The Quest for Ultra-Short X-ray Pulses

1Zholents, 03-09-2011

The Quest for Ultra-Short X-ray Pulses

Alexander ZholentsANL

Fermilab, 03/09/2011

2Zholents, 03-09-2011

J. Levesque and P.B. Corkum, Can. J. Phys. 84: 1–18 (2006)

2010

“…shorter and shorter laser pulses were obtained since discovery of mode locking in 1964. Each advance in technology opened a new field of science and each advance in science strengthened the motivation for even shorter laser pulses.”

Thomson Slicing

SPPS

LCLS

X-ray pulses

Expected

3Zholents, 03-09-2011

Scientific drivers. Why ultra-short pulses?· Phase transitions in solids

· Surface dynamics

· Making and braking of bonds during chemical reactions

· Chemical dynamics in proteins

· Correlated behavior of electrons in complex solids

… for showing that it is possible with rapid laser technique to see how atoms in a molecule move during a chemical reaction

t ~1Å / (speed of sound) ~ 100 fs

Femtochemistry

Ahmed Zewail,Nobel Prize, 1999

The vibration period between the two hydrogen atoms in a hydrogen molecule is about 8 fs

4Zholents, 03-09-2011

Molecule of rhodopsin before …

CH N

1) R.W. Schoenlein, L.A. Peteanu, R.A. Mathies, C.V. Shank, Science, 254, 412 (1991).

… and after absorption of a photon

First step in vision1)

How living systems function using smallest amount of energy?

5Zholents, 03-09-2011

Charles Shank

In 1981 Charles Shank and coworkers at Bell Labs redefined the term “ultrafast” when they reported the demonstration of 90 fs long pulses from colliding-pulse dye laser 1

1) R.L. Fork, B.I. Greene, C.V. Shank, Appl. Phys. Lett., 38, 671(1981).

Probe valence electrons with optical laser

Probe core electrons with x-rays

6Zholents, 03-09-2011

Laser assisted techniques: 90o Thomson scattering1

1) S. Chattopadhyay, K.-J. Kim, C. Shank, Nucl. Instr. Meth., A341, 351 (1994). 2) R.W. Schoenlein et. al., Science, 274(11), 236 (1996).

300 fs

7Zholents, 03-09-2011

90o Thomson scattering (2)

300 fs

1) A.H. Chin, R.W. Schoenlein, T.E. Glover, P. Balling, W. P. Leemans, C. V. Shank, “Ultrafast Structural Dynamics in InSb Probed by Time-Resolved X-Ray Diffraction”, Phys. Rev. Lett. 83, 336 - 339 (1999).

Lattice expansion dynamics in InSb as a function of time delay between laser and x-ray pulses1

X-ray probe before the laser pump

X-ray probe after the laser pump

8Zholents, 03-09-2011Femtosecond X-ray Beamline

First Light2000

In-vacuum Undulator

Femtosecond Laser ~50 W

“Slicing” source of fs x-ray pulses at the ALS1

1) R.W.Schoenlein et al , Science, March 24, (2000)

wiggler

100 fs electron slice

undulator beamline

70 ps electronbunch

100 fslaser pulse

spatial separationby dispersive bend

lW

Laser - ebeam interaction in wiggler

100 fs x-ray pulse

synchronous to laser

9Zholents, 03-09-2011

Electron trajectory through wiggler

22/ zw llWiggler period

Magnetic field in the wiggler

B B0sin(kuz)

Laser wavelengthFEL resonance condition

While propagating one wiggler period, the electron is delayed with respect to the light on one optical wavelength

=

Light interaction with relativistic electron*)

E

k

B

E

k

B V

V 0)(2 dtVEmc

e

Elaser N

S

S

NS

SN

N

e-

lu

light

*) Motz 1953; Phillips 1960, Madey 1971

Energy modulation of electrons in the wiggler magnet by the laser lightenergy modulation over the length of one optical cycle

