Direct Measurement of Planarization Length for Copper Chemical Mechanical Planarization Polishing (CMP) Processes Using
- PDF / 170,815 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 71 Downloads / 280 Views
DIRECT MEASUREMENT OF PLANARIZATION LENGTH FOR COPPER CHEMICAL MECHANICAL PLANARIZATION POLISHING (CMP) PROCESSES USING A LARGE PATTERN TEST MASK Paul Lefevre, Albert Gonzales, Tom Brown, Gerald Martin, International SEMATECH, Austin, TX; Tamba Tugbawa, Tae Park, Duane Boning, Microsystems Technology Laboratories, MIT, Cambridge, MA; Michael Gostein, Philips Analytical, Natick, MA; John Nguyen, SpeedfamIPEC, Phoenix, AZ. Abstract We have used a large pattern test mask and a specific arrangement of structures on a wafer for direct measurement of an average planarization length for copper chemical mechanical polishing (CMP) processes. We propose new minimum, maximum, and average planarization length definitions, based on up and down area measurements as a function of trench width. The average planarization length is useful for qualitatively comparing the planarization capability of copper CMP processes. We have also performed several experiments that show how the average planarization length depends on polish process settings such as down force and relative speed, as well as on consumables such as pad and slurry. Introduction Copper electroplating introduces severe topography on a die with a large pattern range. In the case where “bottom-up” fill plating technique is used, arrays that have sub-micron line widths and line spaces are overplated, arrays that have high layout pattern densities and fine line spaces are under-plated, and arrays that have large line widths and large line spaces are conformal. This topography introduced by the plating process needs to be planarized before the copper overburden is completely cleared, if we are to avoid severely overpolishing certain structures on the die. Severe overpolishing leads to excessive dishing, erosion, and nonuniformity. The planarization length succintly summarizes the planarization capability of a given CMP process. In copper CMP, the planarization length tells us the extent to which a CMP process can planarize the electroplated topography during the overburden copper clearing stage. The larger the planarization length, the better the planarization capability of the process in question, and the more uniform will be the polishing across different densities within the die. Hymes et al. used wafer scale patterns to determine the planarization length for copper CMP processes [1]. However, they did not detail exactly how the planarization length could be directly obtained from the wafer scale patterns. In this work, we have designed a new wafer-scale mask, and used a specific wafer layout scheme to directly measure the average planarization length of a given copper CMP process. The wafer layout scheme takes into account the nonuniform nature of CMP processes. We also present experimental results that show the effect of polish process settings and consumables on the average planarization length. Large Pattern Test Mask The new wafer-scale test mask is referred to as the SEMATECH/MIT 862 mask. It contains square trenches of widths 1 µm to 8 mm, and arrays with lines and s
Data Loading...
