Coherent kev X-Rays from Tabletop Femtosecond Lasers and Applications in Nanometrology

Transcription

1 Member Subscription Copy Library or Other Institutional Use Prohibited Until 2015 Articles published week ending Published by the Coherent kev X-Rays from Tabletop Femtosecond Lasers and Applications in Nanometrology P R L HYSICAL EVIEW ETTERS 22 OCTOBER 2010 American Physical Society Volume 105, Number 17

2 Students and Collaborators Tenio Popmintchev, Ming-Chang Chen, Damiano Nardi, Kathy Hoogeboom- Pot, Dan Adams, Matt Seaberg, Bosheng Zhang, Margaret Murnane, Henry Kapteyn University of Colorado, Boulder Marie Tripp, Sean King, Chris Deeb Intel Rich Haight IBM Tom Silva, Justin Shaw, Hans Nembach NIST Hyung Su Son Samsung Vimal Kamineni Globalfoundries Stefan Mathias, Martin Aeschlimann, Claus Schneider Kaiserslautern and Julich Eric Anderson, Eric Gullikson CXRO, LBL John Miao UCLA Andrius Baltuska Technical University Vienna Bruce Guerney, Olav Hellwig HGST

3 Outline High harmonic generation (HHG) ONLY tabletop source of coherent soft X-rays Recent advance: bright beams from UV to kev Limits? Hard X-ray beams? Commercial systems in EUV Unique nanometrologies using HHG Coherent (lensless) imaging near wavelength limit EUV acoustic and thermal nanometrology Photoemission spectroscopy, spintronics, etc 3D Coherent Imaging Nanoscale spin dynamics Acoustic/thermal nanometrology

4 Nonlinear Optics: microscopic and macroscopic science Second harmonic generation Franken et al, PRL 7, 118 (1961) Ruby laser! Lens! Quartz crystal! Prism! 347nm! 694nm! Photographic plate! Phase Matching Armstrong, Bloembergen et al., PRA 127, 1918 (1962) V phase (2ω) = V phase (ω) " k ω" k ω" k 2ω"

5 Extreme nonlinear optics: microscopic and macroscopic science High harmonic generation (HHG) JOSA B 4, 595 (1987) J Phys B 21, L31 (1988) ~ 100 Bohr radii Phase matched HHG Science 280, 1412 (1998) Science 336, 1287 (2012)

6 High harmonic generation coherent version of the X-ray tube High harmonic generation (JOSA B 4, 595 ( 87); J Phys B 21, L31 ( 88)) 1895 Röntgen X-ray Tube (Roentgen, Nature (1896))

7 Harder challenge: phase-matching in plasmas Net Intensity ~ N atoms 2 PHASE MATCHED Laser field ionization Refractive index of laser is time dependent! Phase matching thought impossible!

8 Dynamic phase matching during small time interval Ionization (t). 0 % Δk = q u 2 λ 11 0 ( % /' * P '(1 η) 2π Δδ η N & 4πa 2 atm r e λ 0 10 ) & λ 0 [ ] ( 20 * 3 ) 40 PLASMA DISPERSION (v laser > c) NEUTRAL ATOM DISPERSION (v laser < c) Dynamic phase matching v laser = c during one half cycle Waves synchronized due to tunnel ionization and phase matching window PHASE MATCHING WINDOW

9 Counterintuitive extreme nonlinear optics 5001 th order 111 th order 11 th order Incoherent x-ray tube Coherent x-ray tube Science 336, 1287 (2012)

10 Bright HHG emission driven by ultrafast mid-ir lasers. 0 % Δk = q u 2 λ 11 0 ( % /' * P '(1 η) 2π Δδ η N & 4πa 2 atm r e λ 0 10 ) & λ 0 [ ] ( 20 * 3 ) 40 kev HHG needs mid IR lasers EUV HHG needs 0.8µm lasers Ti:Sapphire VUV HHG needs UV lasers PNAS 106, (2009) Nature Photonics 4, 822 (2010) PRL 105, (2010) CLEO Postdeadline (2011) Science 336, 1287 (2012) Pat. No. 8,462,824 (2013)

