1 ETH Zurich Ultrafast Laser Physics Prof. Ursula Keller Department of Physics, Insitute of Quantum Electronics, ETH Zurich International Summer School New Frontiers in Optical Technology Tampere, Finland, Aug. 5-9, 2013 Lecture 2: SESAM modelocked VECSELs and MIXSELs Frontiers in ultrafast laser technology Compact ultrafast lasers for “real world application” Telecom & Datacom Interconnects Optical Clocking Frequency comb Multi-photon imaging Modelocking by forcing all modes in a laser to operate phase-locked, “noise” is turned into ideal ultrashort pulses I (ω) φ(ω) 0 I (t) +π -π ~ ~ φ( t) axial modes in laser not phase- locked noise I (ω) I (t) φ(ω) 0 +π -π τ ≈ 1 Δν φ( t) ~ ~ axial modes in laser phase- locked ultrashort pulse inverse proportional to phase- locked spectrum acousto-optic loss modulator needs RF power and water cooling Active Modelocking
Transcript
1
ETH Zurich Ultrafast Laser Physics
Prof. Ursula Keller
Department of Physics, Insitute of Quantum Electronics, ETH Zurich
International Summer School New Frontiers in Optical Technology
Tampere, Finland, Aug. 5-9, 2013
Lecture 2: SESAM modelocked VECSELs and MIXSELs
Frontiers in ultrafast laser technology
Compact ultrafast lasers for “real world application”
Telecom & Datacom Interconnects Optical Clocking
Frequency comb
Multi-photon imaging
Modelocking
by forcing all modes in a laser to operate phase-locked, “noise” is turned into ideal ultrashort pulses
I (ω)
φ (ω)
0
I (t)
+π
-π
~
~φ ( t)
§ axial modes in laser not phase- locked #
§ noise#
I (ω) I (t)
φ (ω)
0
+π
-π
τ ≈ 1Δν
φ ( t)~
~
§ axial modes in laser phase- locked #
§ ultrashort pulse#
§ inverse proportional to phase- locked spectrum#
acousto-optic loss modulator needs RF power and water cooling
Active Modelocking
2
Innovation: before and after
acousto-optic modelocker SESAM modelocker needs RF power and water cooling
SESAM technology – ultrafast lasers for industrial application
Gain!
SESAMSEmiconductor Saturable Absorber Mirror
#
##
self-starting, stable, and reliable modelocking of diode-pumped ultrafast solid-state lasers
Outputcoupler!
U. Keller et al. Opt. Lett. 17, 505, 1992IEEE JSTQE 2, 435, 1996#Progress in Optics 46, 1-115, 2004!Nature 424, 831, 2003#
SESAM solved Q-switching problem for diode-pumped solid-state lasers
20 years of SESAM – looking back
20 years of ultrafast solid-state lasers: invited paper
• Why was it assumed that diode-pumped solid-state lasers cannot be passively modelocked?
• How was the SESAM invented?
• State-of-the-art performance and future outlook
Appl. Phys. B 100, 15-28, 2010
Motivation for semiconductor lasers: Wafer scale integration D. Lorenser et al., Appl. Phys. B 79, 927, 2004
D. J. H. C. Maas et al., Appl. Phys. B 88, 493, 2007
3
MIXSEL wafer scale integration
A. R. Bellancourt et al., “Modelocked integrated external-cavity surface emitting laser” IET Optoelectronics, vol. 3, Iss. 2, pp. 61-72, 2009 (invited paper)
MIXSEL wafer scale integration
A. R. Bellancourt et al., “Modelocked integrated external-cavity surface emitting laser” IET Optoelectronics, vol. 3, Iss. 2, pp. 61-72, 2009 (invited paper)
Development to the MIXSEL
Quantum-Well SESAM 4 GHz, 2.1 W, 4.7 ps A. Aschwanden et al., Opt. Lett. 30, 272 (2005)
Quantum-Dot SESAM 50 GHz, 102 mW, 3.3 ps D. Lorenser et al., IEEE JQE, 42, 838 (2006)
M. Kuznetsov et al., IEEE Photon. Technol. Lett. 9, 1063 (1997)
• Semiconductor gain structure with reduced thickness IEEE JQE 38, 1268 (2002) • Pump: high power diode bar • External cavity
for diffraction-limited output
pump
laser
heat sink
gain structure
output coupler
VECSEL gain structure
pump
laser
heat sink
gain structure
output coupler
gain structure heat sink
pump energy
6
• 7 In0.13Ga0.87As QWs (8 nm) in anti-nodes of standing-wave pattern, designed for gain at ≈ 960 nm
• GaAs spacer layers
• Strain-compensating GaAs0.94P0.06 layers
• Pump at 808 nm
pump energy
VECSEL gain structure Optically pumped semiconductor laser?
