with Intense X- ray Beams at SPring- 8 and WISPs

Size: px

Start display at page:

Download "with Intense X- ray Beams at SPring- 8 and WISPs"

Edward Allen
5 years ago
Views:

1 Results of a Search for Paraphotons 9th Patras Workshop on Axions, WIMPs with Intense X- ray Beams at SPring- 8 and WISPs Toshiaki Inada University of Tokyo T. Namba, S. Asai, T. Kobayashi University of Tokyo/ICEPP Y. Tanaka, K. Tamasaku, K. Sawada, T. Ishikawa SPring-8/RIKEN Harima institute Based on Phys. Le.. B 722 (203) 30

2 Paraphoton/Hidden Sector Photon Photon Both U() charges Paraphoton γ γ! f Gauge bosons of hypothe=cal U() symmetry - Predicted by string- based extensions of Standard Model Tiny mixing with ordinary photons - Neutrino- like flavor oscilla=on Precise test of QED Abnormal heat transfer mechanism in stars One candidate of dark ma.er 2

3 OscillaJon Probability for LSW Experiments Oscilla=on of the probability of γ γ " and γ! γ Axion LSW by S. L. Adler et al (2008) Paraphoton LSW P γ " $ γ = ω + ω 2 2 m ' & γ ' χ ) & ω 2 2 % m ) γ ' ( - For m γ <<ω P γ " 2 $ sin L 2 ω ω 2 2 & ( m γ ' % ) # 2 m γ = 4χ 2 sin 2 γ" % $ 4ω L & ( Oscilla=on length (fixed by setup) ' Mixing angle Paraphoton mass Probed mass scales to the photon energy - Sources with different energies are important for extending the LSW limits ' ) ( Photon energy 3

4 Current Terrestrial Limits Mixing angle Coulomb -4 0 ALPS 0 Rydberg Op=cal BFRT LSW region BMV GammeV LIPPS -2 0 our experiment X- ray LSW region Paraphoton mass m (ev) Op=cal LSW range: mev ev - Below this: Microwave - Above this: X- ray region Purely terrestrial- and intense X- ray source - Synchrotron radia=on facility R. Ba.es= et al ESRF for axion- LSW Extends LSW limits to higher masses 4

in- vacuum undulaot Most intense X- rays available today as a con=nuous beam BL9LXU Value (after monochromator) Undulator Output

5 X- ray Intensity FronJer SPring- 8 and our beamline BL9LXU SPring- 8 (Super Photon ring at 8 GeV) - 62 beamlines around.42 km electron ring - X- rays from sog ( kev) to hard ( 00 kev) SPring- 8 BL9LXU (BeamLine 9 Long X- ray Undulator) m- long in- vacuum undulaot Most intense X- rays available today as a con=nuous beam BL9LXU Value (after monochromator) Undulator Output energy Beam intensity Line width Beam size Pulse width/interval kev kev ev (FWHM) 400 µm (FWHM) 40 ps/24 ns ( CW) 5

6 Beam Energies and Fluxes 9 energies are used Fluxes of higher harmonics are rela=vely weaker We used st/3rd harmonics harmonics used energy (kev) flux (photon/s) n = 7.27,8.00,9.00,5.00,6.00, n =3 2.83,23.00, Measured with a Si PIN photodiode Accuracy be.er than (avg.) 2% 6

7 Overview of Experimental Setup and DefiniJon of the OscillaJon Regions first oscilla=on region second oscilla=on region Photon Paraphoton Photon Light source Oscilla=on Detector 7

8 Overview of Experimental Setup and Beamline Components Beams from the undulator has a con=nuous spectrum Monochromated to Δω/ω 0-4 with a Bragg condi=on Blocked by a 94- mm- thick lead shu.er Only LSW photons are selected by a pair of total reflec=on mirrors Detected with a germanium detector in a experimental hatch 8

9 Overview of Experimental Setup and Beamline Components - Upstream - - Downstream - monochromator DSS DSS the first mirror 277 cm 65.4 cm 9

DetecJon System - Setup - - Inside the experimental hatch - X- ray beam Pb shield Ge detector (Canberra BE2825) φ60 germanium crystal Shielded by 5 cm- thick leads Beams are injected through a φ30