10Zholents, 03-09-2011

Interference term defines energy gain/loss

Alternative approach to find energy modulation

Far field observer

Far field observer sees the field: )()()( tEtEtE sL

laser spontaneous emission

dEEdEdEdE LsLs *222 2

spontaneous emission

)sin(2 sL AAE laser

Electronbeam

11Zholents, 03-09-2011

Selection of fs x-ray pulses

DE

z

100 fs

2sE

Electron beam energy modulation

DX

z

100 fs

2sx

In dispersive placeDX=D•DE/E

Radiation of these electrons is selected

12Zholents, 03-09-2011

Other means to select fs pulse besides coordinate separation

spectral selection (works only with undulator source)2

Nn1)(

0

Bandwidth of the

undulator radiation at a high harmonic number n

1) R.W. Schoenlein, et al., Appl. Phys. B71, 1 (2000).2) H. Padmore, private discussion

Selection of fs x-ray pulses1

time-off-flight selection

-

+

angular selection

modulator dipole magnet

radiator

fs laser pulse fs x-ray pulseelectron bunch

aperturewiggler undulatorbend

magnet

electron bunchlaser pulse

subps x-ray pulse

mask

BESSY: courtesy S. Khan Select before first x-ray mirror

13Zholents, 03-09-2011

bend magnet visible radiation

Ti:Sa laser

Ti:Sa amplifier

BBOe-

slit

hn2 = ~2 eV

hn1 = 1.55 eV

hn =hn1+hn2

Measurement of a short pulse of synchrotron radiation via cross-correlation with a short laser pulse

intermediate focus

PMT

filter

e-beam

fs pulsefs pulse

light

Diagnostics

-60 -40 -20 0 20 40 600

50100150200250300350

# ph

oton

s

delay (ps)

s = 16.6 ps

14Zholents, 03-09-2011

-1200-1000 -800 -600 -400 -200 0 20060

80

100

120

140

160

180

delay (fs)

coun

ts

-1200-1000 -800 -600 -400 -200 0 200

100

150

200

250

coun

ts

-1500 -1000 -500 0 500 1000 1500

2000

2500

coun

tsElectron Density Distribution

time (ps) 0.3-0.3

x/s x

+3sx to +8sx

+4sx to +8sx

-3sx to +3sx

Diagnostics (2)ALS bend magnet

beamline

Modulation amplitude extracted from experimental data is 6.4 MeV

(expected ~ 10 MeV)

15Zholents, 03-09-2011

Diagnostics (3)

Electron loss rate versus scraper position

Deducing energy modulation amplitude from beam lifetime measurements using scraper in dispersive location (BESSY)

Energy modulation amplitude versus laser pulse energy

Approximately two times more laser pulse energy was used than it was predicted (same problem at ALS)

Not explained so far !

theory

experiment

laser pulse energy, (mJ)

LAEE ~D

without laser

withlaser

courtesy S. Khanscraper position, (mm)

16Zholents, 03-09-2011

“Slicing” at BESSY 1)

0 1 2 3 m

x-raypump pulse

e

1) S. Khan, et al., PRL, 97, 074801 (2006)

Detected photon rate per 0.1% bandwidth versus cutoff angle

phot

ons/

(s 0

.1%

bw

)

sign

al-to

-bac

kgro

undwithlaser

without laser

signal/background

angle, (mrad)

17Zholents, 03-09-2011

l1l2

wiggler detuning form

FEL resonance

Gain, %

I ≈ 25A

2EdEd

D

bunch train and a gap

Measurement of FEL gain through wiggler at ALS

Small-signal FEL gain

Madey’s theorem

FEL gain ~

laser spectrum gain

function

laser

e-beam

0.4

0.2

measured with laser oscillator

FEL resonance

-60 -40 -20 0 20 40 60 80

0

1

2

3

4

5

gain

(x10

-3)

delay (ps)

electron bunch laser pulse

s = 16.6 ps

Gain follows peak current distribution

18Zholents, 03-09-2011

first turn

second turn

1.2

time (ps) frequency (THz)

Rel

ativ

e ch

arge

den

sity A

mplitude (arb. units)

0.4

1.04

0.92

Laser-induced coherent fs THz radiation1-4

dtetI ti )(~

e-beam density distribution

Dip in the electron density distribution expands and dries out as the electron bunch travels along the ring

e-beam density distribution

time (ps) 0.3-0.3

x/s x

1) R.W.Schoenlein et al , Appl. Physics, B71, 1 (2000).2) J. Byrd et al , Phys. Rev. Lett., 96, 164801(2006).3) K. Holldack et al , Phys. Rev. Lett., 96, 054801 (2006).4) J. Byrd et al , Phys. Rev. Lett., 97, 074802 (2006).