• Maybe a bad idea coming from semiconductor diode lasers? • But for sure a good idea coming from diode-pumped solid-state lasers: - more flexibility in operation wavelengths - broad tunability - efficient mode conversion from low-beam-quality high-power diode lasers - modelocking possible with SESAMs - waferscale integration - cheaper ultrafast lasers in the GHz pulse repetition rate regime
Increase output power by increasing pump power and mode size
First room temperature VECSEL: 20 µW average power: J.V. Sandusky et al., IEEE Photon. Technol. Lett. 8, 313 (1996) High-power cw operation: 0.5 W in TEM00 beam: M. Kuznetsov et al., IEEE Photon. Technol. Lett. 9, 1063 (1997) 1.5 W: W. J. Alford et al., J. Opt. Soc. Am. B 19, 663 (2002) 30 W: J. Chilla et al., Proc. SPIE 5332, 143 (2004)
High power TEM00 cw-operation
SESAM diamond
heat spreader
808-nm pump
B. Rudin et al., Opt. Lett. 33, 2719-2721 (2008)
material k [W/Km]
GaAs 55
copper 400
diamond >1800
• MBE or MOVPE growth of structure in reverse order
• cleaving of small pieces
• metalization for soldering
• flipping over
• soldering on heat spreader with high thermal conductivity
• substrate removal by selective wet etching
§ Maximum power P = 20.2 W § opt.-to-opt. efficiency up to 43% § M² < 1.1
SESAM diamond
heat spreader 960 nm
808-nm pump
VECSEL pump rad. 240 µm
output coupler (0.7%)
B. Rudin et al., Opt. Lett. 33, 2719-2721 (2008)
High power TEM00 cw-operation OP-VECSEL milestones
• First room temperature VECSEL: 20 µW average power: J.V. Sandusky et al., IEEE Photon. Technol. Lett. 8, 313 (1996)
• High-power cw operation: 0.5 W in TEM00 beam: M. Kuznetsov et al., IEEE Photon. Technol. Lett. 9, 1063 (1997) 1.5 W: W. J. Alford et al., J. Opt. Soc. Am. B 19, 663 (2002) 30 W: J. Chilla et al., Proc. SPIE 5332, 143 (2004) - Coherent 20 W in TEM00 beam: B. Rudin et al., Optics Lett. 33, 2719, 2008
8
2.62 W wafer fused VECSEL at 1550 nm
Opt. Express 16, 21881-21886 (2008)
• Combine advantages of InP-based active medium with GaAs/AlGaAs reflector
• Intra-cavity diamond for good heat dissipation
2.62 W cw
Continuous wave VECSEL § Bandgap engineering § Power scaling
Integration of absorber is now conceptually possible!
