for the APEX Collaboration

The Aʹ′ Experiment (APEX) Searching for New Gauge Bosons in the Aʹ

Experiment at Jefferson Laboratory

Natalia Toro (Perimeter Institute)

1

S. Abrahamyan, Z. Ahmed, K. Allada, D. Anez, T. Averett, A. Barbieri, K. Bartlett, J. Beacham, J. Bono, J.R. Boyce, P. Brindza, A. Camsonne, K. Cranmer, M.M. Dalton,

C.W. de Jager, J. Donaghy, R. Essig (co-spokesperson), C. Field, E. Folts, A. Gasparian, N. Goeckner-Wald, J. Gomez, M. Graham, J.O. Hansen, D.W.

Higinbotham, T. Holmstrom, J. Huang, S. Iqbal, J. Jaros, E. Jensen, A. Kelleher, M. Khandaker, J.J. LeRose, R. Lindgren, N. Liyanage, E. Long, J. Mammei, P.

Markowitz, T. Maruyama, V. Maxwell, S. Mayilyan, J. McDonald, R. Michaels, K. Moffeit, V. Nelyubin, A. Odian, M. Oriunno, R. Partridge, M. Paolone, E. Piasetzky,

I. Pomerantz, Y. Qiang, S. Riordan, Y. Roblin, B. Sawatzky, P. Schuster (co-spokesperson), J. Segal, L. Selvy, A. Shahinyan, R. Subedi, V. Sulkosky, S.

Stepanyan, N. Toro (co-spokesperson), D. Walz, B. Wojtsekhowski (co-spokesperson), J. Zhang and The Hall A Collaboration

0.001 0.01 0.1 110-1110-1010-910-810-710-610-510-40.001 0.01 0.1 1

10-1110-1010-910-810-710-610-510-4

mA' HGeVL

a'êa

APEXêMAMITest Runs

U70

E141

E774

am, 5s

am,±2s favored

ae

BaBar

KLOE

Orsay

HPS APEXDarkLight

VEPP3

HPS

Outline

2

1) The APEX experiment: general setup and rationalea few important details

2) Test run (July 2010)results

3) Full APEXextended targetSciFi optics calibration for

better mass resolutionnew septum

In brief: APEX is a spectrometer-based search, at JLab Hall A, for 50-500 MeV hidden-sector photons decaying promptly to e+e–.

PRL 107:191804,2011, arxiv:1108.2750

JHEP 1102:009,2011, arxiv:1001.2557

APEX Concept and Dark Photon Production

��mA

E

⇥3/2

(wide)

(narrow)

e�

Energy = E

e�

��mA

E

⇥1/2 l+

l�

� mA

EA�

EAʹ′≃Ebeam-mAʹ′

Ee-≃mAʹ′

3

HRS−right

HRS−left

Electron, P = E0/2

Positron, P = E0/2

.

.

Septum

W target

Beam

Nucleus

A�

e+

e�

e�

The High Resolution Spectrometers

4

Lead Glass

Calorimeter S2m

Gas Cherenkov

VDC

S0

0.3<p<4.0 GeV/c -4.5%<Δp/p<4.5% δp/p ≤ 2 10-4 12.5º<θ0<150º 6msr δφ=0.5 mrad (H)

δθ=1 mrad (V) (4.5 msr at θ0=5º with septum)

Range Acceptance Resolution

Distinctive kinematics

Aʹ′ Production and Background Kinematics (mAʹ′≪Ebeam)

0.0 0.2 0.4 0.6 0.8 1.005101520253035

HE++E-LêEbeam

Evtsês

A' Ha'êa=3 10-6L

QED Backgroundê103

A′ Production

Nucleus

A�

e+

e�

e�

5

QED Backgrounds

Aʹ products carry (almost) full beam energy!

