Experimental characterization and dynamical modeling evaluation for enhanced BTA removal by three amino acids in post-Cu-CMP cleaning

Post-chemical mechanical polishing (CMP) cleaning is an indispensable process to remove organic residues and particles on the wafer surface after CMP. Benzotriazole (BTA), a dominant corrosion inhibitor in Cu-CMP slurry, can easily remain on the copper surface after CMP, affecting the performance and yield of devices, and its removal has attracted more attention. This paper revealed the effect and mechanism of three amino acids, lysine, serine, and proline, on BTA removal in tetraethyl ammonium hydroxide-based alkaline solutions during post-Cu-CMP cleaning. Experimental results showed a diminishing enhancement in the cleaning efficiency of BTA after adding lysine (97.1%), serine (96.7%), and proline (94.1%) with almost the same concentration. Ultraviolet–visible absorption spectrum proved the lysine’s complexing ability was slightly better than the other. Besides, quantum chemical parameters of amino acid molecules calculated by density functional theory further suggested that lysine has more reactive sites and mainly bonds with copper through N2 and N8. The adsorption structures of three amino acids were simulated through the molecular dynamics method, and lysine turned out to be parallel absorbed on the Cu (1   1   1) surface with the lowest adsorption energy of −30.24 eV. Finally, the cleaning process simulation revealed the dynamic behaviors of BTA and lysine.