Frequency comb lasers

on the move to extreme wavelengths

Kjeld Eikema

LCVU day17 Dec. 2008

Christoph Gohle, Dominik Kandula, Tjeerd Pinkert

Anne Lisa Wolf, Steven van den Berg (NMI)

Wim Ubachs, Wim Hogervorst

Outline

Introduction frequency comb lasers Goals

First comb excitation of Ca ions in a trap First comb excitation of Helium, at 51 nm Summary & outlook

Addition of waves to pulses

Modelocked laser ~ frequency comb

More than a million lasing modes canwork together to form < 5fs pulses

Laser resonator

Time and frequency picture

frequency

Int.

0

time

Equivalent !!!

Frequency comb structure

frequency

Int.

0

R. Holzwarth et al. PRL 85, 2264 (2000), D.J. Jones et al. Science 288, 635 (2000)

fn = f0 + n.frep

frep = c / 2Lf0 = frep x φCE / 2

Controlled repetition rate + Phase relationship between the pulses = Frequency comb

time

φCE= /2φCE= 0

vg ≠ vφ

Time & frequency consequences

Very precise pulse control

Attosecond timing resolution revolution Reproducible phase coherent optical pulses Absolute long distance measurement

Very precise frequency ruler for THz-PHz

Optical frequency calibration revolution Optical atomic clocks ‘Fingerprinting’ parallel spectroscopy

Ti:Sa Frequency comb implementation LCVU

2 Ti:Sa based comb lasers1 Fiber-based comb laser (Menlo Systems)

frep: 60-300 MHznJ pulse energy500 – 2000 nm

Fiber frequency comb LCVU

600-900 nm & 1100-2000 nm, frep = 250 MHz

Project of Valdas Maslinskas

Measuring with frequency combs

frequency

Int.

0

f0 fr

Atomic clock

f

VIS-NIR comb

12.45239828 MHz

CW ultra-stable laser

Precisionspectroscopy

456 243 268 532 146.7 Hz

Counter

Goal 1: trace attosecond electron dynamics

Attosecond dynamics in mK molecular ions

Tryptofancomplexes

(Trp Ala2

and Trp Leu2)

Remacle & LevinePNAS 103, 6793(2006)

Tools for attosecond dynamics studies

Powerful and precisely controlled electromagnetic waves with sub-femtosecond resolution

Pump-probe via VIS-UV – XUV phase coherent pulses or via re-collision of electron wavepackets (HHG!)

Cold and localized molecules: molecular ions in a trap, sympathetically cooled by atomic ions to mK temperature

Many other things we probably did not think about ...

Goal 2: XUV precision test of QED

1S

2S

Hydrogen

2x 243 nm

1S-2S: 2 466 061 102 474 851 (34) Hz

Lamb shift 1S: 8172.840(22) MHz

M. Fischer et al., PRL 92, 230802 (2004)

Uranium 91+ Lamb shift 1S = 460.2 (4.6) eV

1S1/2

2P3/2

138 576 (6) eV

A. Gumberidze et al., PRL 94, 223001 (2005)

Higher-order QED scales with >= Z4Higher-order QED scales with >= Z4

Tests in Helium and Helium+ ions

Hydrogen Helium Helium+ Lithium 2+

1S

2S

243 nm

243 nm

1S2

1S2S

120 nm

120 nm

1S

2S

60 nm

60 nm

1S

2S

30 nm

30 nm51 nm

1S5P

Searching for a change of alpha

Comparison lab UV spectra vs. IR telescopeabsorption lines in quasar spectra(looking back billions of years)

Direct UV frequency comb excitation

Frequency comb

laser @ 788 nm

2nd Harmonicgeneration

Excitation@ 394

nm

t t

2

5 ns

105 modes

5 ns

190 MHz190 MHz

Timedomain

Frequencydomain

fn=f0 + n fr fn=f0 + n fr fm= 2 f0 + m fr

Ca+ ions

Direct frequency comb spec. of Ca+

Paul trap

Freq. combpulses @ 394 nm

Ions in a trap

Calcium ion trap& Anne Lisa Wolf

The setup

Direct frequency comb signal of Ca+

Mode number determination for Ca+

200 times more accurate than before

Correctionsfor light-shiftsetc. included

Accuracy: 0.5 MHz

Helium groundstate status

Energy of He 1s2 1S0 (+5945204000 MHz)

-400

-200

0Theory: Drake & YanCan. J. of Phys. 86, 45 (2008)

MHz

Eikema et al.PRL 71 (1993)PRA 55, 1866 (1997)

Bergeson et al.PRL 80, 3475 (1998)

Measurements with nanosecond duration VUV (120 nm) or XUV (58.4 nm) single pulses