γ*

(rates before angular cuts)

σ ~ αʹ′/m2 = ε2α/m2 dσ~α2/m3 dm

Symmetric energy, angles in two arms optimize A′ acceptance

E+ ≈ E– ≈Ebeam/2also suppresses e– singles & other pair backgrounds

After rejecting accidental e–π+ (demonstrated in test run), event rate dominated by QED backgrounds above

APEX test run

• Test run performed in Hall A, July 2010

• Verified all key aspects of apparatus performance– VDC tracking performance at 4–6 MHz singles rates– Gas Cerenkov detector in coincidence trigger to reject π+’s– spectrometer optics & mass resolution– measurement of physics backgrounds

• Resonance search on 700K good trident events

6

Many thanks to JLab & Hall A staff for tremendous support!

180 190 200 210 220 230 240 250

-300

-200

-100

0

100

200

300170 180 190 200 210 220 230 240 2500

1000

2000

3000

4000

5000

6000

7000

8000

9000

mass [MeV]-e+e180 190 200 210 220 230 240 250

Resid

ual

-2000

200

Even

ts / 0

.5 M

eV

0

2000

4000

6000

8000

0.1 0.2 0.3 0.4 0.5-710

-610

-510

-410

BaBar

MAMIKLOE+a

TestAPEX

[MeV]A'm100 200 300 400 500

100 300 500_'/

_

-710

-610

-510

-410400200

400200

Accidental

QED (no efficiency

correction)

Data

0.1 0.3 0.510-8

10-7

10-6

10-5

0.1 0.3 0.5

10-8

10-7

10-6

10-5

e+e-HA'LMass HGeVL

a'êaH2s

sensitivityL

Sensitivity of Proposed Run Plan

C AD

B0.1 0.2 0.3 0.4 0.5

-710

-610

-510

-410

BaBar

MAMIKLOE+a

TestAPEX

[MeV]A'm100 200 300 400 500

100 300 500_'/

_

-710

-610

-510

-410400200

400200

Full APEX run plan and sensitivity

7

1 Month Beam Time– 6 days at 1,2,3 GeV– 12 days at 4.5 GeV)

>100x test-run statistics

Approved by JLab PAC 37 with recommendation to run as soon as possibleExplores parameter space with unparalleled efficiency (particularly above ~300 MeV)

0.1 0.2 0.3 0.4 0.5-710

-610

-510

-410

BaBar

MAMIKLOE+a

TestAPEX

[MeV]A'm100 200 300 400 500

100 300 500_'/

_

-710

-610

-510

-410400200

400200

A1 Experiment

Achim Denig Dark Photon Search @ MAMI

Mixi

ng P

aram

eter

ε2

mγ’ [MeV]

This work:MAMI 2012MESA, 10°MESA, 20°

10-8

10-7

10-6

10-5

10-4

10 100

(g-2)µ

|(g-2)µ|< 2σ

E141

E774

KLOE

MAMI

APEX

(g-2)e vs. α

Figure 12. Compilation of existing exclusion limits and our predictions: For a better visualization we restrict ourselves to theregion currently accessible at fixed-target experiments. Only existing limits as published in Refs. [16, 21, 36, 37, 39, 40] areshown, represented by the shaded regions. We do not show the predictions for other experiments [25, 28–30] in this figurewhich are scheduled to probe the same region of parameter space. The limits of MAMI and APEX are those as given in theirpublications [26, 27]. The prediction of this work for the exclusion limit expected for the MAMI 2012 experiment discussedin section IIIB is depicted by the dashed curves. The prediction for MESA obtained in section IV is indicated by the dotted(dashed-dotted) curve for the setups with a central scattering angle of 10! (20!).

signal cross section given in Ref. [21].For comparison we show in Fig. 10 the acceptance integrated cross section depending on me+e! for a proton targetwith a beam energy of E0 = 80MeV. In the left panel the same curves as in Fig. 9 are plotted. In the right panel ofFig. 10 it is demonstrated, that the VCS contribution corresponding with the Feynman diagrams in Fig. 2 are smallerby more than 6 orders of magnitude in the chosen kinematic setting, and can thus be neglected. As indicated by theshape of the curves for !!SL+TL

