Status of EUVL Multilayer Optics Deposition at RIT

Size: px

Start display at page:

Download "Status of EUVL Multilayer Optics Deposition at RIT"

Darleen Martin
6 years ago
Views:

Vladimir Martynov Rigaku Innovative Technologies,

, Auburn Hills, MI 48326, USA Outline RIT

projects New capabilities Conclusion Osmic X-ray

1 Status of EUVL Multilayer Optics Deposition at RIT Yuriy Platonov, Jim Rodriguez, Mike Kriese, Vladimir Martynov Rigaku Innovative Technologies, 1900 Taylor Rd., Auburn Hills, MI 48326, USA Outline RIT introduction ML deposition facility Metrology EUVL projects New capabilities Conclusion Osmic X-ray Products Innovative Technologies, Inc. RIT Auburn Hills, Michigan USA Auburn Hills 18 holes golf course Detroit

2 RIT is part of Rigaku Global Organization X-ray Measurement Systems and Components Manufacturing, R&D, Sales, Service $300M annual revenue

3 former

4 RIT Product optics for hard and soft x-rays multilayer thin films electromagnetic spectrum XUV ~ 2 nm-20 nm range (approx.) EUV = 13.6 nm exactly

5 RIT/Osmic pioneered and commercialized multilayer optics for x-rays Units Shipped: Optics > 50,000 optical subassembly ~ 8000 source modules ~ 200 EUV optics

Multilayers deposition facility Inline Magnetron 7 Carousel Magnetrons Ion Beam Class100 cleanroom with class 10 miniroom Wavelength Range λ = 0.

5 meter 25 years of ML production Materials W/Si, W/C, Ni/Ti, Ni/B 4 C, Ni/C, Cr/C, Cr/Sc, Mo/Si, Mo/B 4 C, La/B, V/C, Ru/B 4 C, Al 2 O 3 /B 4 C, SiC/Si, Si/C, SiC/C,

6 Multilayers deposition facility Inline Magnetron 7 Carousel Magnetrons Ion Beam Class100 cleanroom with class 10 miniroom Wavelength Range λ = 0.2Å 300Å E = 40eV 60keV Multilayer Period d min = 10Å Number of Period N max = 1000 Spectral Resolution λ/λ = 0.2% (high-selective) 20% (depth-graded) Size: ~3mm to 1.5 meter 25 years of ML production Materials W/Si, W/C, Ni/Ti, Ni/B 4 C, Ni/C, Cr/C, Cr/Sc, Mo/Si, Mo/B 4 C, La/B, V/C, Ru/B 4 C, Al 2 O 3 /B 4 C, SiC/Si, Si/C, SiC/C, Fe/Si, Cr/B 4 C, Si/B 4 C, W/Mg 2 Si, V/B 4 C, Ti/B 4 C, etc. Design Uniform or Graded: lateral, radial, bilateral (2D) Depth Graded: supermirror & high-selective Flat or Curved Glancing (<1º) to Normal

5 to 5 mtorr linear ion source 20-100 particles/cm 2 on optical surface Dual Spinning Capability #1: 450mm

7 Inline Magnetron Vacuum (load-locked) 10-8 ultimate 10-9 water 15min from atm to 10-6 Process 5 planar magnetron (RF,DC) 4 process gases 0.5 to 5 mtorr linear ion source particles/cm 2 on optical surface Dual Spinning Capability #1: 450mm dia x 100mm thick #2: 175mm dia x 35mm thick (Compatible with velocity motion control) Mechanical 500 x 1500mm carrier (2) 0.2mm accuracy mm/sec (±0.1%) velocity profiling (6 pts/mm)

8 Deposition of 8 mask blanks (1999) Maximum 12 8 wafers were loaded per deposition run 200mm ~1300mm Coated over 1000 mask blanks ( ) Typical run was 30 blanks/day Typical λ c 13.5 ± 0.02nm within 6 wafer Typical R p = 63 ± 1% within 6 wafer (Si cap layer of ~11nm) Particulates brought down from 55,000/cm 2 to ~50/cm 2 over project (record value ~13/cm 2 ) 20 runs, 9 measurements/blank

9 Rotary cart upgrade. InLine. New Cart Maximum size: Diameter 550mm Thickness 220mm Schedule: Parts delivery - Dec. 10, 2010 Assembling - Dec. 23, 2010 Testing - Jan. 31, 2011 Commissioning - Feb. 1, 2011

EUVL 2-Optic Imaging System (2004) Mirror #1 Range of Data: 13.40-13.

