Asml Holding Nv · EURONEXT:ASML
Asml Holding Nv — Investor Day (November 2003)
50 slides from Asml Holding Nv's investor day presentation, Nov 13, 2003
Slide 1: The Roadmap: Shrink Will Continue Using Cost-Effective Lithography
The Roadmap: Shrink Will Continue Using Cost-Effective Lithography Martin van den Brink Executive Vice President Marketing & Technology Analyst Day, November 13, 2003 / Slide 1
Slide 2: Safe Harbor
Safe Harbor “Safe Harbor” Statement under the U.S. Private Securities Litigation Reform Act of 1995: the matters discussed during this presentation include forward-looking statements that are subject to risks and uncertainties including, but not limited to, economic conditions, product and pricing,
Slide 3: Agenda
Agenda l The shrink roadmap of our customers l Lithography shrink scenarios l Wavelength reduction l Aperture increase enabled by immersion l Lowering the k1 factor l Special topics l Immersion l Maskless l Likely shrink scenarios l Conclusions / Slide 3
Slide 4: Logic and DRAM half-pitch roadmap
Logic and DRAM half-pitch roadmap 250 Logic 200 DRAM ITRS Resolution (nm) 150 100 50 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Time / Slide 4
Slide 5: IC industry will continue to scale transistors
IC industry will continue to scale transistors Source: Intel / Slide 5
Slide 6: The shrink roadmaps of our customers
The shrink roadmaps of our customers l The shrink continues l Significant variation in ambition of customers: node names become marketing names l DRAM continues to take the lead in density scaling driven by cost reduction l Logic will quickly follow DRAM in density and cost scaling but is challenged
Slide 7: Agenda
Agenda l The shrink roadmap of our customers l Lithography shrink scenarios l Wavelength reduction l Aperture increase enabled by immersion l Lowering the k1 factor l Special topics l Immersion l Maskless l Likely shrink scenarios l Conclusions / Slide 7
Slide 8: How to shrink in lithography
How to shrink in lithography l Resolution = k1. = k1. n.sin NA l Resolution: half pitch l k1: process factor, k1 < 0.25 no contrast l : wavelength of the exposure light in vacuum l n: refractive index of the medium above wafer l : maximum angle of the light incident on wafer l NA: numerical aperture
Slide 9: Image contrast reduces by lowering the resolution
Image contrast reduces by lowering the resolution Conventional illumination, NA, k1 ~ 1.25 Iret Mask xret Pupil/Lens Iwfr Wafer xwfr / Slide 9
Slide 10: Image contrast reduces by lowering the resolution
Image contrast reduces by lowering the resolution Conventional illumination, NA, k1 ~ 0.50 Iret Mask xret Pupil/Lens Iwfr Wafer xwfr / Slide 10
Slide 11: Image contrast reduces by lowering the resolution
Image contrast reduces by lowering the resolution Conventional illumination, NA, k1 ~ 0.30 Iret Mask xret Pupil/Lens Iwfr Wafer xwfr / Slide 11
Slide 12: Modulation improves by increasing aperture
Modulation improves by increasing aperture Conventional illumination, 2NA, k1 ~ 0.60 Iret Mask xret Pupil/Lens: 2NA Iwfr Wafer xwfr / Slide 12
Slide 13: How do k1, wavelength, and aperture work together?
How do k1, wavelength, and aperture work together? Resolution = process factor k1 x wavelength / NA (half pitch) NA 365 nm 248 nm 193 nm 157 nm i-line KrF ArF F2 Method: process factor k1 0.65 0.63 > 0.80 0.63 > 0.95 0.70 > 0.95 Conventional Stepper 0.60 335 nm 235 >185 nm + Off axis illumination 0.
