Nitride removal

Park J-G, Katoh T, Lee W-M, Jeon H, Paik U. Surfactant effect on oxide to nitride removal selectivity of nano-abrasive ceria slurry for chemical mechanical polishing. Jap Journ App Phy 2003 42 5420-5425. 

Hansen et al. [23] examined the major factors that influence the (oxidemitride) selectivity. These results were obtained in order to determine the optimum processing conditions for maximizing the (oxidemitride) selectivity. It is important to point out that the oxide removal rate and selectivity (oxidemitride) has a good correlation with the action term (downforce x linear velocity) as shown in Fig. 13.11. The nitride removal rate and nonuniformity as a function of action term is shown in Fig. 13.12. The average nitride removal rate is 83 A/min and is... 

FIGURE 13.12 Nitride removal rate (closed diamonds) and nitride nonuniformity (closed squares) as a function of the action term. The solid line is the average removal rate with a value of 83 A/min. 

The slurry in Fig. 13.19 consists of 2wt% of L-proline and 1 wt% of ceria with pH adjusted to the desired value with either HCl or KOH. The nitride rate is relatively constant from pH 6-8 while the oxide rate gradually increases with pH. The nitride removal rate drops dramatically to a rate of roughly 1 nm/min as the pH increases from 8 to just below 10. The nitride rate rapidly increases from pH 10 to its highest value (with proline additive) at pH 11. Fig. 13.20 shows the selectivity of oxide to nitride in the proline-ceria slurry that has a linear behavior over the pH range of 6-8, has a very large rise from pH 8 to just below pH 10, and then falls quickly from pH 10 to pH 11. 

FIGURE 15.12 The result of CMP field evaluation (a) profiles of PETEOS removal rate (b) profiles of nitride removal rate. ... 

Strip Nitride, Remove Oxide in contact window... 

Figure 3.5a shows the result of CMP field evaluation. Average PETEOS removal rate of slurry A was 2883A/min and B was 672A/min. The within-wafer non-uniformity (WIWNU) shows that ceria slurry B (0.7%) is better than ceria slurry A (1.9%). Average nitride removal rate of slurry A was 5lA/min and B was 44A/min as shown in Figure 3.5b. Thus, oxide-to-nitride selectivity was 56 for ceria slurry A and 15 for ceria slurry B. CMP field evaluation of ceria slurries having different crystallinity showed...

A (nitride removal rate) 74 C (nitride removal rate) 73... 

To enhance the removal selectivity of Cu-to-TaN films wifh suppressing the removal rate of TaN film by selective adsorption, we also optionally added organic polymer (PAM) with the concentration of up to 0.7 wf%. The adsorption of PAM-added slurry on the abrasive particles and the film surfaces corresponds to the differing zeta potential charge. By this zeta potential difference, the removal rate of the Cu and TaN films was more suppressed, and the oxide-to-nitride removal selectivity increased with addition of PAM. 

America, W.G., Babu, S.V., 2004. Slurry additive effects on the suppression of silicon nitride removal during CMP. Electrochem. Solid-State Lett. 7 (12), G327—G330. 

Ceria (Ce02)-based slurries originally used for CMP of silicate glasses in optics for many decades are now applied for semiconductor fabrication with high polishing performance, MRRs and surface quality over the last few years. Compared with silica-based slurries they show a superior Si oxide-to-nitride removal rate selectivity which can be influenced by using additives like amino acids (America and Babu, 2004 Carter and Johns, 2005 Praveen et ah, 2014). 

---