Towards Absorber Integration: “1:1 modelocking”
ΔR
τ p
11
SESAM-VECSEL modelocking and MIXSEL
< 20 GHz
gain
structure
QW-SESAM up to 50 GHz
demonstrated
QD-SESAM gain
structure MIXSEL
QD QW
OP-VECSEL milestones
• First room temperature VECSEL: 20 µW average power: J.V. Sandusky et al., IEEE Photon. Technol. Lett. 8, 313 (1996)
• High-power cw operation: 0.5 W in TEM00 beam: M. Kuznetsov et al., IEEE Photon. Technol. Lett. 9, 1063 (1997) 1.5 W: W. J. Alford et al., J. Opt. Soc. Am. B 19, 663 (2002) 30 W: J. Chilla et al., Proc. SPIE 5332, 143 (2004) - Coherent 20 W in TEM00 beam: B. Rudin et al., Optics Lett. 33, 2719, 2008
• Passive mode locking with SESAM: 20 mW: S. Hoogland et al., IEEE Photon. Technol. Lett. 12, 1135 (2000) 200 mW: R. Häring et al., Electron. Lett. 37, 766 (2001) 950 mW: R. Häring et al., IEEE JQE 38, 1268 (2002) 2.1 W, 4.7 ps, 4 GHz, 957 nm 2.2 W, 6 ps, 1.5 GHz, 957 nm A. Aschwanden et al., Opt. Lett. 30, 272 (2005) Modelocked VECSEL review: Physics Reports 429, 67, 2006
First wafer-fused modelocked VECSEL at 1550 nm
D=34 mmRoC=50 mm
SESAM
D=85mm
D=45mmOutput
Pump 980nm
Gain mirror and diamond heat spreader
RoC=50 mm
• First wafer-fused passively modelocked VECSEL at 1550 nm!
• Combine advantages of InP-based active medium with GaAs/AlGaAs reflector
• Intracavity diamond for good heat dissipation
• Beam-spot diameters: 210 µm on gain chip; 50 µm on GaInNAs-based SESAM
• 600 mW in 16 ps pulses at 1.29 GHz with 10 W pump power
E. J. Saarinen, J. Puustinen, A. Sirbu, A. Mereuta, A. Caliman, E. Kapon, O. Okhotnikov, Optics Letters, 34, 3139 (2009)
1 mW
10 mW
100 mW
1 W
10 W
aver
age
outp
ut p
ower
10 fs 100 fs 1 ps 10 ps 100 pspulse duration
QW-VECSEL QD-VECSEL MIXSEL
Optically pumped ultrafast VECSELs / MIXSELs
M. Hoffmann, O. D. Sieber, V. J. Wittwer, I. L. Kestnikov, D. A. Livshits, T. Südmeyer, U. Keller,
Opt. Express 19, 8108, 2011
12
Femtosecond all quantum dot VECSEL Separate pump mirror DBR separation tuning for maximum absorption
è higher efficiency Active region chirped QD-layer positions
• each layer stack resonant for different laser wavelength
• according to absorption intensity è broader gain
AR section hybrid semiconductor / fused silica
è reduction of the GDD
pump
modelocked laser
CVD-diamond QD-gain structure
output coupler QD-SESAM
heat sink: thinned QD gain structure on CVD substrate
output coupler: 100 mm
output coupler transmission: 2.5%
laser mode radius on QD-VECSEL: 115 µm
laser mode radius on QD-SESAM: 115 µm
heat sink temperature: -20°C
Femtosecond QD-VECSEL
pump
modelocked laser
CVD-diamond QD-gain structure
output coupler QD-SESAM
repetition rate: 5.4 GHz TBP: 1.3 sech2
peak power: 219 W
pulse duration: 784 fs output power: 1.05 W center wavelength: 970 nm
M. Hoffmann, O. D. Sieber, V. J. Wittwer, I. L. Kestnikov, D. A. Livshits, T. Südmeyer, U. Keller, Opt. Express 19, 8108, 2011
Continuous wave VECSEL § Bandgap engineering § Power scaling
First MIXSEL demonstration in 2007 § 40 mW in 35 ps pulses (at -10˚C) and 185 mW in 32 ps pulses (at -50˚C) § complex growth → limited power
D. J. H. C. Maas, A.-R. Bellancourt, B. Rudin, M. Golling, H. J. Unold, T. Südmeyer, U. Keller, APB 88, 493 (2007)
13
MIXSEL concept
MIXSEL cavity • simple linear cavity • optical pumping: inherit good performance OP-VECSEL (power scaling, good beam quality)
