Super Diffusive Self-Aligning Vertical and Lateral Interconnect Formation by Gas Phase Electrodeposition

This presentaion reports self-aligning metallic vertical and lateral intterconnect formation by gas phase electrodeposition. Key operational parameters to fabricate vertical ruthenium and rhodium interconnects (via) with a diameter of 100 nm are discussed. Moreover, airgaps are implemented during the deposition process, which utilizes spark discharge to generate a flux of charged nanoparticles. An inert gas flow transports the nanoparticles through a reactor chamber close to the target substrate. As well, the talk will introduce a super diffusive approach for gas phase electrodeposition, which uses a substrate with pre-patterned resist openings to guide the nanoparticles in the self-assembling manner into the openings by an interplay of drag, Coulomb force, and Brownian motion. Five process parameters were identified, which impact the morphology and conductance of the resulting interconnects: spark discharge power, gas flow rate, micro lens via dimensions, substrate surface potential, and in-situ flash lamp power. This parameter set enables a controlled adjustment of the vertical and lateral interconnect morphology and its minimum feature size. Gas flow rate in combination with spark discharge power contribute significantly to the morphology of the interconnect. Spark power and micro lens via dimensions have the largest influence on the surface potential of the insulating resist cover, which enables a localized micro lensing gas phase electrodeposition of a via with a controlled ratio between conducting diameter and airgap. To get a better insight into the growth characteristics, a statistical model leads to a better understanding of gas phase electrodeposition. The model implements that added to usual standard diffusion with random walks in a certain direction, long flights occur to a particle as well, if there is an additional the electrical field. As a result, the diffusion factor D is not a constant anymore and follows the approach of super diffusion.