Large, medium and

small modulationamplitudes

19Zholents, 03-09-2011

Dip reconstruction from THz spectra1

1) K. Holldack et al , Phys. Rev. Lett., 96, 054801 (2006).

wave number, (1/cm) time, (ps)

P coh

/Pin

c, (a

rb. u

nits

)

elec

tron

den

sity

, (ar

b. u

nits

)

THz signal is now routinely used for tuning of laser e-beam interaction and to maintain it with a feedback on mirrors – BESSY, ALS, SLS

20Zholents, 03-09-2011

Black: measurementRed: theoretical fit

Alternative approach to find energy modulation*

LaserSpontaneous emission An example of a poor

overlapping: wiggler is detuned too far away from the laser frequency

*) A. Zholents, K. Holldack, Proc. FEL’06, Berlin (2006).

In the experiment, wiggler gap was adjusted such as to change the central frequency of spontaneous emission from 200 nm to 1000 nm laser

21Zholents, 03-09-20111) A. Zholents, P. Heimann, M. Zolotorev, J. Byrd, NIM A, 425, 385 (1999).2) M. Katoh, Japan. J. Appl. Phys, 38, L547(1999)

Radiation from tail electrons

angle >> beam divergence + x-ray diffraction

Radiation from head electrons

undulator

Storage Ring

Deflecting cavity delivers a time-dependent vertical kick to the beam

Compression using

asymmetrically cut crystal

Trading brightness to a short pulse

Ideally, second cavity cancels effect of first cavity

~1 ps

High repetition rate source of ps x-ray pulses1,2

22Zholents, 03-09-2011

Application to the Advanced Photon Source1

Part of the APS upgrade pursued in an active collaboration with JLab

Pulse length versus transmission through aperture calculated for 10 keV photons

courtesy M. Borland

6 MV deflection

rf harmonicnumber

Obtaining short x-ray pulses at APS1-3

1) K. Harkay et al, Proc. PAC’05, 668(2005).2) M. Borland, Phys. Rev. ST – AB, 8, 074001(2005).3) M. Borland et al, Proc. PAC’07, (2007).

diffraction limited

Pulse length versus photon energy

emittance limited

rfz

rfy

z

rayx

zy )()(s

ss

sD

saturated at

rf kick

23Zholents, 03-09-2011

Synchronization between laser pump and x-ray probe pulses

LASER OSCILLATOR(passively modelocked)

Dt Dt3.0 GHz

(RF Kick)electron bunch

laser pulse

x-rays

Synchronous bunch

Jitter in the e-beam arrival time is converted into position

jitter of compressed x-ray pulse

Early bunchAsymmetrically cut crystal

Late bunch

Sensitivity to electron bunch timing jitter is significantly reduced

aperture

or intensity jitter behind the aperture

24Zholents, 03-09-2011

add 14-meter chicane in linac at 1/3-point (9 GeV)

1-GeV Damping Ring

sz 1.1 mm sz 40 mm sz 12 mm

sz 6 mmSLAC Linac

30 GeVFFTB

Short Bunch Generation with the SLAC Linac - SPPS

Existing bends compress to 80 fsec

1.5 Å

compression by factor of 500

P. Emma et al., PAC’01

1.5%

28.5 GeV

80 fs FWHM

30 kA

measurement < 300fs

Single pass, low rep. rate

25Zholents, 03-09-2011

New Scientists

Coherent emission of x-rays using Free Electron Lasers: a road to attosecond x-ray pulses

nm30sec10100sec/m103 188 sec10100

137sec/m103 18

8

2 Å

Zholents, Avalon, 10/2010E. Goulielmakis et al., Nature, 466, 739(2010).

