Slide 1: EUV lithography NXE platform performance overview
EUV lithography NXE platform performance overview Rudy Peeters 2014 SPIE Advanced Lithography, San Jose CA, 9048-54
Slide 2: Contents
Contents Roadmap NXE:3100 NXE:3300B Summary and acknowledgements
Slide 3: ASML EUV technology roadmap - extendibility
ASML EUV technology roadmap - extendibility Public Extend NA 0.33 to below 10nm Improved lens and illuminator performance Imaging / Overlay performance match node requirements Increased throughput at higher dose
Slide 4: ASML's NXE:3100 and NXE:3300B
ASML’s NXE:3100 and NXE:3300B NXE:3300B shipments started 2013 NA 0.25 0.33 Illumination Conventional 0.8 Conventional 0.9 , 6 off-axis pupil settings Resolution 27 nm 22 nm Dedicated Chuck Overlay / 4.0 nm / 7.0 nm 3.0 nm / 5.0 nm Matched Machine Overlay Productivity 6 - 60 Wafers / hour 55 - 125 W
Slide 5: Contents
Contents Roadmap NXE:3100 NXE:3300B Summary and acknowledgements
Slide 6: Accumulated wafers exposed on NXE:3100
Accumulated wafers exposed on NXE:3100 0 10000 20000 30000 40000 50000 60000 70000 80000 201026 201030 201034 systems 201038 NXE:3100 201042 201046 in 201050 201102 201106 use 201110 operational 201114 at for 201118 201122 201126 201130 201134 cycles 201138 201142 customers of 201146 201150 201202 2
Slide 7: NXE:3100 shows stable performance
NXE:3100 shows stable performance Data courtesy of imec – EUVL Toyama 2013
Slide 8: Contents
Contents Roadmap NXE:3100 NXE:3300B Summary and acknowledgements
Slide 9: EUV (NXE:33x0B) system status overview
EUV (NXE:33x0B) system status overview Multiple NXE:3300B systems fully qualified systems exposing wafers at the customer sites Install has been started for 3 systems 6 more NXE:3300B systems being integrated 4th generation NXE system (NXE:3350B) integration ongoing EUV cleanroom extension is under
Slide 10: Consistent and reproducible NXE:33x0B lens performance
Consistent and reproducible NXE:33x0B lens performance NXE:3300B NXE:3350B NXE:3350B lens Delivered to ASML Lens Every bar is an individual lens NXE:3300B NXE:3350B Data courtesy of Carl Zeiss SMT GmbH Based on Zeiss input and does not include scanner contributions Lens NCE = Non Correctable Error
Slide 11: Contents – Overlay
Contents Roadmap NXE:3100 NXE:3300B – OVERLAY Summary and acknowledgements
Slide 12: Overlay improvements implemented on NXE:3300B
Overlay improvements implemented on NXE:3300B wafer load grid reticle clamp performance flatness wafer clamp flatness lens performance <1nm
Slide 13: Full wafer dedicated chuck overlay of <1.4nm
Full wafer dedicated chuck overlay of <1.4nm Filtered S2F Chuck 1 (S2F) 8 Lot (1.3,1.3) X Y Chuck [nm] 6 Overlay 1.3 1.2 1.4 Dedicated 1.0 1.4 1.3 99.7% x: 1.3 nm 5 nm y: 1.3 nm 0 1 2 3 Day Layers are exposed on the same wafer on the same chuck over multiple days Dedicated chuck overlay is a measure
Slide 14: Good full wafer matched machine overlay performance
Good full wafer matched machine overlay performance on multiple systems (NXE:3300B to immersion) Public Overlay X Overlay Y 5 4.5 3.5 [nm] 3 Machine 2.5 99.7% 10 nm x: 3.4 nm y: 3.3 nm Overlay 2 Matched 1.5 • Grid matching using 1 different reticles Intrinsic design differences 0.5 Projection optics
Slide 15: Contents – Defectivity
Contents Roadmap NXE:3100 NXE:3300B – DEFECTIVITY Summary and acknowledgements
Slide 16: The mask defect challenge
The mask defect challenge Challenging defect requirements on reflective EUV mask without pellicle Slide 16 DUV Reticles EUV Reticles (193nm) (13.5nm) Reflective multilayer Reticle Reticle Absorber Pellicle pattern Transmitted Particle (µm size) Particle (nm size) illumination Reflected illumination
Slide 17: Current defectivity qualification test for particles per reticle pass
Current defectivity qualification test for particles on the front side of the reticle - particles per reticle pass (PRP) Public Particle per Reticle pass test Reticles and wafers are cycled under Load reticle Take reticle exposure conditions without light therefore executable also in into scanner ou
Slide 18: System level testing for Particles Per Reticle Pass