MIXSEL chip
• gain region with 7 QWs • QD-saturable absorber layer
Þ enables modelocking • intermediate pump mirror
(HR pump, HT laser) Þ avoid pre-saturation by pump
Integration challenges
• spot size on absorber and gain is the same
D. J. H. C. Maas, A.-R. Bellancourt, B. Rudin, M. Golling, H. J. Unold, T. Südmeyer and U. Keller, Appl. Phys. B 88, 493, 2007
3
Towards Absorber Integration
, ,
, ,
0.1sat a sat a a
sat g sat g g
E F AE F A
= <Reduction of saturation fluence
Increase field in absorber
Challenge 1 , ,
, ,
0.1sat a sat a a
sat g sat g g
E F AE F A
= <
Problem: increase of modulation depth
No possibility for uncoupled Fsat and for QW SESAMs
What can we do?
Challenge 2
sat const.F R⋅Δ =
antiresonant SESAM resonant SESAM
Towards Absorber Integration
ΔR
QDs absorbers offer more growth parameters than QWs absorbers
QD size and size distribution QD density
• Stranski-Krastanov growth on MBE • InAs on GaAs substrate • In ML coverage determines density
determines modulation depth
Self-assembled QD formation:
can be tuned with dot density, while Fsat stays constant!
QD growth
determine absorption spectrum
Towards Absorber Integration: Quantum Dots (QDs)
ΔR
14
Resonant design analogue to resonant QD SESAMs
D. J. H. C. Maas et al., Applied Physics B 88, 493-497 (2007)
First MIXSEL demonstration: 35 ps, 40 mW, 2.8 GHz
Sections: • 30 pair bottom mirror for the laser • 1 layer of self-assembled InAs QD • DBR to increase field in absorber • 9 pair mirror for the pump • active region with 7 InGaAs QWs • AR coating
Monday, August 19, 13 54 Ultrafast Laser Physics
mode size on chip: 80 µm optical pumping: 1.5 W @ 45°
Low output power compared to former VECSEL-SESAM modelocking: structure was used as-grown
First MIXSEL demonstration: 35 ps, 40 mW, 2.8 GHz
D. J. H. C. Maas et al., Applied Physics B 88, 493-497 (2007)
1 mW
10 mW
100 mW
1 W
10 W
aver
age
outp
ut p
ower
10 fs 100 fs 1 ps 10 ps 100 pspulse duration
QW-VECSEL QD-VECSEL MIXSEL
Optically pumped ultrafast VECSELs / MIXSELs
B. Rudin, V. J. Wittwer, D. J. H. C. Maas, M. Hoffmann, O. D. Sieber, Y. Barbarin, M. Golling, T. Südmeyer, U. Keller, Opt. Express 18, 27582, 2010
Antiresonant MIXSEL Design
Requirement § QDs with strong saturation § study on QD-growth parameters optimization of
growth temperature and post-growth annealing
Advantages § less variations in absorber enhancement § reduced GDD for shorter pulses § less sensitive to growth errors A.-R. Bellancourt, Y. Barbarin, D. J. H. C. Maas, M. Shafiei, M. Hoffmann, M. Golling, T. Südmeyer, U. Keller, OE, 17, 12, (2009) D. J. H. C. Maas, A. R. Bellancourt, M. Hoffmann, B. Rudin, Y. Barbarin, M. Golling, T. Südmeyer, U. Keller, OE, 16, 23, (2008)
-10
-5
0
5
10
GD
D (1
000
fs2 )
980970960950940wavelength (nm)
10
8
6
4
2
0abso
rber
enh
ance
men
t
980970960950940wavelength (nm)
10
8
6
4
2
0abso
rber
enh
ance
men
t
-10
-5
0
5
10
GD
D (1
000
fs2 )
growth error: layer thickness variations < 1%
15
MIXSEL: improved thermal management
Finite Element (FE) temperature simulations
• exchange the copper with CVD diamond:
reasonable temperatures
• leads to highest output power from a ultrafast semiconductor laser
heat sink material
thermal conductivity (W m-1K-1)
estimated heating power (pump power)
pump/ laser mode radius
temp. rise (FE sim.)
