searching for hidden/dark sectors with the ...10physics department, university of patras, 265 04...

||Paolo Crivelli 26.05.2020

SEARCHING FOR HIDDEN/DARK SECTORS WITH THE NA64 EXPERIMENT AT THE CERN SPSPaolo Crivelli, ETH Zurich, Institute for Particle Physics and Astrophysics

!1


DARK SECTORS - THE VECTOR PORTAL

!2

Standard ModelU(1)’ Kinetic MixingDark

Sector

DARK SECTOR (DS) charged under a new U(1)' gauge symmetry and interacts with SM through kinetic mixing (𝜀) of a MASSIVE VECTOR MEDIATOR (A’) with our photon.  Dark matter with mass (mχ), part of DS. Four parameters: mA’, mχ, 𝛼D ,𝜀

Dark Matter Search in Missing Energy Events with NA64

D. Banerjee,4,5 V. E. Burtsev,2 A. G. Chumakov,13 D. Cooke,6 P. Crivelli,15 E. Depero,15 A. V. Dermenev,7 S. V. Donskov,11

R. R. Dusaev,13 T. Enik,2 N. Charitonidis,4 A. Feshchenko,2 V. N. Frolov,2 A. Gardikiotis,10 S. G. Gerassimov,8,3

S. N. Gninenko ,7,* M. Hösgen,1 M. Jeckel,4 A. E. Karneyeu,7 G. Kekelidze,2 B. Ketzer,1 D. V. Kirpichnikov,7

M.M. Kirsanov,7 I. V. Konorov,8,3 S. G. Kovalenko,14 V. A. Kramarenko,2,9 L. V. Kravchuk,7 N. V. Krasnikov,2,7

S. V. Kuleshov,12 V. E. Lyubovitskij,13,14 V. Lysan,2 V. A. Matveev,2,7 Yu. V. Mikhailov,11 L. Molina Bueno,15

D. V. Peshekhonov,2 V. A. Polyakov,11 B. Radics,15 R. Rojas,14 A. Rubbia,15 V. D. Samoylenko,11 D. Shchukin,8

V. O. Tikhomirov,8 I. Tlisova,7 D. A. Tlisov,7 A. N. Toropin,7 A. Yu. Trifonov,13 B. I. Vasilishin,13 G. Vasquez Arenas,14

P. V. Volkov,2,9 V. Yu. Volkov,9 and P. Ulloa14

(NA64 Collaboration)

1Universität Bonn, Helmholtz-Institut für Strahlen-und Kernphysik, 53115 Bonn, Germany2Joint Institute for Nuclear Research, 141980 Dubna, Russia

3Technische Universität München, Physik Department, 85748 Garching, Germany4CERN, European Organization for Nuclear Research, CH-1211 Geneva, Switzerland

5University of Illinois at Urbana Champaign, Urbana, 61801-3080 Illinois, USA6Department of Physics and Astronomy, University College London, Gower St., London WC1E 6BT, United Kingdom

7Institute for Nuclear Research, 117312 Moscow, Russia8P.N. Lebedev Physical Institute, Moscow, Russia, 119 991 Moscow, Russia

9Skobeltsyn Institute of Nuclear Physics, Lomonosov Moscow State University, 119991 Moscow, Russia10Physics Department, University of Patras, 265 04 Patras, Greece

11State Scientific Center of the Russian Federation Institute for High Energy Physics of National Research Center ’Kurchatov Institute’(IHEP), 142281 Protvino, Russia

12Departamento de Ciencias Fıisicas, Universidad Andres Bello, Sazié 2212, Piso 7, Santiago, Chile13Tomsk State Pedagogical University, 634061 Tomsk, Russia

14Universidad Técnica Federico Santa María, 2390123 Valparaíso, Chile15ETH Zürich, Institute for Particle Physics and Astrophysics, CH-8093 Zürich, Switzerland

(Received 5 June 2019; published 18 September 2019)

A search for sub-GeV dark matter production mediated by a new vector boson A0, called a dark photon,is performed by the NA64 experiment in missing energy events from 100 GeV electron interactions in anactive beam dump at the CERN SPS. From the analysis of the data collected in the years 2016, 2017, and2018 with 2.84 × 1011 electrons on target no evidence of such a process has been found. The most stringentconstraints on the A0 mixing strength with photons and the parameter space for the scalar and fermionicdark matter in the mass range ≲0.2 GeV are derived, thus demonstrating the power of the active beamdump approach for the dark matter search.

