Transcript Tillman-Quasi-phasematching

Quasi-phasematching
A versatile tool for coherent source development
LUMOS Group
K-State University
Karl Tillman
Kansas, U.S.A
Ultrafast Optics Group
Heriot-Watt University
Edinburgh, Scotland (U.K)
Outline
 A Historical Perspective
 The Basic Concepts
- Energy, momentum and the phase condition
- Conversion efficiency
- Birefringent phasematching and Quasiphasematching
 The Modern QPM OPO
 Interesting Adaptations
- The Present
- The Future
A Brief History
 1960 – Theodore Maiman builds the Ruby laser
 Visible spectral region quickly populated by range of
different gain mediums
 Important problems - Atomic / Molecular spectroscopy
 Infrared region is where all the good stuff happens!
- Absorption region for common hydrocarbons (C-H, N-H, O-H)
- Biological fingerprint region (amino acids, proteins, etc)
 Nonlinear materials offered a solution
- 2 – Parametric conversion process
The Motivation
 Parametric frequency conversion offers advantages
over normal laser action
- Generated output wavelength determined by energy and
momentum conservation rather than an atomic energy level
structure
- Level of tuneability not available from traditional laser
sources (based on phasematching conditions)
- Output profiles generally determined by pump pulse
characteristics

Predetermined and predictable spectral and temporal
characteristics
The Basics
 Energy Conservation


3  1  2



 Momentum conservation
k  k3  k2  k1  0
k1
 Phase condition
  3  1  2  
k2
k3

2
 parametric amplification
The Basics
 Parametric conversion efficiency:
 kl 
  l sinc 

 2 
2
2
The Basics
 Efficient parametric conversion requires:
- Appropriate nonlinear medium (i.e. birefringence)
- Interacting waves maintain temporal overlap
- Interacting waves remain phasematched (k ≈ 0)
 Main problems:
- Group velocity dispersion
- Pulse walk-away causes loss of phase condition
 Phase mismatch (k) increases with crystal length
- Limited length  limited gain  high threshold
Birefringent Phasematching
 Initial solution – Birefringence phasematching
- Reduces effect of GVD, increasing interaction length
- Relatively high parametric gain due to crystal length
 Issues:
-
Gain medium must be sufficiently birefringent
Crystals require exact phase-matching angle
Relatively low damage thresholds (initially at least)
Rarely access highest nonlinear coefficient, deff
An Alternative
 Birefringence not a perfect solution
 Alternative originally suggested in 19621
-
Concept of quasi-phasematching (QPM) is born
 Removes need for a birefringent material
 Phasematching condition becomes an issue of
material engineering
 Also not perfect but better in most situations
 Solves some of the key drawbacks of BPM
 Introduces a few different problems
1 J. A. Armstrong et al, Phys. Rev. A, 127 (6), p:1918-1939, 1962
Quasi-phasematching
More Basics
 QPM allows wavelength selection
 Crystal makes momentum contribution
k1
k2
kG
2
kG 
G
k3
k  k3  k1  k2  k0G  0 ki 
2 n  i 
i
, i  3, 2,1
More Basics
 QPM favours collinear phasematching
kG reversal of
k2 • Periodic
k1
lc
electric field
k3
• Regular domain
structure with period:
G  2lc
• k3, k2, k1 co-propagate
 Simple expt. set-ups
G
Summary
 QPM enables phasematching to become a
characteristic of the crystal structure
 Can select phasematched wavelengths by design
 Original concept used stacked single crystals
- High boundary losses  poor conversion efficiency
 Not technologically feasible initially
 1993 lithographic approach used to periodically
reorient electric field in lithium niobate crystal1
 Introduction of the periodically poling technique
1 M. Yamada et al, Appl. Phys. Lett, 62 (5), p: 435-436, 1993
Summary
 Periodic poling requires a ferroelectric medium
 Larger choice of materials than BPM
 Access to highest nonlinear coefficient (i.e. d33)
 Tuning capabilities limited mainly by material
transparency rather than Poynting vector walk-off
 Common materials include:
- Lithium Niobate (LN)
- Potassium Titanyl Phosphate (KTP)
- Associated isomorphs (KTA, RTA, CTA)
The Modern QPM OPO
 The Optical Parametric Oscillator (OPO)
 First successfully demonstrated in 19651
 Versatile parametric device
- Very broad tuning range
- Low operational threshold
- Self-seeding
 QPM allows larger choice of nonlinear materials
- Choose material with highest nonlinearity
- Design the phasematched wavelength
 First successful QPM OPO built in 19952
1. J.A. Giordmaine et al, Phys. Rev. Lett. 14 (24) p. 973-976, 1965
2. L. E. Myers et al, Opt. Lett. 20 (1) p.52-54, 1995
The Modern QPM OPO
 Q-Switched OPO
-
High repetition frequencies (kHz)
Fast pulse durations (ns)
Compact, robust devices
Very high pulse energies (~µJ) – damage risk!
Q-switched
ns Pump
The Modern QPM OPO
 Synchronously pumped (SPOPO)
-
Ultrafast devices (fs/ps)
Very high repetition frequencies (~MHz)
High peak powers (~kW), modest average powers (~W)
Operate well below crystal damage thresholds
fs/ps
Pump
Pump
The Modern QPM OPO
Limitations still exist:
 Temporal overlap still an issue
- GVD more pronounced in QPM OPOs
- Limited useful crystal length, restricted gain
 Limitation on appropriate materials
 Crystal dimensions restricted by fabrication
capabilities
- Only pole Lithium Niobate if wafer ≤500µm
 Restricts aperture size (limits power scaling)
- Could use other materials but fabrication technology not as
mature
 Increases device costs
Adaptations
Ability to control phasematching condition by
engineerable methods allows limitations to be
addressed and other ideas tested
1.
2.
3.
4.
5.
Increase tuning capabilities
Compression techniques
Improve efficiency
Designer pulses
Non-ferroelectric materials – e.g. semiconductors
Adaptations
1. Increasing tuning capabilities
Lithography allows the design of several gratings to be
‘written’ into the structure of a single crystal.
Different grating  different PM 
• Grating tuning
• Temperature tuning
Successfully demonstrated in 19951
1. L. E. Myers et al, Opt. Lett. 20 (1) p.52-54, 1995
Adaptations
1. Increasing tuning capabilities
Lithium niobate is the most common material used to
date in QPM devices. Suffers from photorefractive
effects below Tc~80°C so requires heating.
- Impurities reduce onset of photorefractive damage
- MgO doping allows lithium niobate to operate at room
temperature
 Increased temperature tuning range
Adaptations
2. Compression techniques
 Lithography able to produce aperiodic structure
 Chirped structures used to generate a compressed
SHG output1
Pump
Signal
Longer
wavelengths
converted here
Shorter
wavelengths
converted here
1
1. M. A. Arbore et al, Opt. Lett. 22 (12) p.865-867, 1997