Slide 14: Wavelength reductions are costly because of changed requirements
Wavelength reductions are costly because of changed requirements for: l Photoresist - requiring new resist and process infrastructure after every wavelength change l The optical projection system - requiring new material development: Quartz for 248 nm, CaF2 for 193 and 157 nm l Mask materials - requ
Slide 15: Sub 90-nm ArF ready for production implementation
Sub 90-nm ArF ready for production implementation AT:1200, 193 nm, 0.85 NA, binary mask 80 nm Binary Mask, NA/σ = 0.85/0.88-0.58, 225 nm TArF60111 on ARC28 -0.25 µm -0.20 µm -0.10 µm 0.0 µm +0.10 µm +0.2 µm 0.45 µm focus range 70 nm Binary Mask, NA/σ = 0.85/0.93-0.69, Dipole, 175 nm TArF60111 on ARC
Slide 16: 157-nm equipment available, but processing still needs maturing
157-nm equipment available, but processing still needs maturing Micrascan VII, 157 nm, 0,75 NA, binary mask Dipole, XP2452 resist, 80-nm L&S through focus in 50-nm steps Depth of Focus > 0.4 µm / Slide 16
Slide 17: Agenda
Agenda l The shrink roadmap of our customers l Lithography shrink scenarios l Wavelength reduction l Aperture increase enabled by immersion l Lowering the k1 factor l Special topics l Immersion l Maskless l Likely shrink scenarios l Conclusions / Slide 17
Slide 18: Increase in aperture = increase in lens complexity
Increase in aperture = increase in lens complexity Camera lens complexity scaling with NA at fixed focal length (all 50-mm lenses) F = 1.8 F = 1.4 F = 1.0 NA = 0.28 NA = 0.35 NA = 0.5 6 lens elements 7 lens elements 11 lens elements 130 gram 290 gram 985 gram 100 400 3000 / Slide 18
Slide 19: Increase in aperture = increase in lens complexity
Increase in aperture = increase in lens complexity 12 Volume expected according geometrical scaling Volume for field size 17 mm*33 mm Volume released or expected Quartz Volume (arbitrary units) 10 Air Water 8 6 4 2 0 NA 0.63 0.75 0.85 >0.90 1 >0.90 i 1.1 i 1.2 i 1.3 i 1.43 i n air n water / Slide 19
Slide 20: Technology reduces growth in lens complexity
Technology reduces growth in lens complexity l Macro asphere technology reduces the number of lens elements and their size l Immersion shifts the complexity explosion to NA = n-liquid (the refractive index of the liquid) l Catadioptric lens designs ease color correction l Reducing the field size fur
Slide 21: Agenda
Agenda l The shrink roadmap of our customers l Lithography shrink scenarios l Wavelength reduction l Aperture increase enabled by immersion l Lowering the k1 factor l Special topics l Immersion l Maskless l Likely shrink scenarios l Conclusions / Slide 21
Slide 22: Optimizing illumination improves resolution
Optimizing illumination improves resolution Conventional illumination, NA, k1 ~ 0.30 Iret Mask xret Pupil/Lens: NA Iwfr Wafer xwfr / Slide 22
Slide 23: Optimizing illumination improves resolution
Optimizing illumination improves resolution Non conventional illumination, NA, k1 ~ 0.30 Iret Mask xret Pupil/Lens: NA Iwfr Wafer xwfr / Slide 23
Slide 24: Low k1 imaging leads to small exposure/process tolerances
Low k1 imaging leads to small exposure/process tolerances Iwfr Iwfr Iwfr xwfr xwfr xwfr CD1 CD2 / Slide 24
Slide 25: Low k1 imaging leads to small exposure/process tolerances
Low k1 imaging leads to small exposure/process tolerances Iwfr Iwfr Iwfr xwfr xwfr xwfr CD1 - ∆CD1 CD2 - ∆CD2 Image linearity disappears, ∆CD1 << ∆CD2 DRAM makers can work at 20% lower k1 factor compared to LOGIC makers due to a more repetitive and regular layout / Slide 25
Slide 26: Mask and system integration is needed to enable low k1
Mask and system integration is needed to enable low k1 Source Mask Optimization Illumination DOE Library Customized DOE 6% 6% Mask BIM AttPSM CPL™ BIM AttPSM CPL k1=0.4 Layout Complexity k1=0.3 Scattering Model CPL Mask DDL Mask OPC Bars OPC Decomposition Decomposition / Slide 26
Slide 27: Customized illumination (1)
Customized illumination (1) l NA = 0.8, 248 nm, BIM l CD = 120 nm x 740 nm, pitch = 230 nm x 920 nm / Slide 27
Slide 28: Customized illumination (2)
Customized illumination (2) l NA = 0.8, 248 nm, BIM l CD = 120 nm x 740 nm, pitch = 230 nm x 920 nm 25 % Exposure Latitude 20 Hexapole 15 10 5 0 Annular 0 0.5 1 1.5 Depth of Focus (µm) / Slide 28
Slide 29: Impact of Phase Shift Mask (PSM)
Impact of Phase Shift Mask (PSM) Binary Binary Alternating PSM Chromeless PSM conventional off axis on axis off axis Int. / Slide 29
Slide 30: CPL single exposure versus altPSM double exposure
CPL™ single exposure versus altPSM double exposure Phase mask Trim mask 0 Chrome AltPSM masks layouts (2 required) π Chrome + Clear Residual 0 Phase Edge π 0 Chrome Typical logic pattern 0 π π 0 Chrome CPL mask layout (only 1 needed) 0 / Slide 30
Slide 31: 100-nm SRAM imaging with CPL on an ASML PAS 5500/800
100-nm SRAM imaging with CPL on an ASML PAS 5500TM/800 Expose on ASML 100 nm PAS 5500/800 CPL with NA=0.80 Gate 2 algorithm QUASARTM Gate 1 σin=0.57 σout=0.87 Desired SRAM Pattern Target CD of 100 nm / Slide 31
Slide 32: Even small contacts can be made with CPL
Even small contacts can be made with CPL ASML AT:1200, ArF, 0.85 NA, QUASAR illumination 80-nm contacts, 160-nm pitch 75-nm contacts, 150-nm pitch K1 = 0.29 65-nm contacts, 130-nm pitch / Slide 32
Slide 33: Mask and system integration is needed to enable low k1
Mask and system integration is needed to enable low k1 Scatter Aberrations Illumination Source optimization Dose optimization Mask optimization software / Slide 33
Slide 34: Mask and system integration is needed to enable low k1
Mask and system integration is needed to enable low k1 l Low k1 imaging reduces process windows and overlap of process windows when different features are imaged l DRAM manufacturing expected to reduce down to k1~0.30 using repetitive structures, strong phase shift masks, and customized illumination
Slide 35: Agenda
Agenda l The shrink roadmap of our customers l Lithography shrink scenarios l Wavelength reduction l Aperture increase enabled by immersion l Lowering the k1 factor l Special topics l Immersion l Maskless l Likely shrink scenarios l Conclusions / Slide 35
Slide 36: Why immersion?
