WISP searches by Tokyo tabletop experiments group UTokyo tabletop - - PowerPoint PPT Presentation

WISP searches by Tokyo tabletop experiments group

UTokyo tabletop experiments group Toshio NAMBA

UT tabletop experiments group (only related to today’s talk)

• UTokyo Physics & ICEPP:
• S. Asai, T. Inada, T. Yamaji, T. Yamazaki(KEK), ...

+ S. Knirck(Heidelberg)

• RIKEN, JASRI (SPring-8&SACLA):
• K. Tamasaku, Y. Inubushi, M. Yabashi, T. Ishikawa, ...
• UT ISSP & Tohoku Univ.:
• A. Matsuo, K. Kindo, H. Nojiri
• Fukui Univ. FIR:
• T. Idehara, ..

Core members X-ray experts Pulsed magnet experts Millimeter wave experts

1

WISPs (Weakly Interacting Slim Particles)

• Many extensions of the Standard Model

predicts particles in hidden sectors which are

• nly weakly coupled to our visible sector.
• Usually, relatively light (<~eV) hidden sector

particles are called as ‘WISP’.

• A lot of new exotic these particles can arise, but,

in this talk, two kinds of WISPs are discussed.

• Hidden photons (or paraphotons)
• Axion like Particles (ALPs)

2

Hidden photons

• Extra gauge bosons of hypothetical U(1) symmetry
• Tiny kinetic mixing c with ordinary photons
• Neutrino like flavor oscillation
• Tiny fraction of mixing can be separated by an ordinary mirror
• Can be cold dark matter (CDM), if their parameters are

reasonable

3

Photon Hidden photon Heavy fermions with both U(1) charges

γ ! γ f

ℒint = − ' 2 )

*+,*+

Axion Like Particles (ALPs)

• Originally motivated by the strong CP problem
• CPV caused by q vacuum can be cancelled by SSB of PQ symmetry

→ New NG boson, axion

• More generally, axion like particles are predicted by string

theory or SUSY/SUGRA

(No constraints on mass-coupling relations)

• ALPs interact with two photons (Primakoff process)
• r are converted to photons under EM field
• Also are good candidate of CDM

4

ℒ"##=−

%&'' (

)

*+ ,

)*+- = /"##0 1 2-

• g

g ALPs

Many searches for WISPs

5

• Many searches are based on the WISP-photon conversions
• From the experimental view, the difference between hidden photon and

ALP is the conversion method Hidden photons: Appears through its kinetic mixing ALPs: Additional magnetic field (EM field) is required

• Assumed sources of WISPs should be cared when their comparison

Very ambitions? one: Whole CDM component is WISPs (ADMX, etc.) Stars are well-known: Emitted from astrophysical objects (CAST, etc.) Most rigid one: WISPs are artificially created (ALPS, etc.)

Our WISP searches with various methods

• 1. Hidden photon dark matter search in

millimeter-wave region

• 2. Hidden photon search
• 3. ALP search by pulsed magnets
• 4. ALP search by crystalline electric fields

6

Laboratory search @X-ray facility DM search

Hidden photon dark matter search in milli- wave region

• Hidden photon source: Cold dark matter
• Select tiny fraction of photon

components by using a conductive plate.

• Energy of converted light = Mass of HP

7

Done by S. Knirck from Heidelberg During his half year stay in Japan

!"# = %&'( ⃗ *

Strength

conductor

+,

HP

• '(

Converted light (ordinary EM wave)

• ' ~ 0

0~10,2

Target region: Millimeter wave region

• mHP~meV region corresponds to millimeter wave = A marginal

region between photon and radio wave (a little bit difficult to handle)

• Constraint is not so stringent around this region

8

CMB distortions ADMX LSW Coulomb Solar HP Solar lifetime HB Dish (optical) Millimeter wave

Collaborate with millimeter wave experts in Fukui University

Setup of the HPDM search

9

Conversion Plate 600*600mm，Al Schottky barrier diode mixer(SBD) 155~220GHz !"# = 0.6~0.9 !*+ Corrugated Horn (connected to SBD)

Parabolic mirror

HP Parabolic mirror Φ500mm(area: 0.2!.)， f=1500mm，Al

10

Al conversion plate Parabolic mirror 1500mm SBD LO

• scilloscope

SBD (large view) horn

These setups are placed in a radio dark room in Fukui University.

Data taken from Nov. 2016 to Mar. 2017

Obtained result

• If HPDM existed,

a peak would appear at its mass.

• From the absence of such a peak,

kinetic mixing ! ≿ 10%& is excluded for 0.67<mHP<0.92meV (90%C.L.).

• JCAP 11(2018)031.

