A.J. Verhoef, A.V. Mitrofanov, D. Kartashov, A. Baltuska Photonics Institute, Vienna University of Technology. E.E. Serebryannikov, A.M.

Transcription

1 A.J. Verhoef, A.V. Mitrofanov, D. Kartashov, A. Baltuska Photonics Institute, Vienna University of Technology E.E. Serebryannikov, A.M. Zheltikov Physics Department, International Laser Center, M.V. Lomonosov Moscow State University Julien Lumeau, Leonid B. Glebov University of Central Florida, College of Optics and Photonics/CREOL, Orlando FL, USA Our goal: Use an all optical technique to investigate tunneling dynamics; map attosecond dynamics onto a femtosecond time scale Move away from photoelectron spectroscopy enable experiments on bulk

2 Outline How to read out tunneling ionization with an optical pulse? Observation of recollision-free (Brunel) harmonics: from noble gas from bulk transparent solids Brunel mixing with two-color fields Future: use 1.6 µm IR CEP OPA for bulk

3 Ionization Regimes Multi-photon (MPI) plasma-induced spectral blue-shift Tunnel (TI) harmonics generation γ>1 tunneling rate slower than laser period Multiphoton ionization γ<1 tunneling rate faster than laser period Tunnel ionization ωl γ = 2mW ee b - Keldysh parameter Keldysh (1965)

4 Motivation Real-time observation of tunneling ionization dynamics Uiberacker et al., Nature 446, 627 (27) Attosecond angular streaking using circular polarized light and COLTRIMS P. Eckle et al., Nature Physics 4, 565 (28) We set out to develop an optical read-out technique that can work with bulk solids!

5 log(i q ) Mechanisms of Higher-Order Harmonic Generation χ <3> E 3 χ <5> E 5 χ <7> E 7 Re-collision HHG ω/ω L Plasma wave oscillations High orders (Corkum): Tunnel ionization Acceleration Recollision P.B. Corkum, PRL 71, 1994 (1993) Lowest orders: χ <3> χ <5> χ <7> Time [fs],, dn e /dt [arb. units] hω q Brunel: twice-per-cycle tunnel ionization step-wise plasma concentration increase transverse plasma current J = en e v F. Brunel, JOSA B 7, 521 (199) Origin of new frequency components due to the Brunel mechanism: rapid time-domain phase modulation! Harmonics signal is independent of the final state of electrons! Instantaneous intensity [W/cm 2 ] 6-fs pulse 8 nm γ=1 N e, Δn [arb. units]

6 Previous Work Theory: Classical models for harmonic generation based on high-frequency variation of the tunnel ionization current: F. Brunel, JOSA B 7, 521 (199) S. Rae and K. Burnett, J. Phys. B 26, 159 (1993) These models predict the right magnitude of 3rd and 5th harmonics, but no plateau: N. Burnett, C. Kan, P.B. Corkum, PRA 51, R3418 (1995) Experiment: ω ω 1 2 2ω 1 +ω 2 ω 2ω 1 ω 2 Brunel mixing in gas. C.W. Siders et al., PRL 87, 262 (21) Driven by laser ω 1, n e oscillates at 2ω 1 ω 3 =2ω 1 ±ω 2 for ω 1 =ω 2 predicts THG and THz emission on the leading pulse edge. (Similar to THz emission via 4 wave mixing 2ω ω ω (D. Cook, et al. Opt. Lett. 95, 121 (2), M. Kress, et al. Nat. Phys. 2, 327, (26))

7 Time Dependent Refractive Index Modulation Instantaneous intensity [W/cm 2 ] dn e /dt [arb. units] 1x fs pulse 8 nm γ= Time [fs],16,14,12,1,8,6,4,2, N e, Δn [arb. units] Analytical expression for nonadiabatic tunnel ionization: G. L. Yudin and M. Yu. Ivanov, PRA 64, 1349 (21) 3-D TDSE in good agreement with the Yudin- Ivanov formalism: Uiberacker et al., Nature 446, 627 (27), Suppl. Information Refractive index change: Δn p ω 2 p n e 2 ω p 2 2ω () t Time-dependent phase shift: Δϕ p 1 ω () t 2 Not accessible in the XUV! Can be read out only by an optical field

8 Tunneling Dynamics in Bulk Solids Proof of principle experiment in gas phase Strategic goal: To develop a technique for investigation of TI dynamics in bulk solids Ionization is the starting point for all strong field phenomena It is also starting point for optical breakdown C MPI impurity level Tunneling Photo-electrons cannot be observed from bulk material! V

9 Instantaneous intensity [W/cm 2 ] dn e /dt [arb. units] 1x1 14 Stepwise vs.. Smooth Refractive Index Modulation? 6-fs pulse 8 nm γ= Time [fs],16,14,12,1,8,6,4,2, N e, Δn [arb. units] ωl γ = Instantaneous intensity [W/cm 2 ] dn e /dt [arb. units] 5.x1 12 2mW ee b 6-fs pulse 8 nm - Keldysh parameter γ= Time [fs] 1.x1-11. N e, Δn [arb. units] Attosecond time structure due to strong dependence of ionization probability on the field strength TI γ MPI Smooth ramp due to MPI (follows intensity envelope)