DOI: 10.1103/PhysRevLett.123.121801

The idea that in addition to gravity a new force betweenthe dark and visible matter transmitted by a vector boson,A0, called dark photon, might exist is quite exciting [1–4].The A0 can have a mass in the sub-GeV mass range, andcouple to the standard model (SM) via kinetic mixing with

the ordinary photon, described by the term ðϵ=2ÞF0μνFμνand parametrized by the mixing strength ϵ. An example ofthe Lagrangian of the SM extended by the dark sector (DS)is given by

L ¼ LSM −1

4F0μνF0μν þ

ϵ2F0μνFμν þ

m2A02

A0μA0μ

þ iχ̄γμ∂μχ −mχ χ̄χ −eDχ̄γμA0μχ; ð1Þwhere the massive A0μ field is associated with the sponta-neously broken UDð1Þ gauge group, F0μν ¼ ∂μA0ν − ∂νA0μ,and mA0 , mχ are, respectively, the masses of the A0 and dark

Published by the American Physical Society under the terms ofthe Creative Commons Attribution 4.0 International license.Further distribution of this work must maintain attribution tothe author(s) and the published article’s title, journal citation,and DOI. Funded by SCOAP3.

PHYSICAL REVIEW LETTERS 123, 121801 (2019)Editors' Suggestion

0031-9007=19=123(12)=121801(7) 121801-1 Published by the American Physical Society

𝛼


DARK SECTORS - THE VECTOR PORTAL

!3

Standard ModelU(1)’ Kinetic MixingDark

Sector

I n th i s f ramework DM can be p roduced the r ma l l y i n the ea r l y Un ive rse

OBSERVED AMOUNT OF DARK MATTER TODAY

J. Feng and J. Kumar Phys.Rev.Lett.101:231301,2008

Large range fo r g X and m X


SEARCHES FOR DARK SECTORS AT ACCELERATORS

!4

1) BEAM DUMP APPROACH (MiniBooNE, LSND, NA62…)

𝜒

𝜒e

e

�

e

e

ee

�

ee

e

e

e

e

Wiggly lines don’t always close well. Sometimes you can adjust them by hand.

I don’t have a good solution for this. One option specifically for semi-circles is here: http://bit.ly/1vFCNzi. I think it can be adapted for arbitrary angles. For further discussion, see:http://bit.ly/12wA4kQ.

3 W Diagrams

⌫e

µ

W

⌫e

e

W�

W+

�

3

A0

𝛾

N

e

N

e 𝜀  Flux of X generated by decays of A's 

produced in the dump.  Signal: X scattering in far detector

INVISIBLE DECAY MODE


SEARCHES FOR DARK SECTORS AT ACCELERATORS

!5

2) NA64/LDMX APPROACH

𝜒

𝜒e

e

�

e

e

ee

�

ee

e

e

e

e



3 W Diagrams

⌫e

µ

W

⌫e

e

W�

W+

�

3

A0

𝛾

N

e

N

e 𝜀NA64 missing energy: produced A’s carry away energy form the active dump used to measure

recoil e- energy

INVISIBLE DECAY MODE

||Paolo Crivelli 26.05.2020 !6

EXPLICIT TARGET FOR NA64 (y,mX) DM PARAMETER SPACEAsymmetric Fermion

Elastic Scalar

Inelastic Scalar Hsmall splittingL

Majorana Fermion

Pseudo-Dirac FermionHsmall splittingL

1 10 102 10310-59.

10-57.

10-55.

10-53.

10-51.

10-49.

10-47.

10-45.

10-43.

10-41.

10-39.

10-37.

