GGrantIndex
← Search

A New-Class of Molecularly-Engineered Nanoporous Dielectric Materials for Insulation in Device Wiring for Integrated Circuits

$368,337FY2005MPSNSF

Rensselaer Polytechnic Institute, Troy NY

Investigators

Abstract

This project aims for a new strategy to develop a unique class of nanoporous low permittivity Si-C-H-O dielectric material with built-in thermochemical and mechanical stability, and enable its integration with Cu wiring in integrated circuits without using a separate interfacial layer. Rapid Cu diffusion into adjacent insulating dielectric material and poor Cu-dielectric interfacial adhesion are major issues in device wiring in electronic circuits. Typically, 10-20 nm-thick transition-metal based interfacial barrier layers are used to circumvent the problem. Such thick barrier layers cannot be used in sub-50-nm devices because they decrease the space meant for low-resistivity Cu and neutralize the main advantage of Cu wiring. Ultrathin (e.g., <3 nm) barriers required of these materials are not easy to deposit by conventional methods, especially to conformally coat high depth-to-width aspect ratio features common in multilevel wiring. Intellectual merit: Self- assembled molecular nanolayers (SAMs) offer potential to overcome shortcomings of conventional barriers while enhancing the chemical and mechanical integrity of the Cu-dielectric interface. This project seeks to integrate low-polarizability organosilane SAMs into porous dielectrics during synthesis and develop an understanding of the electrical and mechanical properties, and chemical and thermal stability, to enable the direct integration of this material with Cu. The approach involves incorporating (a) chemical-attack-inhibiting moieties on external and internal surfaces and (b) molecular bridges and ordered pores to mechanically reinforce porous dielectrics. Specific objectives are to: (i) Synthesize barrier-less dielectrics through the incorporation of pore-passivating and adhesion-enhancing molecular moieties into silica-based porous dielectrics; (ii) Mechanically reinforce the porous dielectric by creating ordered pores, and molecular bridging of pore walls by cross-linking low-polarizability organosilanes; (iii) Characterize the effects of the above strategies on dielectric properties, thermal and chemical stability, resilience to Cu-diffusion, and Cu-dielectric interfacial adhesion; and (iv) Understand and optimize the effects of molecular termini, chain length, and fraction of intra-pore molecular crosslinking, and processing parameters on properties. To achieve the above, organosilanes with Cu immobilizing, hydrophobic, or cross-linkable termini will be introduced into micellar templates used in sol-gel synthesis of porous silica, and integrated with the dielectric during gelation and post-treatments. Processing-structure-chemistry-property relationships will be revealed through a combination of electrical tests during thermal annealing in controlled ambients, four-point-bend adhesion tests, nanoindentation, atomic force microscopy, x-ray photoelectron spectroscopy, and transmission electron microscopy. Broader impact: The success of this approach could revolutionize wiring design and fabrication for integrated circuits with sub-50-nm devices by obviating interfacial barrier layers between metals and dielectrics. The new knowledge gained from this study will provide atomistic insights on key properties of molecularly engineered porous dielectrics, aid in combining nanostructure self-assembly with device fabrication, and contribute towards bridging micro- and nano-device technologies. The project will also provide a unique opportunity for interdisciplinary training of graduate and undergraduate students through research in molecular self-assembly, sol-gel processing, materials characterization, and device fabrication and testing. Collaborative interactions with IBM will enrich the students' learning experience. A summer internship for two high-school teachers is planned to create and share demonstrations for use in their classrooms. This new activity will complement ongoing visits and reverse-visits to high-schools in the capital region and Rensselaer, contribute to increasing students' awareness on self-assembly and nanodevices and their connection with applied physics and chemistry, and kindle students' interest in science and engineering. The research will be integrated in the self-assembly and molecular nanostructures of a Nanostructured Materials course taught by the PI through an interactive web-module.

View original record on NSF Award Search →