Immersion Lithography: New Opportunities for Semiconductor Manufacturing Tim Brunner, Dario Gil, Carlos Fonseca and Nakgeuon Seong IBM - SRDC Bob Streefkerk, Christian Wagner and Marco Stavenga ASML
Outline A new paradigm for lithography development DoF trends and likely future Experimental Immersion Litho Imaging Results Immersion Challenges Conclusions 2 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
The NGL Race - Who will make it to the top? EUV EPL Immersion Is Maskless?? 3 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Lithography development as a bicycle race Takes lots of money and effort Overall winning lithography technology must survive many stages Need not win every individual stage! Need both talented individuals and cohesive team to win AMD moving aggressively with this paradigm 4 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
5 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004 01/22/04 Lance Armstrong and the USPS Pro Cycling Team Choose AMD as Technology Sponsor
Immersion Lithography Two Stages NA < 1 NA > 1 Immersion Increases DoF (NA remains same as dry tools) Immersion Increases Resolution Historical lithography evolution applies: What possibilities are enabled compared to dry? Better CD Uniformity Less design restrictions Compared to the current litho-solution, we will get better resolution & lower DoF Simpler masks/ret Better usable resolution? 6 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
First Immersion Systems DoF Increases Aerial Image Measurement (TIS) DRY 7 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
First Immersion Systems DoF Increases Aerial Image Measurement (TIS) Immersion 8 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Future Systems Increased Resolution & NA 9 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Future Systems Increased Resolution & NA 10 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
DoF trends for state-of-the-art Chip Production tools 100000 DoF [nm] 10000 1000 436 365 248 193 193 wet 100 1980 1990 2000 2010 Year 11 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
100000 DoF trends for state-of-the-art Chip Production tools DoF [nm] 10000 1000 436 365 248 193 193 wet 100 10 100 1000 10000 Half-pitch resolution [nm] 12 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Immersion Prototype AT1150i 13 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Imaging Performance of Immersion Scanner Experimental Conditions Exposure System: Full-field scanning immersion prototype (AT1150i) Stage speed: 360 mm/sec Wavelength = 193nm NA = 0.75 4x reduction. Reticle: Alternating-PSM and attenuated-psm Substrate: 300mm Si wafers Materials Stack: ARC: 37nm Resist: 175nm Top-Coat: 41nm 14 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Imaging with Immersion Process Window Improvements 15 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
First IBM Immersion Images Mask: Att-PSM NA = 0.75 Illum: Quasar 117 nm 110nm/220nmpitch 117 nm 110nm/265nmpitch 16 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
altpsm Dry vs. Wet FEM Pitch 250nm, 60nm target DRY WET Critical Dimension (nm) Critical Dimension (nm) Fitted Bossung Curves - Reduced Chi2 = 0.5 100 90 80 70 60-0.2 0 0.2 0.4 Focus (um) Fitted Bossung Curves - Reduced Chi2 = 1.8 110 100 90 80 70 60-0.4-0.2 0 0.2 0.4 Focus (um) Exposure Latitude (%) 12.5 44% DOF increase at 5% exposure latitude 40% increase in total window 15 10 7.5 5 2.5 Process Window DRY WET 0 0 0.1 0.2 0.3 0.4 0.5 DOF (microns) 17 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Theoretical Comparison of Wet and Dry DOF 4 Simplified theoretical treatment Quarter wave OPD, like Rayleigh DOF 2-beam e.g. alt PSM 3-beam e.g. att PSM Biggest improvements appear at smallest grating pitch Wet DOF / Dry DOF 3.5 3 2.5 2 1.5 Wet/Dry Ratio for 193nm Lithography 3beam wet/dry 2beam wet/dry 1 100 200 300 400 Grating Pitch [nm] 18 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Contact Hole Imaging Improvements (130nm Iso) 6 % attenuated Mask, NA/σ=0.63/0.4, 225nm TArF7047 on ARC28 Dry F-0.3 F0.2 Focus [µm] F0.0 F0.50-0.7-0.6-0.4-0.2 0.0 0.2 0.4 0.6 0.7 F-0.4 F0.0 F0.3 F0.0 Wet F0.4 