10 Quasi-Linear MPI Ramp Normalized intensity [arb. units] 1 pump 5-fs pulse Time [fs] i(onization)-spider Plasma-Blue-Shift spectral shear interferometry for characterization of ultimately short optical pulses 3x1 16 2x1 16 1x1 16 N e [cm -3 ] Replica 1 Base harmonic frequency shift Replica 2 SPIDER Interferogram ω Ω ω + Ω 2π/τ A. Verhoef et al., Opt. Lett. 34, 82 (29) ω

11 Attosecond Phase Mask Tunnel ionization case! Advantage: Brunel harmonics do not depend on the final state of electrons Harmonic frequencies N 2ω pump +ω probe, N=1,2,3 Time domain: Attosecond phase mask Showing tunnel ionization dynamics Frequency domain: Harmonic spectra Question: How badly is the attosecond time-domain phase mask distorted by pulse propagation?

12 Spectral Response to Temporal Phase Modulation Formation of time-dependent attosecond phase mask Formation of harmonic spectrum Spectral scattering Ionization loss, mm I(t) Time, fs n e /n Spectral intensity, arb. units I pump = 1.5*1 14 W/cm 2 I pump = 5*1 13 W/cm 2 I pump = 3*1 13 W/cm 2 Input ω/ω

13 Interpreting Spectral Signatures Model: Phase mask Φ(t) with different finite rise time θ This work: A. Zheltikov and E. Serebryannikov, 3-D propagation code for Brunel and Kerr harmonics based on Yudin-Ivanov formalism Phase mask for T/θ = 9 Spectral intensity, arb. units T/θ= ω/ω T/θ=9 T/θ=3.9 Spectra of the resulting laser field for different phase masks Field intensity, arb. units t/t T laser period Power ratio between adjacent harmonics orders depends on the speed of electron density release (Δn step sharpness) Phase, rad

14 Experimental Setup Pump: 5 fs, 2 μj; Probe: ~2 fs, 2 μj. The harmonics are detected in the direction of the weak, cross polarized chirped probe pulse. The pump beam is blocked before the entrance slit of the spectrometer. ω probe may differ from ω pump to see the effect of the phase mask. λ c =75 nm

15 Results experiment Harmonics H7: Channeltron H5: PMT H3: PMT λ (nm) signal detected in the direction of a weak cross polarized chirped probe pulse

16 Experiment vs. Simulation 155 experiment H3 delay (fs) delay (fs) H5 simulation H wavelength (nm) wavelength (nm) wavelength (nm) mbar Kr delay (fs) delay (fs) delay (fs) 16

17 Cross-Correlations Correlations 1. Temporal marginal: Normalized signal Harmonic Probe H3,H5,H7 maps integrated over spectrum. All harmonics follow the same time structure

18 Linear Dependence on Probe Intensity Experiment 3 rd Harmonic, 1mm Argon target Theory 2 Pump-probe direct THG 2 Signal (arb. un.) , Experiment, Theory Probe Energy (μj) Proof of separation of χ (3) from Brunel-harmonics: Linear intensity dependence of H3 on probe intensity! THG spectra measured with pump on are blue-shifted!

19 Time, fs Time, fs Propagation Effects Spatio-temporal profile of the probe pulse nm 78 nm 5 fs.2 mj f=4 cm 5 krypton jet, P= mbar r, μm Filtered THG part r, μm Field intensity, arb. units z = 1. mm z =.5 mm z =. mm z = -.5 mm z = -1. mm Time, fs 19

20 1, E- 3, 194, 1197, 139, 1432, 1566, 1713, 1874, 25, 2242, 2453, 2683, 2934, 321, 3511, 3841, 421, 4596, 527, 5499, 615, 6579, 7197, 7872, 8611, 9419, 13, 1127, 1233, 1349, 1475, 1614, 1765, 1931, 2112, 231, 2527, 2764, 324, 337, 3618, 3957, 4329, 4735, 5179, 5666, 6197, 6779, 7415, 8111, 8873, 975, 162, 1161, 127, 1389, 152, 1663, 1819, 1989, 2176, 238, 4, 2848, 3115, 348, 3728, 477, 446, 4879, 5337, 5838, 6385, 6985, 764, 8358, , 1,E- 3,194,1197,139,1432,1566,1713,1874,25,2242,2453,2683,2934,321,3511,3841,421,4596,527,5499,615,6579,7197,7872,8611,9419,13,1127,1233,1349,1475,1614,1765,1931,2112,231,2527,2764,324,337,3618,3957,4329,4735,5179,5666,6197,6779,7415,8111,8873,975,162,1161,127,1389,152,1663,1819,1989,2176,238,4,2848,3115,348,3728,477,446,4879,5337,5838,6385,6985,764,8358,9142 1, Distortion by Propagation Ar,.3 bar,6 t n n e o, fs Ar,.3 bar,6 t n n e o, fs ,3 1 14, z, mm, -, z, mm, -, Ionization rate, fs -1 -,6 1,5 1,,5 z =,9,6, n n e o Ionization rate, arb -,6 1,5 1,,5 dz. 1mm n n 1 12 e o, arb,, Time, fs, Time, fs

