Winter College on Optics: Trends in Laser Development and Multidisciplinary Applications to Science and Industry February 2013

Transcription

1 Winter College on Optics: Trends in Laser Development and Multidisciplinary Applications to Science and Industry 4-15 February 2013 Fibre Laser S. Taccheo Swansea University UK

2 FIBRE LASER Stefano Taccheo Laser and Photonic Group Nanotechnology Centre - Centre for NanoHealth College of Engineering Swansea University s.taccheo@swansea.ac.uk 1

3 Introduction Summary Fundamentals of light propagation Fiber Lasers Phenomena and limits in Fiber laser Next challenges (UV & Mid-infrared) Conclusion 2

4 AIM: Provide an Overview of Fiber Lasers (high-power) 3

5 Swansea University 4

6 A bit about myself 1964 Born in Trieste 1989 Degree at Politecnico di Milano on laser resonator 1990 Optical Amplifiers (SIRTi & CSELT, Turin) 1991 Joined CRN & Politecnico di Milano (bulk and waveguide lasers) Collaboration with Italtel and IRE-Polus (now IPG) on optical amplifiers 1998 ORC Southampton, UK, on Fibre laser 1999 Lucent Bell Labs, NJ, USA, on supercontinuum sources 2007 Joined Swansea University: Head Laser group and responsible for Photonics Labs. Collaborations with several companies, IXFibers, Perfos, Pirelli, Corning and Marconi/Ericsson on fibre devices 5

7 History First laser was demonstrated in 1960 by T. Maiman First fiber laser was demonstrated in 1963 E. Snitzer 6

8 customized ear implant Blood vessel imaging 30 kw welding of steel. several tens of mm Art Work marking of beetles Cleaning and restoration 7

9 Property of laser light Monochromaticity Directionality (Spatial Coherence) Temporal Coherence Brightness Short Time duration Can concentrate huge amount of energy 1) on small areas (volumes) ( µm 2 / µm 3 ) 2) During short time (10-15 s) 3) Transmit light over long distance (e.g. Earth-Moon distance control, power up satelites) Selectively interact with only specific materials/chemical specimens 8

10 Fundamentals 9

11 Snell s law incident ray θ a θ a θ b air (n a ) water (n b ) n a ( θ ) n sin( θ ) sin = a b b refracted ray n b >n a 10

12 Total Reflection incident ray θ a θ a n a n a sin ( θ ) a = n b θ b refracted ray θ a = arcsin n n b a n b <n a n b 11

13 Fiber Optic n 2 α Core n 1 n θ c NA = sin( α) = n 2 1 n

14 Fiber Optic Single-mode operation (step index profile) λ 2πa = c NA V = 2π NA < λ

15 Fiber Optic (silica) Low Loss: 0.2 db/km 14

16 Fiber Optic (Other Glasses) Lower phonon energy Jiri Orava, Tomas Kohoutek, A. Lindsay Greer, and Hiroshi Fudouzi, "Soft imprint lithography of a bulk chalcogenide glass," Opt. Mater. Express 1, (2011) 15

17 Type of lasers Rod laser Disk laser Fiber laser Waveguide laser 16

18 Elements of a laser 3: Optical cavity 1: ACTIVE MEDIUM 2: Excitation mechanism 17

19 Elements of a fiber laser Optical cavity Active Fiber Excitation mechanism 18

20 Elements of a fiber amplifier Excitation mechanism Active Fiber 19

21 Why fiber devices? Fiber lasers/devices: Structural integration (fiber splicing) High-gain (long interaction distance) Compactness, robustness and reliability Design flexibility and manufacturability High efficiency (up to 40% electric-to to-optical)) High surface/volume ratio (good thermal management) Diffraction-limited operation at multi-kw output powers Photonic structures Exploitation of non-linear effects Fiber lasers/devices: Need of glasses Glasses are very different each other! Small (relatively core size) Constrain on power density (optical damage, non-linear effects (Raman, SBS, ASE)) Constrain on pulse energy ( mj) Limited amount of all-fiber components 20

