SR QPM OPOor

How many acronymscan you put in your title?

Dipl.-Phys. Carsten Langrock

langrock@stanford.edu

Ginzton Lab, Stanford University

AP305 – p.1

Outline

• Motivation:What are optical parametric oscillators?Why not use a laser instead?

• Theory:SVEA, SFG, DFG, OPA, OPG, OPO, QPM

. . . any questions?• Experimental Setup:

What to do when you cannot measure your signaldirectly.

• Experimental Results:Temperature & QPM-Period Tuning,Pump Depletion.

• Conclusion: Where do we go from here?

AP305 – p.2

Outline

• Motivation:What are optical parametric oscillators?Why not use a laser instead?

• Theory:SVEA, SFG, DFG, OPA, OPG, OPO, QPM

. . . any questions?

• Experimental Setup:What to do when you cannot measure your signaldirectly.

• Experimental Results:Temperature & QPM-Period Tuning,Pump Depletion.

• Conclusion: Where do we go from here?

AP305 – p.2

Outline

• Motivation:What are optical parametric oscillators?Why not use a laser instead?

• Theory:SVEA, SFG, DFG, OPA, OPG, OPO, QPM

. . . any questions?• Experimental Setup:

What to do when you cannot measure your signaldirectly.

• Experimental Results:Temperature & QPM-Period Tuning,Pump Depletion.

• Conclusion: Where do we go from here?

AP305 – p.2

Outline

• Motivation:What are optical parametric oscillators?Why not use a laser instead?

• Theory:SVEA, SFG, DFG, OPA, OPG, OPO, QPM

. . . any questions?• Experimental Setup:

What to do when you cannot measure your signaldirectly.

• Experimental Results:Temperature & QPM-Period Tuning,Pump Depletion.

• Conclusion: Where do we go from here?

AP305 – p.2

Outline

• Motivation:What are optical parametric oscillators?Why not use a laser instead?

• Theory:SVEA, SFG, DFG, OPA, OPG, OPO, QPM

. . . any questions?• Experimental Setup:

What to do when you cannot measure your signaldirectly.

• Experimental Results:Temperature & QPM-Period Tuning,Pump Depletion.

• Conclusion: Where do we go from here?AP305 – p.2

Motivation

What are optical parametric oscillators?

• Widely tunable source for coherent radiation.• Efficient nonlinear conversion from “pump” to

“signal” wavelength.• Low threshold for onset of oscillation.

AP305 – p.3

Motivation II

Why not use a laser instead?

• LASER• collect & store

wideband uncollimatedspectral energy.

• center wavelength &linewidth determined byatomic transition.

• pump does notinfluence output oflaser to first order.

• OPO• No energy storage.

Instantaneousnonlinear process.

• center frequency &linewidth determinedby phase mismatch.

• phase coherencebetween signal,idler, and pump isessential.

AP305 – p.4

Theory

Optical Parametric Oscillators (OPO)• Maxwell’s equations in source-free media• Slowly Varying Envelope Approximation (SVEA)

• Examine χ(2) processes, i.e. P (t) ∼ E2(t)

• Three field interaction at ωp, ωs, and ωi

dEp

dz= −iηpωpd(z)EsEie

i∆kz with ωp = ωs + ωi

dEs

dz= −iηsωsd(z)EpE

∗

i e−i∆kz ∆k = kp − ks − ki

dEi

dz= −iηiωid(z)EpE

∗

se−i∆kz

AP305 – p.5

Theory II

Coupled equations describe various nonlinearprocesses.

• Sum Frequency Generation (SFG):strong pump signal at ωi, weak signal at ωs

up-converted signal at ωp

• Difference Frequency Generation (DFG):strong pump at ωp, signal at ωs signal at ωi

• Parametric Generation (OPA, OPG, OPO):strong pump at ωp generation of signals at ωs

and ωi

AP305 – p.6

Theory III

Simple cartoon for optical parametric amplification.

1. Electromagnetic waves at ωp and ωs generatepolarization at ωi.

2. If polarization travels with same speed as freeelectromagnetic wave at ωi, signal at ωi will grow.

3. ωi mixes with ωp and creates polarization current atωs signal at ωs grows.

Same principle holds for OPG and OPO where input

fields at ωs and ωi are created out of quantum vacuum

fluctuations.