11 Predict phase-matched HHG yield bright from VUV to kev. 0 % Δk = q u 2 λ 11 0 ( % /' * P '(1 η) 2π Δδ η N & 4πa 2 atm r e λ 0 10 ) & λ 0 [ ] ( 20 * 3 ) 40 Phase matching pressure increases from EUV to soft x- ray region 30 torr to 40 atm! Gas transparency increases Compensates for low single atom yield PNAS 106, (2009) Nature Photonics 4, 822 (2010) PRL 105, (2010) CLEO Postdeadline (2011) Science 336, 1287 (2012) Pat. No. 8,462,824 (2013)

12 Unique X-ray source coherent supercontinuum to 8Å 0.8µm 1.3µm 2µm Combine > 5000 laser photons efficiently using 4µm lasers!! λ LASER =3.9 µm P = 40 atm B C N O Fe Co Ni Cu ONLY bright coherent tabletop kev X-rays Highest nonlinear and phase matched process at > 5000 orders Phase matching bandwidth ultrabroad since v X-rays c Coherent spectrum spans many elemental x-ray edges simultaneously Science 336, 1287 (2012)

13 High harmonics broad spectral coverage EUV Soft x-ray ΔE 700 ev Δt 210 as Δt 2 as Time (fs) Opt. Express 17, 4611 (2009) Science 336, 1287 (2012)

14 Member Subscription Copy Library or Other Institutional Use Prohibited Until 2015 Articles published week ending Published by the Bright coherent beams from UV to kev 4000 nm 30nm HHG beam 13nm HHG beam High pressure waveguide 3nm HHG beam P R L HYSICAL EVIEW ETTERS 22 OCTOBER nm HHG beam American Physical Society Volume 105, Number 17 Science 280, 1412 (1998) Science 297, 376 (2002) Science 336, 1287 (2012) Pat. No. 8,462,824 (2013)

15 Bright coherent beams from UV to kev 4000 nm 30nm HHG beam 13nm HHG beam Current conversion efficiency: ev: /ev (per 1% band) ev: /ev ev: /ev Laser powers: 10 50W EUV power: µw 0.5mW (per 1% band) 3nm HHG beam 1nm HHG beam Limit not known: Increases in efficiency and photon energy very likely - new results! Science 280, 1412 (1998) Science 297, 376 (2002) Science 336, 1287 (2012) Pat. No. 8,462,824 (2013)

16 Ultrahigh efficiency narrowband UV-driven harmonics λ L = 0.8µm He 3ω = 0.27µm ε = 5x10-4 λ L = 1.6µm Single atom yield λ L -5.5 Ar ω = 0.8µm He 2ω = 0.4µm He ω = 0.8µm Driving HHG with 2ω and 3ω of Ti:sapphire has advantages in VUV/EUV - Ultrahigh 10-3 efficiencies when phase matched! - Harmonics separated by 6.2eV or 9.3eV no need for spectral selection! - Narrow bandwidth around 100meV but still 10fs! - Ideal for imaging and defect inspection at 13nm?

17 Limits of high harmonic generation not yet known! 20 µm mid-ir lasers may generate bright 25 kev beams Quasi phase matching schemes also promising Create designer X-ray waveforms with controlled polarization state Potential for major disruptive technology Incoherent x-ray tube Coherent x-ray tube

18 Basic phenomena broad application: 50 years 1986 XUV up to 17 th harmonic from Xenon gas (MacPherson et al) HHG mechanism explained (Krause and Kulander); classical 3-step model (Corkum) Year Intrinsic phase in HHG (L Huillier); demonstration of phase matched HHG (K/M): potential for HHG as a light source KMLabs commercial HHG source XNLO takes leap into x-rays (K/M) First commercial MRI First MRI imaging (Damadian, Lauterbur) 1961 Varian A-60 commercial NMR spectrometer 1950 Spin echo in NMR (Hahn): central concept for MRI 1945 NMR in a bulk material (Purcell, Torey and Pound) 1937 NMR predicted then immediately observed (Rabi)