!",D+X and !!TL!",D in Figs. 9 and 10, the ratio of these two quantities is equal and thus

the kind of target does not a!ect the exclusion limit concerning the QED background.Fig. 11 shows the calculated ratio !!!/!!TL

! which reaches a value around 8 ! 10 for the proposed settings. Theexpected exclusion limit on "2 as obtained from Eq. (14), to the invariant mass spectra of Fig. 9, is presented onFig. 12, where a mass resolution of 0.125MeV was assumed. The dotted (dashed-dotted) curve on Fig. 12 representsthe settings with central angle of 10! (20!). At very low masses below 10MeV Eq. (14) does not serve as a goodapproximation for the exclusion limit anymore, since Eq. (19) of Ref. [21] overestimates the #" signal cross section byup to 50%.

A compilation of the existing exclusion limits is presented in Fig. 12, which shows the region 5MeV " m!# "600MeV and 10#8 " "2 " 10#4 accessible at fixed-target experiments. Furthermore, existing limits as published inRefs. [16, 21, 36, 37, 39, 40] are also shown, and are represented by the shaded regions. Let us mention that otherplanned experiments [25, 28–30] are scheduled to probe the same region of parameter space. The limits of MAMI andAPEX are those as given in their publications [26, 27]. Our prediction for the exclusion limit expected in the MAMI2012 experiment discussed in section III B is depicted by the dashed curves. The prediction for MESA obtained insection IV is indicated by the dotted (dashed-dotted) curves for the setups with a central scattering angle of 10! (20!).Our calculation shows, that the 2012 experiment is well suited to exclude a large region of the parameter space and

14

Mainz MAMI/A1 experiment– superficially similar

(spectrometer e+e– coincidence)– similar test-run sensitivity

but...– hitting detector rate limits

at least 6mo run needed to match APEX below 200 MeV

– Search above ~200 MeV also hitting beam current and radiation limits

APEX in Context

PRL  106  2518021303.2540

Bottom line: APEX has unique advantage due to high-rate capability and septum magnet (small angle ⇒ higher signal acceptance) 8

0.01 0.110-8

10-7

10-6

10-5

10-40.01 0.1

10-8

10-7

10-6

10-5

10-4

mA' HGeVL

a'êa

APEXfull

A1 Experiment

Achim Denig Dark Photon Search @ MAMI

Mixi

ng P

aram

eter

ε2

mγ’ [MeV]

This work:MAMI 2012MESA, 10°MESA, 20°

10-8

10-7

10-6

10-5

10-4

10 100

(g-2)µ

|(g-2)µ|< 2σ

E141

E774

KLOE

MAMI

APEX

(g-2)e vs. α

Figure 12. Compilation of existing exclusion limits and our predictions: For a better visualization we restrict ourselves to theregion currently accessible at fixed-target experiments. Only existing limits as published in Refs. [16, 21, 36, 37, 39, 40] areshown, represented by the shaded regions. We do not show the predictions for other experiments [25, 28–30] in this figurewhich are scheduled to probe the same region of parameter space. The limits of MAMI and APEX are those as given in theirpublications [26, 27]. The prediction of this work for the exclusion limit expected for the MAMI 2012 experiment discussedin section IIIB is depicted by the dashed curves. The prediction for MESA obtained in section IV is indicated by the dotted(dashed-dotted) curve for the setups with a central scattering angle of 10! (20!).

signal cross section given in Ref. [21].For comparison we show in Fig. 10 the acceptance integrated cross section depending on me+e! for a proton targetwith a beam energy of E0 = 80MeV. In the left panel the same curves as in Fig. 9 are plotted. In the right panel ofFig. 10 it is demonstrated, that the VCS contribution corresponding with the Feynman diagrams in Fig. 2 are smallerby more than 6 orders of magnitude in the chosen kinematic setting, and can thus be neglected. As indicated by theshape of the curves for !!SL+TL

