PhD position in mechanics: Modeling of 3-D chip bonding

Die-to-wafer hybrid bonding is a 3-D vertical integration technique used to build advanced computer chips. Imagine you are carefully placing a tiny tile (the die) onto a large floor (the wafer). Both the tile and the floor have special layers and tiny copper pads that help them stick together and work as one. When making the chip, the small die is brought close to the wafer, and a gentle pressure is applied to start the sticking process. This creates an initial spot of contact, like the first drop of glue holding two pieces together. From there, the bond spreads outwards in a wave, connecting the rest of the surfaces (bond-wave propagation). This method allows manufacturers to choose the best tiny chips and put them together on a bigger surface, making it easier to design powerful and reliable electronics. It’s a bit like assembling a puzzle where each piece can be made perfectly before being joined to complete the picture.

However, the die-to-wafer bonding process presents significant scientific and engineering challenges. Mechanical stresses and bonding dynamics can result in bonding distortions, die stretching, and overlay errors—misalignments that may degrade electrical performance or cause device failures. As device features shrink and dies or wafers become thinner, these effects intensify, making it essential to understand and predict die-to-wafer bonding mechanics to ensure precise alignment and reliability of 3D stacks.

A key challenge in this process is the large number of parameters that impacts distortions, such as: wafer & die properties, process dynamics and tool design. Given the complexity and cost of experimental optimization, simulation-based methods are essential for effective process development.

The purpose of this PhD topic is to deepen the understanding on both the physics of bond-wave propagation and the impact of mechanical boundary conditions on bonding distortions using modeling and simulation techniques. To achieve this, the PhD student will:

1. Develop a 2-D axisymmetric physics-based mechanical modeling environment (coupled fluidic-structural) to study the bonding phenomena.
2. Extend the models to full 3-D die-to-wafer bonding simulations
3. Apply the learnings from physics based models to advance accelerated simulation strategies already available at imec, enabling rapid data acquisition from models
4.  Implement modeling results as training data to efficient machine learning algorithms to create in-situ prediction/optimization toolboxes

This position is best suited for candidates from an engineering background with a keen interest in physics-based mechanical modeling and an affinity for cross-team collaborations.

Reading:

1. Y. Lu et al., "Investigation of Post-Bonding Die Stretching in Die-to-Wafer Hybrid Bonding," 2025 Symposium on VLSI Technology and Circuits (VLSI Technology and Circuits), Kyoto, Japan, 2025, pp. 1-3, doi:
2. O. O. Okudur, S. Iacovo, S. Kang, M. Gonzalez and E. Beyne, "Simulations of Wafer-to-Wafer Bonding Dynamics and Deformation Mechanisms," 2024 IEEE 10th Electronics System-Integration Technology Conference (ESTC), Berlin, Germany, 2024, pp. 1-5, doi: 10.1109/ESTC60143.2024.10712136.
3. https://www.youtube.com/watch?v=ky0-JlfuuM8
4. https://www.youtube.com/watch?v=2ACiuKgYUkI