The optimization of chemical mechanical planarization process-parameters of c-plane gallium-nitride using Taguchi method and grey relational analysis

• Taguchi-based grey relational analysis for the optimization of CMP of c-plane GaN
• Marked improvement in multiple performance characteristic at the optimum condition
• At optimum condition: Surface Roughness ~ 22 nm and Material Removal Rate ~ 142 nm/h
• The strongest influence on the multiple performance was the slurry pH.
• Taguchi-based grey relational analysis optimization superior to Taguchi method only

In this work, Taguchi-based grey relational analysis (TGRA) was adopted for the optimization of chemical mechanical planarization (CMP) process-parameters of c-plane gallium-nitride (GaN), in a potassium-permanganate/alumina (KMnO4/Al2O3) slurry. The TGRA suggests, the combination of process–parameters—slurry pH = 2, KMnO4 conc. = 0.3 M, Al2O3% = 2.50, down-pressure = 38 kPa, and platen RPM = 90—provides the optimum combination of response variables: material removal rate (MRR) ~ 142 nm/h and root-mean-square surface roughness (RMS) ~ 22 nm. The optimum condition provides a marked improvement in the multiple performance, grey relational grade (GRG), as high as ~ 0.86. The strongest influence on GRG was slurry pH, followed by down-pressure, platen RPM, Al2O3%, and KMnO4 conc. The combination of CMP process-parameters suggested by TGRA provides a superior combination of MRR and RMS compared to the combination of process-parameters suggested by Taguchi—larger is better MRR (TMRR) and Taguchi—smaller is better RMS (TRMS). Comparing optimal condition for TGRA to that of TMRR indicates marginal reduction (~ 8%) in MRR, and significant improvement (~ 21%) in RMS. Similarly, comparison of optimal condition for TGRA to that of TRMS indicates increase in RMS by a miniscule ~ 5% and a dramatic increase in MRR by ~ 48%.

Figure optionsDownload as PowerPoint slide