heat sink temperature
output power
GaAs 45 1.5 W (1.7 W) 80 µm 149 K -15 °C 41.5 mW
copper 400 3.2 W (4.3 W) 80 µm 98 K +10 °C 660 mW
diamond 1800 26.6 W (36.7 W) 215 µm 100 K -15 °C 6400 mW
High Power MIXSEL Structure
• MBE-grown in FIRST at ETH Zurich (at 580 °C on 600 µm GaAs wafer) • laser DBR (AlAs/GaAs) at 960 nm • self-assembled InAs quantum dot absorber embedded in GaAs
• grown at 420 °C Stranski–Krastanov
• pump DBR (Al0.2Ga0.8As/AlAs) to reflect the pump at 808 nm • avoids absorber presaturation by the pump
• gain section: 7 In0.13Ga0.87As Quantum wells (5 nm) separated by GaAs • 8-µm structure directly soldered
on a CVD diamond heat sink
1 µm
laser DBR
pump DBR
AR coating
gain absorber
CVD-diamond heat sink
inset at absorber and gain: STEM picture by
High Power MIXSEL
1.0
0.8
0.6
0.4
0.2
0
auto
corr
elat
ion
(arb
. u.)
-100 -50 0 50 100delay (ps)
meas. sech2 fit
τp = 28.1 ps
1.0
0.8
0.6
0.4
0.2
0
spec
tral i
nten
sity
(arb
. u.)
959.4959.2959.0958.8958.6wavelength (nm)
meas.
FWHM: 0.148 nm
-40
-30
-20
-10
0
spec
tral i
nten
sity
(dB
c)
2.4752.4702.4652.460frequency (GHz)
span: 20.0 MHzRBW: 100.0 kHz
Average power 6.4 W Center wavelength 959.1 nm Pulse duration 28.1 ps FWHM spectral width 0.15 nm
§ optical pumping 36.7 W at 808 nm § pump / laser spot radius: ≈215 µm § cavity length: 60.8 mm ð 2.47 GHz § fluence on the MIXSEL : 252 µJ/cm2
B. Rudin, V.J. Wittwer, D.J.H.C. Maas, M. Hoffmann, O.D. Sieber, Y. Barbarin, M. Golling, T. Südmeyer, U. Keller, OE 18, 27582 (2010)
highest average power from an ultrafast semiconductor laser
1 mW
10 mW
100 mW
1 W
10 W
aver
age
outp
ut p
ower
10 fs 100 fs 1 ps 10 ps 100 pspulse duration
QW-VECSEL QD-VECSEL MIXSEL
10 GHz – 2.4 W MIXSEL
1.0
0.8
0.6
0.4
0.2
0
auto
corr
elat
ion
(arb
.u.)
-60 -40 -20 0 20 40 60delay (ps)
meas. fit
τp = 17.0 ps
1.0
0.8
0.6
0.4
0.2
0
spec
tral i
nten
sity
(arb
. u.)