!",D+X and !!TL!",D in Figs. 9 and 10, the ratio of these two quantities is equal and thus

the kind of target does not a!ect the exclusion limit concerning the QED background.Fig. 11 shows the calculated ratio !!!/!!TL

! which reaches a value around 8 ! 10 for the proposed settings. Theexpected exclusion limit on "2 as obtained from Eq. (14), to the invariant mass spectra of Fig. 9, is presented onFig. 12, where a mass resolution of 0.125MeV was assumed. The dotted (dashed-dotted) curve on Fig. 12 representsthe settings with central angle of 10! (20!). At very low masses below 10MeV Eq. (14) does not serve as a goodapproximation for the exclusion limit anymore, since Eq. (19) of Ref. [21] overestimates the #" signal cross section byup to 50%.

A compilation of the existing exclusion limits is presented in Fig. 12, which shows the region 5MeV " m!# "600MeV and 10#8 " "2 " 10#4 accessible at fixed-target experiments. Furthermore, existing limits as published inRefs. [16, 21, 36, 37, 39, 40] are also shown, and are represented by the shaded regions. Let us mention that otherplanned experiments [25, 28–30] are scheduled to probe the same region of parameter space. The limits of MAMI andAPEX are those as given in their publications [26, 27]. Our prediction for the exclusion limit expected in the MAMI2012 experiment discussed in section III B is depicted by the dashed curves. The prediction for MESA obtained insection IV is indicated by the dotted (dashed-dotted) curves for the setups with a central scattering angle of 10! (20!).Our calculation shows, that the 2012 experiment is well suited to exclude a large region of the parameter space and

14

Mainz MAMI/A1 experiment– superficially similar

(spectrometer e+e– coincidence)– similar test-run sensitivity

but...– hitting detector rate limits

at least 6mo run needed to match APEX below 200 MeV

– Search above ~200 MeV also hitting beam current and radiation limits

APEX in Context

PRL  106  2518021303.2540

Bottom line: APEX has unique advantage due to high-rate capability and septum magnet (small angle ⇒ higher signal acceptance) 8

0.01 0.110-8

10-7

10-6

10-5

10-40.01 0.1

10-8

10-7

10-6

10-5

10-4

mA' HGeVL

a'êa

Dashed green is updated sensitivity estimate for ~14 total beam-days (9 settings) [1303.2540  Beranek,  Merkel,  Vanderhaeghen]  

Detector, beam, and radiation limitations ⇒ Mainz probably can only access high-coupling & low-mass portion of APEX sensitivity (without major upgrades)

APEX vs. Mainz

Mixi

ng P

aram

eter

ε2

mγ’ [MeV]

This work:MAMI 2012MESA, 10°MESA, 20°

10-8

10-7

10-6

10-5

10-4

10 100

(g-2)µ

|(g-2)µ|< 2σ

E141

E774

KLOE

MAMI

APEX

(g-2)e vs. α

Figure 12. Compilation of existing exclusion limits and our predictions: For a better visualization we restrict ourselves to theregion currently accessible at fixed-target experiments. Only existing limits as published in Refs. [16, 21, 36, 37, 39, 40] areshown, represented by the shaded regions. We do not show the predictions for other experiments [25, 28–30] in this figurewhich are scheduled to probe the same region of parameter space. The limits of MAMI and APEX are those as given in theirpublications [26, 27]. The prediction of this work for the exclusion limit expected for the MAMI 2012 experiment discussedin section IIIB is depicted by the dashed curves. The prediction for MESA obtained in section IV is indicated by the dotted(dashed-dotted) curve for the setups with a central scattering angle of 10! (20!).

signal cross section given in Ref. [21].For comparison we show in Fig. 10 the acceptance integrated cross section depending on me+e! for a proton targetwith a beam energy of E0 = 80MeV. In the left panel the same curves as in Fig. 9 are plotted. In the right panel ofFig. 10 it is demonstrated, that the VCS contribution corresponding with the Feynman diagrams in Fig. 2 are smallerby more than 6 orders of magnitude in the chosen kinematic setting, and can thus be neglected. As indicated by theshape of the curves for !!SL+TL

