The generation and application of squeezed light in gravitational wave detectors and status of the POLIS project

Transcription

1 The generation and application of squeezed light in gravitational wave detectors and status of the POLIS project De Laurentis* on behalf of POLIS collaboration *Università degli studi di Napoli 'Federico II' INFN- Napoli (Università di Roma Tor Vergata) INFN Roma2 - (Università di Pisa - INFN Pisa) Gruppo Pisa (Università di Roma La Sapienza, INFN Roma1) Gruppo Roma1 - (Università di Napoli Federico II, INFN Napoli, * CNR) Gruppo Napoli (INFN Genova), (INFN Perugia), (INFN Pisa), (Università del Sannio, INFN Napoli) Gruppo INFN (Università di Firenze, INFN Firenze, *CNR) Gruppo Firenze - (Università di Salerno, INFN Napoli) Gruppo di Salerno (Università di Trento, INFN Padova-Trento, *Fondazione B. Kessler) Gruppo di Trento (Università di Camerino, INFN Perugia) Gruppo di Camerino - (Università di Urbino, INFN Firenze) Gruppo di Urbino XXI Conferenza Relatività Generale e Fisica della Gravitazione September

2 Summary Squeezed states of light Generation of SL Squeezed light in IGWD (Interferometric Gravitational Wave Detectors) Ponderomotive Squeezing Status of POLIS project (POnderomotive LIgth Squeezing) 2 De Laurentis - 18 September -September

3 Squeezed State of Light 3 De Laurentis

4 Squeezed State of Light Minimum Uncertainty States Squeezed States Coherent States X2 X2 Y2 X2 Y1 X1 Coherent State= Displaced Vacuum s s X1 Squeezed State = Displaced Squeezed Coherent Vacuum Squeezing angle Orientation of the squeezing axis Squeezing factor De Laurentis 4

5 Squeezed State of Light Minimum Uncertainty States A bright beam ( >0) has the same fluctuation of the vacuum X2 Light as 'sensitive' element X2 Y2 s Coherent States X1 its intrinsic quantum fluctuations Determines the final sensitivity We cannot violate the uncertainty principle but we can squeeze the quantum fluctuations on one quadrature and 'use' that quadrature as sensitive element Y1 s Squeezed States X1 Squeezed States 5 De Laurentis

6 Squeezed States of Light X2 X2 Amplitude Squeezed Light Phase Squeezed Light Y2 Y2 Y1 X1 Y1 X1 6 De Laurentis

7 Concept 1927 and first theory (minimum uncertainty states) ort h S 1970 o ry o t h is n ze ee u Sq es tat S d arxiv:quant-ph/ Squeezed States History Flowering of interest 1980 Optical Communications, detectors noise reduction (Sensitivity Enhancement of sensors) Fervent Discussion about the Standard Quantum Limit Squeezing production, enhancement of the squeezing factor 2000 For detailed references, see for example references of C. M. Caves, Phys. Rev. D 23 ( , 1981 Low frequency squeezing Quantum Optics and Quantum Information Gravitational Waves Interferometers De Laurentis 8

8 Generation of Squeezed Light 9 De Laurentis

9 Generation of Squeezed Light Squeezed states quadrature fluctuations correlated Non linear process in dielectric medium: - Kerr medium - Optical Parametric Oscillator (OPO) below threshold p=2 (2) the 3th and 2th order nonlinear susceptibilities induces the correlation between the phase and amplitude fluctuations Fully degenerate OPO (polarizations and frequencies degeneracy) Empty cavity with suspended mirrors: Ponderomotive Squeezing the radiation pressure on the mirror that is free to oscillate induces a coupling between its position and the intensity of the light beam a b dx dx Correlation between the phase and amplitude quadrature of the output state 10 De Laurentis

10 Squeezing with OPO From M. S. Stefszky, Generation and Detection of Low-Frequency Squeezing for Gravitational-Wave Detection. PhD Thesis. The Australian National University, N.B.: 6 db ~ factor 2 De Laurentis 11

11 Ponderomotive Squeezing At present only realized in MOMS* largest squeezing measured: 1.7 db below the shot-noise 1.54 MHz (T. P. Purdy et al. PHYSICAL REVIEW X 3, ) *MOMS Micro-Opto-Mechanical-System At present the MOMS presented in literature produce squeezing well outside the GWD range and often a bright squeezed light and not the useful squeezed vacuum for GWD De Laurentis 12

12 Squeezing for GWD 13 De Laurentis

13 Why Squeezing? C. M. Caves, 'Quantum noise in an interferometer' Phys. Rev. D 23 ( ), 1981 Noise in the output as beat between the coherent input beam and the vacuum that enters in the unused port of a beam splitter (dark port of interferometer) Laser Coherent Vacuum Vacuum Squeezed in the unused port Noise Reduction (sensitivity improvement) Radiation Pressure Noise (1/f) Shot Noise (phase fluctuation) 14 De Laurentis