962.8962.6962.4wavelength (nm)
meas. fit
FWHM: 0.09 nm
-70
-60
-50
-40
-30
-20
-10
0
spec
tral i
nten
sity
(dB
c)
10.0210.009.989.96frequency (GHz)
span: 100 MHzRBW: 300 kHz
with fused silica etalon (20 µm thick)
Repetition rate 10 GHz Average output power 2.4 W Pulse duration 17.0 ps Center wavelength 963 nm
§ Optical pumping 25.4 W at 808 nm § Pump / laser spot radius: ≈193 µm § Cavity length: 15.0 mm ð 10.0 GHz § Output coupling: 0.5% (ROC 1.0 m)
V. J. Wittwer, M. Mangold, M. Hoffmann, O. D. Sieber, M. Golling, T. Südmeyer, U. Keller, Electronics Letters 48, 1144 (2012)
16
Continuous wave VECSEL § Bandgap engineering § Power scaling
F. X. Kärtner and U. Keller, Optics Letters 20, 16, 1995 F. X. Kärtner, I. D. Jung, U. Keller, IEEE J. Sel. Topics in Quantum Electron. (JSTQE) 2, 540, 1996
soliton modelocking
SPM (n2>0) + negative GDD ➠ solitons
Soliton modelocking: GDD negative, n2 > 0
Master equation:
TR∂∂T
A T , t( ) = iD ∂ 2
∂t 2− iδ A T , t( ) 2⎛
⎝⎜⎞⎠⎟A T , t( ) + g − l + Dg
∂ 2
∂t 2− q t( )⎛
⎝⎜⎞⎠⎟A T , t( ) = 0
Aout T , t( ) = e−q(t )Ain T , t( )Aout T , t( ) = 1− q t( )⎡⎣ ⎤⎦ Ain T , t( )⇒ ΔA T , t( ) = −q t( )A T , t( )
R. Paschotta, R. Häring, A. Garnache, S. Hoogland, A.C. Tropper, and U. Keller, Applied Physics B 75 (2002), 445
10
5
0
-5
-10phas
e ch
ange
(mra
d)
-2 0 2time (ps)
pulse (a. u.) SPM
10
5
0
-5
-10phas
e ch
ange
(mra
d)
-2 0 2time (ps)
pulse (a. u.) SESAM gain total
Δϕ(t) = γP(t)
Δϕ(t) = −α
2g(t) 10
8
6
4
2
0
puls
e du
ratio
n (p
s)
-4 -2 0 2 4GDD (1000 fs
2)
with phase shifts without phase shifts
10
8
6
4
2
0
puls
e du
ratio
n (p
s)
-4 -2 0 2 4GDD (1000 fs
2)
with phase shifts
10
8
6
4
2
0
puls
e du
ratio
n (p
s)
-4 -2 0 2 4GDD (1000 fs
2)
SPM coefficient pulse power
γ P(t)
linewidth enhancement factor α power gain g(t)
α g(t)
F. X. Kärtner and U. Keller, Optics Letters 20, 16, 1995 F. X. Kärtner, I. D. Jung, U. Keller, IEEE J. Sel. Topics in Quantum Electron. (JSTQE) 2, 540, 1996
soliton modelocking
Experimental Verification
output coupler
SESAM GTI
25 µm etalon
gain
chi
p
-10
-5
0
5
10
GD
D (1
000
fs2 )
970965960955950945wavelength (nm)
20
15
10
5
0
puls
e du
ratio
n (p
s)
-20 -15 -10 -5 0 5 10GDD (1000 fs
2)
953 nm 954 nm 955 nm 956 nm 957 nm 958 nm
➠ positive GDD supports shorter pulses
➠ minimum achieved with slightly positive GDD
➠ confirms soliton-like pulse shaping mechanism in passively modelocked VECSELs
➠ pulse duration limited to picosecond regime due to etalon
10
8
6
4
2
0
puls
e du
ratio
n (p
s)
-20 -10 0 10 20GDD (1000 fs
2)
measurement simulation
M. Hoffmann, O. D. Sieber, D. J. H. C. Maas, V. J. Wittwer, M. Golling, T. Südmeyer, and U. Keller, Opt Express 18, 10143 (2010)
Low-Dispersion Top Coating
• top coating optimized for minimal reflection
• strongly wavelength dependent GDD
• also high negative GDD of down to -5000 fs2
old VECSEL structures new low-dispersion top coating
• 6 AlAs/AlGaAs pairs with fused silica (quarter-wave layer) top coating
• Monte Carlo simulation • local GDD optimization