10 EUVL 2-Optic Imaging System (2004) Mirror #1 Range of Data: nm Average Reflectivity: 61.6% Mirror #2 Range of Data: nm Average Reflectivity: 61.4% 200mm toroidal R~ D non-radial gradient Ru/B 4 C topcoat (best R p 67.1%) Achieve < ±1% wavelength on all four optics (2 sets of 2) M1#1 Final - Peak Position M2#2 Final - Peak Position

005nm λ c in CA) Relative Variation of CWHM or thickness 1.01 1 C1 C2 C3 M1 M2 0.99-1 -0.5 0 0.

11 6-Optic Condensor/Imaging (2005) Reticle Imaging Microscope:Tinsley/Exitech RIM 4 condensor (1 Ru, 3MoSi) 2 imaging (MoSi) Added Figure Error in imaging optics: M1: 0.015nm (± 0.018nm λ c in CA) M2: <0.010nm (± 0.005nm λ c in CA) Relative Variation of CWHM or thickness C1 C2 C3 M1 M Relative distance along diameter of clear-aperture H.Glatzel et al. Characterization of prototype optical surfaces and coatings for the EUV Reticle Imaging Microscope, Proc. of SPIE, Vol (2005),

Deposition Simulations Practical simulation of the deposition, using

substrate Predict coating results, capability, y trajectory sensitivity;

travel cylindrical, spherical and aspherical optics (radial & 2D) optic x

12 Deposition Simulations Practical simulation of the deposition, using distributed array of point sources Accommodate geometry of system & substrate Predict coating results, capability, y trajectory sensitivity; reduce calibration time x R r αx φ Successfully used for flat optics, z travel cylindrical, spherical and aspherical optics (radial & 2D) optic x p Target Optic target αy ψ Non-radial on sphere Linear on sphere Uniform on cylinder

In-House surface characterization 2 δ PSD( f ) = N N e n= 1 2 π ι ( n 1) δ f z( n) PSD is surface roughness power per unit spatial frequency.

5x Zygo 5x Zygo 10x Zygo 20x Zygo 40x AFM 10 x 10 microns Line approximation Spatial Frequency (1/microns) rms 2.

Installed in class 100 clean room 25 x 10 mm, Filter 50µm 5 mm High Spatial Frequency roughness (HSFR) Spatial periods ~10nm - 5µm Instrument -AFM manufactured by

13 In-House surface characterization 2 δ PSD( f ) = N N e n= 1 2 π ι ( n 1) δ f z( n) PSD is surface roughness power per unit spatial frequency. σ RMS = f 2 f 1 PSD( f ) d( f ) 1D PSD (nm^3) Zygo 1.25x Zygo 2.5x Zygo 5x Zygo 10x Zygo 20x Zygo 40x AFM 10 x 10 microns Line approximation Spatial Frequency (1/microns) rms 2.2 sec, 2 nm Mid Spatial Frequency Roughness (MSFR) Spatial periods ~1µm 5mm Instrument -Interferometric Microscope manufactured by Zygo, model New View Installed in class 100 clean room 25 x 10 mm, Filter 50µm 5 mm High Spatial Frequency roughness (HSFR) Spatial periods ~10nm - 5µm Instrument -AFM manufactured by Veeco, model #DI3100. Installed in class 100 clean room Mid Spatial Roughness Instrument Contact Profilometer Talysurf Spatial periods >10µm Max scan range -200mm, accuracy - 0.5µm LTP ZYGO PSD AFM Figure Ripple Roughness mid frequency low frequency high frequency µm

Mo/Si multilayers surface roughness 0.5 dia. Si substrate before coating Same after depositing 40 bi-layers of Mo/Si with d 7nm 1.10 5 1.10 5 1.10 4 1.10 4 1. 10 3 1. 10 3 1D PSD (nm^3) 100 10 1 0.

