ETH Zurich Ultrafast Laser Physics

Ursula Keller / Lukas Gallmann

ETH Zurich, Physics Department, Switzerland www.ulp.ethz.ch

Chapter 10: Ultrafast Measurements

Ultrafast Laser Physics!

Ultrafast laser physics (ULP)

1as!

Time

1am!

Length

1 picosecond = 1 ps = 10–12 s 1 femtosecond = 1 fs = 10–15 s 1 attosecond = 1 as = 10–18 s

Measurement with !s time resolution

Harold E. Edgerton, MIT"1903-1990"

!

Flash photography: !Flash lights driven by electronics !

triggered flash lights " #µs time resolution "

(already available 1935)"limited by flash duration "(„light pulse duration“)!

The problem

•$ Straightforward: Measure slow event with fast event

•$ However, all detectors are time-integrating on these time scales

•$ Solution: Map dynamics/time axis to static observable!

E. Muybridge: Animal Locomotion (1887)

The solution

•$ The classical pump-probe approach:

•$ Map time to translation in space: •$ Therefore •$ 1 nm resolution in yields 7 as resolution in •$ Delay is equivalent to real time if duration of probe pulse is

negligible and process is perfectly reproducible •$ This idea can be generalized to other mappings of time to

time-independent quantities

! = 2"xc

# "xS(!x)" S(# )

!x !

Pump-Probe Measurement

2 !z = c!t!

!t

!z =1 µm " !t # 2 $ 3.3 fs

Differential Transmission Spectroscopy

Laserbeam splitter chopper

!c"t

probe

pump

device under test

photo detector (PD)

!!c

JPD noise of probe

signalsignal = [T("t, Ipump) – T(Ipump = 0)] Iprobe

Transmission of device under test with pump on pump off

Ultrafast Pump-Probe Techniques

• Why a chopper?

• Why not the chopper in the probe pulse?

• Why do you use a lock-in amplifier?

Differential Transmission Spectroscopy

Ultrafast measurements need some kind of nonlinearities in the measurement system (i.e. intensity dependent transmission)

Laserbeam splitter chopper

!c"t

probe

pump

device under test

photo detector (PD)

!!c

JPD noise of probe

signalsignal = [T("t, Ipump) – T(Ipump = 0)] Iprobe

Transmission of device under test with pump on pump off

Ultrafast Pump-probe Techniques

Different arrangements

•$Noncollinear degenerate pump-probe measurements

•$Collinear degenerate pump-probe measurements

Noncollinear degenerate pump-probe

PUMP

PROBE

SAMPLEPolarisation

k1

k2

"Langsamer" Detektor (zeitlich gemittelt)

!t

LINSECHOPPER

Noncollinear: pump and probe beam not collinear good for signal-to-noise because pump power is not on detector

Degenerate: pump and probe pulse have the same central wavelength

Collinear degenerate pump-probe

PUMP

PROBESAMPLE

Polarisation

!t

LINSE

CHOPPER bei f

PUMPPBS

PROBE

PBS

1

LOCK-IN VERSTÄRKER bei f1

What is the reason for the PBS (polarizing beam splitters) in the set-up?

Collinear degenerate pump-probe

PUMP

PROBESAMPLE

Polarisation

!t

LINSE

CHOPPER bei f

PUMPPBS

PROBE

PBS

1

LOCK-IN VERSTÄRKER bei f1

Potential problem?

PUMP

PROBE

SAMPLE

Polarisation

!t

LINSE

CHOPPER bei f

PUMP

STRAHL- TEILER

PROBE

1

LOCK-IN VERSTÄRKER bei f œ f1

CHOPPER bei f2

2

Detector can be saturated by strong pump beam.

-

Degenerate four-wave mixing

PUMP

PROBESAMPLE

k

k2

1

E

E 1

2

"Langsamer" Detektor

Polarisation => Beugungsgitter

Why is this set-up a degenerate four-wave mixing experiment?

Degenerate four-wave mixing

PUMP

PROBESAMPLE

k

k2

1

E

E 1

2

"Langsamer" Detektor

Polarisation => Beugungsgitter

Parallel polarization creates a transient diffraction grating inside the sample.

This grating exists as long as there is a coherent excitation (i.e. within the dephasing time)

Review articles: K.-H. Pantke und J. M. Hvam, "Nonlinear quantum beat spectroscopy in semiconductors," Int. J. of Modern Physics B, 8, 73-120, 1994 E. O. Göbel, "Ultrafast Spectroscopy of Semiconductors," Festkörperprobleme, Advances in Solid State Physics, 30, S. 269-294, 1990 J. Shah, "Ultrafast Spectroscopy of Semiconductors," Springer-Verlag

Degenerate four-wave mixing

PUMP

PROBESAMPLE

k

k2

1

E

E 1

2

"Langsamer" Detektor

Polarisation => Beugungsgitter

PUMP

PROBE

SAMPLE

Polarisation

k1

k2

"Langsamer" Detektor

!t

LINSE

k22 - k1

Optical Gating

SIGNAL

PROBE

NICHTLINEARER KRISTALL

k1

k2

LANGSAMER DETEKTOR

!tLINSE

Application: time resolved femtosecond luminescence measurement

T. C. Damen and J. Shah, "Femtosecond luminescence spectroscopy with 60 fs compressed pulses," Applied Phys. Lett. 52, 1291, 1988

J. Shah, "Ultrafast Luminescence Spectroscopy using sum frequency generation," IEEE JQE, 24, 276-288, 1988

Optical Gating: Time-of-flight imaging

Application of optical gating for “time-of-flight” imaging

M. R. Hee, J. A. Izatt, J. M. Jacobson, J. G. Fujimoto, "Femtosecond transillumination optical coherence tomography," Optics Lett., vol. 18, pp. 950-952, 1993

Biologisches Gewebe

Abtasten

Starke Streuung durchgelassener

Laserpuls

Laserpuls

Lichtanteil mit wenig Streuung => Abbildung

SIGNAL

PROBE

NICHTLINEARER KRISTALL

k1

k2

LANGSAMER DETEKTOR

!tLINSE

Optical coherence tomography (OCT)

How does this work?

