Application Notes - CMP Process Development on Oxide Wafers

Five 2” coupons of similar patterned STI wafers have been test-polished on the same polishing pad. Four process parameters were continuously monitored during polishing, coefficient of friction (COF), contact acoustic emission (AE), pad temperature (PT) and pad wear (PW). Summary plots for all the STI samples are presented in Fig. 1. 

One can see that both friction coefficient (COF) and acoustic emission (AE) data are very repeatable for all the wafers, and the COF signal allows for clear end-point detection for this type of wafers. Pad temperature was quite similar for all the samples, though it reduced slightly from wafers 1 and 2 (26.8 ºC) to wafer 5 (25.9 ºC). Correspondingly, pad wear was higher for wafer 1 (58 micron) than for the rest of the samples (35 to 47 micron).

Two 2” coupons of similar blanket TEOS wafers have been polished, too, on the same pad. The data summary plots are presented in Fig. 2. The data for the blanket wafers are also very repeatable. The AE produced a clearer signal for end-point detection than COF. 

Conclusions

1. The CMP tests have proven the uniquely high repeatability of polishing results on the bench-top polisher mod. CP-4.
2. The CMP tests have proven that simultaneous multisensor monitoring of the friction, acoustic, temperature and wear parameters allows for uniquely sensitive and  comprehensive characterization of polishing process and materials.
3. The CMP tests have proven that the direct monitoring of friction and acoustic parameters with proprietary CETR sensors allows for repeatable and reliable end-point  detection in oxide polishing on both patterned STI and  blanket TEOS wafers.

Fig. 1 Summary CMP Data for Five STI Wafters

Fig. 2 Summary CMP Data for Two TEOS Wafers