MIT X-ray Laser Project An X-ray Laser at the Transform Limit: Technical Challenges and Scientific Payoff David E. Moncton Massachusetts Institute of Technology NSLS February 20, 2004 MIT X-ray Laser Project MIT X-ray Laser R&D Proposal Contact: David E. Moncton, Director Telephone: 617-253-8333 E-mail: email@example.com website: http://mitbates.mit.edu/xfel/indexpass.htm Co-Principal Investigators Science Collaborators William S. Graves Simon Mochrie Keith A. Nelson Franz X. Kaertner Gregory Petsko Dagmar Ringe Richard Milner Henry I. Smith Andrei Tokmakoff Bates Senior Staff Contributors Manouchehr Farkhondeh William M. Fawley James Fujimoto Jan van der Laan Ian McNulty Erich Ippen Christoph Tschalaer Jianwei Miao Denis B. McWhan Fuhua Wang Mark Schattenburg Michael Pellin Abbi Zolfaghari Gopal K. Shenoy Townsend Zwart MIT X-ray Laser Project A Unique Opportunity– An X-ray Laser User Facility •30 or more independent beamlines •Fully coherent milli-Joule pulses at kHz rates •Wavelength range from 200 nm to 0.1 nm The Scientific Impact Femtosecond pulse duration Chemistry Full transverse coherence Biology Milli-volt bandwidths Condensed Matter Physics Full quantum degeneracy Atomic Physics High electric field Fundamental Physics MIT X-ray Laser Project Photon pulses would be “transform-limited” satisfying the Uncertainty Principle in all six phase-space dimensions Transverse Phase Space Dx , Dy ~ 100 microns Dkx , Dky ~ 10-5 nm-1 Longitudinal Phase Space Dt ~ 1fs - 1ps Dw ~ 2eV - 2meV Note that all 1011 to 1014 photons in each pulse would occupy the same quantum state MIT X-ray Laser Project Inelastic X-ray Scattering (IXS) Photon-in – photon out Probes charge-neutral excitations: q=ki-kf • Phonons, diffusive modes, orbitons, w=wi-wf superconducting gaps… • Excitons, plasmons, particle-hole creation, interband transitions… Complements existing techniques: • ARPES measures A(q,w), surface sensitive. • INS does not couple to charge, and requires large sample. • EELS cannot measure to large q, or in fields. • s(w), Raman are restricted to q=0. MIT X-ray Laser Project Existing 3rd Generation IXS Beamlines ESRF ESRF APS Spring-8 APS •3 Beamlines with meV resolution, 2 more soon •1 Beamline with 100 meV resolution •2 Beamlines with eV resolution MIT X-ray Laser Project IXS with 3rd Generation Synchrotron Sources • Data taken on APS ID3 TiOCl • 2.2 meV resolution • 109 incident flux • Data shown took 6 hrs E. Isaacs, Y. Lee, D. Moncton MIT X-ray Laser Project Large IXS Signal Gain with Bandwidth Seeded X-ray Laser 3 x 1011 photons/pulse at 1 kHz = 3 x 1014 ph/sec Bandwidth seeding: 100 fs = 20 meV (l = 0.1nm) 1013 ph/sec at 1 meV resolution Bandwidth seeding: 1 ps = 2 meV 1014 ph/sec at 1 meV Compare with 109 ph/sec at 3rd Gen Sources An intensity gain of 4-5 orders of magnitude will revolutionize the study of the dynamics of condensed matter MIT X-ray Laser Project Transient Grating Spectroscopy or Time-Dependent IXS Keith Nelson, Chemistry Department, MIT MIT X-ray Laser Project Next-generation backscattering analyzers (Yuri Shvyd’ko, DESY) •Variable resolution down to 0.1 meV •Decouple the incident energy and the resolution •Less demanding temperature stability •Using sapphire, any incident energy can be selected MIT X-ray Laser Project Accessible phase space Inelastic Scattering Time-Dependent Methods MIT X-ray Laser Project The Potential of a Transform-Limited X-ray Laser for Inelastic X-ray Scattering Third Generation IXS •1010 ph/sec in 1 meV bandwidth •S/N still too low for many experiments •Phenomena with DE < 1 meV not resolved by IXS •Generally phenomena with 1 ns > Dt > 1 ps are inaccessible Bandwidth Seeded X-ray Laser •Up to 1014 ph/sec in meV bandwidth •Transform-limited pulses reach Heisenberg limit Dt Dw ~ p •Time-dependent (or pump-probe) IXS for full t/E coverage MIT X-ray Laser Project Photon pulses would be “transform-limited” satisfying the Uncertainty Principle in all six phase-space dimensions Transverse Phase Space Dx , Dy ~ 100 microns Dkx , Dky ~ 10-5 nm-1 Longitudinal Phase Space Dt ~ 1fs - 1ps Dw ~ 2eV - 2meV Note that all 1011 to 1014 photons in each pulse would occupy the same quantum state MIT X-ray Laser Project To realize such a source, the most sophisticated laser and accelerator technology must be integrated together. The laser generates the coherent signal MIT Ultrafast Laser Group Franz Kaertner, Erich Ippen, et al An accelerated electron beam amplifies and frequency shifts the laser radiation MIT Bates Laboratory William