MIT X-ray Laser Project

A true x-ray laser will have enormous impact

No x-ray source is coherent No laser has much power for l < 30 nm

The number of photons per quantum state, the photon degeneracy is less than 0.1

Murnane and Kapteyne produced l=31nm light pulses with a nano-Joule per pulse

X-ray Lasers: Promise to be a comprehensive probe of all spatial and temporal scales and resolutions relevant to condensed matter Spatial Scales Temporal Scales

MIT X-ray Laser Project

Unique opportunity to integrate:
Accelerator technology (MIT/Bates Lab) Fast laser technology (MIT Ultrafast Group)

Self-Amplified Spontaneous Emission (SASE)

SASE Radiation has full Transverse Coherence

APS Demonstrates Self-Amplified Spontaneous Emission (SASE)

SASE Radiation is not Transform Limited

N e / N lc = 10 3

N lc / N e 10

A SASE FEL is an amplifier of electron density modulations

SASE Radiation is Powerful, But Noisy

t (fs)

Dw/w (%)

Seeding to Limit Fluctuations

DUV-FEL Facility at BNL

50 m
Copper cathode photoinjector Bunch compressor S-band linac 266 & 89 nm FEL output

S-band linac

Modulator undulator Buncher

Radiator undulator

800 nm laser seed line 30 mJ, 100 fs Ti:Sapphire laser Electron Beam Parameters
Energy Peak current RMS Emittance Pulse length RMS dE/E RMS Emittance 200 MeV 500 A 3 mm-mrad 300 fs 0.1% 3 mm-mrad

FEL Output Parameters

Max output energy Max 3rd harmonic Seed input energy Repetition rate Pulse length 130 uJ @ 266 nm ~1 uJ @ 89 nm 100 uJ @ 800 nm 5 Hz 200 fs

Data from BNL DUV-FEL experiment

Bandwidth and Pulse Length

Seeded beam Output wavelength 100 nm 10 nm FEL param SASE beam SASE Dtmin (fs) 100 100 SASE DEmin (meV) 110 500

rFEL

Dtmin (fs) at max BW 20 5

DEmin (meV) at 1 ps FWHM 2 2

9.e-3 4.e-3

1 nm
0.1 nm

1.5e-3
0.2e-3

1
0.8

2
2

100
100

1900
2500

Seeded beams limited only by

uncertainty principle and seed properties. SASE properties determined

Df Dt

1 2

Df f

by ebeam.

= rFEL

MIT X-ray Laser Project

Provide full transverse and longitudinal coherence
get rid of the SASE noise

Provide wide spectrum coverage: 100 nm > l <0.1 nm

integrate the laser VUV and x-ray communities

Implement a large number of beamlines (10-30)

to host a large and diverse research community to be much more cost effective

Produce x-ray beams that exceed 3rd Generation sources

in flux and brilliance in peak flux and peak brilliance and therefore, in coherence and photon degeneracy

MIT X-ray Laser Project

How to reach wavelengths below 1 nm?
Must get the shortest wavelength seeds using High Harmonic Generation methods, --30nm available now, possible 10 nm or below Then use cascaded High Gain Harmonic Generation methods in FEL,
--factors of >30 are possible

MIT Ultra-fast GroupHHG seeding methods

J. Fujimoto, H. Haus, E. Ippen, F. Kaertner

= =/

x-ray harmonic emission

-4

-2

0 Time, fs

See current issue of Physics Today

High-Harmonic Generation
Noble Gas Jet (He, Ne, Ar, Kr)
100 mJ - 1 mJ @ 800 nm XUV @ 3 30 nm h = 10-8 - 10-5

t
0

Propagation

Recombination

wXUV
Ionization

-Wb Energy

tb

Laser electric field

High Gain Harmonic Generation

Method to reach short wavelength FEL output from longer wavelength input seed laser.
Input seed at w0 overlaps electron beam in energy modulator undulator. Energy modulation is converted to spatial bunching in chicane magnets. Electron beam radiates coherently at w3 in long radiator undulator.

Modulator is tuned to w0.

Electron beam develops energy modulation at w0.

3rd harmonic bunching is optimized in chicane.

Radiator is tuned to w3.

Cascaded HGHG
Output at 3w0 Output at 9w0 Final output

seeds 2nd stage

Input seed w0 1st stage

seeds 3rd stage

at 27w0

2nd stage

3rd stage

Number of stages and harmonic of each to be optimized during study.

Factor of 10 30 in wavelength is reasonable without additional acceleration between stages. Seed longer wavelength (100 10 nm) beamlines with ~200 nm harmonic from synchronized Ti:Sapp laser. Seed shorter wavelength (10 0.3 nm) beamlines with HHG pulses.