10-35.

mDM

se

Thermal and Asymmetric Targets for DM-e Scattering

FIG. 17: Direct annihilation thermal freeze-out targets and asymmetric DM target for (left)non-relativistic e-DM scattering probed by direct-detection experiments and (right) relativisticaccelerator-based probes. The thermal targets include scalar, Majorana, inelastic, and pseudo-dirac DM annihilating through the vector portal. Current constraints are displayed as shaded ar-eas. Both panels assume mMED = 3mDM and the dark fine structure constant ↵D ⌘ g2D/4⇡ = 0.5.These choices correspond to a conservative presentation of the parameter space for accelerator-based experiments (see section VIG).

dump experiments, the mediator can be emitted by the incoming proton, or if kine-matically allowed, from rare SM meson decays, while detection could proceed throughDM-nucleon scattering. Thus, proton beam-dump experiments are uniquely sensitiveto the coupling to quarks. On the other hand, leptonic couplings can be studied inelectron beam-dump and fixed target experiments, where the mediator is radiated o↵the incoming electron beam. The DM is identified through its scattering o↵ electronsat a downstream detector, or its presence is inferred as missing energy/momentum.

C. Experimental approaches and future opportunities

The light DM paradigm has motivated extensive developments during the last few years,based on a combination of theoretical and proposed experimental work. As a broad orga-nizing principle, these approaches can be grouped into the following generic categories:

• Missing mass: The DM is produced in exclusive reactions, such as e+e� ! �(A0 !��̄) or e�p ! e�p(A0 ! ��̄), and identified as a narrow resonance over a smoothbackground in the recoil mass distribution. This approach requires a well-known initialstate and the reconstruction of all particles besides the DM. A large background usuallyarises from reactions in which particle(s) escape undetected, and detectors with goodhermeticity are needed to limit their impact.

70

Probed

recent review https://arxiv.org/pdf/1707.04591.pdf

Solid lines  predictions from DM 

relic abundance

Cross sections DM -> SM annihilation is ~ Y,  useful variable to compare exp. sentivities

||Paolo Crivelli 26.05.2020 !7

EXPLICIT TARGET FOR NA64 (y,mX) DM PARAMETER SPACEAsymmetric Fermion

Elastic Scalar

Inelastic Scalar Hsmall splittingL

Majorana Fermion

Pseudo-Dirac FermionHsmall splittingL

1 10 102 10310-59.

10-57.

10-55.

10-53.

10-51.

10-49.

10-47.

10-45.

10-43.

10-41.

10-39.

10-37.

10-35.

mDM

se

Thermal and Asymmetric Targets for DM-e Scattering

FIG. 17: Direct annihilation thermal freeze-out targets and asymmetric DM target for (left)non-relativistic e-DM scattering probed by direct-detection experiments and (right) relativisticaccelerator-based probes. The thermal targets include scalar, Majorana, inelastic, and pseudo-dirac DM annihilating through the vector portal. Current constraints are displayed as shaded ar-eas. Both panels assume mMED = 3mDM and the dark fine structure constant ↵D ⌘ g2D/4⇡ = 0.5.These choices correspond to a conservative presentation of the parameter space for accelerator-based experiments (see section VIG).

dump experiments, the mediator can be emitted by the incoming proton, or if kine-matically allowed, from rare SM meson decays, while detection could proceed throughDM-nucleon scattering. Thus, proton beam-dump experiments are uniquely sensitiveto the coupling to quarks. On the other hand, leptonic couplings can be studied inelectron beam-dump and fixed target experiments, where the mediator is radiated o↵the incoming electron beam. The DM is identified through its scattering o↵ electronsat a downstream detector, or its presence is inferred as missing energy/momentum.

C. Experimental approaches and future opportunities

The light DM paradigm has motivated extensive developments during the last few years,based on a combination of theoretical and proposed experimental work. As a broad orga-nizing principle, these approaches can be grouped into the following generic categories:

• Missing mass: The DM is produced in exclusive reactions, such as e+e� ! �(A0 !��̄) or e�p ! e�p(A0 ! ��̄), and identified as a narrow resonance over a smoothbackground in the recoil mass distribution. This approach requires a well-known initialstate and the reconstruction of all particles besides the DM. A large background usuallyarises from reactions in which particle(s) escape undetected, and detectors with goodhermeticity are needed to limit their impact.