Why immersion? dry immersion Convert existing lens to immersion • Resolution remains same • Increased depth of focus • Larger process window => increased yield • Decrease double exposure layers, lower cost per wafer dry immersion Develop new NA>1 immersion lens • Increased resolution • Delay wavelen
Slide 37: Dry versus wet focus sensing
Dry versus wet focus sensing Air Water Resist Resist Wafer Wafer Process-induced focus shift (nm) 100 0 dry -100 -200 -300 wet -400 -500 -600 0 100 200 300 400 500 600 700 800 Resist thickness (nm) / Slide 37
Slide 38: AT:1150i, 90-nm dense lines and spaces
AT:1150i, 90-nm dense lines and spaces Dry Focus (um) Wet -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 / Slide 38
Slide 39: Evaluation: immersion
Evaluation: immersion l Immersion most likely does not need infrastructural changes in materials, resist, mask, and pellicles l ASML can quickly adopt their equipment for immersion given the dual stage/dry metrology option and adaptable lens technology, giving an acceptable introduction risk l First
Slide 40: Agenda
Agenda l The shrink roadmap of our customers l Lithography shrink scenarios l Wavelength reduction l Aperture increase enabled by immersion l Lowering the k1 factor l Special topics l Immersion l Maskless l Likely shrink scenarios l Conclusions / Slide 40
Slide 41: Spatial light modulator in maskless scanner
Spatial light modulator in maskless scanner Micronic SLM chip 8 mm x 33 mm active area (512 x 2048 pixels) 1.048 Mpixels each mirror 16 m x 16 m2 Lithography pattern generation is controlled by reflecting laser light from individual pixels towards or away from the wafer / Slide 41
Slide 42: How optical maskless helps to reduce time to market
How optical maskless helps to reduce time to market Design introduction Mask respin Start Design Release Mask making Release Mask Respin >18 months Start Design Release Release < 12 months / Slide 42
Slide 43: Maskless practically eliminates fixed mask cost
Maskless practically eliminates fixed mask cost Even at modest throughput OML is economical Maskless Scanner Maskless more economical Cost per Mask Run Scanner ∝ 1/Tput(mask) = 1/100 Fixed cost of masks With mask Small Data ∝ 1/Tput(maskless) = 1/5-10 more economical Prep Cost Breakeven #Wafers # Wa
Slide 44: Evaluation: maskless
Evaluation: maskless l Optical MaskLess (OML), unlike non-OML, uses the existing process and process simulation infrastructure l Productivity scaling is not limited by physics but by the evolution of data path improvements l Resolution scaling is compatible with mainstream mask based lithography, 19
Slide 45: Agenda
Agenda l The shrink roadmap of our customers l Lithography shrink scenarios l Wavelength reduction l Aperture increase enabled by immersion l Lowering the k1 factor l Special topics l Immersion l Maskless l Likely shrink scenarios l Conclusions / Slide 45
Slide 46: Technology choices: k1 values
Technology choices: k1 values Relaxed pitches required Half pitch (nm) 110 90 65 45 32 22 KrF 0.35 0.29 ArF 0.40 0.31 ArF-I 0.44 0.30 0.22 F2 0.39 0.27 F2-I 0.37 0.26 EUV 0.64 0.44 / Slide 46
Slide 47: 300-mm product roadmap and decision points
300-mm product roadmap and decision points NA >>1 157 nm Wet Wet F2 >1 >0.85 Dry 0.85 Dry 0.75 MSVII Decide Mid-04 >>1 Wet 193 nm ArF >1 Wet Decide >0.85 Q403 Wet 0.85 Wet 0.75 1150i PROTO TOOL ASML >0.85 Dry 0.85 1200 1250 2003 2004 2005 2006 2007 2008 / Slide 47
Slide 48: Conclusions (1)
Conclusions (1) l For our customers, shrink continues to drive costs down and performance up. l High resolution capability challenges the costs of lithography, however, new technology such as macro aspheres, catadioptric designs, immersion, and field size reduction, in combination with productivity
Slide 49: Conclusions (2)
Conclusions (2) l ASML’s lens and dual-stage technology enables a quick and cost effective transition to immersion tools completely compatible with current dry tool metrology. l Immersion will likely become an 193-nm extension pushing 157 nm and EUV backwards. Qualified customer integration expected
Slide 50: Commitment
Commitment / Slide 50