System level testing for Particles Per Reticle Pass (PRP) is done simulating production use during integration at ASML Public [#] pass Multiple NXE:3300B systems Measured at >92nm sensitivity reticle During integration phase actions are per done on the system which can lead to variations in test res
Slide 19: New imaging PRPi test focuses on printability of reticle defects
New imaging PRPi test focuses on printability of reticle defects on wafer level and excludes outside scanner reticle handling Public Expose PRPi Expose PRPi pre reticle cycle wafers post reticle cycle wafers Reticle/wafer Cycles Expose 32nm line- Expose 32nm line-space defectivity space defectivity
Slide 20: Early NXE:3300B test results from PRPi test are promising
Early NXE:3300B test results from PRPi test are promising Early test performed on two systems A using a test reticle (not a production reticle) System Repeated defects (blue dots in the figures) observed already present during initial PRPi exposure Repeated defects are a combination of reticle absor
Slide 21: Progress continues in EUV pellicle development
Progress continues in EUV pellicle development Public Slide 21 EUV Reticles (13.5nm) Reflective multilayer Reticle Absorber pattern Pellicle Particle (µm size) Reflected illumination First prototype full size free-standing pSi pellicle in progress See Carmen Zoldesi 9048-58
Slide 22: Contents – Productivity
Contents Roadmap NXE:3100 NXE:3300B – PRODUCTIVITY Summary and acknowledgements
Slide 23: MOPA Prepulse Source Architecture
MOPA Prepulse Source Architecture Vessel Public Key factors for high source power are: With Collector, Droplet Slide 23 High input CO2 laser power Generator and Metrology High conversion efficiency (CO2 to EUV energy) High collection efficiency (reflectivity and lifetime) Advanced controls Focusing
Slide 24: Technology Demonstration: 70W power using MOPA+PP
Technology Demonstration: 70W power using MOPA+PP Dose margin >3.5x reduced using advanced dose control Slide 24 74 73 72 [W] Power 70 69 Plasma and gas-dynamic forces distort droplet trajectories causing 67 energy instability Controls that compensate for all forces enable closed loop 0 1 2 3 4 5 6
Slide 25: Industrialization: 30W EUV power on a production source
Industrialization: 30W EUV power on a production source NXE:3300B MOPA+PP source operated in automated mode Slide 25 [%] 100 33 90 1.5 80 32 [%] [W] [%] 70 [W] 60 31 1 (Exposures) Power Repro Repro ies (Exposures) 50 Power 30 D ies 40 Dose EUV Dose 0.5 Good 30 Repro 28 10 1.0% Repro 99.9% 0 Calculat
Slide 26: Industrialization: Stable source operation is key
Industrialization: Stable source operation is key Public Dedicated test run:XE:3300B MOPA+PP source operating in automated mode for 6hrs Slide 26 continuously within dose spec. Fully automated for Fab operations 100 +4% Max Error 80 0.5% Error (exposures) 60 99.9% Error 0 Dies Dose 40 4% 20 - Calcul
Slide 27: In-situ Collector Cleaning: Technology Proven
In-situ Collector Cleaning: Technology Proven Significant improvement to COO and Availability Public Reflectivity after cleaning compared to ‘as new’ Measurements on different reflectometers and sample locations Full collector in-situ cleaning capability demonstrated with NXE:3300B compatible config
Slide 28: Contents – Focus & Imaging
Contents Roadmap NXE:3100 NXE:3300B – FOCUS & IMAGING Summary and acknowledgements
Slide 29: NXE:3300B full wafer focus uniformity <12nm
NXE:3300B full wafer focus uniformity <12nm [nm] 20 10.4nm 11.5nm 8.1nm 9.6nm Uniformity 10 Focus 5 The focus uniformity performance test 1 2 3 4 is done in resist System # • Focus sensitive marks are being exposed to obtain full wafer coverage information
Slide 30: 3-day Focus Stability of 2.4nm on NXE:3300B
3-day Focus Stability of 2.4nm on NXE:3300B 10 8 Focus Stability: (nm) 4 Day 1 Day 3 • 3-day Mean Focus range over both 2 chucks Focus 0 Chuck 1 • Per wafer 16 fields are exposed Mean -2 Chuck 2 -4 Day 2 • 3-day Focus Stability of 2.4nm -6 -8 -10 0 1 2 3 4 5 6 Wafer
Slide 31: Multiple systems show NXE:3300B CDU below 1.6nm
Multiple systems show NXE:3300B CDU below 1.6nm dense and iso lines exposed at 16mJ/cm2 Public System A System B System C System D 22 nm DL Full wafer CDU <1.2nm 22 nm IL Full wafer CDU < 1.6nm CD measurements performed with YieldStar
Slide 32: 22nm dense and iso L/S with exposure latitude >14%