11

HP mass (meV)

1 −

10 1 10

χ Kinetic mixing

9 −

10

8 −

10

7 −

10

6 −

10

5 −

10 Helioscope (Xe) LSW H e l i

• s

c

• p

e s ( v a c u u m ) Coulomb FIRAS

Now working for MADMAX project

Hidden photon search at X-ray facility

• (Maybe just my biased opinion), we, particle

physicists want to search new particles without cosmological/astrophysical assumptions.

• Create WISPs by ourselves & detect them by

using strong light sources.

• Our source: SPring-8 synchrotron facility
• BL19LXU: One of the most brightest hard X-ray source
• 1013~1014 X-rays/s for 7.2~30keV

12

• Dr. T. Inada

M thesis

SPring-8 Undulator @BL19LXU

Light Shining through a Wall (LSW) method

• Convert photons to WISPs ⇒ Shield photons at a wall ⇒ Re-

convert WISPs to photons ⇒ Detect photons

13

Photon

Photon-WISP conversion WISP

Light source Photon Detector

Wall to shield photons

Light Shining through a Wall (LSW) method

• Convert photons to WISPs ⇒ Shield photons at a wall ⇒ Re-

convert WISPs to photons ⇒ Detect photons

• For hidden photon ⟺ photon conversion

Oscillation in vacuum

14

Photon

Photon-Hidden photon conversion Hidden photon

Light source Photon Detector

Wall to shield photons

P

γ→ " γ = 4χ 2 sin2 m " γ 2

4ω L # $ % % & ' ( (

Search setup in the beamline

• Permanent apparatuses in the beamline were used for

the LSW setup

• The search was performed on June in 2012 for 2 days.
• X-ray energies were changed 9 times from 7.27 keV

to 26 keV.

• A germanium detector was used to detect re-

converted X-rays.

15

• Conversion

Re-conversion Wall

Pb shield

Ge detector (Canberra BE2825)

X-ray beam

Search Results

• No significant signals were observed in all data.
• The spectra for other energies had also no peaks.

16

energy (keV) 7 7.5 8 8.5 9 9.5 10 10.5 11 counts / sec / 0.125 keV

• 1
• 0.5

0.5 1

• 3

10 ×

signal region data 95% C.L. upper limit

9.00 keV measurement

Example of the measured spectrum (9keV)

Search Results

• No significant signals were observed in all data.
• Constraints of ! < 8.06×10)* for (0.04 eV < mHP< 26 keV) were
• btained (95% C.L.)
• Phys. Lett. B722(2013)301.

17

The most sensitive laboratory search at 0.1~100eV region

Next: ALP search

• We want to do ALP search similar as hidden photon search, but,

dedicated magnets are required for the conversion.

• Since the magnetic field should be applied perpendicular to the

light path, and the conversion depends on (BL)2, usual solenoid magnets are not suitable.

18

Light source Detector Photon Photon ALP Wall N S N S P = gαγγBL 2 sinθ θ ! " # $ % &

2

, θ = mα

2l

4ω

• Dr. T. Inada designed and constructed coils & power supplies

dedicated for the ALP search

Our magnet for ALP search

• Racetrack shape coils made of copper wire
• Its length is L=20cm
• Capsulated in a stainless frame to endure the magnetic field stress
• Designed to be operated at pulse mode, 14.1T max, ~1ms duration

(Good for S/N separation)

• Cooled by Liq. Nitrogen

19

• Coil

X-ray path B field 20cm

Power supply for the pulsed magnet

• Total 3mF capacitance (0.25mFx12capacitors) is charged to

4.5kV (30kJ power).

• The rapid charging system enables 0.5Hz repetition rate.
• Total 2ton weight (can be carried by motortrucks).
• NIM A 833(2016)122

20

2m 1.7m

• Charging unit

Capacitors

time [ms]

0.5 1 1.5 2 2.5 3

B [T]

2 4 6 8 10 12 14

Typical excited B field 14.1T ~1ms

ALP search in BL19LXU

• Performed in Nov. 2015
• 4 coils were placed at the X-ray

path in the experimental hatch

• 2 for (X-ray → ALP) conversion
• 2 for( ALP → X-ray) re-conversion
• X-ray energy was set to 9.5 keV
• Net 2 days operation

(total 28,000 excitations)

21

Conversion coils Re-conversion coils Lead shield ! → # # → !

Event distribution

• No significant events correlated to the magnet excitations were
• bserved.
• BG rate is consistent with the one observed event.

22

Energy [keV]

2 4 6 8 10 12 14 16

Count / s / 0.4 keV

0.2 0.4 0.6 0.8

• 3

10 ×

• 100

Black: All events Red: Events when magnet excited x 100

Search result

• Most stringent X-ray LSW limit gagg <2.5110-4 GeV-1 (95%C.L.) was
• btained below ma~0.1eV
• More sensitive search will be performed with upgraded magnets or

using SACLA (XFEL).

