon VP fab operations and process step R&D at imec: “Through the collaboration with Riber, imec can integrate the power of UHV-systems into state-of-the-art semiconductors production equipment on large diameter wafers. This will allow the application of powerful in-situ surface analysis tools in CVD and ALD equipment which so far was not feasible due to the gas phase environment. Also, MBE-like UHV-techniques for interface control and passivation come in reach which will benefit the development of next generation CMOS technologies.”