5x Zygo 5x Zygo 10x Zygo 20x Zygo 40x AFM 20x20 microns Go_Round_Si Spatial Frequency (1/microns) σ = 2.2Α for 1.25x σ = 2.6A for 2.5x σ = 2.8A for 5x σ = 1.2A for 10x σ = 0.5A for 20x σ = 0.

14 Mo/Si multilayers surface roughness 0.5 dia. Si substrate before coating Same after depositing 40 bi-layers of Mo/Si with d 7nm D PSD (nm^3) D PSD (nm^3) Zygo 1.25x Zygo 2.5x Zygo 5x Zygo 10x Zygo 20x Zygo 40x AFM 20x20 microns Go_Round_Si Spatial Frequency (1/microns) σ = 2.2Α for 1.25x σ = 2.6A for 2.5x σ = 2.8A for 5x σ = 1.2A for 10x σ = 0.5A for 20x σ = 0.7A for 40x σ = 1.0A for AFM 20x20 microns Zygo 1.25x Zygo 2.5x Zygo 5x Zygo 10x Zygo 20x Zygo 40x AFM 20x20 microns Go_Round_Si Spatial Frequency (1/microns) σ = 2.3Α for 1.25x σ = 2.1A for 2.5x σ = 2.8A for 5x σ = 1.7A for 10x σ = 0.5A for 20x σ = 0.8A for 40x σ = 1.0A for AFM 20x20 microns AFM and Zygo interferometer images after depositing Mo/Si multilayer structure

$diffractometers$ (λ=1.54å) 3

15 In-House XUV Characterization Cu-K α diffractometers (λ=1.54å) 3 instruments for ML reflectivity testing UV reflectometer (λ=150nm 350nm) Maximum samples size 300mm XRF spectrometer (λ > ~5Å) MF/CMF tester Microsource characterization

Substrates recovery Reflectivity 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 10-Period MoSi on Superpolished Fused Silica Fresh Substrate: Rp = 0.252 λ m = 13.515nm fw = 1.

5 15 Wavelength, nm Wet Chemical Stripping Convex optics were stripped, recoated, re-stripped and re-recoated (N=10): 1 st optics 1 st recoat: 15.2-15.9% Rp no change 2 nd recoat: 6.3-13.

16 Substrates recovery Reflectivity Period MoSi on Superpolished Fused Silica Fresh Substrate: Rp = λ m = nm fw = 1.170nm Etched Substrate: Rp = λ m = nm fw = 1.164nm Etched Substrate: Rp = λ m = nm fw = 1.164nm Wavelength, nm Wet Chemical Stripping Convex optics were stripped, recoated, re-stripped and re-recoated (N=10): 1 st optics 1 st recoat: % Rp no change 2 nd recoat: % Rp 2 nd optics 1 st recoat: % Rp no change 2 nd recoat: % Rp EUV measurements were done at NIST original FS100 stripped FS Multilayer Etched Virgin 1D-Iso PSD, nm Frequency, 1/µm

17 High selective % bandpass: 13.51nm 2% bandpass: 12.63nm 13.37nm λ/λ 1% SiC/Si structures. 5 o off normal Test coating: 5 o off normal EUV Spectra SiC/Si ML structure λ c = 13.51nm R p = 29.7% fwhm = 0.149nm Wavelength, nm All measurements were done at CXRO Reflectivity at θ = 30º from normal λ/λ 2% Si/B4C structures. 30 o off normal Test Coating: 30 o EUV Spectra λ c = 12.63nm λ, nm Test Coating: 30º EUV Spectral Scan λ C = nm R p =0.56 fwhm = 0.275nm R P = fwhm = 0.288nm Si/B 4 C ML structure International Workshop on EUV sources, University college Wavlength, Dublin, nmnovember 13-15, 2010

