Influences of Solvent and Abrasive in Polishing Slurry on Material Removal, Surface Roughness, and Acoustic Emission Signal of 4H-SiC Si-Face in Chemical Mechanical Polishing

Abstract

This study comprehensively investigates the effects of solvent type (deionized water, ethylene glycol, and their mixture), abrasive content (0.05-0.45 wt.%), and abrasive size (W0.5 to W14) on the material removal rate (MRR), surface roughness (Sa), and AE signal (root mean square, RMS) during the CMP of 4H-SiC Si-face. The experimental approach was complemented by ReaxFF molecular dynamics (MD) simulations to elucidate chemical interactions and finite element analysis (FEA) to model mechanical indentation and material removal. The findings demonstrate that AE signal RMS is sensitive to key slurry parameters and can be correlated with polishing performance under specific conditions. Ethylene glycol, owing to its high viscosity and strong chemical reactivity, yielded the highest MRR (32.5 nm/min) but also the maximum surface roughness (Sa = 2.30 nm). An optimal abrasive content of 0.25 wt.% was identified as the peak concentration, maximizing MRR, Sa, and RMS simultaneously. Increasing abrasive size monotonically increased MRR and Sa, while the AE RMS exhibited a non-linear trend, peaking at W7 size. The correlation coefficient obtained by the fitted curves between AE RMS and MRR under varying abrasive content was R² = 0.94, and between AE RMS and Sa was R² = 0.93. The correlation between AE RMS and polishing performance (MRR, Sa) was strong and positive for variations in abrasive content but was complex and balanced by chemical-mechanical interactions for solvent and abrasive size variations.