WOMBAT 2015 Tutorial
Aspelmeyer et al., Rev. Mod Phys. 86 (2014)
Cavity optomechanics
Eggleton et al,. Adv. Opt. Photon 5 (2013)
Stimulated Brillouin scattering Applications:
• Fiber sensing
• Narrow linewidth laser
• Slow light/ delay line
• RF and optical filtering
• Microwave oscillator
• Microwave signal processing
Applications:
• Quantum optical measurement
• Displacement sensing
• Tunable optical filter
• Slow light/ delay line
• Optomechanical oscillator
Microwave photonic applications of Brillouin scattering
Bao et al,. Sensors 11 (2011) Metcalfe, App. Phys. Rev 1 (2014)
Fundamentals of Microwave Photonics
Stimulated Brillouin Scattering
• Bandpass and bandstop filters
• Tunable delay lines and phase shifters
• Low noise microwave oscillators
Applications:
Future of SBS microwave photonics
Microwave photonics
Microwave photonics (MWP): manipulation of RF signals using
photonic techniques/components
Capmany and Novak, Nat. Photon 1 (2007)
Seeds and Williams, J. Lightwave Technol.24 (2006)
Yao, J. Lightwave Technol. 27 (2009)
Marpaung et al., Laser Photon. Rev. 7 (2013)
vs.
• Heavy (copper, 567 kg/km)
• High loss(190 dB/km @ 6 GHz)
• Rigid and large cross section
• Lightweight
• Low loss(0.25 dB/km)
• Very flexible
• Radio over fiber
• Antenna remoting
• Filtering
• Phase shifter, tunable delay
• Ultra-wideband (UWB)
• Low phase noise
synthesizer
• Spectrum analyzer
• IFM receiver
Optical frequency
Intensity modulation (IM)
Optical frequency
Phase modulation (PM)
f = 0 f =p
LS
Optical frequency
Single sideband (SSB) modulation
f = 0
E/O conversion
Optical frequency
Complex modulation
f = 0 f =Df
E/O and O/E conversion losses
Laser phase and intensity noise (RIN)
Nonlinear distortion
Photodetector shot and thermal noise
Challenges
O/E conversion
E/O conversion O/E conversion
Link “gain” RF to RF loss (typical:-30 dB, good: ~ 0 dB )
Noise figure SNR in/SNR out (typical: 30 dB, good: <10 dB)
Dynamic range margin of noise and distortion (typical: 80-90 dB, good >110 dB)
Figures of merit
Functionalities • filtering • delay • frequency conversion…
Spectrally crowded environments
• Wireless communications (5G)
• Radar and EW
Application
Functionalities • filtering • delay • frequency conversion…
Interference mitigation and filtering
Frequency agile
interferer
Radio frequency
Po
wer
Signal
Interferer Tunable MWP
filter
Spectrally crowded environments
• Wireless communications (5G)
• Radar and EW
Application Solution
Functionalities • filtering • delay • frequency conversion…
Satellite communications
• On-board wifi and live television
Application
Phased array
antenna
Solution
Tunable true time delay
Optical beamformer (U.Twente & LioniX)
Si3N4 Passive WGs, thermal tuners
Discriminator filters (UPV)
InP WGs, thermal tune, BPD
AWG (Purdue)
Silicon modulator, WGs
OEO (OEWaves)
LiNbO3 WGMR, electronics
Marpaung et al, Laser Photonics Rev. 7, No. 4, 506–538 (2013)
Stimulated Brillouin
Scattering
14
• One of the strongest nonlinear optical effects
• Results from a coherent interaction between vibrations and electromagnetic waves
The fundamental physical effects of the interaction are:
Electrostriction:
Electric field causes
material compression
The photo-elastic effect:
Compressive strain
causes change in refractive index
[Light influences sound] [Sound influences light]
Robert W. Boyd, “Nonlinear Optics”, San Diego, CA: Academic press 2001.
Stimulated Brillouin Scattering (SBS)
11/12/12 15
Pump 1 w1
Eggleton et al,. Adv. Opt. Photon 5 (2013)
Robert W. Boyd, “Nonlinear Optics”, San Diego, CA: Academic press 2001.