DOF: DOF: 1.2 1.2 µm µm > 2x 2x larger than than dry dry equivalent 19 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Imaging with Immersion CD Uniformity Improvements 20 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Measurement of CD Uniformity Through Focus Dose 10mm 10mm Focus 19X21 chip array exposure measured not measured 10x10mm chips 16 CD-unif. sites per chip Total 147 chips measured 2,352 measurements/wafer 21 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Methodology for Analysis 10mm 10mm 1- Measure 16 sites per chip (pitch = 250nm) 2- Obtain FEM for each of the 16 sites 3- Build model with measured FEM data 4- Use model to calculate CD-variation in the presence of focus errors and offsets. 95 90 Sample fitted FEM for Monte Carlo Model Experiment Sample experimental FEM with corresponding model Printed CD (nm) (nm) 85 80 75 70 65 22 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004 60 0.1 0.2 0.3 0.4 0.5 0.6 Focus (um) Focus (µm)
Contour CD Uniformity Maps as a Function of Mean Focus and Focus Errors DRY nm 20 WET nm 20 Focus Errors (3σ, µm) 0.25 0.2 0.15 0.1 0.05 15 10 5 Focus Errors (3σ, µm) 0.25 0.2 0.15 0.1 0.05 15 10 5 0-0.2-0.1 0 0.1 0.2 Mean Focus (µm) 0 0-0.2-0.1 0 0.1 0.2 Mean Focus (µm) 0 Immersion provides better CD uniformity in the presence of Focus Errors and Mean Focus Deviations 23 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Contour CD Variation Line-Plots as a Function of Mean Focus and Focus Errors CD ACLV Variation (3σ, (3σ, nm) nm) 45 40 35 30 25 20 15 10 DRY DRY 0µ m (3σ) 0.1µm (3σ) 0.15µ m (3σ) 0.20µ m (3σ) 0.30µ m (3σ) CD ACLV Variation (3σ, (3σ, nm) nm) 45 40 35 30 25 20 15 10 WET WET 0µ m (3σ) 0.1µ m (3σ) 0.15µ m (3σ) 0.20µ m (3σ) 0.30µ m (3σ) 5 5 0-0.2-0.1 0 0.1 0.2 Mean Focus (µm) 0-0.2-0.1 0 0.1 0.2 Mean Focus (µm) 24 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
CD Uniformity Control vs Focus Errors CD Variation (3σ, nm) 16 14 12 10 8 6 4 CD Variation vs. Focus Errors dry wet Percent Improvement (%) % CD-Control Improvement w/ Immersion 100 80 60 40 20 2 0 0 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 Focus Errors (σ, µ m) 0 0.1 0.2 0.3 Focus Errors (σ, µm) 25 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Imaging with Immersion Can we reduce RET complexity? 26 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Immersion vs. Dry Assist Performance Simulation 40 70nm Isolated Line 10S PC lithography with att PSM Conclusion Immersion is roughly as effective as SRAF in achieving focus tolerance Error Assumptions : Focus: varied Dose: 3% ( 3σ ) Mask CD: 12nm ( 3σ @ 4X ) Imaging: Water Immersion (n=1.4366) ArF, 0.75NA System Annular illumination Att PSM CD variation [3 sigma in nm] 35 30 25 20 15 10 5 0 SRAF dry No SRAF dry SRAF wet No SRAF wet 0 100 200 300 400 500 Focus variation [3 sigma in nm] 27 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Immersion vs. Dry Assist Performance Experimental Data 110 100 Isolated Line Through Focus Immersion no assists Dry no assists 110 100 Isolated Line Through Focus Immersion w/ assists Dry w/ assists SEM CD (nm) 90 80 70 SEM CD (nm) 90 80 70 60 60 50-0.4-0.3-0.2-0.1 0 0.1 0.2 0.3 0.4 Focus (µm) 110 100 Isolated Line Through Focus Immersion no assists Dry w/ assists 50-0.4-0.3-0.2-0.1 0 0.1 0.2 0.3 0.4 Focus (µm) SEM CD (nm) 90 80 70 60 50-0.4-0.3-0.2-0.1 0 0.1 0.2 0.3 0.4 Focus (µm) 28 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Immersion: Expanded Design Options CD Variation Comparison Dry/Wet K1 0.7 0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 Litho Trend Change k1 NA NA_i 1 0.95 0.9 0.85 0.8 0.75 0.7 0.65 0.6 NA CD Variation (nm) Mask/focus/dose variation 0.25 130 90 65 45 0.55 tech node Pitch (nm) Effective lower NA provides less restriction in design space 29 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Challenges Ahead 30 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Immersion To-Do List Tooling development First production-worthy manufacturing tools Track (new process requirements) Immersion defects (YIELD) Bubbles Particles Processing defects Imaging resist materials dev. Novel top-coats Novel resists/compatibility Leaching and