22 Fiber devices (MOPA) 60W module pump LDs 21

23 Industry 22

24 Maximum power (1 micron) Source: IPG 7/

25 Active Material 24

26 Emission Wavelength Pr Dy Ho Tm Er Bi Yb Nd Raman/ OPOs/SC Wavelength [µm] 25

27 Example of Importance of right λ: Tissue absorption 26

28 ACTIVE IONS Direct Pumping

29 ACTIVE IONS Pumping Schemes 28

30 Pr Doping 2 micron pumping? 29

31 Dy Doping 1.8 micron pumping? 30

32 Tb Doping 3 micron pumping? 31

33 Dy Doping Low gain: long length (50 mm is enough?) 32

34 4 I 7/2 Rate-equations (3-level) 4 I9/2 (i) (i) 4 I11/2 ( f) ( g) (h) (i) 4 I13/2 980 nm pump 1550-nm a) (b Er 3+ 4 I15/2 We can introduce co-operative upconversion and many others.. Like excited state absorption 33

35 Full Rate-equations (3-level) system below threshold dn dt dn dt dn dt N = σ = σ + N 2 a13 a12 = σ F a13 + N 3 p F N s F N p = 1 1 N σ σ 1 N e31 e21 + σ = 1 F F N e31 p s F N p 2 3 N N τ N τ a12 e σ N τ N + τ F N s 1 + σ F N s 2 N + τ 2 21 N + τ From second equation, neglecting ASE photons and decay from level 3 to level 1 and imposing steady-state condition (derivative=0) we have N 0 = τ N τ N = τ 3 32 N + τ 3 32 therefore To avoid to waste inversion in an useful level the pump level should rapidly decay to the upper laser level: τ 32 <<τ 21 N 3 τ = τ N

36 35 Rate-equations above threshold: steady-state ( ) ( ) T c a N N N condition N N B V q N N N qb Wp N = + = = : τ τ Photons within resonator (from solving first equation) ( ) ( ) Δ = x N N V q c th T A τ τ

37 Rate-equations (3-level) above threshold Laser output power (assuming output only from Mirror #2) P L x = VA = 2τ P P P p, th 2 ( N + ΔN ) hν ( x 1) T th γ 2 γ P L = VA γ q γ 2τ ( N + Δ ) ν 2 ( 1) = ( ν ) 2 γ τ h T Nth h x c γ x represent the ration between pump power and power at threshold. If x=1 laser power is obviously zero 36

38 Host Materials Silica (Al-silicate) glass is the most common host glass Pros: Excellent quality, Telecom Know-how, Thermal conductivity Cons: Low doping level (below 1%), Phosphate glass: Pros: High-doping level (5% and above), high phonon energy (efficient codoping). No Photodarkening Cons: Difficult splicing with silica fiber, low thermal conductivity Unsuitable for Telecom grade C-band amplifiers Tellurite Glass, ZBLAN Pros: transparent in the 1-5 micron window, low phonon energy. (good for IR transition above 2 micron). Cons: Thermal dissipation, splicing Chalcogenaide Glass Pros: transparent up to 10 micron High-non linear coefficient Cons: Absorb 1 micron radiation, soft glass 37

39 Pumping: fficiency Optical Optical-to to-optical efficiencies (typical): 1 micron: Yb-doped fused silica fibers: 70% 80% 2 micron: Tm-doped fused silica fibers: 50%- 75% 1.5 micron: Er and Er/Yb doped silica fibers: 20% - 40% Electrical Electrical-to to-optical efficiencies: Typical ~25%; maximum reported 40% Pumps 9xx nm (-60%) Pump 1480 nm (-40%) Pumps 780 nm (40%-50%) 38

40 Pumping scheme and Heat Removal Quantum defect = λ p /λ L Lasing Pump QD Heat removal from 10kW laser Yb: % 1.1 KW Er: % 3.7 kw Er: % 0.4 kw Tm % 2.8 kw Tm %* 1.1 kw (fiber pumped) Ho % 1.4 kw (fiber pumped) 39

41 Components 40

42 Components 41

43 Pumping system Need to deliver pump power into active fiber: Power level from 100s mw into SMF to kws into double cladding fibers (+100 microns diameter) Source: DILAS 42