Intensity
compresses material
(electrostriction)
Pump 2
w2=w1
Compression creates index grating
(photoelasticity) Excites acoustic wave
frequency W
Pump 2
w2 = w1 - W Doppler effect:
Pump reflected,
down-shifted to w2
waveguide
Stimulated Brillouin Scattering (SBS)
The main effect of SBS is to resonantly excite an acoustic grating,
which back-reflects the pump at exactly the acoustic frequency W.
11/12/12 16
• SBS leads to a narrow Stokes peak in the counter-propagating direction
w wp
W
Gain (Stokes)
Loss
(anti-Stokes)
W
Slow light
Fast light
W ~ 7-11 GHz
GB ~ 15-50 MHz Typical values (Silica)
GB
• The linewidth is determined by the acoustic lifetime (~ 9 ns for silica)
• The Brillouin shift W is determined by the acoustic wave frequency
• Kramers-Kroenig relation: gain resonance refractive index change
sharp amplitude and phase (delay) responses
High-Q resonators Waveguides with large Brillouin gain
1. Brillouin, Annals of Physics 17, 88, (1922)
2. Mandelstahm, Rus. J. Phys. Chem (1926)
3. Chiao et al. Phys. Rev. Letters 12, 592 (1964).
4. Brewer et al. Phys. Rev. Letters 13, 334 (1964).
5. Hagenlocker et al. Appl. Phys. Letters 7, 236 (1965)
6. Ippen et al. Appl. Phys. Lett. 21, 539 (1972)
7. Hill et al., App. Phys. Lett, 28 (1976)
8. Dainese et al. Nature Physics 2, 388 (2006)
9. Grudinin et al. Phys. Rev. Lett. 102, (2009)
10. Pant et al. Opt. Exp. 19, 8285 (2011)
11. Lee et al. Nat. Photon. 6, 369 (2012)
12. Shin et al. Nature Comm. 4, (2013).
SBS in optical fibres6
SBS in silicon12
SBS in WGM
resonators9 SBS in wedge
resonators11
On-chip SBS10
First theoretical predictions1,2
First Brillouin laser7
First demonstration
of SBS3
SBS in liquids4
SBS in PCF8
Year of
discovery
Invention
of the laser
SBS in gases5
1920 2000 2010 1970 1980 1960
SBS on chip-scale devices
Eggleton et al,. Adv. Opt. Photon 5 (2013)
How to get enough gain in a chip scale device?
On-chip SBS is challenging because the waveguides are very short.
g0 = Brillouin gain coefficient
Pp = Pump power
Leff = Waveguide length
Aeff = optical mode area
The gain is
1) Material with high refractive index
2) Small mode area
4) Good opto-acoustic overlap
3) Low loss optical waveguides
cladding
Core
x
y
z
Guiding/confinement of acoustic mode
Determined by acoustic velocity in materials
Poulton et al. JOSA B 30 (2013)
Pant et al., Opt. Express 19 (2011)
vIPG ~ 1500 m/s
vchalc ~ 2600 m/s
vsilica ~ 6000 m/s
7 cm
Chalcogenide waveguide:
• High index material As2S3 (n~2.45, g0~n8)
• Small mode area (Aeff ~ 2.3 µm2)
• Low propagation loss (~0.2 dB/cm)
• Large overlap of acoustic and optical modes
Eggleton et al., Nature Photonics, (2011)
Key parameters:
• GB ~ 34 MHz
• g0 ~0.74*10-9 m/W (~100 x silica)
• W ~ 7.7 GHz
• 16 dB gain for 300 mW pump
Pant et al., Opt. Express 19 (2011)
(va ~ 8000 m/s)
(va ~ 6000 m/s)
(n =3.48)
(n =1.45)
Phonon leakage
Si
SiO2
SOI waveguide • Silicon has high refractive index
• But no acoustic confinement in Si core
• Phonon lifetime is very short for small
waveguides Poulton et al. JOSA B 30 (2013)