Probing intra-atomic electron motion by attosecond absorption spectroscopy.

80 attoseconds !

June 19, 2008

Attosecond XUV pulses had been obtained

27Zholents, 03-09-2011

Attosecond x-ray pulses is a powerful tool for addressing Grand Challenges in Science and BES Research Needs

• How do we control materials and processes at the level of electrons?

• How do we design and perfect atom-and energy-efficient synthesis of new forms of matter with tailored properties?

• How do remarkable properties of matter emerge from complex correlations of atomic and electronic constituents and how can we control these properties?

• Can we master energy and information on the nanoscale to create new technologies with capabilities rivaling those of living systems?

• How do we characterize and control matter away—especially very far away—from equilibrium?

2007G. Fleming

Zholents, Avalon, 10/2010

Short EUV/x-ray pulses are routinely produced at FLASH (10 – 70 fs), SCSS (~ 30 fs) , LCLS (<10 – 80 fs)

All future x-ray FEL projects consider ultra-short x-ray pulse capabilities

Zholents, Avalon, 10/201029

20-pC bunch operation at LCLS1

Photo-diode signal on OTR screen after BC21) Y. Ding et. al, PRL 2009

FEL gas detector measurements

BL signal

+1deg-1 deg

soft x-rayh=840 eV

Simulated signal

Simulated bunch length

X-ray pulse duration should be <10 fs, but no direct

measurement yet possible

(9/29/2009)

~ 1mm

Soft x-rays at 1.5 nm (simulations for LCLS)1

Under-compression: +1 deg off Full-compression Over-compression: -1 deg off

z = 40m z = 40mz = 40m

Actual measurements qualitatively confirm simulations. Direct measurement of ultra-short x-ray pulse duration remains to be difficult.

At undulator entrance, 4.3 GeV, Laser heater off1) Y. Ding ,LBNL workshop, 08/2010

~ 3 fs ~ 3 fsFEL peak power is ~ 3 orders lower

Too narrow pulse: slippage larger than pulse duration

31

~1 fs

at 70 m~150 mJ

L1 = -22 deg

Simulation at 13.6 GeV At 1.5 Å FEL performs well at full compression (slippage just right)

Not short enough for FT limited pulse

Y. Ding ,LBNL workshop, 08/2010

Gain length =2.74 m

140 mJ FEL energy

J. Frish et. al., talk at FEL’09

Measurement

deflecting angle

SUB-FEMTOSECOND X-RAY PULSES USING THE SLOTTED FOIL METHODP. Emma, M. Cornacchia, K. Bane, Z. Huang, H. Schlarb ,G. Stupakov, D. Walz , PRL, 2004

BC24.3 GeV

14 GeVBC1250 MeV

135 MeV

gun 4 wirescanners to FEL

h=8 keV

Linac Coherent Light Source, SLAC

200 fs

7 M

eV

Only bright spot will lase

z

DE

energy chirp

X-ray pulse width ~ 400 asecBunch arrival time jitter ~ 50 fs~ 109 photons/pulse -> 7 GW

Courtesy P. Emma

Simulation using Elegant

time (fs)

Pow

er (G

W)

0

10

5

0-150 fs

2 fs

2-Pulse Productionwith 2 slots

0-6 mm

0.25 mm

pulses not coherent

Double X-Ray Pulses from a Double-Slotted Foil

Courtesy P. Emma

FEMTOSECOND X-RAY PULSES IN THE LCLS USING THE SLOTTED FOIL METHOD

Precise controlled time delay between x-ray pump and x-ray probe pulses

Slotted Foil in at -6000 um

Slotted Foil in at -34000 um

Slotted Foil in at -37000 um

Over-compressed bunch introduces e- energy chirp on dump screen… ~time

~time

~time

time

dumpscreen

sing

le s

lot

doub

le s

lot

doub

le s

lot

HEAD

TAIL

Courtesy P. Emma

Initial results (very preliminary), June 17, 2010

X-ray pulse energy

slots e-bunch

Wider slot gives more energy

Zholents, Avalon, 10/2010

Pump – probe studies using ultra-fast x-ray pulses

Stop-motion photography E. Muybridge, 1878

Pellet hits a strawberry

delay

x-ray probe

visible pump

36Zholents, 03-09-2011

Options for experiment utilizing synchronized pump and probe signals when electron bunch arrival time has a “large” jitter