• PRL 118(2017)071803

23

[eV]

a

m

• 2

10

• 1

10 1 ]

• 1

[GeV

γ γ a

g

• 4

10

• 3

10

• 2

10 ESRF SPring-8

(2010) (2015)

2.5110-4 GeV-1

CAST ALPS ADMX Axion models

Another X-ray LSW by X-ray diffraction

• Periodic electric fields in crystals (1010V/m~1011V/m)

can be used for photon-ALP conversion (similar power as 102~103 T).

• Relatively heavier ALPs can be converted if the

incident angle of X-rays is tuned.

24

• Dr. T. Yamaji

Doctor thesis

• m2

a − m2 γ − 2qT

• kγsinθγ

T − qT

2

• <

∼ 4kγ L , dhkl λγ λγ θB θB λa>λγ θB θB ~Δθ ~Δθ

• ALP
• Depending on the incident angle

qT=2p/dhkl: Reciprocal lattice spacing

qTg=qB+Dq

Calculation for Laue case diffraction

• T. Yamaji et al., Phys. Rev. D 96(2017)115001

25

• In the case of Silicon: 600µm, 17keV X rays
• ET=4.1x1010V/m
• Leff=488µm
• Until ma10keV can be converted

Pa→γ = 1 2gaγγET LeffcosθB 2 , Leff = 2Latt

• 1 − exp
• −

L 2Latt

The silicon crystal for the search

• Two blades cut from a single silicon crystal were used for
• search. (Lattice planes of two blades are guaranteed as

parallel)

• Dq is scanned by rotating the whole system.

26

27

• µ

Attached to a goniometer, and controlled 1pulse=0.17µrad precision Rotation

• Performed in Oct. 2017, for 2 days.
• Scanned from Bragg angle to Dq=4.6mrad

(Corresponding to 0<ma<1keV)

Placed in BL19LXU

Scan result

• Scans were repeated 4 times.
• ALP signals are expected to distribute for 27.2µrad, but no

significant excesses were found.

28

[mrad] θ ∆ 0.5 1 1.5 2 2.5 3 3.5 4 4.5 Energy [keV] 16 16.5 17 17.5 18 [mrad] θ ∆ 0.5 1 1.5 2 2.5 3 3.5 4 4.5 X rays within signal window [photon] 1 2 3 4 5 6 7

• ..
• Dq (=ma)

Distribution of all events Signal region Events integrated for the interval of the expected angle (27.2µrad) Dq (=ma)

Obtained limit

• Hatched region, relatively heavy ALPs, are excluded.
• Phys. Lett. B 782(2018)523.

29

gaγγ < 4.2 × 10−3 GeV−1 (for ma < 10 eV), gaγγ < 5.0 × 10−3 GeV−1 (for 46 eV < ma < 1020 eV).

CAST ALPS ADMX Axion models

Summary

• UTokyo tabletop group searches WISPs with various methods.

(In niche? region)

• Hidden photon DM search @ millimeter-wave region
• Hidden photon laboratory search @ X-ray facility
• ALP search @ X-ray facility with pulsed magnets
• ALP search @ X-ray facility with crystal diffraction
• Unfortunately, we have not found WISPs yet, but we will

continue to search WISPs with our original ideas & methods.

• (In my personal opinion, rapid progress of intensity frontier of

photon (including X-ray facility) will open new experimental window for the particle physics.)

30

OMAKE

31

Experiment Setup ②

• Corrugate Horn

(Coupling with condensed signal ~ -2dB)

• SBD mixes millimeter-wave(!

"#) with local oscillator

signal(!

$%), downconverts to ! &#(conversion loss ~ -40dB)

• Amplify SBD output(~36dB×2，total gain: ~ 72dB)
• After amplified，FFT by oscilloscope (region: 0~4GHz)

LO !

$% = 20~27,-.

!

&# = |! "# − 8×! $%| < 4,-.

SBD !

"#

(155~220GHz) Amplifier

• scilloscope

32

• Input power !"#， output power of SBD !$%&

Conversion Loss '. ). =

+,-. +/0

SBD conversion loss

BWO Conical horn SBD SBD ~ lenses 260.0 ~ 100.0 mm 光学ステージ Teflon lenses R50mm, Φ80mm Horn ~ lenses 256.0 mm(fixed)

Peak Search

• HPDM signal reflects its velocity distribution(Maxwell-

Boltzman dist.)

