All-optical Control of Magnetism I (including pump-probe techniques) - PowerPoint PPT Presentation

Double pump coherent control Second pump Second pump First pump First pump DyFeO 3 DyFeO 3 10 K 10 K Faraday rotation 0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 Delay time (ps) Delay time (ps) 55

AOS of Ferrimagnetic Metals 56

Femtosecond laser reversal of magnetization? GdFeCo 57

Polarization microscope + pulsed laser Circularly polarized 40fs laser pulses 20 nm GdFeCo film Magneto-Optical microscope 58

GdFeCo 40 fs pulses, 1 kHz H ext = 0 s s Reversal by 40fs laser pulsen! 59

switching of magnetization by single pulse! Sweeping the pulsed laser beam at high speed across the sample C.D. Stanciu et al., PRL 99,047601 (2007) 60

Femtosecond Magnetic Recording in GdFeCo! C.D. Stanciu et al., patent #P77323PC00, PRL 99 , 047601 (2007) 61

AOS: role of light helicity/intensity Helicity- Multidomain All-optical Pulse independent state reversal profile: reversal σ + +M σ - 50 μm Pulse intensity Helicity dependent AOS in narrow (~few%) intensity range 62

Excitation wavelength: 700nm Switching No Switching Size of window is 1.5%. Exactly equal to difference in absorption ! 63

Femtosecond laser reversal: role of exchange? Gd Fe Fe Gd J Fe-Gd ~ 30-50T 64

2-Temperature model Free-electron Lattice bath (T e ) (T L =300 K) Gd Fe Fe 2-afm coupled sublattices! 65

Temperature dominated: T >> T Curie ~100 fs Bloch relaxation 40 fs Dynamics scales with magnetic moment Distinct dynamics of sublattices! 66

Exchange dominated: T < T Curie t~1 ps S 2 S 1 ? Conservation total angular momentum Ground state AFM, transient FM! 67

How to probe? Element specific view: XMCD! ? Fe TEY (a.u.) k L 3 L 2 775 780 785 790 795 800 805 BESSY II Photon energy (eV) fs-Laser pump – X-ray probe X-rays 400-1400 eV 10-50 ps (FWHM) (for more details: see lecture Prof. Luning) 68

Femtosecond-XMCD! BESSY II fs-Laser pump – X-ray probe X-rays FEMTO-SLICING! 400-1400 eV 100 fs (FWHM) 69

Laser heat induced magnetization reversal! S 2 Fs-XMCD, BESSY S 1 T. Ostler et al, Nature Comm.3, 666, 2012 J.H. Mentink et al., PRL 057202, 2012 Radu et al, Nature 472, 205-208 (2011) reversal of magnetization driven by exchange!!! 70

Ultrafast electrical pulse reverses magnetization! 10 ps laser (heat) pulse No hot, spin polarized or spin-orbit coupled electrons! 71

AOS of Ferromagnets 72

AOS of ferromagnetic CoPt (FePt)? [Co/Pt] n What’s the mechanism? 73

HD-AOS in Co/Pt multilayer Pt 2.0 nm Pt 0.7 nm x3 Co 0.4 nm Pt 5.0 nm Ta 5.0 nm Glass Substrate 2.2 ps 1.6 ps 0.1 ps Y. Tsema et al, APL 2016, R. Medapalli et. al., Phys. Rev. B 96, 224421 (2017). 74

HD-AOS in Co/Pt multilayer Pt 2.0 nm Pt 0.7 nm x3 Co 0.4 nm Pt 5.0 nm Ta 5.0 nm Glass Substrate 2.2 ps 1.6 ps 0.1 ps Y. Tsema et al, APL 2016, R. Medapalli et. al., Phys. Rev. B 96, 224421 (2017). 75

HD-AOS in CoPt: fs single-shot imaging 1) No single shot switching MECHANISM? 2) Stochastic + deterministic 3) ~100 pulses 76

Multi-pulse induced HD-AOS No pulse 5 10 30 50 1 ms 35 μm Stochastic nucleation & growth!!! 77

Deterministic displacement of domain walls 2 nm/pulse @ 0.4mJ/cm 2 takes many pulses! R. Medapalli et. al ., arXiv: 1607.02505, PRB 96, 224421 (2017). 78