19 Tested in many university research labs worldwide First commercial ultrafast coherent EUV source for scientific market Operated at CLEO exhibit in May 2009 Commercial, integrated, UHV-compatible system installed in Germany (4), Israel (2), MIT(1), Caltech (1), China (1) and Bulgaria (1) for applications in materials science Used successfully by many groups

20 Developing robust HHG platforms Driving Laser: Dragon/Wyvern EUV Generator: XUUS Monochromator/spectrometer NSF 2-yr awards for 6nm and 13nm HHG - Identify best driving laser, gas medium, waveguide designs DOE STTR 2-yr award - Develop compact EUV monochromator for scientific applications

21 Next generation lasers for 6.Xnm and beyond Next generation mid-ir lasers based on chirped optical parametric amplifiers Robust fiber laser front end (briefcase size) Addresses 6.Xnm node at 190eV Scalable to 100kHz DARPA 5-yr award for 1 6nm HHG ($7.5M between 5 groups) - Tabletop microscope with 5nm spatial resolution

22 Unique nanoscience applications of EUV HHG Rohwer#et#al.,#Nature#471,#490#(2011)# Understanding complex materials and nanosystems - Explore correlations, many body dynamics, non-equilibrium electrons/spins, little theory - HHG and other new tools uncover new information and enable benchmarking with theory - Important technologically - data storage, nanoelectronics, energy, catalysis Nanoscale spin dynamics Electrons in quantum dots Correlated materials Electronic)) Excita5on) CDW) Gap, PLD) Na Graphene PRX 2, (2012); PNAS 109, 4792 (2012); Nat. Commun. 3, 1037 (2012); PRL 110, (2013) Nano Letters 13, 2924 (2013) In prep (2013) The# closing# of# the# gap# can# clearly# be# monitored# at# the# ΓGpoint# and# behaves# iden suppression#of#the#se#4p#band#on#short#omescales.#on#longer#omescales,#the#dynamica can#be#idenofied#with#the#amplitude#mode#of#the#periodic#lauce#distoroon.#addioonally, the#amplitude#mode#of#the#periodic#lauce#distoroon#is#fluencegdependent,#i.e.#oscillaoo pump#intensioes.#the#extracted#frequency#of#the#amplitude#mode#is#about#2#thz." Nature 471, 490 (2011); Nat. Comm 3, 1069 (2012); Submitted (2013)

23 Unique nanoscience applications of EUV HHG Understanding nanoscale materials requires new capabilities - 3D non-destructive imaging with λ spatial resolution (next generation lithography, nanoelectronics) - Understanding nanoscale energy/charge/spin flow, no theory (thermal, strain, metamaterials) Acoustic nanometrology Nanoscale coherent imaging Nanoscale heat flow Gate Source Channel Drain Nature Materials 9, 26 (2010); Nano Letters 11, 4126 (2011); PRB 85, (2012) Nature 463, 214 (2010); Op. Ex. 19, (2011); Op. Ex. 17, (2012); Opt. Ex. 21, (2013) Submitted (2013) Nature Materials 9, 26 (2010); Nano Letters 11, 4126 (2011); PRB 85, (2012)

24 Coherent Diffractive X-Ray Imaging (XCDI) Diffraction-limited imaging λ/2να! Image thick samples in 3D Inherent contrast for X-rays Robust to vibrations Needs a coherent beam and isolated sample Sayre, Acta Cryst 5, 843 (1952) Miao et al., Nature 400, 342 (1999) Miao, Nature 463, 214 (2010) Miao, Nature 483, 444 (2012) Miao, Nature 496, 74 (2013)