HHG: High-Harmonic-Generation

IR pulses – mJ level

XUV pulses – nJ level

1014 W/cm2

lens/mirror

Required for 1s2 – 1s5p transtion: 51.56 nm

gas jet

Hoge-harmonische generatie

Model van P. Corkum, PRL 1994

High-harmonic generation

T

Train of x-ray bursts

Generation of 170 attosecond pulses:

Lopez-Martens et al, PRL 94, 033001 (2005)

~ 5 fs

Problem: single pulse frequency chirp

‘Clean’ or ‘Fourier limited’well defined central frequency

‘Chirped’Unclear frequencydue to amplification & harmonic conversion

Solution: two or more pulses

time frequency

Direct UV frequency comb excitation

Frequency comb

laser @ 788 nm

2nd Harmonicgeneration

Excitation@ 394

nm

t t

2

5 ns

105 modes

5 ns

190 MHz190 MHz

Timedomain

Frequencydomain

fn=f0 + n fr fn=f0 + n fr fm= 2 f0 + m fr

Direct XUV frequency comb excitation

Frequency comb

laser @ 773 nm

Two-pulseamplifier

15th Harmonicgeneration

Excitation@ 51 nm

t t t

15

7 ns

106 modes 104 modes

7 ns

150 MHz150 MHz

Timedomain

Frequencydomain

fn=f0 + n fr fn=f0 + n fr fm=15 f0 + m fr

Frequency comb up-conversion

IR DUV

frequency

VUV XUV X-RAYharmonicconversion

IR pulses with CE control

XUV

1014 W/cm2

UV

Xe or Ar jet

How accurate does it have to be?

time frequency @ 51 nm

!CE~ /100

CE~ 0

=20 MHz

T = 6.6 ns

2 identical pulsesAccuracy requiredCE~ /200 !

Delay 6.6 ns, accuracy ~ 10 attoseconds (10∙10-18 s)

While amplifying more than 10 000 000 xat 10 GW/cm2 power levels, over the full6 mm beam diameter !

Amplification of the comb with a NOPCPA

s

p

i

The good: High gain (=short path length) Broad bandwidth (non-collinear) No thermal effects Phase preserved of signal (but ...)

The tricky: Picosecond pump laser Synchronization < 1 ps Angle sensitivity

pk

p

pump

idler

signal(2)

fluorescence cone

seedsks

Setup for He spectroscopy at 51 nm

regen amp.2 mJ

power amp.2 x 200 mJ

SHGsync.

7 ps osc.1064 nm

comb laser ~780 nm

frep=150 MHz NOPCPA

HHG

51 nm pulses, 6.6 ns apart1s5p 1P1

1s2 1S0

Helium

51 nm

2x 2 mJ200 fs

Relay imaging

Phase measurement setup

BS 50%

BS 5%

NOPCPA

Pump laser

BS

NG

Oscillator

PC2

PC1

G

BS

BS 5%

D

fiber

Computer

CCD Camera

PBS'

2)0()(

02dz

f

fkz

z

sss

Phase accuracy between 2 pulsesAt the moment for pulses 6.6 ns apart:

Interferometric measurement to 1/600th of IR

! Phase flatness across the IR beam: ~ 1/200th of IR

Many tricks involved to control and measure this!

Christoph GohleDominik Kandula

Photos of the lab

Photos of the lab

XUV comb generation and He excitation

HHG

Spherical grating

XUVdetector

Kryptonjet

XUV Beam

Monochromatorslit

2 times >2 mJ, 200 fs

Heliumbeam 90o

!

IR ionizationbeam

HHHG: Holey-High-Harmonic-Generation

IR pulses – mJ level

XUV pulses – nJ level

1014 W/cm2f=50 cm

Harmonic generation double pulses

In Krypton @ ~772nm

7

911

13

15

Harm. : (nm)

7 : 1109 : 8611 : 7013 : 6015 : 51

HHG with double NOPCPA pulses

15th harmonicin Krypton ~ 51 nmnJ pulses

Fundamentalat 795 nm

First helium 1s5p signal at 51 nm ...

comb laser repetition rate scan

He-ionsignal

FFT

Better Helium 1s5p signal at 51 nm

150 MHz

Total scan of pulse delay < 0.5 fs !

8 attoseconds change per scan step

Photo He experiment

Xe two-pulse 125 nm excitation

Short delay for mode identification, long delay for accuracy

Decreasing contrast is caused by the residual 10 MHz Doppler width

Best fringe position fitting result: 40 kHz

Phase errors reduce to <1 MHz at longer delays

13 times better than CW/ns pulsed experiments.

Zinkstok et al., PRA 73, 061801(R) (2006)

Summary

First demonstration of direct frequency comb excitation of calcium ions in a trap at UV wavelengths

First XUV frequency comb excitation of helium: on route to MHz precision

In preparation: Calibration He ground state, two-photon excitation etc.Sympathetic ion cooling of atomic and molecular ions Attosecond electron dynamics in biomolecules Extension of XUV combs for helium+ ions

@ 51 nm