Adaptations
s
Time,
 Spatial localisation of
conversion uses GVD as a
temporal control
 Shorter wavelengths
generated earlier
 Travel faster than longer
wavelengths generated
later in crystal resulting in a
pulse compression
t
2. Compression techniques
b
p
r
Crystal length,
lc
1. T. Beddard et al, Opt. Lett. 25 (14), p.1052-1054, 2000; 2. P. Alverez et al, JOSA B, 16 (9), p. 1553-1560, 1999
s
Adaptations
3. Improving Efficiency
Conversion efficiency dependant on available gain
determined by crystal length
- Longer crystals have a narrower conversion bandwidth
- GVD  pump/signal pulse walk-away increased
- Both reduce conversion efficiency
Chirped grating could maintain conversion bandwidth as
crystal length is increased
Adaptations
3. Improving Efficiency
Pump
Signal
Longer
wavelengths
converted here
Shorter
wavelengths
converted here
1
 Allows the use of longer crystals

Adaptations
3. Improving Efficiency
 Longer crystals


More parametric gain
Better conversion
efficiency
 Lower threshold[1,2]
 Smaller OPOs possible
as less pump power
needed
 Longer pump pulses reduce effect of GVD
1.
K. A. Tillman et al, Opt. Lett. 28 (7) p.543-545, 2003
2.
K. A. Tillman et al, J. Opt. Soc Am B. 20 (6) P.1309-1316, 2003
Adaptations
3. Improving Efficiency
 Photon recycling
 Generate two identical
pulses using one
pump
 Cascaded process
 3 tuneable outputs
 Improves quantum
efficiency
p
i
2nd conversion step
1st conversion step
s
1. K. A. Tillman et al, J. Opt. Soc.Am. B. 21 (8) p.1551-1558, 2004
i
i2
Adaptations
4. Designer Pulses



Each grating has well defined phase response
Possible to design arbitrary aperiodic grating
structures based on the overall phase response
Careful design can lead to generation of pulses with
desirable temporal profile
a)
b)
c)
d)
Double and triple pulses
Triangle pulses
square pulses
stepped profiles
1. U. K. Sapaev et al, Opt. Exp. 13 p.3264-3276, 2005
Adaptations
5. New Materials
Semiconductors can have very high nonlinearities in
comparison to current QPM materials
e.g.
-
LN: d33 = 27pm/V
KTP: d33 = 18pm/V
GaAs: d14=170pm/V
InSb: d14=307pm/V
Organic molecules can be even higher (~105)
The Present
5. New Materials
Semiconductors not ferroelectric so alternative poling
method needed
1. Growth techniques (MBE & HPVE)1
Stanford group (M. Fejer & co)
OP - GaAs
2. Ion implantation2
St Andrews group (W. Sibbett & co) PNS - GaAs
1.
L. Eyres et al, Appl. Phys. Lett. 79 (7) p. 904-906, 2001
2.
D. Artigas et al, IEEE J. Quan. Opt. 40 (8) p.1122-1130, 2004
The Future
 Directly diode pumped QPM OPOs
 Dual colour OPOs for phase stabilized operation 1
- Heriot-Watt, Edinburgh
 Compact semiconductor QPM devices for operation
at GHz repetition rates 2
- St Andrews
 QPM devices for THz frequency generation 3
- Stanford
 Organic QPM OPOs ??
1.
J. H. Sun et al, Opt. Lett. 31 (13) p.2021-2023, 2006
2.
T. C. Brown et al, New Journal of Physics, 6 Art.175, 2004
3.
K. L. Vodopyanov et al, Appl. Phys. Lett. 89 (14) 141119, 2006
CEO phase-locking of Cr:Forsterite
 Joined LUMOS group 9/25
 Project: CEO phase
stabilization of
Cr:Forsterite laser
 In the process of locking
frep and fceo
 Next task – Lock laser to
GPS signal then use it to
make absolute spectral
measurement of acetylene
P(11) – P(16) absorption
lines at ~ 1.53mm
PPLN
The End