High Precision Wide Dynamic Range Nonlinear Reflectivity Measurement - PowerPoint PPT Presentation

Diploma Thesis High Precision Wide Dynamic Range Nonlinear Reflectivity Measurement Christian Walther October 2004–March 2005 Supervised by Deran Maas, Rachel Grange, and Markus Haiml Ultrafast Laser Physics Group Prof. Dr. Ursula Keller Institute of Quantum Electronics ETH Zurich

High Precision Wide Dynamic Range Nonlinear Reflectivity Measurement Outline Motivation Theory Setup 1: AOM Setup 2: Cube Measurement Results Remaining Challenges Conclusion

Motivation SESAM reflectivity Theory Setup 1: AOM Setup 2: Cube Measurement Results Remaining Challenges 100% Conclusion ! R ns R ns Reflectivity ! R R lin 0 1 2 3 10 10 F sat 10 10 Pulse Energy Fluence [ µ J/cm 2 ] small modulation depths ∆ R < 1% are desired

Motivation Previous measurement setup Theory Setup 1: AOM Setup 2: Cube Measurement Results Remaining Challenges Conclusion incident and reflected power are measured separately

Motivation Errors Theory Setup 1: AOM Setup 2: Cube Measurement Results Remaining Challenges Conclusion ±5% ±10% OUT = c 1 P R c 1 P (1–R) IN = c 2 P ±5% c 2 P ±10% ±7% ±14% c 1 /c 2 R c 1 /c 2 (1–R) R = 0.99 δ R = 7% · 0.99 δ R = δ (1–R) = 14% · 0.01 = 0.069 = 0.0014

Motivation Basic setup Theory Setup 1: AOM Setup 2: Cube Measurement lock-in amplifier Results SIG REF Remaining Challenges Conclusion L + ! L 100 MHz pulsed sample laser source beam splitter L high reflector power on detector t no interference

Motivation Fourier transform Theory Setup 1: AOM Setup 2: Cube S ( t ) S ( t ) = P ( t ) ∗ ( D ( t ) + R · D ( t − τ )) Measurement t Results = Remaining Challenges FT P ( t ) P ( f ) Conclusion = t f · * P ( f ) T 1/T FT D ( t ) D ( f ) f · t f + D ( f ) + R f · f 0 Re · Re R 1/ " · Im 1/ " e 2 ! if " ! FT D ( t – ! ) f 1 + R·e 2 ! if " · t Im f 0 R 1 D ( f ) f f � � � 1 + R · e 2 π i f τ � S ( f ) = ˆ ˆ D ( f ) + R · e 2 π i f τ · ˆ ˆ = ˆ P ( f ) · ˆ P ( f ) · D ( f ) D ( f ) · 1 – R

Motivation Lock-in amplifier output Theory Setup 1: AOM Setup 2: Cube Measurement Results y Remaining Challenges Conclusion small sample arm signal variation x combined signal with shifted variation reference arm signal power ↔ amplitude scaling delay ↔ phase rotation exact adjustment of power ratio and delay is not needed

Motivation Setup design Theory Setup 1: AOM Setup 2: Cube Measurement Results polarizing laser 3 detector cubes 6 0 2 . Remaining Challenges 2 AOM high reflector 14 mm 0th order 1 Conclusion glass wedge Faraday 1 s t o ! ⁄ 2 r d e r sample beamsplitter rotator 0.99 4 1.32 6 5 f = 30 cm 7 2.66 2.80 3 0 . 3 high reflector f = 40 cm f = 12.5 cm 5 5 3 . Beam diameters at specific places [mm]: 1.3 1.6 0.3 6.5 1 2 3 4 position along beam [m] 0.03 7.0 0.15 0.99 5 6 7 Design considerations include: • fitting beams through apertures without clipping, errors by beam displacement • effect of lenses on beam size, adjustment simulated in flexible Matlab program

Motivation Setup using an AOM Theory Setup 1: AOM Setup 2: Cube Measurement Results polarizing laser detector cubes Remaining Challenges AOM high reflector 0th order Conclusion glass wedge Faraday 1 s t o ! ⁄ 2 r d e r sample beamsplitter rotator high reflector • wide dynamic range attenuation by half wave plate and polarizer • isolator to prevent destabilizing feedback into the laser • focus in AOM • AOM splits beam into 0th and 1st diffraction order • separation of beams by a mirror • glass wedge beamsplitter deflects reflected light onto detector Failure because of • too wide beam and too small diffraction angle • 50/50 beam splitting ratio difficult to stabilize

Motivation Setup using a cube beamsplitter Theory Setup 1: AOM Setup 2: Cube detector Measurement 2.60 1.10 5.8 mm Results Remaining Challenges laser Conclusion ! ⁄ 2 20 µ m isolator beamsplitter cube fiber f = 21 mm sample 22 µ m f = 38.1 mm 0.42 f = 75 mm 3.2 mm 5.6 mm high reflector 0.16 mm 0 0.30 1.08 1.17 • large mode area holey fiber (single-mode) for beam cleaning • fiber acts as λ /5 plate because of bend-induced birefringence • cube beamsplitter, in contrast to coated glass plate beamsplitter, is completely symmetric for accurate 50/50 beam splitting • lens creates focus on reference arm mirror • wide beam on detector makes additional attenuation unnecessary • move sample and focusing lens to adjust arm length difference • move high reflector to adjust power ratio on detector surface by changing beam diameter