25 Record tabletop full field light microscope: 22nm NA HHG wavelength = 13 nn Resolution of 1.6 l or 22 nm BUT isolated object in transmission mode 22nm PRL 99, (2007); Nature 449, 553 (2007); PNAS 105, 24 (2008); Nature Photon. 2, 64 (2008); OL 34, 1618 (2009); Optics Express 19, (2011)

26 3D high numerical aperture (angle) imaging reconstruction Nature 463, 214 (2010) Optics Express 19, (2011)

27 Scanning, non-isolated object, transmission mode CDI SEM CDI Semi-transparent background can extract thickness Non-destructive imaging compared with AFM Thickness map 50nm hole not completely drilled through: 48nm (CDI) vs 52nm (AFM) Opt. Express 21, (2013)

28 First general, scanning, reflection mode, non-isolated object, coherent imaging on a tabletop (30nm light) Raw CCD Data sample Optical Microscope High Resolution Shadowgram EUV multilayer Pinhole EUV CDI height map in nm Ptychographic reconstruction recovers 31nm object height ( 1nm precision) Spatial resolution limited by NA and 30nm wavelength in this preliminary work Next steps: increase spatial resolution to 2λ! Increase spatial resolution to 30nm using 13nm harmonics

29 First general, scanning, reflection mode, non-isolated object, coherent imaging on a tabletop (30nm light) Raw CCD Data sample Optical Microscope High Resolution Shadowgram EUV multilayer EUV CDI height map Pinhole Ptychographic reconstruction recovers 31nm object height ( 1nm precision) Spatial resolution limited by NA and 30nm wavelength in this preliminary work Next steps: increase spatial resolution to 2λ! Increase spatial resolution to 30nm using 13nm harmonics

30 First general, scanning, reflection mode, non-isolated object, coherent imaging on a tabletop (30nm light) Raw CCD Data sample Optical Microscope High Resolution Shadowgram HHG CDI EUV multilayer AFM image Pinhole Ptychographic reconstruction recovers 31nm object height ( 1nm precision) Spatial resolution limited by NA and 30nm wavelength in this preliminary work Next steps: increase spatial resolution to 2λ! Increase spatial resolution to 30nm using 13nm harmonics

31 First general, scanning, reflection mode, non-isolated object, coherent imaging on a tabletop (30nm light) Raw CCD Data sample Optical Microscope High Resolution Shadowgram HHG CDI with position correction EUV multilayer Pinhole HHG CDI without position correction Ptychographic reconstruction recovers 31nm object height ( 1nm precision) Spatial resolution limited by NA and 30nm wavelength in this preliminary work Next steps: increase spatial resolution to 2λ! Increase spatial resolution to 30nm using 13nm harmonics

32 Dramatic XCDI advances using XFEL Single-shot 3D structure determination using femtosecond XFEL pulses at 5.4 kev (SACLA, Japan) 5.5nm spatial resolution, 10fs time resolution XFEL Miao, Nature 483, 444 (2012) Miao, Nature 496, 74 (2013) Miao, submitted (2013)

33 Importance of nano-to-bulk heat transfer Gate Source Channel Drain MOSFET(IBM) TE generator (BMW) Quantum dots in photovotaics Nano-thermal therapy for cancer Nano patterned hard drive (Hitachi)

34 Understanding nanoscale heat flow in 1D Heat is carried by phonons In the macroscopic world, Fourier Law applies What happens when a nanostructure is smaller than the phonon mean free path? Existing theories of nanoscale heat dissipation disagree Fourier law over-estimates the heat flow - need to compare interface dimension to phonon mean free path Nature Materials 9, 26 (2010)

35 Understanding nanoscale heat flow in 1D Heat is carried by phonons In the macroscopic world, Fourier Law applies Λ 2nm What happens when a nanostructure is smaller than the phonon mean free path? Λ 120nm Existing theories of nanoscale heat dissipation disagree Fourier law over-estimates the heat flow - need to compare interface dimension to phonon mean free path Nature Materials 9, 26 (2010)

