Analyses of Diamond Disk Substrate Wear and Diamond Microwear in Copper Chemical Mechanical Planarization Process

Abstract

Diamond disk substrate wear and diamond microwear in the copper chemical mechanical planarization process were investigated in this study. Three types of disks (D1, D2, and D3) made by three different manufacturers were analyzed. For each type of disk, 24 h static etch tests were performed with Fujimi PL-7103 and Cabot Microelectronics Corporation iCue 600Y75 slurries at 25 and . Scanning electron microscopy (SEM) analysis showed that there was no appreciable microwear on the diamond after the static etch tests for all three types of disks. Disks D1 and D3 showed no appreciable corrosion on the diamond disk substrate for both slurries at both temperatures. In comparison, disk D2 showed apparent surface corrosion using the Fujimi PL-7103 slurry at 25 and and the Cabot Microelectronics Corporation iCue 600Y75 slurry at . Inductively coupled plasma-mass spectroscopy (ICPMS) analysis was performed before and after the static etch tests to investigate metal concentration increases in the slurry due to diamond disk substrate corrosion. The ICPMS analysis was consistent with the SEM images, showing a significant Ni concentration increase in the slurry for disk D2 with the Fujimi PL-7103 slurry at 25 and and the Cabot Microelectronics Corporation iCue 600Y75 slurry at . In addition to the above static etch tests, 24 h wear tests were performed on each type of diamond disks with Fujimi PL-7103 and Cabot Microelectronics Corporation iCue 600Y75 slurries at two different platen temperatures (25 and ). SEM analysis was performed on selected aggressive and inactive diamonds as well as on the surrounding disk substrate before and after the wear tests. SEM images showed that there was microwear on the cutting edges of the aggressive diamonds for disks D1 and D3 with both slurries at 25 and . For disk D2, there was microwear on the cutting edges of the aggressive diamond with the Fujimi PL-7103 slurry at and with the Cabot Microelectronics Corporation iCue 600Y75 slurry at 25 and , and the aggressive diamond broke off from the disk substrate with the Fujimi PL-7103 slurry at . The SEM images also showed that there was no microwear on the inactive diamond for all three types of disks with both slurries at 25 and , confirming that the inactive diamonds did not participate in regenerating pad asperities during conditioning. The pad thickness profile was measured after the wear tests, and the effect of platen temperature on pad wear rate was investigated.