-6
-4
-2
0
2
4
6
GD
D (f
s2 )
980960940920wavelength (nm)
old structure
-6
-4
-2
0
2
4
6
GD
D (f
s2 )
980960940920wavelength (nm)
old structure new structure
-40
-20
0
20
40
GD
D (f
s2 )
980960940920wavelength (nm)
> 30 nm
±30 fs2
GD
D (1
000
fs2 )
19
Modelocked QD-VECSEL - Setup
laser
QD gain structure
CVD diamond
QD-SESAM
pump output coupler
output coupler radius: 100 mm output coupler transmission: 2.5% laser mode on QD-VECSEL: 115 µm
laser mode on QD-SESAM: 115 µm heat sink temperature: -20°C
P. Kreuter, B. Witzigmann, D. J. H. C. Maas, Y. Barbarin, T. Südmeyer, U. Keller, Appl. Phys. B 91, 257 (2008)
Y. Barbarin, M. Hoffmann, W. P. Pallmann, I. Dahhan, P. Kreuter, M. Miller, J. Baier, H. Moench, M. Golling, T. Südmeyer, B. Witzigmann, U. Keller, IEEE J. Selected Topics in Quantum Electronics (JSTQE) 17, 1779 (2011) P. Kreuter et al., Appl. Phys. B, 91, 257, 2008
p
n p
n
Current spreading layer
Electron mobility > hole mobility
Simulations for EP-VECSEL Design
Carriers distribution in the QW layers
Distance from the center (µm)
p-DBR design favorable for large output beam with
fundamental transverse
mode
20
ETH Zurich EP-VECSEL design
Design guidelines: P. Kreuter, B. Witzigmann, D.J.H.C. Maas, Y. Barbarin, T. Südmeyer and U. Keller, Appl. Phys. B, 91, 257, 2008
Suitable for modelocking • Relatively low GDD: AR section • Confined current injection for good beam profile
• 6 µm current spreading layer • bottom p-doped, top n-doping • small bottom disk p-contact
Power scalability • Wafer removal • Large apertures possible
Trade off between electrical and optical losses • Optimized doping profile
• High doping → high free carrier absorption • Low doping → high resistivity
• Intermediate n-DBR for increased gain
top contact
bottom contact
CuW wafer
p-DBR
current spreading
layer
AR section
n-DBR
SiNx
SiNx
active region
SEM
14 µm
11
➠ shortest pulse duration from an electrically pumped VECSEL so far
experimental results: pulse duration: 9.5 ps
average output power: 7.6 mW
center wavelength: 975.1 nm FWHM spectral width: 0.43 nm
W. P. Pallmann, C. A. Zaugg, M. Mangold, V. J. Wittwer, H. Moench, S. Gronenborn, M. Miller, B. W. Tilma, T. Südmeyer, U. Keller, Optics Express, vol. 20, pp. 24791-24802 (2012)
Laser prototype for noise characterization
V. J. Wittwer, C. A. Zaugg, W. P. Pallmann, A. E. H. Oehler, B. Rudin, M. Hoffmann, M. Golling, Y. Barbarin, T. Südmeyer, and U. Keller, IEEE Photonics Journal 3, 658 (2011)
free-running laser 212 fs rms [100 Hz, 1 MHz]
rms amplitude noise 0.45% in [1 Hz, 40 MHz]
timing jitter Laser prototype for noise characterization
V. J. Wittwer, R. van der Linden, B. W. Tilma, B. Resan, K. J. Weingarten, T. Südmeyer, U. Keller IEEE Photonics Journal 5, 1400107 (2013)
free-running laser
stabilized laser 58 fs rms [1 Hz, 100 MHz]
rms amplitude noise 0.45% in [1 Hz, 40 MHz]
timing jitter
21
S/N in nearly all applications is limited by available power per comb line Increasing repetition rate leads to larger line spacing
