| 摘要 |
It is arranged such that the least common multiple of two numbers m and n of which one is prime to the other, is made as large as possible where the number m is the rotational speed (rpm) of a platen with a polishing pad affixed thereto and the number n is the rotational speed (rpm) of a carrier with a wafer mounted thereon. As a result of such arrangement, it is not until the platen completes m revolutions that a point on the polishing pad that comes into contact with a fixed point on the wafer returns to the original contact point with the fixed point at the start of polishing, and the resulting trajectory is therefore spread uniformly over the polishing pad. Each point on the wafer is brought into contact with most surface regions of the polishing pad, therefore preventing the wafer from undergoing deterioration in planarity uniformity due to a particular point on the wafer, on one hand, frequently coming into contact with low polishing-rate regions in the polishing pad and due to the other points on the wafer, on the other hand, less frequently coming into contact with the regions.
|