44 Fiber Single-mode fiber (core pumped) are telecom type. 43

45 Fiber For high-power we need to increase active volume and optimize pump absorption: Double cladding pumping scheme Core diameter 15 micron SM, 20 (max 25 micron) slightly MM High-order suppression techniques (e.g. bending) Source: ORC Southampton 44

46 Microsctructured Fiber P. Russell, Science 17 January 2003: Endlessy single-mode, dispersion management, management of non-linearity (LMA Fibers) 45

47 Microsctructured Fiber: LMA J. Limpert et al. Opt. Express 12, (2004). 46

48 Fiber Air-cladding region. Bridges width is small than laser wavelength. High N.A. Low non-linearity 40 µm core (7 missing holes) and 200 µm air-cladding. About 1000 µm 2 area Use: Amplifier stage: 10 ps to 60kW Peak power J. Limpert et al. Opt. Express 12, (2004). 47

49 Fiber (rod-type) In Q-switched laser pulsewidth is proportional to cavity length (fiber are not so good therefore.). Typical value of pulsewidth in Q-switched fiber laser is 100 ns. Rod type. High Volume with guided light STIFF Fiber 58 µm MFD (19 missing holes) and 200 µm air-cladding. Possibility of large volume (high-energy) in short fiber. 2 mj, 100 W pulsewidth down to 7 ns Pump cladding diameter: 200 mm N.A. =0.6 Outer cladding: 1.7 mm MFD 58 mm O. Schmidt, Jet al., "Millijoule pulse energy Q-switched short-length fiber laser," Opt. Lett. 32, (2007) 48

50 49

51 50

52 J. K. Ranka et al.,, Opt. Lett. 25 (1), 25 (2000) 51

53 52

54 Pump combiner (US patent # 5,864,644) GTWave technology (credit: D. Payne) (Goldberg, Opt. Lett. 1999) 53

55 Laser combiner 54

56 cw lasers Source: IPG 55

57 cw lasers Source: IPG 56

58 DESIGN LIMITS 57

59 Nonlinear Schrodinger Equation z Loss Dispersion SPM α β A + A + i A i A A = t γ 0 A(z,t) is the complex slowly varying pulse envelope in a frame of reference moving with the pulse at the group velocity v g = 1/β 1. α accounts for linear absorption at carrier frequency ω 0. γ = 2πn 2 =(λaeff) Raman di s dz di dz p = g R I = g R p I s ω ω α I p s I p s I s s α I p p 58

60 What are the limits for fiber lasers? 1: Available pump power (especially to pump active ions other than Yb!) 2: Output wavelength: only few are covered by standard ions 3: (Pulsed) Length: looking for new glasses 4: (HP) Heat dissipations (water cooling, etc. ) CW bulk ~100-kW (anticipation) 5: (HP and pulsed) Non-linear effects ( (3) nonlinearity) Self-Phase-Modulation (SPM) Stimulated Raman Scattering (SRS) management increase core size Stimulated Brillouin Scattering (SBS) for narrow-linewidth management increase core, SBS suppression techniques 6: Photodarkening: management: low doping (silica), alternative glasses 59

61 What are the limits for fiber lasers? 1: Available pump power (especially to pump active ions other than Yb!) 2: Output wavelength: only few are covered by standard ions 3: (Pulsed) Length: looking for new glasses 4: (HP) Heat dissipations (water cooling, etc. ) CW bulk ~100-kW (anticipation) 5: (HP and pulsed) Non-linear effects ( (3) nonlinearity, silica is symmetric) Self-Phase-Modulation (SPM) Stimulated Raman Scattering (SRS) management increase core size (SF) Stimulated Brillouin Scattering (SBS) for narrow-linewidth management increase core, SBS suppression techniques 6: Photodarkening: management: low doping (silica), alternative glasses 60