14 What kind of squeezing? Which quadrature must be squeezed depends on the interferometer type Es.: Michelson with Fabry-Perot in arms and homodyne detector Shot Noise (SN) related to the input vacuum field phase quadrature; Radiation-Pressure Noise (RPN) related to the input Amplitude quadrature High frequencies (SN dominated) Low frequencies (RPN dominated) Phase squeezed vacuum s s Amplitude squeezed vacuum s IGW detectors require a frequency-dependent squeezing quadrature: squeezing angle s= s( ), with detection frequency 15 De Laurentis

15 Typical Squeezer Set-Up for GW interferometer generation losses GW detector Laser PLL0 Squeezing Injection (1) (2) SFI OPO Pump MC SHG MZ GW Output FI OPO Pump (5a) (3) Main Laser (5b) LO MCr OMC +/- Bright Alignment Beam Main Laser and their stabilization And mode cleaner system: Green OPO pump Homodyne Local Oscillator Alignment beam OPO PLL1 Slave Laser BS50:50 (1a) Coherent Control (squeezing angle control) De Laurentis Homodyne Detector (detection of squeezing) LEGENDA LO: homodyne Local Oscillator MC=Mode Cleaner cavity SHG: Second Harmonic Generator FI: Faraday Isolator OMC:output Mode Cleaner cavity 16

16 Experimental Sensitivity Enhancement GEO600 broadband noise reduction of up to 3.5 db (red trace) in the shot-noise-limited frequency band. Losses 10 db produced, 3.5 db injected LIGO Scientific Collaboration et al. A gravitational wave observatory operating beyond the quantum shot-noise limit. In: Nature Physics 7.12 (2011), pp De Laurentis LIGO H1 detector up to 2.15 db in the shot-noise-limited frequency band Losses 10.3 db produced, 2.15 db injected J Aasi et al. Enhanced sensitivity of the LIGO gravitational wave detector by using squeezed states of light. In: Nature Photonics 7.8 (2013), pp

17 Squeezing in AdV Virgo Actually Virgo has a group that are engineering a squeezing set-up similar to that one used in GEO600 Priority in Virgo is AdV The activity is non an R&D and it mainly involves people not hardly involved in the main AdV activities 18 De Laurentis

18 Ponderomotive Squeezing 19 De Laurentis

19 What is interesting in ponderomotive squeezing? Squeezing generation in MOEMS (Micro-Opto-Electro-Mechanical Systems) Communication and integrated sensors; on chip devices OPO integration in the well consolidate silicon technology (mixed technology with KTP and LiNb on Si) is more expensive and does not allow the same integration factor Study of the coupling between macroscopic opto-mechanical objects and their quantum mechanical behavior Theoretical interest Low frequencies and frequency independent squeezing GW detectors PS is a completely suspended system we expect that could be, a regime, a more robust system In long period operation OPO squeezer seems experimentally frequency limited due to losses mechanism in the medium (photo-thermal fluctuations) even if progress in material engineering and dedicate mechanical design promises to overcome these limits. 20 De Laurentis

20 POnderomotive LIght Squeezing POLIS project Project to realize a completely suspended low frequencies ponderomotive squeezer, moving from the pioneers work made in the LIGO laboratory at the MIT [Corbitt et al. Phys. Rev. A 73 (2 Feb. 2006), p ] and taking the advantage of the available Virgo Super Attenuator Facility at EGO, SAFE to control the main noises source in the low frequencies range 21 De Laurentis

21 SAFE LIGO VIRGO The resonance frequency of the first stage is 40 mhz 22 De Laurentis

22 POnderomotive LIght Squeezing POLIS project Ponderomotive Squeezing: large squeezing values without use high laser power and/or very high cavity finesse requires very small suspended mirrors mass very critical point is the mass suspension A relative larger mass allows us to use the available well consolidate technologies A higher chance of success of the project 23 De Laurentis

23 Ponderomotive Squeezing Use of radiation pressure as the squeezing mechanism Intensity fluctuations of laser field creates test mass motion a Test mass motion creates phase shift of reflected light b Phase shift is proportional to intensity fluctuations s dx dx This frequency-dependent shift couples the phase quadrature with the amplitude quadrature and generates the squeezing ( )( )( ) ba 1 0 aa = b P K (Ω) 1 a B Coupling Factor Frequency Dependent Frequency Dependent Squeezing 24 De Laurentis

24 Optical Spring for frequency independent Ponderomotive Spring Detuned Cavity (Optical Spring) detuning increases (cavity becomes longer) the intra-cavity power decreases mirror is pushed back to the detuned point detuning decreases (cavity shorter) power in the cavity increases and the mirror is pushed back 25 De Laurentis