Laser System & Synchronization

Fiberlink + Synchronization

Photo-Injector:
1-10 ps Pulses 1-10 mJ 1-20 kHz @ 266 nm (conv. NLO)

~200 m

High Harmonic Generation

1 nJ 10 nJ 100 as 10 ps 1-20 kHz @ 1 - 30 nm

10 fs Timing Jitter

E-beam

LINAC

FEL

Output: Three highly synchronized pulse streams E-beam, EUV 1 - 30 nm and @ 800 nm driver pulse

MIT X-ray Laser Concept

Main oscillator Fiber link synchronization

Seed laser

UV Hall
Undulators 100 nm 30 nm

Pump laser

Seed laser

X-ray Hall

Pump laser

Undulators 1 nm 0.3 nm

Injector laser

10 nm

SC Linac

0.3 nm

SC Linac

0.1 nm

1 GeV

2 GeV

4 GeV
10 nm
3 nm 1 nm Undulators Seed laser Pump laser

Upgrade: 0.1 nm at 8 GeV

Nanometer Hall

Nanometer Hall
to master oscillator for timing sync

Direct seeded or cascaded HGHG undulators 10 nm Ti:Sapp + BBO = 200 nm seed

Pump
lasers

Seed
lasers

~20 m length 10 GW peak Cascaded HGHG undulators 3 nm

Ti:Sapp + HHG = 10-30 nm seed Tune by OPA or harmonic number

Cascaded HGHG undulators Ti:Sapp + HHG = 10-30 nm seed Tune by OPA or harmonic number ~30 m length 4 GW peak

1 nm

Pulse Structure (Quasi-CW)

RF Gun Pulse 0.1% Duty Factor 8 Pulses 8 Beamlines ~500 pC / Pulse 1 us spacing 20

-10

Time (us)

10

~90 Warm RF Gun Pulses

1 ms spacing

0 Time (ms)

100

SC Linac Pulse

@1Hz

10% Duty Factor

500

1000

1500

2000

2500

Time (ms)

Seeding for short pulse

Output time profile
2

Time profile (log plot)

10 10
10

Spectrum
1000 800 600 400 200

2
Power (GW)

1.5

Power (GW)

1.5 1 0.5 0

Power (W)

1 0.5 0 24.5

10 10 10 10

25

25.5 26 Time (fs)

26.5

27

Power (kW/bin)
0 10 20 30 Time (fs) 40 50

0 0.2995

10

20 30 Time (fs)

40

50

0.3 0.3005 Wavelength (nm)

0.301

GINGER simulation of seeded FEL at 0.3 nm.

Same ebeam parameters as SASE case.

Seed laser parameters FWHM Power Pulse energy 0.5 fs 10.0 MW 5 nJ

FEL output parameters Saturation FWHM Saturation power Saturation energy FWHM linewidth Undulator length 0.75 fs ~2.0 GW 1.5 mJ 6.0E-4 20 m

Seeding for narrow linewidth

Output time profile
2

Time profile (log plot)

10

Spectrum
500

10
Power (MW/bin)
8

Power (GW)

1.5

10

400 300 200 100 0 0.2995

Power (W)
0 10 20 30 40 50

1 0.5 0 Time (fs)

10
4

10
2

10
0

10

10

20 30 Time (fs)

40

50

0.3

0.3005

0.301

Wavelength (nm)

GINGER simulation of seeded FEL at 0.3 nm.

Same ebeam parameters as SASE case.

Seed laser parameters FWHM Power Pulse energy 50 fs 0.1 MW 5 nJ

FEL output parameters Saturation FWHM Saturation power Saturation energy FWHM linewidth Saturation length 30 fs ~2.0 GW 0.1 mJ 1.0E-5 28 m

Comparison of SASE and Seeded Sources with APS Undulator A

Cost Basis
Fixed Costs
(Gun, X-ray Beamlines, Buildings, Cryoplant, Controls)

80 M$

Linac Systems (20 MeV/m, ~0.4M$/m) Undulator Systems (0.2 M$/m) 20M$/100m
Total Undulator Length = 4 x longest saturation length

20 M$/GeV

Contingency

25%

Example
4 GeV Linac 50 m Saturation Length
Costs: 80 M$ 80 M$ 40 M$ 50 M$ -----------250 M$ Fixed Linac Undulators Contingency Total

1000

Saturation Length (m)

100

10

1 0 5 10 Electron Energy (GeV) 15 20

1000

u = 18 mm

u = 23 mm

u = 30 mm

Saturation Length (m)

0.1 nm

100
0.3 nm 1 nm 0.15 nm (LCLS)

