In this work, a number of simple oxy-anions have been identified as effective surface-modifying agents to support low-P CMP of both Ta barriers and Cu lines. All these anions tested here were capable of selectively polishing Cu and Ta within the non-alkaline range, pH = 3-7. The Ta polish rates were independent of pH while the Cu polish rates were influenced by the anionic species used and therefore Ta:Cu polish rate tunability was achieved. The Cu removal was facilitated by proton-induced dissolution of Cu. The high Ta polish rates were because of the incorporation of these anions into the Ta2O 5 film, making it structurally weak and removable with low-P. The electrochemical and EDS measurements were performed in order to understand the mechanism of anion-incorporation into Ta2O5 films/particles. The sulfate system was studied in more detail for low-P CMP using both disk and wafer samples.Glycine-oxalic acid-DBSA-H2O2 based 1st step and sulfate-H2O2 based 2nd step slurries were analyzed using 300mm blanket and patterned wafers, which compare favorably to commercially available slurries. A very good dishing performance was observed with 1st step slurry (1wt% glycine + 0.021M oxalic acid + 5wt% H2O2 + 3.5mM DBSA + 3wt% SiO2) at pH 3 with a dishing range of 30nm for 200 x 200 microm features, compared to a dishing of &sim 50 -- 60 nm with a commercially available Cu slurry. This slurry also showed a 2x improvement in the polish time compared to the commercial slurry. A dishing correction of &sim 20 -- 30 nm was observed when the wafer was polished with the 2nd step slurry (0.065M K2SO4 + 1wt% H2O 2 + 8wt% SiO2 at pH 4) after polishing the bulk Cu with commercial slurry. When the 1st and 2nd step slurries were used together for polishing, a dishing value of 15 nm was obtained for 200 x 200 microm features. The optical inspections, SEM inspection, electrical testing results with these slurries are comparable to those of the commercial slurries. However, there is a need to minimize defectivity levels in order to make this slurry combination valuable. This study proved that a careful design of slurry chemistries can remarkably improve performance parameters such as dishing and erosion.2" wafers were polished using sulfate-based slurries varying the abrasive concentration, sulfate content and H2O2 concentration. A dispersion of potassium sulfate and Nexsil 35A (colloidal silica d m &sim 35nm) could be used to achieve tunable Ta:Cu polish rate selectivity of 1 to 62 at pH = 4. The polish rates of Cu vary significantly in the presence of H2O2. The maximum Ta polish rate achieved is &sim270nm min-1 with slurry consisting of 8wt% Nexsil 35A + 0.065M K 2SO4 + 1wt% H2O2 at pH 4 at 2 psi. A reasonable SiO2 polish rate of &sim170nm min-1 was obtained with the same slurry and under the same conditions mentioned above. The post polish surface roughness of Cu, Ta and SiO2 wafers polished with the sulfate-based slurries were in the acceptable range. The anion based dispersions studied here, and particularly the sulfate system, can be potentially useful for 2nd step polishing of Cu/Ta/SiO 2 structures.
展开▼