25 Ponderomotive Squeezing Optical Spring modification of the cavity dynamics the mechanical resonance of the mirror is completely given by the optical spring characteristic frequency ±Θ x/f If the pendulum frequency can be neglected (Ωp << Ω, Θ ) f Detuning normalized to the cavity linewidth When constant 1) frequency independent squeezing 2) The Optical Spring resonance gives the frequency independent squeezing band 26 De Laurentis

26 Optomechanical parameter PS Fixed the and key parameters are: Input power, Cavity Finesse, Cavity detuning, Suspended mirrors mass Parameters value optimization to have A large enough squeezing factor and an useful frequency band 27 De Laurentis

27 POLIS Optical Parameters: cavity detuning, squeezing factor and band Cavity detuning: trade-off between squeezing value/band High level increases band Low level increases squeezing We fix: =0.3 => 18 db and =2 khz By considering the losses, this will assures more than 10 db of potential injected squeezing in the interferometer It covers the GWD band 28 De Laurentis

28 POLIS Optical Parameters: cavity Finesse enters directly in the reachable squeezing coupling with the cavity losses Large values Large Optical spring frequency; Reduce effective intracavity losses; losses of 10 ppm, if we impose that squeezing not degraded by more than 60% (to reach the anyway remarkable experimental value of 12 db of squeezing) Low values Reduce optical spring instability; Higher circulating power could damage mirrors F Recently cavities with Finesse as high as has been realized with commercial, low cost mirrors [F Della Valle et al. New Journal of Physics 15.5 (2013), p ] 29 De Laurentis

29 POLIS Optical Parameters Input power Large value Large Optical spring frequency Low values Use of available lasers; No problem with mirrors damage threshold Fixed the detuning, the finesse and the desired optical spring stiffness, we can derive the value for the input power. With high finesse we can relax the input power P=2.5 W N.B.: The most important constraint on the power circulating in the cavity: for powers higher than 0.2 MW the thermal effects start to degrade the behavior of the cavity, For beam waist of mm size, so a circulating power of 0.1 MW will be our conservative constraint. 30 De Laurentis

30 POLIS Optical Parameters Suspended Mirrors Mass Low value High mass ease of construction; ease to sense and actuate motion; use commercial size Large optical Spring resonance (frequency independent squeezing Band) Virgo interferometer is more sensitive than LIGO in the low frequency range, thanks to better seismic isolation ( SAFE), we can chose slightly higher mass m=10 g A standard 25.4 mm mirror in fused silica with a 6.35 mm thickness has a mass of about 7.8 g, while with a 10 mm of thickness it can reach a mass of 11.1 g. It can be suspended with the available technology in Virgo Moreover in this case the gain with respect to the case of 1 g mass is remarkable, because the requirements on the interferometer sensitivity follow as 1/f 2 that is to say, that are far less demanding in the low frequency region, for a given interferometer. 31 De Laurentis

31 POLIS status: cavity parameters Constrain for the cavity length: internale SAFE diameter L = 440 mm Negative stability factor in the cavities with low suspended masses reduce the angular instability Available Radii of Rurvature of the commercial substrates gi = 0.76 RoC = RoCE = 250 mm This value of the RoC assures the maximum spot size on the mirrors with the available commercial radii (large w => reduce thermal deformation of the coating) wi = mm N.B.: Other commercial values: RoCI = RoCE = 300 mm are wi = mm RoC = 200 mm unstable cavity Gouy phase = 2.43 rad = π well stable cavity, avoiding the couplings with thermal deformation or imperfections, mis-match problems and the higher order mode resonance 32 De Laurentis

32 POLIS set-up sketch Configuration simpler than the OPO squeezer set-up we are able to suspend the masses!!!! IF EM1 600 Power and frequency stabilized laser IM1 50x50 MML 50x LEGENDA 440 STR 80x PDH1 Homodyne BS PDH2 13 EM2 IM1: input mirror Is the mirror that brings the Beam from the top on this stage MML: Mode Matching Lens STR: Beam Steering Mirrors To alline the interferometers BS:Beam Splitter EMi: End Mirror IMi:Input Mirror 33 De Laurentis

33 POLIS design of the optical bench Mechanical Design by Andrea Conte Suspended optical bench at the place of the Virgo suspended mirror 34 De De Laurentis Laurentis --18 September September -September th

34 Pendulum Resonance FEM simulation by Andrea Conte 35 De Laurentis

35 POLIS next steps To build the mechanical system (marionette and suspended bench) To experiment the Virgo monolithic suspensions on the 1 inch/10 g mirrors To provide the mirrors for the high finesse suspended cavity (super polished substrates, coating..) 36 De Laurentis

36 Thank you! 37 De Laurentis