70

Probed

recent review https://arxiv.org/pdf/1707.04591.pdf

Solid lines  predictions from DM 

relic abundance

NA64e TARGET  

h igher mass reg ion cou ld  be covered by NA64mu (p i l o t

run i n 2021) PLB796 , 117 (2019)

matter (DM) particles, χ, which are treated as Diracfermions coupled to A0μ with the dark coupling strengtheD of the Uð1ÞD gauge interactions. The mixing term of (1)results in the interaction Lint ¼ ϵeA0μJμem of dark photonswith the electromagnetic current Jμem with a strength ϵe,where e is the electromagnetic coupling and ϵ ≪ 1 [5–7].Such small values of ϵ can be obtained in grand unifiedtheories from loop effects of particles charged under boththe dark UDð1Þ and SM Uð1Þ interactions with a typicalone-loop value ϵ ¼ eeD=16π2≃10−2–10−4 [7], or fromtwo-loop contributions resulting in ϵ≃10−3–10−5. Theaccessibility of these values at accelerator experimentshas motivated a worldwide effort towards dark forcesand other portals between the visible and dark sectors;see Refs. [4,8–17] for a review.If the A0 is the lightest state in the dark sector, then it

would decay mainly visibly to SM leptons l (or hadrons);see, e.g., [18–23], and also [17]. In the presence of lightDM states χ with the masses mχ < mA0=2, the A0 wouldpredominantly decay invisibly into those particles providedthat eD > ϵe. Various dark sector models motivate theexistence of sub-GeV scalar and Majorana or pseudo-DiracDM coupled to the A0 [13,14,17,24–28]. To interpret theobserved abundance of DM relic density, the requirementof the thermal freeze-out of DM annihilation into visiblematter through γ −A0 mixing allows one to derive a relation

αD ≃0.02f!10−3

ϵ

"2!

mA0100 MeV

"4!10 MeV

mχ

"2

; ð2Þ

where αD ¼ e2D=4π and the parameter f depends on mA0and mχ [29]. For mA0=mχ ¼ 3, f ≲ 10 for a scalar [24], andf ≲ 1 for a fermion [25]. This prediction provides animportant target for the (ϵ, mA0 ) parameter space, whichcan be probed at the CERN SPS energies. Models intro-ducing the invisible A0 also may explain various astro-physical anomalies [30] and are subject to variousexperimental constraints leaving, however, a large areathat is still unexplored [24,31–40].In this work we report new results on the search for the A0

mediator and light dark matter (LDM) in the fixed-targetexperiment NA64 at the CERN SPS. In the following weassume that the A0 invisible decay mode is predominant,i.e., ΓðA0 → χ̄χÞ=Γtot ≃1. If such invisible A0 exists, manycrucial questions about its coupling constants, mass scale,decay modes, etc. arise. One possible way to answer thesequestions is to search for the A0 in fixed-target experiments.The A0s could be produced by a high-intensity beam ina dump and generate a flux of DM particles through theA0 → χ̄χ decay, which can be detected through the scatter-ing off electrons in the far target [24,25,31,34,41,42]. Thesignal event rate in the detector in this case scales asϵ2y∝ϵ4αD, with one ϵ2associated with the A0 productionin the dump and ϵ2αD coming from the χ particle scattering

in the detector, and with the parameter y defined asy ¼ ϵ2αDðmχ=mA0Þ4. Another method, discussed in thiswork and proposed in Refs. [43,44], is based on thedetection of the missing energy, carried away by the hardbremsstrahlung A0 produced in the process e−Z → e−ZA0;A0 → invisible of high-energy electrons scattering in theactive beam dump target. The advantage of this type ofexperiment compared to the beam dump ones is that itssensitivity is proportional to ϵ2, associated with the A0

production and its subsequent prompt invisible decay.The NA64 detector is schematically shown in Fig. 1. The

experiment employed the optimized H4 100 GeV electronbeam [45]. The beam has a maximal intensity ≃107electrons per SPS spill of 4.8 s produced by the primary400 GeV proton beam with an intensity of few 1012protonson target. The detector utilized the beam defining scintil-lator (Sc) counters S1–4 and veto V1;2, a magnetic spec-trometer consisting of two successive dipole magnetsMBPL1;2with the integral magnetic field of ≃7Tm anda low-material-budget tracker. The tracker was a set of twoupstream Micromegas chambers MM1;2, and four MM3–6,downstream stations, as well as two straw-tube ST1;2 andGEM1;2 chambers allowing the measurements of e−