Chicane Undulator

IR pump, ~100 MW

X-ray probe

Single color x-ray pump – x-ray probe experiment

Split single x-ray pulse into two and adjust delay

Create pump signal using cohrent undulator radiation and adjust delay1 (in case of an ultra-short e-bunch, ~ 1 fs)

1) U. Fruhling et al., Nature photonics, 3, 523(2009); F. Tavella et al., Nature photonics, 5, 162(2011)

Use double slotted foil

37Zholents, 03-09-2011

Precision synchronization of pump and probe pulses Seeded FELs naturally posses precise synchronization

if electron bunch length > laser pulse + jitter

Current-enhanced SASE FEL --> same conclusion (Zholents, 2004) BunchingAccelerationModulation

Pea

k cu

rren

t

t

20-25 kA 10 fsAfter bunching

This part will lase

~50 fs

laser

e-bunch

10 fs

Require synchronization of the seed and pump lasers 1)

1) J. Kim et. al., Nature Photonics, 2, 733, 2008;

R. Wilcox, Opt. Lett. 34, 3050, 2009

38Zholents, 03-09-2011

e-beam ~ 50 fs

e-beam ~ 50 fs

Laser pulse ~ 5 fs

)t(E

Attosecond pulse generation via electron interaction with a few cycle carrier-envelop phase stabilized

laser pulse

Basic idea:Take an ultra-short slice of electrons from a longer electron bunch to produce a dominant x-ray radiation

Small jitter in the electron bunch arrival time is not important – good for pump-probe experiments using variety of pump sources derived from initial laser signal

39Zholents, 03-09-2011

Electron trajectory through wiggler

22/)2/21( ll KwL Wiggler period

Laser wavelengthFragment of the electron bunch

While propagating one wiggler period, the electron is delayed with respect to the light on one optical wavelength

Light interaction with relativistic electron

E

k

B

E

k

B V

V 0)(2 dtVEmc

e

Elaser N

S

S

NS

SN

N

e-

lu

light

Laser pulse: 1 mJ, 5-fs

at 800 nm wave length with CEP stabilization

sE

Sin-wave Cos-wave

40Zholents, 03-09-2011

Energy modulation induced in the electron bunch during interaction with a ~1 mJ, 5 fs, 800 nm wave length laser pulse in a two period wiggler magnet with K value and period length matched to FEL resonance at 800 nm

Current enhancement

Pea

k cu

rren

t This spike gives a dominant signal

41Zholents, 03-09-2011

Energy modulation of electrons produced in interaction with two lasers

Current enhancement method *)

Combined field of two lasers:increases one laser bandwidth

t (fs)

Elec

tric

fiel

d

*) Zholents, Penn, PRST-AB, 8, (2005); Y. Ding et al., Phys. Rev. ST-AB, 12, (2009).

800 – 1000 nm laser; 7.5 – 25 fs; 0.2 – 0.5 mJ

Undulator with ~ 1% taper

1300 – 1600 nm laser; 10 – 45 fs; 0.07 – 0.2 mJ

42Zholents, 03-09-2011

one lasertwo lasers

Selection of attosecond x-ray pulse:regions with higher peak current reach saturation earlier

Contrast ≈ 1 (assuming 100 fs long bunch ) Nearly Fourier transform limited pulse

t (fs)

curr

ent,

(kA

)

t (fs)

10 mJ1010ph

Pow

er, (

GW

)

250 asec

X-ray wavelength = 0.15 nm

Current enhancement method (2)