10 DetecJon System - Setup - - Inside the experimental hatch - X- ray beam Pb shield Ge detector (Canberra BE2825) φ60 germanium crystal Shielded by 5 cm- thick leads Beams are injected through a φ30 collimator Energy spectrum is recorded by a peak- holed ADC for energy cuts Component Ge crystal Detector window (CFRP plastic) Pb shield Beam collimator Value diameter 60 mm thickness 25 mm thickness 0.6 mm thickness 50 mm diameter 30 mm 0

11 Energy ResoluJon of Ge Detector and DefiniJon of Signal Region counts / sec / 0.04 kev Co spectrum - Fe X- rays 4.4 kev energy (kev) (kev) Fe RI source Fitting 57 Co ω 24 Am 2 / ndf.696 / 3 Prob p ± p 0.23 ± energy (kev) Measured with RI sources - 57 ev 4.4 kev from 57 Co Interpolated by the func=on of σ = p 0 E + p (kev) Defined beam energy ±2σ as a signal region

12 DetecJon Efficiency efficiency A.enua=on in the window Ge K- edge ω data MC (best fit, 7.7 µm) MC (with 30 cm air attenuation, 8.6 µm) energy (kev) Also measured by RI sources Thickness of surface dead layer is crucial for inefficiency around Ge K- edge GEANT4 simula=on with the thickness as a free parameter Conserva=ve curve of σ devia=on including air a.enua=on (dashed red) is used in the analysis Typical efficiencies: 7 kev and 26 kev 2

13 Setup of Main Measurements with Beam ON and OFF Beam ON - Change beam energy for 9 =mes - Live=me on each measurements: 5-9 hours Beam OFF - Completely the same setup except for closing the main beam shu.er hours of live=me Paraphoton signal - Sta=s=cally significant difference of the detector count rates between ON and OFF 3

14 Background Spectrum Arrows show the signal regions of 9 measurements - No overlaps => commonly used for subtrac=on counts / sec / 0.25 kev lead X-rays energy (kev) 0.5 kev and 2.7 kev peaks: Lead X- rays from shields and a collimator - Avoided for the choice of beam energy Except for this, normal con=nuous spectrum BG rate in each signal region is a few mhz 4

15 Background- Subtracted Spectrum One example of 9 kev measurement Bars are sta=s=cal errors ( sigma) and signal region data with red points Paraphoton- like signal over +2 standard devia=ons was not detected! (Also with the other energy measurements) counts / sec / 0.25 kev kev measurement signal region data 95% C.L. upper limit energy (kev) Hereager, focus on the discussion of constraining the mixing angle Dashed line: a signal upper limit(95% C.L.) calculated from total counts in the signal region 5

16 Background- Subtracted Spectrum The Other Energy Measurements kev measurement kev measurement kev measurement kev measurement counts / sec / 0.25keV signal region data counts / sec / 0.25keV signal region data counts / sec / 0.25keV signal region data counts / sec / 0.25keV signal region data 95% CL upper limit 95% CL upper limit 95% CL upper limit 95% CL upper limit energy (kev) energy (kev) energy (kev) energy (kev) kev measurement kev measurement kev measurement kev measurement counts / sec / 0.25keV signal region data counts / sec / 0.25keV signal region data counts / sec / 0.25keV signal region data counts / sec / 0.25keV signal region data 95% CL upper limit 95% CL upper limit 95% CL upper limit 95% CL upper limit energy (kev) energy (kev) energy (kev) energy (kev) Significant excess was not observed 6

17 CalculaJon of Mixing Angle Ver=cal direc=on with y axis Detec=on efficiency ɛi P (y) P 2 (y) ρ(y) dy = N 95%C.L. Photon flux (s - ) LSW- photon flux (s - ) Signal upper limit (s - ) ρ(y): Beam profile along the y direc=on - Normalize its area to a unit Neutrino- like conversion probability P i (y) = [ 2χ sin ( m 2 γ L i (y) )] 2 4ω (for low masses) - Depends on L, the oscilla=on region length! 7