22nm dense and iso L/S with an exposure latitude of >14% Conventional illumination BE = 16 mJ/cm2 LWR = 3.8nm Exposure Latitude > 14% DoF > 100nm Conventional illumination BE = 16 mJ/cm2 LWR = 3.4nm Exposure Latitude > 14% DoF > 100nm No post processing done
Slide 33: 24nm regular contact holes with large process window
24nm regular contact holes (~18nm node DRAM) with conventional illumination Large process window and 1.2nm FWCDU Process Window Full wafer CD uniformity Exposure latitude 18% Full wafer CDU: 1.2nm 3 DoF > 120nm Systematic fingerprints dominated by reticle and process Yieldstar measurements
Slide 34: NXE:3300B proximity matching performance
NXE:3300B proximity matching performance 2 2 System 1 - H System 2 - H System 1 - V System 2 - V 1.5 1.5 System 3 - H System 4 - H System 3 - V System 4 - V 1 1 [nm] 0.5 [nm] 0.5 delta 0 delta 0 CD -0.5 CD -0.5 -1 -1 -1.5 Horizontal 22nm L/S -1.5 Vertical 22nm L/S -2 -2 44 57.25 66 88 154 44 57.25 6
Slide 35: NXE platform offers off-axis illumination to enhance process window
The NXE platform offers new concept off-axis illumination to enhance process window Public (%) 8 latitude 6 Exposure 4 0 20 40 60 80 100 120 140 Flexibel and programmable illumination setting Depth of focus (nm) mechanism using fly’s eye integrator Simulations by Tachyon SMO NXE See Xiaofeng Liu 904
Slide 36: 16nm dense lines with >10% exposure latitude on NXE:3300B
16nm dense lines with >10% exposure latitude on NXE:3300B (dipole-90 setting) Public 16nm H dense lines 8 (%) EL 6 16nm L/S 4 Dipole 90Y 0 20 40 60 80 100 120 DoF (nm) Focus [nm] 50 25 0 -25 -50
Slide 37: Good logic tip-to-tip and tip-to-line performance at low doses
Good logic tip-to-tip and tip-to-line performance at low doses Public Tip-to-tip Tip-to-line 60 60 55 Conventional 55 Conventional 50 50 Quasar-45 [nm] 45 Quasar-45 45 40 [nm] 40 CD 35 CD 35 30 31nm gap 30 25 27nm gap 25 T2T 20 T2L 20 20nm 15 15 19nm 10 10 12 14 16 18 20 22 14 16 18 20 22 24 Dose [m
Slide 38: EUV single-exposure enables aggressive shrink on 2D logic
EUV single-exposure enables aggressive shrink on 2D logic shrink possible beyond 10nm Logic node requirement 27nm Tip to Tip Tip to Line 19nm Line/Spaces 16nm Logic Node Actual Requirements per node
Slide 39: Random logic metal layer use case
Random logic metal layer user case NXE:3300B enables single exposure with large DoF -80nm EUV Node: Logic 10nm -60nm Single Exposure -40nm Conventional illumination -20nm DoF 120nm 0nm focus 20nm 40nm Position in the exposure slit -12mm 0mm +12mm 60nm Good print performance over the full exposure sl
Slide 40: EUV single exposure 10nm Logic metal layer
EUV single exposure 10nm Logic metal layer 45nm pitch, Conv. 45nm pitch, Conv. Public Dose ~45mJ/cm2 Dose ~20mJ/cm2 2013 • Low dose resist focus • Conventional setting Feb • DOF of 60-80nm
Slide 41: EUV single exposure 10nm Logic metal layer
EUV single exposure 10nm Logic metal layer 45nm pitch, Conv. 45nm pitch, Conv. Public Dose ~20mJ/cm2 Dose ~20mJ/cm2 Low dose resist Tachyon NXE OPC+ focus Conventional setting DOF of 100nm
Slide 42: EUV single exposure 10nm Logic metal layer
EUV single exposure 10nm Logic metal layer 45nm pitch, Conv. 45nm pitch, Q45 Public Dose ~20mJ/cm2 Dose ~20mJ/cm2 Low dose resist NXE OPC+ Quasar 45 setting DOF 120nm
Slide 43: Excellent process window for 10nm Logic metal layer
Excellent process window for 10nm Logic metal layer Low dose, off axis illumination, NXE OPC+ 45nm pitch, Q45 Public 45nm pitch, Conv. Dose ~45mJ/cm2 Dose ~20mJ/cm2 > factor 2 dose focus reduction
Slide 44: Contents
Contents Roadmap NXE:3100 NXE:3300B Summary and acknowledgements
Slide 45: Summary
Summary 27nm 19nm Slide 45 NXE:3100 in use for process and device development at customers NXE:3300B exposing wafers at customer sites Multiple NXE:3300B system qualified and showing good imaging and overlay performance Performance fit for customer development 10nm Logic 70W source power and sub-20n
Slide 46: Acknowledgements
Acknowledgements The work presented today, is the result of hard work and dedication of teams at ASML, Cymer, Zeiss, and many technology partners worldwide including our customers Special thanks to our partners and customers for allowing us to use some of their data in this presentation