43Zholents, 03-09-2011

[1] A. A. Zholents and W. M. Fawley, Phys. Rev. Lett. 92, 224801 (2004).[2] E.L. Saldin, E.A. Schneidmiller, M.V. Yurkov, Opt. Commun. 237,153 (2004).[3] E.L. Saldin, E.A. Schneidmiller, M.V. Yurkov, Opt. Commun. 239,161 (2004).[4] A. A. Zholents and G. Penn, Phys. Rev. ST-AB 8, 050704 (2005).[5] E.L. Saldin, E.A. Schneidmiller, M.V. Yurkov, Phys. Rev. ST-AB 9, 050702 (2006).[6] A. A. Zholents and M.S. Zolotorev, New J. Phys. 10, 025005 (2008).[7] W.M. Fawley, Nucl. Inst. and Meth. A 593, 111(2008).[8] Y. Ding, Z. Huang, D. Rather, P. Bucksbaum, H. Maerdji, Phys. Rev. ST-AB 12, 060703 (2009).[9] D. Xiang, Z. Huang, G. Stupakov, Phys. Rev. ST-AB 12, 060701 (2009).[10] A. A. Zholents and G. Penn, Nucl. Inst. and Meth. A 612, 254(2010).

Publications exploring generation of attosecond x-ray pulses using a few-cycle laser pulse with a carrier envelop phase stabilization:

Zholents, Avalon, 10/2010

~ 200 as

Contrast ≈ 1

Energy modulation

lx=0.15 nm

cdtd

KK

dzKd

u

x ll ln

2/2/1ln

2

2

Tapered undulator method1

Energy chirp is compensated by the undulator taper in the central slice

1) E.L. Saldin, E.A. Schneidmiller, M.V. Yurkov, Phys. Rev. ST-AB 9, 050702 (2006).

2)/( tt s

Frequency chirp definition

Wigner transformof the on-axis far field

-2.5 0-5t, fs

7.9

7.6

7.8

Wav

elen

gth,

nm Chirp

≈ 0.57.7

2.5

W.M. Fawley, Nucl. Inst. and Meth. A 593, 111(2008).

Fourier transform limited pulse ~ 1.5 fs (FWHM)

Hard x-rays Soft x-rays

With two laser one can manipulate the energy chirp and, thus, the frequency chirp

Zholents, Avalon, 10/2010

Photoelectron probability

Pho

toel

ectro

n m

omen

ta

“Sudden” photoionization creates a coherent superposition of electronic states, G. Yudin, PRL, (2006)

The figure-of-merit is broad bandwidth of attosecond pulses

Zholents, Avalon, 10/2010

Artist’s (Denis Han) view of excited electron wavepackets in molecule created by core excitation with attosecond x-ray pulses (courtesy S. Mukamel)

Intense attosecond x-ray pulses from FELs provide the opportunity to probe the matter on atomic scale in space and time

Stimulated X-ray Raman spectroscopy *)

X-ray pump, X-ray probe;element specific

ener

gy

v-bandEF

core level

phot

on in

photon out

DE~10eV

hnout, kout

300 asec -> 6 eV, i.e. “sudden” excitation reveals multi-electrondynamics

*) Schweigert, Mukamel, Phys. Rev. A 76, 012504 (2007)

In molecules all electrons move in a combined potential of ion core and other electrons

x-rays

UndulatorWigglerLinac

e-beam2 mm laser

TEM00

TEM10

Selection of attosecond x-ray pulses via angular modulation of electrons*)

*) Zholents, Zolotorev, New Journal of Physics, 10, 025005 (2008).

Hermite-Gaussian laser modes

)cos()( 0 zkxzx DD

)sin()( 0 zkEzE DD2 mm laser

10-1

0 Px-

ray

(W)

t (fs)

X-ray peak power as a function of time

Combining angular and energy modulations for improved contrast of attosecond x-ray pulses

100 GW 115 as, 10 mJ~100mJ/cm2

Central peak

Contrast > 100

X-ray wavelength = 0.15 nm

Obtaining attosecond pulses at 1 nm using echo effect*)

*) D. Xiang, Z. Huang, G. Stupakov, Phys. Rev. ST-AB 12, 060701 (2009).

