This PhD topic is part of the 3D System Integration Program at imec. In 3D integration, integrated circuits are vertically stacked to overcome limitations arising from ever continuing transistor scaling. Wafer-to-wafer hybrid bonding of patterned copper-dielectric surfaces is one of the key technologies, targeting vertical interconnect pitches from 5 down to 1 µm. Following dielectric bonding at room temperature, bonding of the copper pads is achieved through annealing to induce thermal expansion, deformation and diffusion of the copper. The success of this process critically depends on the topography of the copper pads, which is tailored in a polishing step prior to bonding. Because of their small size, each copper pad consists of only a few grains, and the intrinsic anisotropy of the crystals impacts the different mechanisms involved.
The goal of this PhD is to model and experimentally characterize the influence of crystal anisotropy on the different mechanisms (topography from polishing, thermal expansion, deformation and diffusion), leading to a proposal of (an) ideal crystal orientation(s) for wafer-to-wafer hybrid bonding pads. The properties and dimensions of the dielectric are important parameters in the mechanical and chemical models you will develop. The experimental techniques will include atomic force microscopy, electron backscatter diffraction and nano-indentation. With the support of the experts at imec, you will work out suitable processes to evaluate your proposal through bonding of 300 mm wafer pairs.
Required background: materials science, physics. The candidate must have a keen interest in modeling. Prior experience with crystallography is appreciated.
Type of work: 50% modeling, 30% experimental work, 20% literature
Supervisor: Ingrid De Wolf
Daily advisor: Joke De Messemaeker
The reference code for this PhD position is STS1712-12. Mention this reference code on your application form.