62 Stimulated Brillouin Scattering (SBS) Stimulated Brillouin Scattering (SBS) is the dominant non-linear effect for narrowlinewidth laser (below 10 MHz). SBS ha narrow gain bandwidth. Scattering of light from acoustic waves. Becomes a stimulated process when input power exceeds a threshold level. P th 21 g eff 1 xl L = L( 1 e ) B L A eff eff x G. P. Aggarwal, Non-linear Fiber Optics, Academic Press, 2006 Management: Short Fiber, Large Mode Area. Linewidth Broadening, Fiber straining (broadening SBS gain). Currently: Threshold at 100s W SF laser (limit 1 kw) Broadening at 10 GHz for kw class laser A. Liem, et al. "100-W single-frequency master-oscillator fiber power amplifier," Opt. Lett. 28, (2003) 61

63 Stimulated Raman Scattering (SRS) Scattering of light from vibrating (silica) molecules. Since we are in glass the vibrational state is a band Gain extends to tens of THz with peak (silica) at 13 THz (i.e. SRS at 1.5 mm is about 100 nm) P th 16 g R A L eff eff Telecom fiber(50 m 2 A eff ): 1 W Active Fibers (10 m, 20 micron core): 10 kw (100 ns, 1 mj!), Affects; Q-Swicthed laser for marking! Pulsed laser in general, HP CW lasers Management: increase core size, short fiber (high-doping) 62

64 SRS: only Bad? Bad for laser and amplifiers Good for Raman Laser and Raman Amplifiers : allow to shift wavelength and cover neighborough wavelength intervals where no active ions are available! di s dz di dz p = g R I = g R p I s ω ω α I p s I p s I s s α I p p 63

65 Self-phase modulation (SPM) Based on Kerr effect: refractive index depends on optical intensity. () φ t k n2 L eff A P eff ( t) Limits: 100 ns pulses: 10W on LMA fiber: 1 mj 10 ps: 1 µmj 100 fs: 10 nj Management: increase core size, short fiber (high-doping) Good or Bad: Bad for pulse amplification Good for supercontinuum generation 64

66 Bulk damage Bulk damage inside fiber core: depends in on Intensity and effective area For ps to ns pulses. Thermal damage by melting/heating. Below tens of ps ablation occurs. Process becomes almost insensitive from pulsewidth E P P Max 2 mj A p Max µ Expected 100kW threshold for CW lasers eff τ W m 4 Management: Increase core size. Interplay pulse duration and pulse peak power. B. C. Stuart, et al. Nanosecond-to-femtosecond laserinduced breakdown in dielectrics, PRL, ,

67 Surface damage Weakest point. Threshold can be 1 order of magnitude below bulk damage threshold. About W µm 2 at 1 µm wavelength (pure silica). Decrease for doped glass Limits for single fiber lasers (10 kw) Management: Undoped coreless endcaps. 66

68 Other real-world limits Self-pulsation: An amplifier left without seed will generated large pulse destroying (even evaporate) the fiber core and damaging inline components (Yb-doped amplifiers) Fiber Fuse Effect: Travelling hot-spot generate by defects or even dirty connectors. Threshold of few Ws. Generated by local defect. Absorption enhanced temperature rise D. D.Davis et al, SPIE Vol. 2966, X, 1997 Courtesy: P. Andre, Univ. of Aveiro, Portugal 67

69 Fuse fibre effect Courtesy: P. Andre, Univ. of Aveiro, Portugal 68

70 Photodarkening Self-similarity Quadratic vs. Doping Level Normalized Loss 0,8 0,6 0,4 0,2 Y050 Y090 Y110 Y135 Y180 α eq [db/m] ,0 0,5 1,0 1,5 2,0 Normalized Time Yb concentration [wt%] α eq = 3 k N N k N N Yb 2, Yb Yb 2, Yb 2 ~ = 69

71 Overall trends Does it help if we increase: Length A eff Doping level Stimulated Raman Scattering (SRS) no yes yes SBS no yes yes SPM - yes - Thermal Management yes no no Photodarkening yes yes no Detrimental effect (cluster) - - no High Power yes yes yes Q-switching short pulse no - yes High Quality Beam - no - Single-Frequency operation no - yes A Fiber laser design is always a compromise! 70

72 What are the limits for cw fiber lasers? J. Limpert et al. Fiber Lasers, Jena University 71

73 How far we are from limits? Yb-doped laser (silica) are.almost close to physical limits! 72

74 Type of Devices 73

75 Fibre devices CW Lasers Single-frequency, Linearly Polarized, High-power Pulsed lasers ns regime ps regime ultrashort laser Amplifiers CW Pulsed Continuum generation Waveguide laser 74