18 Beam Size Space structure - Measured with a slit scan along horizontal/ver=cal direc=on with a 0 μm pitch - Ver=cal width 400 μm (FWHM) ) - (x) (mm beam intensity horizontal ) - (y) (mm μm 400 μm bean intensity ver=cal x (mm) y (mm) 8

19 TilJng Edges and Beam Width Effect Both edges of oscilla=on region (i.e. monochromator and first mirror) have shallow angles along the beam axis Length of oscilla=on region changes with respect to the local y posi=ons! y ρ(y) ΔL variances from beam width ( 400 μm) - First oscilla=on region: a few mm <= Bragg angle 00 mrad - Second oscilla=on region: 0 cm <= total reflec=on angle few mrad Integrate over each y contribu=ons 9

20 Limits on the Mixing Angle 95% C.L exclusion limit (upper side is excluded) One example from a single 9.00 kev measurement (a) (b) m (ev) Region (a): spiky structure from the sin func=ons of oscilla=on probabili=es Region (b): smeared out due to the integra=on - For heavy masses: oscilla=on length < ΔL variance 20

21 CombinaJon of the Results Obtain a combined result by the same procedure using χ 4 distribu=ons and mul=plying each others Spiky structures of the region (a) are compensated with 9 measurements Magnified view kev 9.00 kev combined 0.07 (a) (b) The worst value appears at.39 ev: χ worst = Represents our result m (ev) 2

22 SystemaJc Errors Uncertain=es of the beam intensi=es and detec=on efficiencies - Already taken into account by using σ decreased conserva=ve values(*) The other uncertain=es: energy scale and oscilla=on region length Factor Contribution to χ worst Beam intensities (avg.)±0.40 % Detection efficiencies (avg.) % Absolute energy scale ( ω = 8 ev) < ±0.0 % Oscillation lengths (L =277± 2cm, L 2 =65.4 ± 0.5 cm) % Appear in the phase of sin func=on - Cause a shig of the whole limit line along the mass axis Traced χ worst by changing the two parameters and listed maximum devia=ons χ worst % represents our final result: χ < (95% C.L.) 22

23 Comparison of the Results Probed mass region up to 26 kev - 4- order- heavier than op=cal LSW ev 0 - Rydberg our experiment Coulomb ALPS BFRT Op=cal LSW region BMV GammeV LIPPS Most stringent as a LSW limit for this region m (ev) 23

Further Prospects /2 Paraphoton Search New X- ray source: Free Electron Laser, SACLA - In public use since last year, and reaches to the designed performance in next year - The same flux (s - )

24 Further Prospects /2 Paraphoton Search New X- ray source: Free Electron Laser, SACLA - In public use since last year, and reaches to the designed performance in next year - The same flux (s - ) with SPring- 8 and pulsed beam Pulse width 0 fs Time window of detector coincides with beam pulse - Zero background count 2- order- improvements of S/N for one week measurement SACLA SPring- 8 24

25 Further Prospects 2/2 Axion- like ParJcle Search Introduce photon- ALP conversion magnets to the paraphoton setup Shield Photon N N Photon ALP Light source Already maintains eight dipole S S Conversion magnets Detector magnets decommissioned from KEKB - 2 T 2. m 8 Improvement by factor 40 from NOMAD (2000) is expected ) - (GeV g Coupling constant ESRF ALPS (200) (200) Expected SACLA NOMAD ALP mass (2000) 2 0 m (ev) 25

26 Summary Paraphoton search using intense X- ray beams was performed at SPring- 8. LSW method was applied and wall- penetra=ng LSW photons were searched with Ge detector. From the absence of paraphoton signals, a new experimental constraint was obtained: χ < (0.04 ev < m γ < 26 kev, 95% C.L.) Probed mass region is 4- order- heavier than op=cal LSW searches 26

Axion Searches Overview. Andrei Afanasev The George Washington University Washington, DC

Axion Searches Overview. Andrei Afanasev The George Washington University Washington, DC Axion Searches Overview Andrei Afanasev The George Washington University Washington, DC Andrei Afanasev, Intense Electron Beams Workshop, Cornell University, 6/17/2015 Introduction to a Dark Matter problem