!",D+X and !!TL!",D in Figs. 9 and 10, the ratio of these two quantities is equal and thus

the kind of target does not a!ect the exclusion limit concerning the QED background.Fig. 11 shows the calculated ratio !!!/!!TL

! which reaches a value around 8 ! 10 for the proposed settings. Theexpected exclusion limit on "2 as obtained from Eq. (14), to the invariant mass spectra of Fig. 9, is presented onFig. 12, where a mass resolution of 0.125MeV was assumed. The dotted (dashed-dotted) curve on Fig. 12 representsthe settings with central angle of 10! (20!). At very low masses below 10MeV Eq. (14) does not serve as a goodapproximation for the exclusion limit anymore, since Eq. (19) of Ref. [21] overestimates the #" signal cross section byup to 50%.

A compilation of the existing exclusion limits is presented in Fig. 12, which shows the region 5MeV " m!# "600MeV and 10#8 " "2 " 10#4 accessible at fixed-target experiments. Furthermore, existing limits as published inRefs. [16, 21, 36, 37, 39, 40] are also shown, and are represented by the shaded regions. Let us mention that otherplanned experiments [25, 28–30] are scheduled to probe the same region of parameter space. The limits of MAMI andAPEX are those as given in their publications [26, 27]. Our prediction for the exclusion limit expected in the MAMI2012 experiment discussed in section III B is depicted by the dashed curves. The prediction for MESA obtained insection IV is indicated by the dotted (dashed-dotted) curves for the setups with a central scattering angle of 10! (20!).Our calculation shows, that the 2012 experiment is well suited to exclude a large region of the parameter space and

14

–no publishable result expected from 2012 run

–present dark-force plans focus on lower Aʹ′ mass, including MESA APEX

full

1303.2540

9

Readiness for Full Run

• APEX extended target– target built for test run is at JLab

• SciFi detector for optics calibration– design and production complete; first arm

assembled and tested

• Septum– design nearly finalized & vendor quotes obtained

• HRS detector maintenance is proceeding

10

Target Design: Minimizing Multiple Scattering

schematic overhead view

beam5º

15µm W

Goals: • σ(θ)mult scat≤0.5 mrad

⇒ typical e+e– pair must only go through 0.3% X0 (2-pass) • Target thickness 0.7–8% X0 (depending on Ebeam)

• High-Z target (reduce π yield for given QED rates)• Stable under currents up to ~100 μA

long target ⇒ wider single-run mass coverage 11

Target designed and built by SLAC APEX group for the test run (but not installed), currently at JLab.

Target Design: Minimizing Multiple Scattering

11

Target designed and built by SLAC APEX group for the test run (but not installed), currently at JLab.

Magnetic Spectrometer Optics

12Top view

Measuring Contributions to the Mass Resolution(dominant: angular resolution + mult. scatter)

Test Run Optics Calibration

13

Removable sieve plate is inserted upstream of septum.

Use surveyed locations of sieve holes to calibrate magnetic optics.

Use reconstructed hole sizes to measure resolution.

...this method only works for negative polarity, and requires running at different beam energy.

Mass resolution≈1 MeV~0.5%

Test Run Optics Calibration

13

Removable sieve plate is inserted upstream of septum.

Use surveyed locations of sieve holes to calibrate magnetic optics.

Use reconstructed hole sizes to measure resolution.

...this method only works for negative polarity, and requires running at different beam energy.

Mass resolution≈1 MeV~0.5%

After Calibration

Sieve H. Pos [m]

Siev

e V.

