Full Scale Multi-physics CMP Modeling

The science behind CMP is not well understood which makes it tough to predict the multi-scale defects that are common during any industrial polish. This project studies the wafer scale phenomena, namely the slurry hydrodynamics, the contact mechanics between the wafer and the pad, and abrasive wear due to the particles, to predict the wafer-scale defects, such as within wafer non uniformity, waviness and high TTV, through experimentally validated computational modeling. The simulation shows the in silico evolution of a wafer during polishing where wear is shown on the left and fluid pressure contours are shown on the right, as the wafer seeks force and moment equilibrium.
PFTL Research Assistant(s):   Gagan Srivastava
Method(s) Employed:   Mixed Lubrication Modeling, Chemical mechanical polishing, Optical profilometry,
Rig(s) and/or Software(s) Employed:   GnP Poli300 Polisher, Zygo New-View 7300 Optical Interferometer, Amscope 520T Microscope, Radwag XS220 Microbalance, MATLAB, Mathematica, Paraview, QtiPlot

Sample Results:

Select PFTL References:

Srivastava, G., and Higgs III, C. F., "A Fast, Experimentally-validated, Particle Augmented Mixed Lubrication Framework to Predict CMP", Materials Research Society Symposium Proceedings, 2013

Srivastava, G., and Higgs III, C. F., "A wafer-scale particle augmented mixed lubrication modeling approach for chemical mechanical polishing", STLE Annual Meeting and Exhibit, 2013