! " = $%& " − ()* ×2 " − ()*

• "%

./ 0 exp − " − ()4

"% +6"0 + 7" + 8%

$%: Power of HPDM signal &: Step function "%: dispartion const. (()*×4e-7)

(Calculate from standard Halo model velocity dist. of HPDM)

()*: mass of HPDM Peak search for each ()* using ! "

($%, a,b, 8%:fitting parameters)

34

BG "%~4()*×10.SmeV

"

power

()*

• Probed mass scales to the photon energy
• Sources with different energies are important for extending the

LSW limits

• Oscillation of the probability of

and

γ → " γ ! γ →γ

P

γ→ " γ = ω + ω 2 − mγ ' 2

ω 2 − mγ '

2

χ $ % & & ' ( ) )

2

sin L 2 ω − ω 2 − mγ '

2

( )

$ % & ' ( )

P

γ→ " γ = 4χ 2 sin2 m " γ 2

4ω L # $ % % & ' ( (

Oscillation length (fixed by setup) Photon energy Mixing angle Paraphoton mass

• For mγ’<<ω
• Axion LSW by S. L. Adler et al (2008) ⇒ Paraphoton LSW

Oscillation Probability for LSW Experiments

35

X-ray Intensity Frontier SPring-8 and our beamline BL19LXU

• SPring-8 (Super Photon ring at 8 GeV)
• 62 beamlines around 1.42 km electron ring
• X-rays from soft (1 keV) to hard ( 100 keV)
• BL19LXU (BeamLine 19 Long X-ray Undulator)
• 30-m-long in-vacuum undulaot

ÞMost intense X-rays available today as a

continuous beam

BL19LXU Value (after monochromator) Output energy 7.2–51 keV Beam intensity 1013–1014 photon/s @7.2–30 keV Line width ∼eV (FWHM) Beam size ∼400 µm (FWHM) Pulse width/interval 40 ps/24 ns (∼CW)

SPring-8

36

Undulator

x (mm)

• 1 -0.8 -0.6 -0.4 -0.2

0.2 0.4 0.6 0.8 )

• 1

(x) (mm ρ 0.5 1 1.5 y (mm)

• 1 -0.8 -0.6 -0.4 -0.2

0.2 0.4 0.6 0.8 )

• 1

(y) (mm ρ 0.5 1 1.5 2 2.5

800 μm 400 μm beam intensity bean intensity horizontal vertical

Beam Size

• Space structure
• Measured with a slit scan along horizontal/vertical direction

with a 10 μm pitch

• Vertical width 400 μm (FWHM)

37

1/2

4.5kV 15kV/A

• 4.5kV9T
• 10mΩ/
• 3kV
• 3kV26T

3mF

9 6 3

• 3
• 6
• 9

38

z [m]

0.15 − 0.1 − 0.05 − 0.05 0.1 0.15

B/I [T/kA]

0.1 − 0.1 0.2 0.3 0.4 0.5 0.6 0.7

• 20 cm
• 5.3 mm
• à
• 0.2%
• SUS
• 2

39

gaγγ

time [ms] 1 2 3 4 5 I [kA]

• 6
• 4
• 2

2 4 6 8 10

• B/I [T/kA]

B/I [T/kA]

∝ g2

aγγ

B(z,t)dz

∫

( )

2

• à P(t))

N 2.96 @95%CLη 89% F X3.01013 photon/s

• B(z,t)=B/I(z)I(t)

P: /

• 40

: LSW(Light Shining through a Wall)

• ALPS (

(5Tx8.8m))

(ma<10-3eV)

41

• 42/44
• α

β

• 10

15 20 25 30 35 40 45 50

15

• 10

14

• 10

13

• 10

12

• 10

11

• 10

10

• 10

9

• 10

8

• 10

7

• 10

simplified approx. this result

1

• 0.8
• 0.6
• 0.4
• 0.2
• 0.2

0.4 0.6 0.8 1

15

• 10

14

• 10

13

• 10

12

• 10

11

• 10

10

• 10

9

• 10

8

• 10

7

• 10

simplified approx. this result

• Δθ=4.6 mrad()
• ma-1keV [eV]
• 1 keV

44/44

• 45/44

10-1 101 103 103 104 ma [eV] normaliz 102

200 400 600 800 1000 1200 0.2 0.4 0.6 0.8 1

a → γ γ → a

Δθ [mrad] DcosθT /cosθB 4.6 mrad =1 keV 510 mrad =10 keV Si(220)+17 keV (θB=10.95 deg) θγT=90 deg 510mrad Δθ [mrad]

• (X)
• Bragg(qB=10.95)

X2

• Dq/RR2%

46

rad] µ [ q D 40

• 30
• 20
• 10
• 10

20 30 40 RR efficiency [%] 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2

• : 10.8µrad

X: 6.1µrad

• X: 2.1eV

• 47/44
• qT

[keV] θB [deg] ET [V/m] Latt EL (Ge: 1) C(220) 9.85 16.8 6.81010 7.7mm 266 Si(220) 6.46 10.9 4.41010 650μm 14.5 Ge(220) 6.20 10.5 7.31010 27μm 1

(220)

• 48/44
• gaγγ [GeV-1]

ma [eV]