Shin et al, Nature Communications. 4 (2013)
• Si3N4 membrane for acoustic confinement
• Forward SBS • Low gain ( <1 dB)
• Breakthrough in SBS on chip • Under etched silicon • Forward SBS with ~4 dB of gain
Application: filtering
11/12/12 22
Probe
RF out
Po
wer
RF Frequency
f = 0 Df = ±p
Optical frequency
SBS gain
Phase
modulator
RF in
Pump
Gain Loss
SBS
medium
Zhang et al., IEEE Photon. Tech. Lett. 23 (2011)
6 MHz
3-dB width
Pagani and Shania., (unpublished)
> 50 dB
extinction
On chip SBS bandpass filter
• 2-12 GHz tuning
• 20 dB extinction
• 20-40 MHz tunable bandwidth
Byrnes et al., Opt. Express 20, (2012)
11/12/12 23
Stern et al., Photon. Res. 2 (2014)
• Broad reconfigurable bandwidth
(tens of MHz-to GHz)
• Flat passband
• Sharp and high extinction
• Polarization pulling to enhance
filter suppression
• Pump sweeping for broad SBS
• Result : 44 dB selectivity, 250
MHz -1 GHz tunable bandwidth
• 3 dB passband flatness
11/12/12 24 Wei et al., IEEE Photon. Tech. Lett.. 27 (2015)
• Electrical comb for SBS pump
• Digital feedback for shape control
• Non-uniform pump spacing to mitigate
FWM improve flatness
• Dual fiber stage to limit SRS and
FWM improve selectivity
Wei et al., Opt. Express 22 (2014)
• 50 MHz to 4 GHz tunable bandwidth
• > 40 dB suppression up to 2 GHz width
• ~ 1 dB passband flatness
• Improve SNR
11/12/12 25
Probe
RF out
SSB
modulator
RF in
Pump
Gain Loss
Optical frequency
SBS loss (anti-Stokes)
Po
wer
RF Frequency
Notch
Morrison et al., Opt. Comm. 313 (2014)
SBS
medium
• 2-8 GHz tuning
• 20 dB extinction
• 120 MHz 3-dB width (FWHM)
• High pump power (350 mW)
RF Frequency
3-dB
Bandwidth
No
tch
att
en
uati
on
Desired properties
• High peak attenuation (>50 dB)
• High resolution (FWHM ~ 10 MHz)
• Large frequency tuning (tens GHz)
• Bandwidth reconfigurability
• Attenuation >50 dB
• Bandwidth ~ 10 MHz
• Tuning: 3-4 GHz
State-of-the-art RF filter
B. Kim, IEEE Trans. Elect. Dev. (2013)
• SOI ring
• Rejection 30 dB
• FWHM 910 MHz
• Tuning 12 GHz
IMWP filter
M. Rasras, J. Lightwave Technol. (2009)
EO
modulator
SBS gain
filter Laser Photodetector
Input
RF signal
Po
wer
RF frequency
Output
RF signal
Po
wer
RF Frequency
Novel MWP filter Notch
LS US
Optical frequency
Phase and amplitude control
f = 0 f =Df
US LS
f = 0 Df = ±p
Phase and amplitude filter
Optical frequency Amplitude matching Phase cancellation Filter response
G = 1 dB
Rejection: 1 dB
Conventional SSB
G = 1 dB
Rejection: 1 dB
Conventional SSB
G = 20 dB
Rejection: 20 dB
Pump = 350 mW
Conventional SSB
G = 1 dB
Rejection: 1 dB
Conventional SSB
G = 20 dB
Rejection: 20 dB
Pump = 350 mW
Conventional SSB
G = 0.8 dB
Rejection: 55 dB
Pump = 8 mW
Novel filter
D. Marpaung et al, Postdeadline paper Frontiers in Optics 2013 FW6B
D. Marpaung et al., Optica, 2, 76-83 (2015)
2900% fractional tuning
Q= 375
at 30 GHz
D. Marpaung et al., Optica, 2, 76-83 (2015)
Conventional filter Cancellation filter
Application: delay and phase shift
Zadok et al., IEEE Photon. Tech Lett. 19 (2007)
230 ps delay 1 GHz bandwidth
Analog applications
Extreme broadening: 25 GHz