momenta with the precision δp=p≃1% [46]. To enhanceelectron identification, synchrotron radiation (SR) emittedin the MBPL magnetic field was used for their efficienttagging with a SR detector (SRD), which was an array of aPbSc sandwich calorimeter of a very fine segmentation[43,47]. By using the SRD the initial admixture of thehadron contamination in the beam π=e−≲ 10−2was furthersuppressed by a factor ≃103. The detector was alsoequipped with an active dump target, which is an electro-magnetic calorimeter (ECAL), a matrix of 6×6Shashlik-type modules assembled from Pb and Sc plates formeasurement of the electron energy EECAL. Each modulehas≃40 radiation lengths (X0) with the first 4X0 serving asa preshower detector. Downstream of the ECAL, thedetector was equipped with a large high-efficiency vetocounter VETO, and a massive, hermetic hadronic calorim-eter (HCAL) of ≃30 nuclear interaction lengths intotal. The modules HCAL1−3 provided an efficient vetoto detect muons or hadronic secondaries produced in thee−A interactions in the target. The events were collectedwith the hardware trigger requiring an in-time cluster inthe ECAL with the energy EECAL ≲ 80 GeV. The searchdescribed in this paper uses the data samples ofnEOT ¼ 0.43 × 1011, 0.56× 1011 and 1.85 × 1011 electronson target (EOT), collected in the years 2016, 2017, and2018 with the beam intensities in the range≃ð1.4–6Þ × 106,≃ð5–6Þ × 106, and≃ð5–9Þ × 106 e− per spill, respectively.Data corresponding it total to 2.84 × 1011 EOT from thesethree runs (hereafter called respectively runs I, II, and III)were processed with selection criteria similar to the oneused in Ref. [38] and finally combined as described below.

PHYSICAL REVIEW LETTERS 123, 121801 (2019)

121801-2


The NA64 method to search for A’ → 𝛘�̅�

!8

𝜒

𝜒e

e

�

e

e

ee

�

ee

e

e

e

e



3 W Diagrams

⌫e

µ

W

⌫e

e

W�

W+

�

3

A0

𝛾

N

e

N

e 𝜀

ELECTROMAGNETIC  CALORIMETER (ECAL)

TAGGED 100 GeV Requested ECAL ENERGY < 50 GeV

“BREMSSTRAHLUNG” OF A’

Active Dump


Signature for the invisible decay A’ → 𝛘�̅� - large missing energy

!9

𝜒

𝜒e

e

�

e

e

ee

�

ee

e

e

e

e



3 W Diagrams

⌫e

µ

W

⌫e

e

W�

W+

�

3

A0

𝛾

N

e

N

e 𝜀

ELECTROMAGNETIC  CALORIMETER (ECAL)

HADRONIC CALORIMETER (HCAL)

STANDARD MODEL: EECAL+EHCAL = 100 GeV

A’→ MISSING ENERGY:  ECAL < 50 GeV HCAL < 2 GeV


The Electromagnetic Calorimeter (ECAL)

!10

Active target

✦ High hermeticity (∼40 X0) ✦ PbSc sandwich, 6x6 matrix, cells 38x38x490 mm3 ✦ WLS fibers in spiral→ suppress energy leaks ✦ Energy resolution ~ 9%/√(E[GeV]) ✦ Longitudinal (Pre-shower) and lateral segmentation → shower profiles (hadron rejection)

100 GeV electrons (tagged with S1,2,3)


The Hadronic Calorimeter (HCAL)

!11

High hermeticity  due to Lorentz boost

✦ High hermeticity : 4 HCAL (∼7 λ/module)✦ FeSc sandwich 3x3 matrix, cells 19.4x19.2x150 cm3 ✦ WLS fibers in spiral→ suppress energy leaks ✦ Energy resolution ~ 60%/√(E[GeV])


The magnetic spectrometer

!12

Tracking system:  XY multiplexed resistive micromegas, GEM and straw.