Motivation Reproducing sample position Theory Setup 1: AOM Setup 2: Cube crosshair target Measurement laser (tilt) diode Results crosshair target Remaining Challenges (position) collimating lens f = 30 cm Conclusion sample 1.5 m f = 12.5 cm folding mirror • lateral movement of the beam ( ≈ sample displacement) is converted to angular movement by lens, which can be amplified by distance • reflex off the lens enables tilt control • not independent, but still uniquely identify sample orientation • collimating lens controls spot size � positional reproduction accuracy of 20 µm (Rayleigh range: 300 µm)

Motivation Lock-in amplifier coherent pickup Theory Setup 1: AOM • small part of reference signal is picked up at measurement input Setup 2: Cube and added to the measured signal Measurement • not constant, has to be measured for every point Results • computer-controllable motorized beam blocker needed Remaining Challenges • no unused motor controller available – build our own, controlled Conclusion by the lock-in amplifier’s auxiliary output

Motivation Beam blocker motor Theory Setup 1: AOM Setup 2: Cube Implementation problems: Measurement • current through relay affected lock-in amplifier measurement Results • moving away did not help Remaining Challenges Conclusion • transistors were damaged by induction voltage

Motivation Attenuation calibration Theory Setup 1: AOM Setup 2: Cube 1.0x10 -3 Measurement 0.8 Results Signal [V] 0.6 Remaining Challenges Conclusion 0.4 0.2 0.0 -30 -20 -10 0 10 20 Waveplate Angle [°] -3 -4 log 10 (Signal [V]) -5 -6 -7 -8 -30 -20 -10 0 10 20 Waveplate Angle [°] • attenuation to zero is not possible • measurement by adaptive algorithm • fit to model function

Motivation Measurement Theory Setup 1: AOM Setup 2: Cube Measurement Results Remaining Challenges Conclusion • record lock-in amplifier x and y signal for a logarithmically spaced set of powers • 3 measurement runs: • SESAM • no sample (reference arm only) • high reflector

Motivation Evaluation Theory Setup 1: AOM Setup 2: Cube y Measurement reconstructed Results sample arm signal Remaining Challenges Conclusion x measured combined signal measured reference arm signal For both SESAM and HR: • sample arm signal is reconstructed by subtracting reference signal from combined signal • reference signal and reconstructed sample signal are projected onto their main axes to get scalar values • division yields values proportional to reflectivity ● = c 1 · P · R ● = –c 2 · P c 2 R = – ● ● – ● = 1 – ● ● c 1 ● = 1 – c 1 = – c 2 R ● ● ● ≈ 1

Motivation ES134 ( ∆ R = 7%) Theory Setup 1: AOM Setup 2: Cube Measurement 102 Results 101 Remaining Challenges 100 Conclusion 99 98 0.3% 97 Reflectivity [%] 96 95 94 93 92 91 HR SESAM 1 90 SESAM 2 89 previous 88 ! 1 0 1 2 10 10 10 10 Fluence [ µ J/cm 2 ] ± 11.5 %

Motivation ES160 ( ∆ R = 0.3%) Theory Setup 1: AOM Setup 2: Cube Measurement Results 100 Remaining Challenges Conclusion 99.9 0.03% 99.8 99.7 Reflectivity [%] 99.6 99.5 99.4 HR 1 SESAM 1 99.3 SESAM 2 HR 2 99.2 0 1 2 10 10 10 Fluence [ µ J/cm 2 ] ± 11.5 %

Motivation ES160 ( ∆ R = 0.3%) Theory Setup 1: AOM Setup 2: Cube Measurement Results 100 Remaining Challenges Conclusion 99.9 99.8 99.7 Reflectivity [%] 99.6 0.02% 99.5 HR 1 99.4 SESAM 1 SESAM 2 99.3 HR 2 previous 99.2 0 1 2 10 10 10 Fluence [ µ J/cm 2 ] ± 11.5 %

Motivation Remaining challenges Theory Setup 1: AOM Setup 2: Cube Measurement Results Remaining Challenges Conclusion • nonlinearity in amplitude and phase • reproducibility only intermittently achieved

Motivation Conclusion Theory Setup 1: AOM Setup 2: Cube Measurement Results Remaining Challenges precise SESAM measurements achieved Conclusion • dynamic range of 4–5 orders of magnitude • accuracy below 0.1% over 2–3 orders of magnitude experimental difficulties overcome • precise calibration of wide dynamic range attenuation • beam quality improvement by fiber • sample positioning reproducibility • coherent pickup correction by motorized beam blocker