Slurries for Copper Damascene Patterning: Similarities and Differences
- PDF / 120,967 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 2 Downloads / 210 Views
F6.8.1
Slurries for Copper Damascene Patterning: Similarities and Differences A. Jindal, G. Rajagopalan, M. Gupta, J.-Q. Lu, K. Rose, and R.J. Gutmann Focus Center-New York, Rensselaer: Interconnections for Gigascale Integration Rensselaer Polytechnic Institute, Troy, New York-12180 ABSTRACT Two first-step copper damascene slurries and one commercial second-step slurry are characterized in terms of their intrinsic properties and CMP performance. A prototype first-step slurry with high static etch rate (~150 nm/min) yielded higher dishing in the copper lines (~200 nm in 100 µm lines) compared to a commercial first-step slurry with negligible static etch rate. In both the cases, dishing in copper lines is observed to be a strong function of line width and radial position on the wafer. High static etch rate of the prototype slurry is believed to be responsible for the high dishing. Non-selective second-step polishing removes the liner layer while maintaining planarity. INTRODUCTION Single or dual damascene patterning is commonly used for copper-based multilevel interconnects. Two main issues involved with copper damascene patterning are dishing of the copper lines and erosion of the dielectric in the field region. To overcome these issues, copper CMP is performed in two separate steps. The first-step CMP removes copper overburden at high rates (>300 nm/min). The second-step CMP could be either selective (higher removal rates for liner than for copper and dielectric) or non-selective (comparable removal rates for copper, liner, and dielectric) depending upon the specific application [1-2]. The overburden copper is completely removed in the first-step if a selective second-step slurry is used to remove the liner layer and to buff the oxide. Most, but not necessarily all, of the copper overburden is removed in the first-step CMP if a non-selective second-step slurry is used. A non-selective second-step slurry can maintain better planarity at the expense of overall thickness control. In this paper two first-step slurries (P1 and C1) have been evaluated, along with one nonselective second-step slurry (C2), for their performance in single level copper damascene patterning. A non-selective slurry is chosen as it has been a baseline second-step slurry for our 3D via-chain test structure patterning [3, 4]. Wafers with single level copper damascene structures in tetraethyl orthosilicate (TEOS) oxide and a proprietary low-k dielectric material have been used (wafers obtained from Texas Instruments and International SEMATECH). The metrics on which the slurries have been evaluated are slurry (particle size, size distribution, zeta potential, iso-electric point (IEP), and long and short term slurry stability) and CMP (removal rates, dishing, surface quality and non-uniformity) characteristics. The primary objective of this paper is to evaluate intrinsic slurry properties and blanket film removal rates and characterize single level copper damascene patterning. EXPERIMENTAL PROCEDURES A brief description of slurry characterization te
Data Loading...