10

10 nm

100 nm

1 0 5 10 Electron Energy (GeV) 15 20

Hybrid Undulator Parameters VISA: = 18 mm, K=1.4, B=0.8 T 23mm: = 23 mm, K=2.4, B=1.1 T LCLS: = 30 mm, K=3.9, B=1.4 T

Electron Bunch Parameters Q = 0.5 nC E/E = 0.02% T = 250 fs = 1.5 m

Superconducting Undulator = 14 mm K = 1.3

Better Gun = 0.75 m

1000
0.1 nm

Hybrid Undulator Parameters VISA: = 18 mm, K=1.4, B=0.8 T 23mm: = 23 mm, K=2.4, B=1.1 T LCLS: = 30 mm, K=3.9, B=1.4 T

Saturation Length (m)

100

0.3 nm

1 nm

10

10 nm

100 nm

1 0
Superconducting Undulator Miracle Gun = 0.1 m

10 Electron Energy (GeV)

15

20

Electron Bunch Parameters Q = 0.5 nC E/E = 0.02% T = 250 fs = 1.5 m

Essential to Improve e-Gun Performance

In linacs, electron emittances scale inversely with energy

Electron beam emittance is born at the electron gun

Electron gun emittances today are ee = 0 .5 nm / E (GeV)

Photon emittances for full transverse coherence ep = lp /4

To couple a given electron beam most effectively to a coherent photon field, we should have:

ee = ep

MIT X-ray Laser Concept

Main oscillator Fiber link synchronization

Seed laser

UV Hall
Undulators 100 nm 30 nm

Pump laser

Seed laser

X-ray Hall

Pump laser

Undulators 1 nm 0.3 nm

Injector laser

10 nm

SC Linac

0.3 nm

SC Linac

0.1 nm

1 GeV

2 GeV

4 GeV
10 nm
3 nm 1 nm Undulators Seed laser Pump laser

Upgrade: 0.1 nm at 8 GeV

Nanometer Hall

The MIT X-ray Laser Project

A National User Facility: 10-30 beams Wavelength range 100-0.1 nm Integrated laser seeding for full coherence Pulses: Dt=1-1000 fs; Dw=3-0.003eV Pulse power of up to 1 mJ Pulse rates of 1 kHz or greater

MIT/ Bates Laboratory

Science: single molecule imaging, femtochemistry, nanometer lithography Technology: superconducting FEL, Ti:Sapp HHG seeding technology

Education: accelerator science curriculum, synergy with CMSE programs

Cost/Schedule: $300M; design: FY04-FY06; construct: FY07-FY10

MIT Commitment
MIT has embraced the x-ray laser concept exclusively for the future of Bates Laboratory Deans of Science and Engineering and the VP of Research provided over $400K in seed support President Vest asked a key CEO to chair a corporation-level advisory committee to secure support of business and political leaders in MA

Charge to
MIT X-ray Laser Accelerator Science Advisory Committee

September 18-19, 2003

The proposed MIT x-ray laser facility is at an early stage of conceptual design. The goals of the design are to produce fully coherent x-ray pulses with the stable and reliable operations required of a user facility. We seek guidance and constructive criticism regarding the technical choices that are being made.

The ASAC committee should:

Review laser and accelerator sections of proposal to NSF and technical presentations at committee meeting. Evaluate the appropriateness of chosen technologies and suggest alternatives. Identify the primary technical challenges for each system and for the facility as a whole.

Respond to NSF reviewer comments.

Evaluate the potential for a facility based on the Bates linac to demonstrate laser seeding and cascaded HGHG, and selected scientific applications

MIT X-ray Laser Design Proposal

3-year duration, $15M total request
Contact: David E. Moncton, Director Telephone: 617-253-83333 E-mail: dem@[Link] website: [Link]

Co-Principal Investigators William S. Graves Franz X. Kaertner Richard Milner Simon Mochrie Gregory Petsko Henry I. Smith

Science Collaborators Keith A. Nelson Dagmar Ringe Andrei Tokmakoff

Bates Senior Staff Manouchehr Farkhondeh Jan van der Laan Christoph Tschalaer Fuhua Wang Abbi Zolfaghari Townsend Zwart William M. Fawley Hermann Haus Ian McNulty Jianwei Miao Mark Schattenburg

Contributors James Fujimoto Erich Ippen Denis B. McWhan Michael Pellin Gopal K. Shenoy

Existing Technology
Electron Guns
Adequate performance has been demonstrated. Room for continuing R&D and improvement. Not a cost driver.

Existing Technology
Linac
Successful operation at Tesla Test Facility, JLAB. Capital cost driver.

Existing Technology
Undulator
Well established. Successful experience at LEUTL, TTF. Make use of investment in LCLS design. Capital Cost driver.

3-year study plan