HD-AOS in CoPt: mechanism Magnetic recording in Co/Pt requires multiple pulses. The first pulses form (stochastically) domains with reversed magnetization. The following pulses cause helicity dependent domain wall motion. 79

AOS of Dielectrics 80

Photo-magnetism of Co-substituted iron garnet (Y 2 CaFe 3.9 Co 0.1 GeO 12 / GGG (001)) laser CW: E II[110 ] E II[1-10] Co 2 Co 3 + Fe 3 + Y 3+ + Co 2 Co 3 + + Fe 3 + Light-induced slow (~  m/sec) motion of domain wall Recording? A.Chizhik et al. PRB , 57 (1998). Heating? A.Stupakiewicz et al. PRB , 64 (2001). Speed? 81

AOS in iron garnet Y 2 CaFe 3.9 Co 0.1 GeO 12 on GGG (001) thickness d=7.5 μm) 50 fs pulse 50 fs pulse 200 × 200  m 2 82

AOS in iron garnet  single pulse  repeatable switching  zero applied field  room temperature A. Stupakewiecz et al, Nature 542 , 71 (2017). 83

All-Optical Switching @ the nanoscale! 84

But……… Femtosecond Present magnetic magnetic recording recording 100 nanometer! 10 micron 100 x smaller 85

All-Optical Switching @ the nanoscale? All-Optical Switching @ the nanoscale! 86 PEEM experiment SLS; L. Le Guyader et al, APL 2012, Nature Comm. 2015

All-Optical Switching @ the nanoscale? All-Optical Switching @ the nanoscale! 87 PEEM experiment SLS; L. Le Guyader et al, APL 2012, Nature Comm. 2015

Can’t we go smaller? plasmonic antenna! 88

Nanoscale switching with plasmonic antennas (with Bert Hecht, Wuerzburg) 40 nm Switching!! Tian-Min Liu et al, Nano Letters, 2015

Outlook: speed and energy consumption in data storage (10 ns/bit) (>pJ/bit) 0,05 $/GB 0,65 $/GB (>nJ/bit) (2 ns/bit) (~ ps/bit) ?? $/GB (~fJ/bit) (20x20nm) 90

Outlook: Spintronic-Photonic Integrated Circuit 10-100 times more energy efficient! With: Aarhus University, IMEC, CEA SpinTEC, QuantumWise 91

Towards more complex nanostructures: Laguerre-Gaussian Beams 92

Donut Switching with L-G beams 22.558 Fluence(mJ/cm 2 ) 34.036 30.968 25.308 12.953 Fluence(mJ/cm 2 ) 93 19.928 17.337 15.663

Towards stable complex nanostructures: Neel skyrmion Bloch skyrmion 94

Opto-magnetic generation of Skyrmions Single pulse illumination (  = 800 nm) Sample: Tb 22 Fe 69 Co 9 through microscope objective (NA = 0.4) + = Read-out with Near-field microscopy (  = 532 nm, Resolution 80 nm) 95 Finazzi, et al. Phys. Rev. Lett. 110, 177205 (2013)

Opto-magnetic generation of Skyrmions Low fluence High fluence FeTb 96

Skyrmion generation model Bloch skyrmion Exchange Energy From the model, R 0 /R 1 is sample & fluence Anisotropy Energy Dipolar Energy independent at H z = 0: R 0 /R 1 = 1.87 Experimentally R 0 /R 1  2 Zeeman Energy Minimizing the energy we obtain the Skyrmion radius Single Skyrmion Skyrmionium: Sk + + Sk - 97 M. Finazzi et al, Phys. Rev. Lett. 110, 177205, (2013)

End of smaller and faster? I. End of “Moore”: too much heat II. Higher density = too much energy 0101010 101010 III. von Neumann bottleneck: transfer information back and forth Create a new paradigm, beyond von Neumann 98

Supercomputer versus Brain: 10 MW 10 W Processing and storage Processing and storage Separated and serial Integrated and parallel and 2D and 3D!

Supercomputer versus Brain: 10 MW 10 W Processing and storage Processing and storage Separated and serial Integrated and parallel and 2D and 3D!