21 Interaction Length vs.. Pressure z, mm,6,3, -,3 Electron density buildup, fs-1 Ar,.3 bar Time marginal + dt.3 bar Ar Ar density.3 bar Ar arb.units arb.units -, full <-.1mm >-.1 mm Time, fs Time, fs full <-.1 mm >-.1 mm 1 arb.units Interaction length drops with pressure increase

22 Chirp of Probe Pulse 3 rd Harmonic, 1mm Argon target 2 2 short chirped 2 short probe chirped probe Normalized Signal Normalized Signal short chirped The spectral marginal does not depend on the chirp of the probe pulse.

23 Coherent Control with Pump Pulse 3 rd Harmonic, 1mm Argon target 2-2μm FS 2-5μm FS 2 dω Cross-correlation μm FS μm FS μm FS μm FS dτ Spectrum μm FS 2 55μm FS 2 75μm FS

24 Chirp of Pump Pulse 3 rd Harmonic, 1mm Argon target spectral marginal vs. pump chirp Each marginal is normalized to its own maxima cross correlation vs. pump chirp net FS wedge insertion (μm) Pump-probe delay (fs) net FS wedge insertion (μm) the signal Spectrum decreases broadens fast even with for increasing very pump little chirp

25 Investigation of TI in bulk solids 2 Experiment in bulk: Brunel type harmonic in glass ω pump 2 ω pump + ω probe ω probe For measuring several harmonics: Use an OPA system at 1.5 μm

26 Are We There Yet? Linear Polarized Pump Circular Polarized Pump Low Probe Power/Intensity High Probe Power/Intensity direct probe THG (background) Target: fused silica Q: Why is there a signal with circular polarization?

27 Two Color Experiment 23 Linear pump polarization 23 Circular pump polarization Wavelength [nm] ω pu +ω pr 27 ω pu +2ω pr 3ω pr Wavelength [nm] Delay [fs] Delay [fs] ω pump ω probe ω 1 ω ω ω 1 2

28 Looking for Beat Modes THG, probe chirped with 3 mm of FS THG, probe chirped with 6 mm of FS 4ω pu -ω pr Wavelength, nm Wavelength, nm 2ω pu +ω pr Delay, fs Delay, fs

29 Ionization in Quartz Intensity TW/cm ρ c γ min =.6 5 fs 16 fs 1 fs 5 fs 16 fs 1 fs electron density, cm -3 τ/τ p Red curves: with avalanche ionization Dotted blue: without = + ρ σ 2 ρ WPI ( ) ρ t Ui τ r

30 CEP stable 2-Hz IR OPCPA DAZZLER stretcher O.D. Mücke, A. Alisauskas, D. Sidorov A. Pugzlys, A. Verhoef compressor OPA th KTP to filamentation 3 rd KTP Nd:YAG pump 12.5 mj FWHM~1 1.5 µm phase drift (rad) wavelength (nm) frame number CEP drift single frame intensity average

31 2nd OPA Stage (1 khz) non-collinearity angle: walk-off angle of ~2 m M 2 = 1.13±.4 (2nd-stage signal) M 2 < 1.2 (Yb:KGW pump)

32 Pulses after the 4 th stage IR OPCPA (2 Hz)

33 4-Fold Self-Compression of mj IR Pulses Pulses after the 4 th stage Self-compressed 1.5-µm pulses: >1.5 mj, 3 optical cycles Gas: Ar, 5 bar, Cell length: 14 cm Filament length: cm total beam after filament Optimal compression: E in : 2.2 mj, E out =1.5 mj Throughput: 66%

34 Summary First direct experimental observation of Brunel harmonics in gas and bulk. Attosecond ionization dynamics can be mapped onto a spectral response that is free of recollision contribution. Attosecond phase mask is not intuitive but quite robust. It is feasible to develop an optical technique instead of registering photo-ionization fragments attoscience in bulk. Future experiments on bulk: use the CEP 1.5(signal)/3.5(idler) µm OPCPA. Funding: Die Österreichische Forschungsförderungsgesellschaft mbh

35 Skirmantas Alisauskas The Group Audrius Pugzlys Aart Verhoef Daniil Kartashov Oliver MückeM Alex Mitrofanov

36 Thank You! Acknowledgements O. Smirnova M. Ivanov P. Corkum R. Hörlein R. Kienberger