Pos

[m]

HRS optics“Active sieve slit”: tagging by a Sci Fiber detector1 mm fibers with 1/16” pitchProjected rate: 1-3 MHz per fiberOff-line time window < 5 nsNearing completion

Allows optics calibration at production beam energy & for both polarities

14

New HRS Septum Magnet

Designed for parallel field capability(minimize fringe field near beamline)

Optimized for full angular acceptance

High density coils used to enable high-energy use15

New HRS Septum Magnet

16

Design to be completed in next few weeks quotes obtained from vendors partially funded by NSERC DAS

New extension box (as for PREX) and vacuum connection to spectrometers needed

Requires 2kA for high-energy settings (same as SBS magnet)

Summary

17

APEX can explore important range of mass and coupling most efficiently and before other experiments

Opportunity for immediate science impact – even with commissioning-quality beam.

APEX has demonstrated feasibility and power of spectrometer searches for hidden-sector photonsStrong physics impact already from test run (most cited Hall A result in last 5 years)

0.1 0.3 0.510-8

10-7

10-6

10-5

0.1 0.3 0.5

10-8

10-7

10-6

10-5

e+e-HA'LMass HGeVL

a'êaH2s

sensitivityL

Sensitivity of Proposed Run Plan

C AD

B0.1 0.2 0.3 0.4 0.5

-710

-610

-510

-410

BaBar

MAMIKLOE+a

TestAPEX

[MeV]A'm100 200 300 400 500

100 300 500

_'/_

-710

-610

-510

-410400200

400200

18

Thanks!

19

BACKUP SLIDES

Sieve Slit Method

20

After Calibration

Sieve H. Pos [m]

Siev

e V.

Pos

[m]

Before Calibration

Siev

e V.

Pos

[m]

Left HRS calibration used 35 holes, Right HRS calibration used 38 holes

HRS optics for APEX

21

Angular Resolutions

22

LHRS (mrad)LHRS (mrad) RHRS (mrad)RHRS (mrad)Δφ 0.10 Δφ 0.10

Δθ 0.24 Δθ 0.20

σφ_width 0.26 σφ_width 0.43

σθ_width 1.81   σθ_width 1.75

σφ 0.29 σφ 0.44

σθ 1.86 σθ 1.77

Optics calibration precisionTracking precision

φ/θ – hor / vert angles

Averages weighted according to statistics

Final resolutions

Mass Resolution

23

Mass (MeV) 180 195 210 225 240 AverageUsing different angular resolutions for each event

0.833 0.965 1.026 1.061 1.037 1.005Using angular resolutions listed in above table for all events 0.822 0.962 1.023 1.054 1.043

-

Using angular resolutions from "Total" column in above table for all events 0.869 0.965 0.995 0.994 0.966 0.977

Mass (MeV) 180 195 210 225 240 AverageLeft theta (mrad) 1.95 1.87 1.89 1.93 1.88 1.86Left phi (mrad) 0.26 0.3 0.32 0.33 0.33 0.29Right theta (mrad) 1.69 1.74 1.81 1.85 1.85 1.77Right phi (mrad) 0.38 0.43 0.46 0.5 0.53 0.44

Mass resolutions (MeV) determined for different masses using 3 different methods

Angular resolution averages (mrad) determined for different masses

Coincidence trigger and particle ID performance

24)/psh + Eps

(E0 0.5 1 1.5 2 2.5

Even

ts

0

100

200

300

400

500

600Electron detection eff.Pion rejection eff.Pion leakage

0.9670.9690.044

Electron detection effPion rejection effPion leakage

0.9560.9810.027

= 765 kHzscinf

sample from GCπe sample from GC

Calorimeter

Amplitude0 200 400 600 800 1000 1200 1400 1600 1800 2000

Even

ts

1

10

210

310

Positron detection eff.Pion rejection eff.

0.9640.979

sample from LGπ sample from LG+e

Gas Cherenkov

raw TDC channels (0.5 ns)470 480 490 500 510 520 530 540 550 560 570

Eve

nts

0

500

1000

1500

2000

2500

containing coincident events10 ns timing gate

A on Tantalum target+ with 56 - e+Trigger level timing of e

coincidence peak for two-arm X–e+ trigger (requires coincident GC signal in positive-polarity arm)