contamination Image quality Process window (DoF improvements) CD Uniformity, simpler RET, DFM Overlay, image placement Modeling High-NA imaging for future immersion systems. Index fluid development Purity and contamination control High-index fluids (for higher resolution) 31 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Bubbles Yes, They do Exist! Sources of bubble formation: Exposure tool Topographic steps between wafer and stage Appropriate fluid flow to prevent cavitation Adequate degassing Surface characteristics of imaging material Minimize outgassing Hydrophobic vs. hydrophilic Imaging effect of a micro-bubble 2 µm Impact of bubbles on imaging: Extent of effect depends on bubble size, location & density. If bubble ~ size of λ Scattering (flare) If bubble >> λ Micro-uniformity effects Pattern magnified by the lower optical path in the bubble 32 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Particles Clean-air vs. clean-water control (state-of-the-art): Air vs. Water Filtration: Meas. Criteria: Particles > 0.1 µm Water is ~ 2,800 time worse Water is ~1000x more dense than air More fluid momentum for particulate transport Evaporation: How to deposit 6,000 100nm-particles on a 300mm wafer: DI water with 100ppt of Silica (not uncommon) Leave 1 µm layer of water left behind Liquid containment method can cause particle formation. 33 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Imaging Materials Challenges Top-coats (if required for immersion): New top-coats must be developed that are nonwater soluble Removal of top-coat needs to be defect free. Water-proof? Solvent strippable Not all 193nm resists work for immersion! Dry exposures exhibit square profiles Water-resistant? Aqueous developer strippable Resist: Resist component extraction should be minimized Loss of CD control Induce lens degradation Absorption of water Minimize LER 1 µm 193nm resist imaged w/ water presoak Low Blur Resists for sub 45 nm resolution Development of high index resists for ultra high NA imaging tools 34 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Water/Imaging Layer Interactions Water Contamination (Leaching) Sub ppm levels of resist components detected at long contact times (60 sec) Water Uptake of 193nm Resist by visible reflectance spectroscopy Thickness = 200 nm Resist Contamination (Water absorption) ppth levels water absorption measured in resists at < 10 sec times intervals What levels of Leaching and Absorption are acceptable for both imaging performance and lens preservation? 35 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
The Challenge of Defects No lithography technology is production worthy until it demonstrates manufacturable defect levels Typically takes several years and cycles of learning 193nm litho: at least 3 years from first beta tool to good wafers out-the-door Learning requires tooling located in a clean environment with integrated tracks. Why so difficult? Modeling random defects is not well understood Statistics of defect measurement requires lots of data Are defects a problem? We will know next year 36 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Conclusions Immersion litho is real, and is here to stay. Simulations and experiments indicate DOF improvement even better than index of water Better CD uniformity through focus Great for device performance and yield! Immersion can simplify RET choices Roughly as effective as SRAF for DOF Less design restrictions Active experimentation in progress. Focus should be on defects & materials development Materials development will be key: Non-water soluble top-coats need to be developed. Resist needs to be optimized for low-defect imaging. DI water quality is very important. 193 immersion litho equipment making very rapid progress: Early tools, suitable for development available this year. Manufacturing tools to be available in 2005. 37 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004
Acknowledgements At IBM: Mike Lercel, George Gomba Carl Larson, Greg Wallraff, Bill Hinsberg & Bob Allen At ASML: Remco Rombeek, Martin Chaplin, Gerard van Reijen & Hans Jansen 38 Intnl. Symposium on Immersion & 157 Lithography, Vancouver, Aug. 2004