Frequency combs: need for high repetition rates
Spectroscopy Information Metrology
Astronomical spectrograph calibration
Molecular spectroscopy
Tunable laser calibration
High speed communication Precision
ranging
Arbitrary waveform generation
Optical clocks Low noise
microwaves
Timing synchronization Frequency
transmission LIDAR
FTIR
frequency
intensity
Frequency comb with larger spacing: • higher power per mode • easier to access individual lines • more compact system
4
S/N in nearly all applications is limited by available power per comb line Increasing repetition rate leads to larger line spacing
Frequency combs: need for high repetition rates
Spectroscopy Information Metrology
Astronomical spectrograph calibration
Molecular spectroscopy
Tunable laser calibration
High speed communication Precision
ranging
Arbitrary waveform generation
Optical clocks Low noise
microwaves
Timing synchronization Frequency
transmission LIDAR
FTIR
5
Passively modelocked laser
• compact/robust laser system
• repetition rate >1 GHz
• pulse width <200 fs
• peak power >6 kW SC fCEO
N = τ p2 ⋅Ppeak ⋅γ|β 2|
⋅0.322
Coherence requirement for SCG:
J.M. Dudley et al., Rev. of Modern Physics 78, 1135 (2006)
< 10
soliton order N:
GHz oscillators without compression or amplification
6
>1 GHz
<200 fs
>6 kW
N=10
coherent octave possible
GHz oscillators without compression or amplification
>1 GHz
<200 fs
>6 kW
N=10 PCF1
N=10 PCF2
octave SC
octave SC
22
GHz oscillators without compression or amplification
7
>1 GHz
<200 fs
>6 kW
N=10
octave-spanning spectrum
N=10 PCF1
octave-spanning spectrum
20-fold better fractional frequency stability
for Diode-Pumped Solid-State Laser (DPSSL)
1 10 100 1000
Er-fiber laserEr:glass DPSSL
10-6
10-8
10-10
10-12
10-13
10-15
10-17
10-19
!(s)
"#f
/ f C
EO(!)
"#f
/ f n(!)
MHz-DPSSL vs. MHz-fiber laser
fCEO = 20 MHz
fn ≈ 200 THz
GOAL: GHz-DPSSL frequency comb
S. Schilt et al., Opt. Express 19, 24171 (2011)
coherent octave possible
Conclusion
18
N=10
22.7 kW 125 fs
First CEO-beat detection of a GHz DPSSL without pulse compression
A. Klenner et al., Opt. Express 8, 10351 (2013)
• Further investigations on the requirements for stabile frequency comb generation • Full stabilization of the 1 GHz Yb:KGW laser
Spec
tral p
ower
(dBm
)
Frequency (GHz)
100 kHz RBW unaveraged
b)
0 0.2 0.4 0.6 0.8 1.0 1.2
!70
!60
!50
!40
!30
!20
!10
fCEO frep-fCEO
-500 fs2
L1L2
M1M2
SESAM
M3M4
pump
Yb:KGW
-500 fs2
30 dB
coherent octave possible
Outlook
19
N=10
22.7 kW 125 fs
Promising alternatives:
for more compact and low noise frequency combs
Vertical External
Cavity Surface
Emitting Laser
VECSEL
Modelocked Integrated
External-Cavity Surface
Emitting Laser
MIXSEL
V. W. Wittwer, et al., IEEE Photonics Journal, Vol. 5, No. 1, pp 1400107 (2013)
More info on the web ...
• Web page of Prof. Ursula Keller at ETH Zurich: http://www.ulp.ethz.ch
• All papers are available to download as PDFs: http://www.ulp.ethz.ch/publications/paper
