Critical issue of non-topcoat resist for ultra low k 1 lithography

Transcription

1 Critical issue of non-topcoat resist for ultra low k 1 lithography 1 Hirokazu Kato, 1 Tomoya Oori, 1 Koutaro Sho, 1 Kentaro Matsunaga, 1 Eishi Shiobara, 1 Tsukasa Azuma, 2 Yukio Nishimura, 2 Hiroki Nakagawa, 2 Yoshikazu Yamaguchi, 3 Naoko Shirota, 3 Osamu Yokokoji, 4 Tomoharu Fujiwara, 4 Yuuki Ishii and 1 Shinichi Ito 1 Toshiba Corporation 2 JSR Corporation 3 Asahi Glass Co., Ltd. 4 Nikon Corporation 5th_International Symposium on Immersion Lithography Extensions

2 Background Effect of hydrophobic additives Basic study scan performance Practical study imaging performance Conclusion 2

3 Background Effect of hydrophobic additives Basic study scan performance Practical study imaging performance Conclusion 3

4 Beyond the red brick wall 120 Single Exposure Double Patterning k 1 >0.25 <0.25 Exposure Once Twice or more 100 λ0 HP = k1 NA Cost Half pitch (nm) Logic Flash k 1 = 0.25 High Index Fluid 20 0 Double Patterning (Ultra low k 1 ) NA NA = 1.44 TOO EXPENSIVE! 4

5 Issue of low cost process Throughput of scanner Throughput 15% UP Scan speed of wafer stage Current immersion process Meniscus broken Blister defect Hydrophobicity is not enough High scan speed Film surface should be more hydrophobic for high-speed scan! 5

6 Candidates of immersion process Developer soluble Current Topcoat process Developer insoluble Non-topcoat process Best candidate Hydrophobicity (Throughput) Process steps? Lithography performance Non-topcoat process is the best candidate! 6

7 Issues of non-topcoat process Trade-offs? Scan performance Hydrophobic additives Imaging performance Resist Trade-offs? Non-topcoat resist Defectivity Trade-offs? Can non-topcoat process satisfy all issues? 7

8 Procedure of our study Scan performance (Including immersion defectivity) Imaging performance Basic study Dynamic RCA Hysteresis Sliding angle Surface roughness Leaching Practical study Blister defects Bubble defects Pattern profile LWR Pattern collapse Sensitivity DOF Defectivity Zeta potential Satellite defects 8

9 Background Effect of hydrophobic additives Basic study scan performance Practical study imaging performance Conclusion 9

10 Sample conditions Base resist + Hydrophobic additive Non-topcoat resist No. Additive Mw Conc. CA Low A High Low B High C Unit of concentration: arb. units 10

11 Measured parameters Pin nozzle Scan Scan #1 Dynamic receding contact angle (DRCA) θ Scan #2 Tail length of water drop (Tail) 11

12 Measured parameters PAG θ Scan #3 Leaching Scan #4 Hysteresis (Hys) Scan #5 Sliding angle (SA) Scan #6 Surface roughness (Ra) Defectivity #1 Zeta potential (Zeta) 12

13 Dynamic RCA Higher concentration Higher molecular weight DRCA (degree) Concentration of additives (arb. units) Current target DRCA > 60 deg. A (Low Mw) A (High Mw) B (Low Mw) B (High Mw) C 13

14 Zeta Potential 0-5 Additive B -10 Zeta potential (mv) Additive C Additive A Concentration of additives (arb. units) Low defectivity A (Low Mw) A (High Mw) B (Low Mw) B (High Mw) C 14

15 Correlation of parameters Decision Parameter R 2 (square of correlation parameter R) Strong correlation Tail length Leaching CA Group A DRCA SA Hys Group B Ra Zeta Group C ~ ~ ~0.6 <0.5 Weak correlation Measurement of DRCA is recommended. (Group A) Hysteresis can be smaller with larger roughness. (Group B) Middle or weak correlation with group A. (Group B) Zeta potential is independent. (Group C) 15

16 Summary of basic study Base resist + Hydrophobic additive Non-topcoat resist No. Additive Mw Conc. CA Low A High Low B High C No. 4 shows the best balance in scan performance! Also in imaging performance? 16

17 Background Effect of hydrophobic additives Basic study scan performance Practical study imaging performance Conclusion 17

18 Cross sectional profile 43 nm L/S Attenuated PSM A2 A4 B2 B4 C2 Conc. up TC Large pattern deformation Conc. up pattern height Additive No. Sample TC Conc. Structure Mw (arb. units) 3 A2 w/o A High 2 4 A4 w/o A High 4 7 B2 w/o B High 2 8 B4 w/o B High 4 10 C2 w/o C TC w None - - by higher concentration by additive A or B 18

19 Litho. performance DOF & Sensitivity No. Sample DOF Sensitivity at 12%EL (mj/cm 2 ) (um) 3 A2 > A4 > B2 > B4 > C2 > TC > Change in DOF and sensitivity was small LWR (nm) LWR nm L/S Attenuated PSM A2 A4 B2 B4 C2 TC Trade-off exists between scan and imaging particularly in pattern profile and LWR. 19

20 Model C2: 5.71 nm A4: 8.53 nm 65 DRCA (degree) A2: 6.46 nm B4: 8.62 nm Additive C 45 B2: 6.59 nm Concentration of additives (arb. units) A (Low Mw) A (High Mw) B (Low Mw) B (High Mw) C Additive A or B Is segregation a key parameter? 20

21 Defectivity No. Sample Dynamic Zeta RCA potential 10 C TC Defectivity is almost at the same level. C2 DD 0.31 cm -2 TC DD 0.28 cm -2 21

22 Summary Non-topcoat resist process is a candidate for ultra low k 1 lithography. There are three issues, scan performance, imaging performance and defectivity for non-topcoat resist. Segregation of additives may be a key parameter for trade-off between scan performance and imaging performance. It is possible to keep balance on three issues. Non-topcoat resist process is a powerful candidate for ultra low k 1 lithography. 22

23 23