76 Fibre devices CW Lasers Single-frequency, Linearly Polarized, High-power Pulsed lasers ns regime ps regime ultrashort laser Amplifiers CW Pulsed Continuum generation Waveguide laser 75

77 cw fiber lasers 76

78 High-power cw fiber lasers High power cw fiber laser divide into three categories Laser fiber (Only a fiber laser): power is limited to about 1 kw MOPA configuration (seeder + amplification stages): power up to 5 kw, possibility of linear polarization. Single mode. M 2 close to 1. MOPA configuration offers also possibility of single-frequency and linearly polarized output, as well as pulsed operation Combining of several laser into the same fiber or in free space (beam combining). Power up to 100 kw. M 2 approaching 5. 77

79 cw lasers 60W module pump LDs Source: IPG 78

80 cw lasers Source: IPG 79

81 Beam Shape Pulsed laser Source: IPG 80

82 High-power fiber lasers Source: IPG 81

83 cw lasers Source: IPG 82

84 cw lasers Results 1,9 kw fundamental mode, dual side pumped fiber laser oscillator: Source: ROFIN 83

85 84

86 ISLA project aims to expand applications of 2-micron fiber laser technology 85

87 Pulsed lasers 10 1 Hybrid ML Rep. Rate 10 0 Active ML Nonlinear compression GHz 10-1 Passive ML Pulse duration [ps] For fs pulses Basic Osillator 10 pj-1 nj Direct Amplification 1nJ 1 µj Chirped Pulsed Amplification 1µJ 1 mj 86

88 Pulsed lasers Source: ORC 87

89 Pulsed fibre Laser Objective Femtosecond MOPA (master-oscillator amplified system) Parameter Average power 200 W Pulsewidth Pulseenerg y/rep.rate Beam quality (M 2 ) < 500 > 100 < 1.2 fs 2 MHz (TEM 00 ) Source: LIFT 88

90 pulsed fibre Laser Challenge Femtosecond MOPA (master-oscillator amplified system) High-energy pulses out of fibre amplifier Nonlinear effects Damage threshold of bulk material Dispersion management (temporal pulse shape) Beam quality (TEM 00 at high average power) Source: LIFT 89

91 pulsed fibre Laser Results Femtosecond MOPA (master-oscillator amplified system) Master-oscillator (MO) output Power-amplifier (PA) output Source: LIFT 90

92 pulsed fibre Laser Femtosecond MOPA (master-oscillator amplified system) Results Power characteristics 8.2 MHz Source: LIFT 91

93 LIMITS (Visible/UV) 92

94 cw visible lasers Wavelengths for specific medical treatments (532nm, 577nm, 633 nm, ) Apply the advantages of fibre lasers to the medical area (beam quality, higher power, new wavelengths accessible, low maintenance lasers, ) New laser tools for Ophthalmology (1 to 5 Watt range) and Dermatology (10 to 20 W range) Challenges & tasks Laser Concept IR linearly polarized CW Fiber laser (narrow linewidth or single frequency) Non linear stage Architecture based on Fiber laser in the infrared then frequency-doubled using periodically poled crystals (non-linear stage) Single pass configuration for SHG (much better stability) Source: LIFT 93

95 UV/Visible Continuum generation From J. Hecht, LFW 2011 Webcast A. Efimov et at, Nonlinear generation of very high-order UV modes in microstructured fibers, Opt. Express, 11 (918) (Los Alamos National Laboratory and University of Bath) 94

96 UV/Visible Continuum generation -15 Power [dbm, arbitrary unit] nm 750 nm 790 nm 800 nm Wavelength [nm] 95

97 Confocal laser scanning microscopy measurement RGB Confocal image of rat neuronal cells Astrocytes Neuronal cell body a) b) BLUE: glial cells excited at 630 nm, RED: neuronal cellular body excited at 550 nm. GREEN: astrocytes excited at 458 nm Glial cells Constant pump power. Different excitation wavelength chosen by a monochromator 96

98 THANK YOU 97