Tuesday (9.00 h - 9.20 h)

New Materials and Concepts for Nano Transistors
J. Efavi (Sp), M.C. Lemme, H.D. Gottlob, T. Mollenhauer, T. Wahlbrink, H. Kurz, AMO GmbH, Aachen (Germany)

The unprecedented advances of CMOS device technology for the past decades have been based on the use of bulk-Si, poly-Si, and silicon dioxide that enabled scaling of device geometric dimensions to improve performance. In the near future, however, their use is limited and it has become inevitable to integrate new materials and device concepts in order to continue the aggressive scaling of CMOS into the nanometer regime in line with semiconductor international association SIA road map [1].
The need to integrate new materials is the result of: high gate leakage current, boron penetration in p+ MOSFET, poly-silicon gate depletion and channel mobility issues for submicron devices [2]. Elimination of these device problems requires the introduction of new materials such as high-k dielectrics, metallic gate electrodes and strained silicon that can provide equivalent operational reliability [3].
Silicon dioxide (SiO2) has been used in the past decades as the sole transistor gate insulator. In the nano-transistor era, the physical thickness of SiO2 is only a few atomic layers thick and is therefore losing its inherent physical properties as an insulator due to gate direct tunneling leakage current. To overcome this problem, there are numerous efforts to integrate high-k materials which can provide the same or better device performance as silicon dioxide but at a higher physical thickness [3].
Another area of material change in nano-transistors is the replacement of the poly-silicon gate electrode by a metallic gate. Poly silicon gates are usually doped to change their semi-conducting properties into metallic properties. Dopant concentrations at the Si/SiO2 interface are not high enough to realize this after the activation process step. This causes the poly silicon gate electrode to be depleted when devices are biased in inversion adding an unacceptable thickness of about 0.5nm to the effective gate oxide and as a consequence a degraded inversion capacitance. Further increase of the dopant concentration in poly-silicon gate electrodes is not possible since dopant concentration has reached its limit in silicon. Boron which is used for forming p+ MOSFETs has high diffusivity in SiO2 during thermal processing and also results in threshold voltage fluctuation and oxide degradation. Metallic electrodes are now intensely being investigated to address these issues [4].
A key criteria for improve performance in integrated circuits is the drive current. Increased circuit speed can be enhanced with the use of strained silicon in the MOSFET channel [5]. By introducing strain in the channel, in principle the carrier mobility is increased resulting in higher drain currents and switching speed finally.

The introduction of new materials requires new device concepts since conventional planar MOSFTs suffer from severe short channel effects (SCE) in nano-transistors. To suppress effectively SCE in high performance device the interaction of new materials has to be accomplished within novel concepts with non-planar gate structures (Finfet, Trigate, Double gate FET built on silicon-on-insulator) [3].

In this presentation, emerging and promising high-k materials, metal gate electrodes together with novel non-planar device structures are discussed. Integration related challenges as well as efforts to overcome them are addressed.

REFERENCES
[1] The International Technology Roadmap for Semiconductors 2002 Update, pp. 9-10, http://public.itrs.net/Files/2002Update/2002Update.htm
[2] C. Hu, “MOSFET Scaling in the Next Decade and Beyond”, Semiconductor International, June 1994.
[3] H. S. P. Wong, “Beyond the conventional transistor”, IBM J. Res. & Devices, 46(2/3), pp. 133-168, March/May 2002.
[4] G. A. Brown, et al., “Scaling CMOS Materials & Devices”, Materials Today January 2004, pp 20-25.
[5] H. R. Huff, P. M. Zeitzoff, “A Perspective on Enhancing Mobility” Solid State Technology, January 20

Wednesday (9.00 h -9.20 h)

SiO2-Glass Fibres with Complex Cross-Sectional Geometry

J. Heiber (Sp), M. Wegmann, T. Graule, Eidgenössische Materialprüfungs- und Forschungsanstalt, Dübendorf (Switzerland); D. Hülsenberg, Ilmenau Technical University (Germany)

Today a variety of possibilities are available to produce silica glass fibres with circular cross-sections. These include melting and drawing technologies as well as technologies based on either sol-gel processing or powder extrusion.
The aim of the current study was to fabricate fibre substrates with at least one flat face which can be used as microelectronic substrates for integrating computing power into textiles. To this end, the preform drawing and powder extrusion technologies were investigated to produce silica fibres with either triangular or rectangular cross-sections.
In the case of the perform-drawing method, commercially-available circular SiO2-rods were ground and polished to yield preforms with the desired non-circular cross-sections. These rods were then heated up to temperature >1600°C and drawn to fibres with the desired dimensions. The cross-sectional fibre dimensions could be adjusted by varying the feed rate of the rod into the furnace or the drawing rate of the fibre out of the furnace. Rounding of the flat faces and the sharp corners was investigated as a function of drawing velocity and temperature.
For the extrusion process, a nano-sized and a micron-sized SiO2 powder were blended with a thermoplastic binder system for 3 hours at 150°C and subsequently extruded through a die with a 500 mm diameter circular die land. The effects of the two different particle size distributions as well as the influence of varying powder loading (between 38 and 58 vol%) on the rheological properties of the feedstocks were analysed using capillary rheometry. Green fibres were debound at 500°C and sintered 1 h at 1100°C in air. The debinding and sintering behaviour was investigated using mercury intrusion porosimetry, thermal gravimetric analysis and dilatometry. Because only the green extrudates consisting of nano-sized powder could be sintered crack-free to glass fibres, these feedstocks were used to extrude and sinter fibres with a non-circular cross-section.

Thursday (9.00 h -9.20 h)

Metal Foams - Airy Materials
N. Babcsan (Sp), F. Garcia-Moreno, J. Banhart, Hahn-Meitner-Institut Berlin GmbH (Germany); D. Leitlmeier, ARC Leichtmetallkompetenzzentrum Ranshofen GmbH (Austria)

Airy materials can be made from metallic systems by foaming. They even float on the water but still remain strong for several application. One can foam melts directly by gas injection (beer route) or can apply indirect foaming by using gas releasing agents (bread route). Metal foams have similar structure to a slice of bread. The shiny surface and the shape of the metallic bubbles are catching the eyes and have motivated industrial artists. However, their properties have high importance in industrial applications. Some products have already been realized as part of lifting arm of light trucks or sound barriers for highways and bridges. The volume fraction of gas bubbles in the metal is a simple variable to tailor the properties for the costumer requirements. By the help of material designers, the wide range of possible properties can lead to innovative applications and multifunctional materials. The main challenge of recent research is to achieve more uniform cellular structure, to control foam architecture and to improve reproducibility of part production. It requires new research directions (high temperature colloid chemistry), multidisciplinary knowledge and new tools (real time X-ray radiography) for the materials scientists and engineers. As the life of the liquid metal foam is strongly affected by the gravity, one of the aims of the research community is to understand the fundamentals even in low gravity environment.