18 Next generation EUVL? 1 st Generation EUVL Next Generation EUVL λ=13.5nm, Mo/Si LSM R(calc) ~ 73% λ(calc) ~ 0.54nm λ=6.7nm, La/B4C LSM R(calc)~73% λ(calc)~0.064nm R(exp) ~ 70%, λ(exp) ~ 0.54nm R(exp)=43% (Year ) λ(exp)~0.044nm 0.7 Y.Platonov, L.Gomez, D.Broadway, SPIE Proc. (2002), p152 Reflectivity Mo/Si with Ru Cap 200mm curved optic λ P = nm Rp = Reflectance deg 72.5 deg La/B4C (XRO # ) als deg 0.1 θ ~ 20.7 deg Wavelength, nm λ (A) Measured at ALS by Eric Gullikson. Jan.2001

19 Normal Oct Experimental results Experimental reflectivity of La/B4C structures. October CXRO measurements. R p =46.2% fwhm=0.0452nm λ c =6.68nm 5 o off normal R p =38.8% #34142 #34150 #34158 #34165 #34155 fwhm=0.0505nm λ c =6.92nm Wavelength, nm R max (exp) = 46.2% vs - 74% (calc) Calculated performance Ideal La/B4C structure Calculated reflectivity of La/B4C structures 6.66nm R(pek), % fwhm=0.0623nm Normal incidence 6.9nm fwhm, nm Wavelength, nm Reflectivity versus number of bi-layers "Ideal" La/B4C structures R, % λ(exp) = nm vs σ(eff) 0.56nm nm (calc) 50 R =73.5% p R =74.6% N=150 N= N

20 Water window Experimental reflectivity of Cr/C structure. NIST, degrees off normal R p =12% fwhm~0.048nm λ c ~4.94nm XRO#23071 XRO# Wavelength, nm R(exp) = 12% vs 23% (calculated for an ideal Cr/C structure) fwhm = 0.048nm vs 0.051nm (calc.) σ(eff.) 0.33nm

21 Optics throughput vs wavelength Periodical structures λ=13.5nm R=0.7 fwhm=0.52nm R*fwhm=0.364nm λ=9.5nm R=0.6 fwhm=0.21nm R*fwhm=0.126nm λ=6.7nm R=0.7 fwhm=0.062nm R*fwhm=0.0434nm λ=4.5nm R=0.47 fwhm=0.0323nm R*fwhm=0.0152nm Assuming 10 mirrors optical system 7.3E-3 nm 6.4E-4 nm 8.8E-4 nm 8.5E-6 nm Lower wavelength Lower optics throughput Peak reflectivity is the most valuable parameter for maximizing multi mirror optics throughput The most promising wavelength for the next generation EUVL is ~6.7nm due to highest expected peak reflectivity from La/B 4 C ML structures

22 Conclusion Capabilities 25 years experience in ML X-ray optics X-Ray performance modeling Deposition flux simulation Ray-trace illumination modeling Surface roughness characterization EUV Projects completed 2-Optic imaging system (1999) >1000 Mask blanks ( ) 360mm Condensor (2002) 2-Optic imaging system (2003) 2-Optic toroidal imaging system (2004) 6-Optic condensor/imaging system (2005) Variety of flat normal incidence and 45 deg. EUVL mirrors X-Ray performance characterization Clean room environments Magnetron and ion-beam sputtering deposition of multilayers on up to 1.5m long or up to 400mm in diameter substrates Future EUVL activities New rotary cart for 550mm optics deposition Deposition technology for a large-sag optics Continue La-based for a better 6.7nm ML for 8nm to ~10nm wavelengths In-house Soft X-Ray Reflectometry

23 Acknowledgement James Wood Gary Fournier Jerry Hummel Calvin Coffel Tony Camitan Olga Faytlin Ella Sherstinskaya Nathan Frank And all other members of RIT team

24 Thank you Osmic Products

EUV Multilayer Fabrication

EUV Multilayer Fabrication Rigaku Innovative Technologies Inc. Yuriy Platonov, Michael Kriese, Jim Rodriguez ABSTRACT: In this poster, we review our use of tools & methods such as deposition flux simulation