Science, 254, 1178, 1991!

Optical coherence tomography (OCT)

Michelson Interferometer Interference only within coherence length

Science, 254, 1178, 1991!

Prof. J. G. Fujimoto, MIT, USA"

30 fs pulse duration -> 10 µm axial resoluton "

10 fs pulse duration -> 3 µm axial resolution "

Optical coherence tomography (OCT)

Time resolved four-wave-mixing

PUMP

PROBE

SAMPLE

Polarisation

k1

k2

"Langsamer" Detektor

!t

LINSE

k22 - k1

How do you do time resolved four-wave mixing?

Time resolved four-wave-mixing

2

PUMP

PROBE #1

SAMPLEPolarisation

k1

k2!t

LINSE

k2 - k1 NICHTLINEARER KRISTALL

!tLINSE

PROBE #2

1

2

LANGSAMER DETEKTOR

Photoconductive Switching

V

Laserpuls

Output

Photoconductive switch or Auston switch: D. H. Auston, "Picosecond optoelectronic switching and gating in silicon," Appl. Phys. Lett., vol. 26, pp. 101-103, 1975 D. H. Auston, P. Lavallard, N. Sol, D. Kaplan, "An amorphous silicon photodetector for picosecond pulses," Appl. Phys. Lett., vol. 36, pp. 66-68, 1980

Photoconductive Switching

OUT

+VI (t)

Z 0Z 0

I (t + !)

Photoconductive sampling gate D. H. Auston, A. M. Johnson, P. R. Smith, J. C. Bean, "Picosecond optoelectronic detection, sampling, and correlation measurements in amorphous semiconductors" Appl. Phys. Lett., vol. 37, pp. 371, 1980

High-order harmonic generation in gases

Classical electron-trajectories ! Multiple trajectories with same recombination energy but different excursion time exist

P. B. Corkum, Phys. Rev. Lett. 71, 1994 (1993) Harmonic order

Log(

Stre

ngth

)

Plateau Cutoff

Spectrum of harmonics

ti

tr

!1 ! 2Short

trajectory Long

trajectory

HHG

+

Laser-based HHG: “a success story”

Challenges/Problems of laser based HHG: Low pulse repetition rates and low pulse energy! limits signal-to-noise ("5 orders of magnitude reduction)

HHG and attosecond science

Laser-based HHG

Intense ultrafast Ti:sapphire CPA (!800 nm, > 300"J) pulse repetition rate: "1 kHz (moving towards 10 kHz) pulse energy center: up to 100 eV pulse energy of attosecond pulses: < nJ pulse duration: "100 as

Femtosecond domain: nJ pulses at 100 MHz 100 mW average power

Attosecond domain: nJ pulses at 1 kHz 1 !W average power

Streaking techniques instead of pump-probe attosecond resolved measurements strongly signal-to-noise limited use phase sensitive techniques instead:

energy streaking: mapping time to energy angular streaking: mapping time to angular momentum

attosecond pulse synchronized with strong infrared field strong infrared field can be used for streaking

[1] R. Kienberger et al., Science, 207, 1144 (2002) [2] R. Kienberger et al., Nature, 427, 817 (2006) [3] E. Goulielmakis et al., Science, 305, 1267 (2004)

energy streaking: mapping time to energy (linear polarized streaking field)

Attosecond streak camera •$ Most versatile and most successful technique to date:

Attosecond streak camera Hentschel et al., Nature 414, 509 (2001)

1.$ The attosecond pulse and an intense, short infrared pulse are overlapped in/on a medium being studied – they can be delayed with respect to each other

2.$ The attosecond pulse ionizes the medium

3.$ The vector potential of the infrared pulse shifts the resulting electron spectrum in energy as a function of the relative delay

Measured at ETH, 2012

Streaking techniques instead of pump-probe attosecond resolved measurements strongly signal-to-noise limited use phase sensitive techniques instead:

energy streaking: mapping time to energy angular streaking: mapping time to angular momentum

attosecond pulse synchronized with strong infrared field strong infrared field can be used for streaking

angular streaking: mapping time to angular momentum (circular polarized)

5 fs

time measurement =

angle measurement

no as pulses!

P. Eckle, A. Pfeiffer, C. Cirelli, A. Staudte, R. Dörner, H.-G. Muller, M. Büttiker, U. Keller, Science 322, 1525, 2008

Delay in tunnel ionization

30!

A. S. Landsman, M. Weger, J. Maurer, R. Boge, A. Ludwig, S. Heuser, C. Cirelli, L. Gallmann, U. Keller Optica 322, 1525 (2008)

How long does it take for an electron to traverse the tunneling barrier in tunnel-ionization of helium?

and much more ….