AP305 – p.7

Theory IV

Phase Matching and Quasi Phase Matching (QPM).

Solutions to coupled equations predict efficientfrequency conversion only if

∆k = kp − ks − ki = 0 Phase Mismatch

Due to dispersion, this is not given, in general.

Solution:• Birefringence:

works, but we are stuck with material properties.• Quasi Phase Matching:

engineer appropriate phase mismatch for desiredsignal wavelength; take advantage of largestnonlinear coefficient. AP305 – p.8

Theory V

Quasi Phase Matching (QPM).

Problem:• ∆k 6= 0

conversion only over crystal length Lc = π/∆k.• Polarization and free wave slip out of phase.

kp

ki ks Kg

kp

ki ks

Kg

Solution:• Reset phase after Lc by changing sign of

nonlinear coefficient (QPM)• Equivalent to introducing a grating k-vector

Kg = 2π/Λg with Λg = 2Lc.

AP305 – p.9

Theory V

Quasi Phase Matching (QPM).

kp

ki ks Kg

kp

ki ks

Kg

Solution:• Reset phase after Lc by changing sign of

nonlinear coefficient (QPM)• Equivalent to introducing a grating k-vector

Kg = 2π/Λg with Λg = 2Lc.AP305 – p.9

Theory VI

Quasi Phase Matching (QPM).

0

5

10

15

20

25

30

1 2 3 4Propagation Distance (L )c

SH

G P

ower

(a.

u.)

AP305 – p.10

Experimental Setup

Q-switched Nd:YAG @ 1.06 µmvariable

Attenuator

HR @ 1.5 µmHR @ 1.5 µm

PPLN

HR @ 0.632 µm HT @ 1.06 µm

Si Photodiode

InGaAs Photodiode

Lens

Oven

Translation Stage

Signal and idler wavelength are determined by phasemismatch:

1

Λg=

np

λp−

ns

λs−

ni

λi

Tuning is achieved by varying either Λg, np, ns, or ni,i.e. change temperature or QPM period.

AP305 – p.11

Experimental Results

Temperature Tuning.

Temp er at ur e ( o C)

80 100 120 140 160 180

Wa v

el en

gt h

(n m

)

1500

2000

2500

3000

3500

Signal: 1529.2 - 1 607. 0 nm Id le r: 349 7. 5 - 3148. 7 nm

Measured ωp + ωs ωs and ωi via ωp = ωs + ωi

and ωp = 1.06µm.AP305 – p.12

Experimental Results II

QPM-Period Tuning.

Grating Po si ti on (a .u .)

0 1 2 3 4 5 6 7

Wa v

e len

gth

(nm

)

1500

2000

2500

3000

3500

Signal: 1769.8 - 1 481. 8 nm Id le r: 266 8. 0 - 3774. 0 nm

15 QPM periods on chip starting at 31.5 µm,decreasing in 0.5 µm decrements.

AP305 – p.13

Experimental Results III

Pump Depletion = Increased Conversion Efficiency.

Time (n s)

0 2 0 4 0 6 0 8 0

Pum

p A

mpl

itude

(a.

u .)

0

1

2

3

In creasi ng Pump Conv ersi on

As the pump energy is being increased, the conversionefficiency increases and the pump gets depleted. AP305 – p.14

Conclusion

In this talk we presented the

• theory behind OPA, OPG, OPO, and other χ(2)

processes.• need for phase matching to achieve efficient

conversion.• advantage of quasi phase matching as compared

to birefringent phase matching.• generation of 1.6 µm, 3 µm, and 636 nm radiation

using an OPO.• possibility of temperature and QPM-period tuning.

AP305 – p.15

Outlook

Future research will involve• direct detection of signal wavelength at 1.6 µm.• complete characterization of conversion efficiency.• modification of setup to lower oscillation threshold

(DRO instead of SRO).• implementation of continuously tunable

QPM-period by means of “fanned” grating structure

AP305 – p.16

Acknowledgements

I would like to acknowledge contributions to thisresearch by the following people:

• Arun Sridharan• Karel Urbanek• Supriyo Sinha• David S. Hum• Eleni Diamanti• Mathieu Charbonneau-Lefort

AP305 – p.17