Adding energy chirp via interaction with ultra short laser pulse; needs sub-fs synchronization

radiator has only 12 periods

x-ray pulse

20 asec

2 2

wiggler1chicane1

radiator11

wiggler2 wiggler3 radiator2chicane2 chicane3

Two color attosecond pump and attosecond probe x-ray pulses*)

DT

*) Zholents, Penn, Nucl. Inst. and Meth. A 612, 254(2010).

z/l

DE/s

E

Fragment of the longitudinal phase space

z/l

DE/s

E

Adjustable time delay with better than 100 asec precision

zoom

Zoom into the main peak shows microbunching at 2.28 nm

DT

ħ=410 eV ħ=543 eV

7.2 eV FWHM

6.4 eV FWHM

543 eV210 asec FWHM

410 eV480 asec FWHM

DT

Simulation results using 1D code and GENESIS for two color scheme

time, (fs)

photon energy, (eV)

adjustable

Molecule structure of 5-quinolinol

Zholents, 03-09-2011 52

Three synchrotron light sources ALS, BESSY, SLS routinely operate with 100 – 200 fs x-ray pulses

SOLEIL, SPring8, APS pursuing plans to add ulrafast x-ray science capacities

X-ray FELs FLASH, SCSS and LCLS routinely work with ultra-short XUV/x-ray pulses.

Summary

53Zholents, 03-09-2011

Summary: FELs – the light fantastic

We are at the threshold of a new era of science, where for the first time, the new instruments, the x-ray FELs, are capable to study the matter with a single atom time and space resolution.

Remarkably, adding attosecond capabilities to existing FELs require rather modest modifications.

P. Bucksbaum, Science, 317, 766(2007).

Conical intersection

pump

Attosecond probe

Expected soon:X-ray pulses million times brighter and thousand times shorter than anything can be done using spontaneous emission

54Zholents, 03-09-2011

Acknowledgement

I greatly benefited from discussions with many people including:

M. Borland, J. Byrd, J. Corlett, P. Emma, W. Fawley, E. Glover, P. Hiemann, K. Holldack, Z. Huang, S. Khan, M. Martin, B. McNeil, C. Pellegrini, G. Penn, V. Sajaev, F. Sannibale, R. Schoenlein, J. Staples, G. Stupakov, M. Venturini, J. Wurtele, M. Zolotorev

,

Zholents, Avalon, 10/2010

Thank you for your attention

Zholents, Avalon, 10/2010

Back-up slides

57Zholents, 03-09-2011

Recent studies show that a smaller emittance (by a factor of 10), and very short, ~ 1 fs or less, electron bunches can be produced if the electron bunch charge is dropped to 1 -10 pC

Ultra-short X-ray pulses with low charge electron bunches 1- 10 pC *)

UCLA

*) J.B. Rosenzweig et al., NIM A 593, 39 (2008); C. Pellegrini, seminar at LBNL, 07/11/08

peak current

s 2

IB

Normalized slice emittance Slice energy spread

Electron beam brightness can be increased by a factor 10 - 100

Road to compact FELs

58Zholents, 03-09-2011

Example of simulations for LCLS with 1 pC*)

*) J.B. Rosenzweig, talk at FEL workshop, LBNL, November 2008

Improved brightness -> reduced saturation length

saturation at ~70 m

~350 asec

Difficult to synchronize to external source due to electron bunch arrival time jitter

59Zholents, 03-09-2011*) Saldin, Schneidmiller, Yurkov, Optics, Com., 239 (2004)

Chicane

x-rays

UndulatorUndulatorWigglerLinac

e-beam x-rays

0 + 0

• seed signal interacts with “virgin” electrons in the second undulator

• second undulator is tuned to favor SASE for electrons at the offset frequency 0+

• Contrast >> 1 (assuming 100 fs long electron bunch)

no saturation

~300 asec, ~5-25 mJ

Attosecond x-ray pulse

Spectral separation plus seeding *)

seed 0 +

X-ray wavelength = 0.15 nm

Selection of attosecond x-ray pulse

60Zholents, 03-09-2011

Echo effect for harmonic generation of x-rays*)

DE/sE

z/l*) G. Stupakov, PRL, 102, 074801(2009).

FEL undulator