S. Graves et al MIT X-ray Laser Project Self-Amplified Spontaneous Emission (SASE) e- ~100 fs Argonne APS first demonstrates SASE at optical wavelengths Gain of 107 LCLS project at SLAC aims to demonstrate SASE at 0.15 nm MIT X-ray Laser Project A SASE FEL amplifies random electron density modulations t (fs) Dw/w (%) The SASE radiation is powerful, but noisy! One solution: Impose a strong coherent modulation with an external laser source MIT X-ray Laser Project Brookhaven laser seeding technology e- Laser output 800 nm 266 nm Modulator Buncher Radiator High Gain Harmonic Generation (HGHG) HGHG •Suppressed SASE noise •Amplified coherent signal SASE x105 •Narrowed bandwidth •Shifted wavelength MIT X-ray Laser Project To Produce Transform-Limited Pulses below 10 nm •Must get powerful short-wavelength seeds using High Harmonic Generation methods •Then use ―cascaded‖ High Gain Harmonic Generation methods in FEL Stage 1 output at Stage 2 output at …Nth stage 5w0 seeds 2nd stage 25w0 seeds 3rd stage output at 5Nw0 Input seed w0 1st stage 2nd stage …Nth stage MIT X-ray Laser Project Two different seeding regimes Seeding for short (1 fs) pulses—bandwidth of a few eV Seeding for narrow (meV) bandwidth—pulse lengths 0.1 – 1.0 ps MIT X-ray Laser Project Superconducting linac Developed technology is essential at DESY High pulse rates provide for many independent beamlines CW operation provides much greater beam stability Most important is minimizing electron arrival time jitter 6.0 Copper linacs like Bates or LCLS have jitter of 100’s of femtoseconds Phase: Klystron - MOA (deg) 5.5 5.0 Seeded FEL requires 10 fs timing stability at short wavelengths 4.5 σ = 0.14° (150 fs) Measured inside 10 s window 4.0 CW Superconducting cavities will have 0 100 200 300 400 much less phase jitter Time (s) MIT X-ray Laser Project MIT X-ray Laser Project MIT X-ray Laser Project Main oscillator Fiber link synchronization Seed UV Hall Pump X-ray Hall Seed Pump laser laser laser laser Undulators 200 nm Undulators 30 nm 1 nm Injector laser 10 nm 0.3 nm SC Linac 0.3 nm SC Linac 0.1 nm 1 GeV 2 GeV 4 GeV 10 nm Upgrade: 0.1 nm at 8 GeV 3 nm 1 nm Undulators Seed Pump laser laser Nanometer Hall MIT X-ray Laser Project Narrow bandwidth performance for MIT and LCLS 12.4 eV 124 eV 1.24 keV 12.4 keV LCLS performance from SLAC website parameter MIT UV table. Dt = 200 fs MIT beamlines are 1 kHz. Dw = 10 meV LCLS is 120 Hz. MIT covers wide spectrum MIT X-ray simultaneously with Dt = 200 fs multiple undulators. Photons per second Dw = 10 meV LCLS limited by undulator lattice to spectrum shown, must tune energy for LCLS different wavelengths. Dt = 200 fs Note steep falloff at short Dw = 4 eV wavelength for MIT due to gun performance and 4 GeV energy. MIT Change in performance at 3rd harmonic 5 nm is due to beam energy change from 1 LCLS GeV at longer Dt = 200 fs wavelengths to 4 GeV. Dw = 10 meV This is conservative spectral flux density for MIT. A 2 ps long pulse would have 10 times the Wavelength (m) flux in 1/10 the bandwidth. MIT X-ray Laser Project Short-pulse performance for MIT and LCLS 12.4 eV 124 eV 1.24 keV 12.4 keV Photons per second assuming FEL output from the short pulselengths 30 fs MIT UV shown. LCLS uses electron beam slicing, MIT uses short seed pulse. MIT beamlines are 1 kHz. 10 fs LCLS is 120 Hz. Photons per second Dots are minimum FWHM MIT pulselengths using FEL gain X-ray bandwidth. 1 fs CPA pulselength accounts 3 fs CPA for bandwidth and slippage. CPA can be used at longer wavelengths also. Slippage LCLS limits the min pulse length to be near the values shown at each wavelength. The pulse 1 fs intensity would be increased by 1-2 orders of magnitude. 0.5 fs Bandwidth ranges from 5e-4 at 0.3 nm to 1e-2 at 200 nm. Wavelength (m) MIT X-ray Laser Project Chirped Pulse Compression l = 2d E 20 meV 2eV lf lf i li l 100 fs t For compression: 100 fs to 1 fs •Energy chirp/bandwidth > compression df di •Reflectivity/layer< pulse chirp •But extinction depth~ few pulse lengths •Crystal chirp > pulse chirp MIT X-ray Laser Project MIT Ultrafast Laser Group is developing: Overall laser timing and synchronization below 10 fs MIT X-ray Laser Project MIT Ultrafast Laser Group is developing: RF phase control and stabilization MIT X-ray Laser Project MIT Ultrafast Laser Group is developing: Powerful short wavelength seed lasers Phase Controlled 5fs, 5mJ, 1 kHz MIT X-ray Laser Project MIT Ultrafast Laser Group is developing: High-Harmonic Generation with Noble gas jets (He, Ne, Ar, Kr) Phase Controlled XUV @ 3 – 30 5fs, 5mJ, nm 1 kHz h = 10-8 - 10-5 t Propagation Recombination 0 wXUV x -Wb tb Ionization Energy Laser electric field MIT X-ray Laser Project A focused, concept-driven R&D program is a pre-requisite to an X-ray Laser User Facility incorporating many beamlines and seeding for full coherence Execute the critical laser related R&D to achieve necessary seed power, wavelength, pulse duration, and timing synchronization Work in collaboration with ANL, BNL, DESY to demonstrate seeding and cascaded HGHG. Establish a facility at 100 nm for experimental use Work in collaboration with DESY, Jlab, Cornell and others to optimize SC RF technology for CW applications w/ 10-5 amplitude control Explore a new concept pioneered at Bates for greatly simplifying RF systems and significantly reducing costs Develop high rep-rate, high-brightness photoinjector and drive laser in collaboration with LBNL Collaborate with ANL and NHFML to optimize the LCLS undulator design for variable gap performance MIT X-ray Laser Project Critical DUVFEL Experiments 1. Cascaded HGHG experiment. Future x-ray FEL facilities that generate fully coherent radiation will require multiple cascaded HGHG stages starting from a long wavelength seed laser. 2. Chirped pulse amplification using HGHG. Seed FEL with frequency-chirped laser, amplify, and compress optical pulse to produce high power, short time duration output. 3. Start-to-end simulation using measured parameters. Include beam-based measurements of injector RF fields, thermal emittance, photocathode laser time profile, undulator fields, and seed properties. Test codes including parmela, MAD, elegant, and ginger. 4. Seeding with HHG. The Quantum Optics group at MIT is developing high harmonic generation from conventional lasers for use as a short wavelength (10- 100 nm) seed. This has advantages over seeding with low harmonics including requiring fewer HGHG stages, and generating pulse lengths approaching 1 fs. MIT X-ray Laser Project Develop a risk-based prototyping program for all critical components Plan for a broad and inclusive User Program appropriate for a National Facility Educate the scientific community to develop beamline concepts and execute the necessary R&D to support 10 initial beamlines Leverage this R&D program with MIT educational programs to involve graduate students, undergraduates, K-12 students and teachers And finally…. Develop the overall conceptual design, cost and schedule data necessary for a decision to construct We propose a 3-year $15M collaborative effort centered at MIT MIT X-ray Laser Project Conclusions: Technical A multi-beamline X-ray Laser User Facility can be conceived based on existing technology combined with a focused 3-year R&D program. A modular approach with 2 or more stages with increasing linac energy would be a systematic approach, establishing capabilities and proving technology at various cost/performance points versus wavelength. Lower emittance electron guns would have enormous impact, enabling x-ray wavelengths to be reached at conventional (6-8 GeV) energies. Seeding technology would greatly improve performance with highly synchronized transform-limited pulses, and seeding reduces undulator gain lengths and associated costs. CW SC RF is probably essential for synchronization stability, and since cryogenic costs rise rapidly with linac energy and gradient, the lower electron energies achieved with lower emittance guns will be very important. Possible pulse structures are strongly influenced by this choice. MIT X-ray Laser Project Conclusions: Scientific Transform-limited beams from seeded sources will enable science well beyond SASE, for example . . . Seeding for narrow bandwidth will enable pulses as long as 1 ps to have a bandwidth of 2 meV. SASE sources monochromated to meV levels have large fluctuations and low photon flux. Seeding with for short pulses using CPA combined with “compression optics” may allow femtosecond pulses containing 1011 photons or more. This is significantly higher than SASE and of crucial importance for molecular imaging, and chemical dynamics studies. Finally, I believe that virtually all experiments carried out at 3rd generation sources are easily accomplished on such a source at lower facility cost. This will not be an exotic facility for only niche experiments, but represents a source of extraordinary power and flexibility with which all x-ray experiments possible can be done.
Pages to are hidden for
"PowerPoint Presentation"Please download to view full document