36 Nanoscale energy flow in 2D even more ballistic Thermal decay in 2D slower than 1D on short time scales - stronger ballistic effects! At large time scales, thermal decay similar in 1 and 2D - dominated by substrate Thermal modeling in progress - no current theory available Decreasing size 1D nanostructures Ni#on#Sapphire# Decreasing size 2D nanostructures Nature Materials 9, 26 (2010) Nano Letters 11, 4126 (2011) PRB 85, (2012) Nano Letters 13, 2924 (2013) Submitted (2013)

37 Nanoscale energy flow in 2D even more ballistic Thermal decay in 2D slower than 1D on short time scales - stronger ballistic effects! At large time scales, thermal decay similar in 1 and 2D - dominated by substrate Thermal modeling in progress - no current theory available Energy transport comparable in 1D and 2D for 500nm structures! 2D dots" Ni#on#Si# 1D lines" Time delay (ps)" Energy transport slower in 2D than 1D for 560nm structures!! 2D dots" Nature Materials 9, 26 (2010) Nano Letters 11, 4126 (2011) PRB 85, (2012) Nano Letters 13, 2924 (2013) Submitted (2013) 1D lines" Time delay (ps)"

38 Nanoscale energy flow: acoustics Characterizing nanoscale mechanical properties very challenging < 100nm EUV HHG is proven to work for < 10nm films! Sensitive to pm displacements! Demonstrated sensitivity to sub-monolayer! ζ!10nm! 6000 SAW penetration depth (nm) E"="200"GPa" 5000 Nature Materials 9, 26 (2010) Nano Letters 11, 4126 (2011) PRB 85, (2012) Nano Letters 13, 2924 (2013) Submitted (2013) SAW velocity (m/s) SAW wavelength (nm) Fundamental SAW 2nd-order SAW Softer" SiC:H"films" E"="13"GPa"

39 Nanoscale energy flow: acoustics Characterizing nanoscale mechanical properties very challenging < 100nm EUV HHG is proven to work for < 10nm films! Sensitive to pm displacements! Demonstrated sensitivity to sub-monolayer! How do bulk properties develop on layer-by-layer basis? Nature Materials 9, 26 (2010) Nano Letters 11, 4126 (2011) PRB 85, (2012) Nano Letters 13, 2924 (2013) Submitted (2013) Reflectivity LAW measurements Time (ps) Ta 6nm 6nm 6nm Ta thickness 4nm 3nm Ta it has n 0nm 1nm 2nm 3nm 3.3nm 3.6 4nm 6nme2 6nme4 6nme3 3.6nm 3.3nm 3nm 2nm 1nm 0nm

40 Surprising ultrafast spin dynamics No complete microscopic theory of magnetism exists on fs time scales High harmonics enable ultrafast, element-specific, spin dynamics to be probed at multiple sites simultaneously Even in a strongly exchange-coupled Fe- Ni ferromagnetic alloy, the dynamics of the individual spin sublattices can be different on timescales faster than that characteristic of the exchange interaction energy (10 80 fs) HHG light fundamental light CCD detector Al Ru Al filter Ni Fe Si Si 3 N 4 grating Large, superdiffusive, spin currents can be launched by a femtosecond laser through magnetic multilayers, to enhance or reduce the magnetization of buried layers, depending on their relative orientation PUBLICATIONS PRX 2, (2012); PRL 110, (2013) PNAS, 109, 4792 (2012) Nature Commun. 3, 1037 (2012) NEWS ARTICLES ABOUT WORK Physics 5, 11 (2012) Physics Today 65 (5), 18 (2012) Physik Journal 11, Nr. 6, page 26 (2012)

41 Combine tabletop coherent X-rays with coherent imaging