bandwidth slow light
10.9 ps
delay
Song et al., Opt. Lett. 32 (2007)
Song et al., Opt. Lett. 30 (2005)
Ultra-long delay: High gain SBS
Pant et al., Opt. Lett. 37 (2012)
On chip SBS slow light
Problem:
• Applications require tunable large delays (~ns), large bandwidths
(~GHz), high carrier frequency (microwave, mm-wave)
• But… it is difficult to achieve a (1) tunable, (2) large slope (3) wide band
linear phase response
)(w
w
Ideal phase
response
Real phase
response
“True-time-delay”
bandwidth
cw
RFc ww cw
RFc ww cw
RFc ww
Desired phase
Actual phase
Burla et al., Opt. Express 19(22) (2011)
Chin et al., Opt. Express 18 (2010)
• 100 MHz delay bandwidth
• 0.03 ns to 9.9 ns tunable delay
• 300o carrier phase tuning
Ca
rrie
r
Sid
eba
nd
Optical signal
spectrum:
SBS pump
spectrum:
𝜔𝑐 − 𝜔𝑅𝐹 𝜔𝑐 𝜔
𝜔 𝜔𝑝1
Ω𝐵
𝜔𝑝2
Ω𝐵
𝜔
SBS phase
shift:
Amplitudes cancel
Phases add up
RF frequency
Mag
nitud
e
RF frequency
response
RF frequency
Ph
ase
• Full tuning range (360o)
• 3 dB amplitude fluctuations
• Bandwidth limited to 2WB
Loayssa & Lahoz, IEEE Photon. Tech Lett. 18 (2006)
Pagani, et al., Opt. Lett., 39 (2014)
• Two degrees of freedom: amplitude
and phase
• Ultra-wideband operation: 1 – 31 GHz
• Record-low amplitude fluctuations
(< 0.5 dB)
Application: signal generation
H. Lee, et al., Nat. Photon, 6 (2012)
J. Li, et al., Nat. Commun. (2013)
• 1st and 3rd Stokes beating to generate
microwave frequency (~ 21 GHz)
• Electronic frequency division to
achieve lower frequencies
• Low phase noise, comparable to
commercial RF synthesizers
• Silica on silicon wedge resonator
• Q = 875 million
• FSR matched to SBS shift
• Narrow linewdith SBS laser
Future direction
Computer-controlled smart RF filter
with high performance
1-30 GHz continuous tuning
Tunable filter resolution
Tunable bandpass
Tunable notch extinction
1-30 GHz frequency tuning
Marpaung et al., “Nonlinear integrated microwave photonics”,
Journal of Lightwave Technol. 32 (invited, 2014)
• AOM operating at 10 GHz
• Reconfigurable filtering (?)
• Link and interaction with SBS
• Potential for wider comb (?)
• Miniaturizing high quality light
and RF source
Microwave photonics
• Manipulation of RF signals using photonic techniques
• Promise: reduced footprint and weight, wide bandwidth
• Challenge: conversion losses, noise, distortion
SBS applications in MWP
• Tunable filtering with performance unmatched by any
technology
• RF phase shifter with record-low amplitude fluctuation
• RF synthesizer with low phase noise
What the future holds:
• High SBS gain in CMOS compatible chip
• Functional SBS circuit (modulator, detectors, SBS engine)
• Chip scale optical and RF sources
Christopher G. Poulton, Christian Wolff
University of Technology Sydney (UTS)
Duk-Yong Choi, Steve J. Madden, Barry Luther-Davies
Australian National University
Alvaro Casas-Bedoya, Amol Choudhary, Irina Kabakova, David Marpaung, Birgit Stiller, and
Benjamin J. Eggleton
University of Sydney
Main contributors
Mattia Pagani
Blair Morrison
Shayan Shania
Hengyun Jiang
Iman Aryanfar
Thank you