Carbon-based materials are heavily investigated as future CMOS-like devices and in interconnect applications. While much of the interest has been devoted to the device aspects in competition to conventional CMOS transistors, the talk here will focus on some less known applications of carbon. Proposed are interconnect applications like on-chip and DRAM capacitors, gate material or through-silicon vias (TSV), novel non-volatile memories, sensors or superior diodes. The quality and resistivity of the carbon layer depends on the deposition method, precursor and temperature, but optimum conditions can be found to achieve a benefit over competing materials in terms of temperature budget, resistivity, stress and ease of integration . The mere properties of the carbon layer is as important as their interaction with different interfacing materials like high-k materials, semiconductors and metals. Some important question which needs to be answered are, is there diffusion of carbon or from the interfacing materials? How does this impair breakdown behaviours? What are the effective work-functions for different interface materials? The answers to these questions will lead us to applications of graphene-like layers in MIM capacitor structures, DRAM capacitors and mid-gap gate-material, attractive for low power CMOS. The question on how these graphene-like layers interface with silicon will give us in the end powerful low-barrier Schottky diodes which can deal with the high temperature requirements of a front-end process. Therefore, for the first time, carbon-silicon Schottky-diodes can be incorporated in a CMOS flow. Stress, built during deposition of the many layers during a process has serious impact on the integration. Thin layers of carbon can reduce the overall stress and can ease the integration. For through-silicon vias the most economic approach of “via-first” was limited up to now to c-Si or poly-Si approach due to the high temperature requirement. Carbon layers can fill very high aspect ratio vias (up to 400) and offer a much better alternative to Si in terms of resistivity, stress and cost. If modifications of the carbon layers can be achieved at reasonable temperatures the overall resistivity can be dropped down to 10 uOhmcm which makes it already attractive for competition with W and Cu wiring. The study of maximum possible current densities in the carbon layers results in the finding of a new non-volatile memory based on the conductivity of different carbon configurations. This will not only enable cross-point memory architectures but could also be implemented in configuring FPGAs. Finally, the spin transport properties of carbon may be beneficial to solve the problems of the high current densities in spin-torque magnetic memories and for the ubiquitous GMR sensors where Neel coupling limits the sensitivity. Overall, the properties of carbon promise applications in a wide range of possible “More than Moore” scenarios for microelectronics.
«
Carbon-based materials are heavily investigated as future CMOS-like devices and in interconnect applications. While much of the interest has been devoted to the device aspects in competition to conventional CMOS transistors, the talk here will focus on some less known applications of carbon. Proposed are interconnect applications like on-chip and DRAM capacitors, gate material or through-silicon vias (TSV), novel non-volatile memories, sensors or superior diodes. The quality and resistivity of t...
»
Login
BibTeX