Answers are grounded in this paper and cite their sources.
Chemical mechanical planarization (CMP) of copper-filled through-glass vias (TGVs) requires sequentially removing the Cu overburden and Ta barrier while controlling Cu dishing and glass roughness. Electrochemical, dissolution, CMP removal-rate, direct Cu–Ta galvanic, and patterned-TGV polishing measurements were combined to examine how slurry chemistry affects these evolving interfaces. Increasing oxalic acid concentration strongly activated Cu, whereas Ta remained passive, with corrosion current densities below approximately 0.01 μA cm⁻² throughout the acidic ligand-series experiments, providing a basis for Cu/Ta selectivity. Benzotriazole (BTA) inhibition depended strongly on pH. At pH 5, BTA suppressed Cu corrosion and dissolution, whereas at pH 10 it substantially reduced dissolution without decreasing the Tafel-derived corrosion current. Direct coupling measurements demonstrated that isolated Cu–Ta corrosion-potential differences did not reliably predict galvanic current. At pH 10, 10 mM BTA reduced galvanic-current magnitude by approximately 61%. Increasing hydrogen peroxide concentration enhanced bulk Cu removal while Ta removal remained comparatively low. Patterned-TGV polishing revealed a trade-off: alkaline barrier finishing yielded the lowest glass RMS roughness (0.56 nm), whereas acidic barrier finishing yielded the lowest mean Cu dishing (0.37 μm). Effective TGV planarization therefore requires stage-specific control of slurry chemistry and polishing conditions rather than optimization using a single electrochemical or removal metric.
No takes yet. Share an insight, caveat, or question.
Jani et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: