Effect of Retaining Ring Slot Design & Polishing Conditions on Bow Wave Fluid Dynamics

Transcription

1 Effect of Retaining Ring Slot Design & Polishing Conditions on Bow Wave Fluid Dynamics Xiaoyan Liao 1, Yasa Sampurno 1,2, Adam Rice 1, Fransisca Sudargho 2, Yun Zhuang 1,2, Ara Philipossian 1,2 and Christopher Wargo 3 1 University of Arizona, Tucson, AZ USA 2 Araca, Inc., Tucson, AZ USA 3 Entegris, Inc., Billerica, MA USA 1

2 Outline 1. Background and Objectives 2. UV Fluorescence Technique 3. Experimental Setup 4. Slurry Film Thickness Calibration 5. Data Analysis 6. Summary 2

3 Background Retaining rings are widely used in CMP processes to hold the wafer in place, facilitate slurry transport into pad-wafer interface through slots, and also improve RR WIWNU Bow wave is formed when using rings, which can decrease slurry utilization efficiency Our previous study indicates that ring slot designs have significant effect on the slurry distribution in the pad-wafer interface 3

4 Objectives Develop and use a UV fluorescence technique to measure slurry flow dynamics at the bow wave (UVIZ 100 by Araca Inc.) Investigate the effect of retaining ring slot design and polishing conditions including pad/retaining ring sliding velocity, retaining ring pressure and slurry flow rate on fluid dynamics at the bow wave 4

5 UV Fluorescence Technique Excited State S 1' d 2 Excited State S 1 Energy hv AB 1 Energy hv EM 3 S o Ground Electronic State Jablonski Energy Level Diagram UV Fluorescence Technique 5

6 Experimental Setup Camera UV Lights Polisher = Araca APD 800 (Araca Inc.) Camera shutter time = 0.02 sec Frequency = 5 Hz No. of frames per run = 50 No. of runs for each condition = 2 Carrier Head Politex Pad Retaining Ring 6

7 Example Image Collection Process 7

8 Image Analysis Procedure Carrier Head Raw Image Bow wave Pad Grey Scale Image Cropped Image 8

9 Example of Cropped Image Analysis Procedure Brightness Data of One Image Average Brightness Calculation Mean brightness: 44.5 Standard deviation: 6.4 9

10 Slurry Film Thickness Calibration Curve Slurry Thickness (mm) y = x x R 2 = Mean Brightness 10

11 Experimental Conditions Retaining Ring Designs PEEK-1 and PEEK-2 Sliding Velocities 0.6 and 1.2 m/s Ring Pressures 1.4 and 2.8 PSI Slurry Flow Rates 150 and 300 ml/min Pad Dow Electronic Materials Politex REG Pad Conditioning In-situ conditioning at 3 lb f by 3M PB32A brush Slurry 1 part of Fujimi PL-7103 slurry + 4 parts of DI H 2 O g/l Coumarin (4-Methyl-umbelliferone) Polishing Time 20 seconds 11

12 Retaining Ring Slot Designs PEEK-1 PEEK-2 Retaining Ring for 300-mm Wafer Process 12

13 Example Effect of Retaining Ring Designs 0.6 m/s, 2.8 PSI, 300 ml/min PEEK-1 Mean slurry film thickness: 0.71 mm PEEK-2 Mean slurry film thickness: 0.18 mm 13

14 Bow Wave Slurry Film Thickness Comparison Effect of Retaining Ring Design V1 = 0.6 m/s, V2 = 1.2 m/s, P1 = 1.4 PSI, P2 = 2.8 PSI, Q1 = 150 ml/min, Q2 = 300 ml/min 14

15 Bow Wave Slurry Film Thickness Comparison Effect of Sliding Velocity PEEK-1 PEEK-2 P1 = 1.4 PSI, P2 = 2.8 PSI, Q1 = 150 ml/min, Q2 = 300 ml/min 15

16 Bow Wave Slurry Film Thickness Comparison Effect of Retaining Ring Pressure PEEK-1 PEEK-2 V1 = 0.6 m/s, V2 = 1.2 m/s, Q1 = 150 ml/min, Q2 = 300 ml/min 16

17 Bow Wave Slurry Film Thickness Comparison Effect of Slurry Flow Rate PEEK-1 PEEK-2 V1 = 0.6 m/s, V2 = 1.2 m/s, P1 = 1.4 PSI, P2 = 2.8 PSI 17

18 Summary Slurry film thickness at the bow wave was successfully measured by the UVIZ system using UV fluorescence methods. Results indicated that the retaining ring with the sharp angle slot design (PEEK-1) generated significantly thicker slurry film at the bow wave than its counterpart with the round angle slot design (PEEK-2), as such, we believe PEEK-2 is more efficient in slurry transport. For PEEK-1, slurry film thickness at the bow wave increased with increasing flow rate and ring pressure while it decreased with increasing pad/retaining ring sliding velocity. For PEEK-2, slurry film thickness at the bow wave did not change significantly under different polishing conditions indicating an apparent robustness of the PEEK-2 design to various operating conditions. 18

19 Acknowledgement We wish to express our sincere gratitude to SRC/SEMATECH Engineering Research Center for Environmentally Benign Semiconductor Manufacturing and Fujimi Corporation for their generous support 19