Modeling and Simulation on 3M Chemical-Mechanical Planarization Products

Modeling and Simulation on 3M Chemical-Mechanical Planarization Products
Jiadi Fan, Kazuma Nuno, David Buckley, Jaimie Stomberg, Matt Fritz, Junqing Xie, Fay Salmon, Uma Rames Krishna Lagudu

3M Company, St. Paul, MN USA

In the chemical-mechanical planarization (CMP) process, the complex interaction between conditioner, pad, and wafer results in a system that can be difficult to predict and relies heavily on large empirical data sets and historical knowledge to achieve an optimized process. To address these challenges, 3M developed the modeling and simulation capability to introduce science into this empirical CMP world so we can accelerate data driven process development, with applications across various 3M CMP products. Among these methods, the finite element method (FEM) simulation and kinematic modeling are introduced, which are essential in understanding the CMP process and enhancing the overall performance and efficiency of CMP operations.
The FEM simulation focuses on understanding the interaction between the CMP pad conditioner tips and the pad material. The simulation investigates how different pad conditioner tip designs affect the stress distribution on the pad. Additionally, the material properties of the tip, such as hardness or elasticity, also influence the force applied to the pad and, in turn, the polishing performance.
On the other hand, kinematic modeling considers how the pad conditioner moves over the surface of the pad. By accumulating the motion of all active tips, the simulation provides insights into the overall pad coverage and uniformity. The kinematic model enables the simulation of various processing parameters—referred to as the “recipe”—including the rotational speed of the pad and conditioner, sweep range, sweep speed, process duration, and the choice of pad conditioners. It helps predict how these parameters influence the evenness of the conditioning process, which is crucial for ensuring uniform planarization across the wafer surface.
The combination of these simulation methods allows for a more robust approach for CMP products development. The FEM simulation offers insights into localized stress and wear patterns, while the kinematic modeling ensures that the conditioning process is optimized for overall uniformity and coverage.
In this conference, we will present the modeling and simulation works on 3M CMP products. We will demonstrate that by accurately predicting the effects of various parameters and conditions, customers can quickly identify optimal settings for different products, accelerating the time it takes to screen and refine CMP recipes. This not only improves the performance of the CMP process but also reduces the time and resources required for testing, leading to faster and more efficient semiconductor manufacturing.