Two bending magnets in series → 7 T.m field

Reconstruction of e-  incoming momentum

D. Banerjee et al., Advances in HEP, 105730 (2015)

D. Banerjee et al., NIMA881 (2018) 72-81


The Synchrotron radiation detector (e- tagging)

!13

Particle identification SR emission ~ 1/m4

Granularity

e- 𝜋-

e-efficiency > 95%  Suppression π, K>10-5

E. Depero et al., NIMA 866 (2017) 196-201.

BGO → PbSc


The NA64 search for A’ → 𝛘�̅� - results (July + October 2016, 5 weeks)

!14

→ exclusion of most of g-2 muon favored region

NA64 collaboration, Phys. Rev. Lett. 118, 011802 (2017) and Phys. Rev. D 97, 072002 (2018)g-2 closed completely by BABAR results corresponding mixing strength ϵ were determined from the

90% C.L. upper limit for the expected number of signalevents,N90%A0 by using the modified frequentist approach forconfidence levels (C.L.), taking the profile likelihood as atest statistic in the asymptotic approximation [70–72]. Thetotal number of expected signal events in the signal box wasthe sum of expected events from the three runs:

NA0 ¼X3

i¼1NiA0 ¼

X3

i¼1niEOTϵ

itotniA0ðϵ; mA0 ;ΔEeÞ ð26Þ

where ϵitot is the signal efficiency in the run i given byEq. (23), and the niA0ðϵ; mA0 ;ΔEA0Þ value is the signal yieldper EOT generated by a single 100 GeV electron in theECAL target in the energy range ΔEe. Each ith entry in thissum was calculated by simulating the signal events forcorresponding beam running conditions and processingthem through the reconstruction program with the sameselection criteria and efficiency corrections as for the datasample from the run-i. The expected backgrounds andestimated systematic errors were also taking into account inthe limits calculation. The combined 90% C.L. exclusionlimits on the mixing strength as a function of the A0 masscan be seen in Fig. 15. In Table V the limits obtained withthe ETL and WW calculations for different mA0 valuesare also shown for comparison. One can see that thecorrections are mostly relevant in the higher mass regionmA0 ≳ 100 MeV. The derived bounds are the best for themass range 0.001≲mA0 ≲ 0.1 GeV obtained from directsearches of A0 → invisible decays [15].The limits were also calculated with a simplified method

by merging all three runs into a single run as describedpreviously by Eq. (26). The total error for the each NiA0

value includes the corresponding systematic uncertaintiescalculated by adding contributions from all sources inquadrature, see Sec. VII. In accordance with the CLsmethod [72], for zero number of observed events the90% C.L. upper limit for the number of signal events isN90%A0 ðmA0Þ ¼ 2.3. Taking this and Eq. (26) into accountand using the relation NA0ðmA0Þ < N90%A0 ðmA0Þ resulted inthe 90% C.L. limits in the (mA0 ; ϵ) plane which agreed withthe one shown in Fig. 15 within a few %.

X. CONSTRAINTS ON LIGHT THERMALDARK MATTER

As discussed previously, the possibility of the existenceof light thermal dark matter (LTDM) has been the subject of

FIG. 14. The sensitivity, defined as an average expected limit,as a function of the ECAL energy cut for the case of the A0

detection with the mass mA0 ≃ 20 (blue) and 2 (green) MeV.FIG. 15. The NA64 90% C.L. exclusion region in the (mA0 ; ϵ)plane. Constraints from the BABAR [39], E787 and E949 experi-ments [34,35], as well as the muon αμ favored area are also shown.

Here, αμ ¼gμ−22 . For more limits obtained from indirect searches

and planned measurements see e.g., Ref. [13,14].

TABLE V. Comparison of upper bounds on mixing ϵ at90% CL obtained with WW and ETL calculations for the Pb-Sc ECAL target for Emiss > 0.5E0 at E0 ¼ 100 GeV.

mA0 ,MeV

90% C.L. upper limiton ϵ; 10−4, no k -factors

90% C.L. upper limiton ϵ; 10−4with k -factors

1.1 0.22 0.192 0.23 0.245 0.43 0.4916.7 1.25 1.3320 1.29 1.6100 5.5 8.2200 13.0 22.6500 38.7 97.8950 94.20 362.0

SEARCH FOR VECTOR MEDIATOR OF DARK MATTER … PHYS. REV. D 97, 072002 (2018)

072002-17

CO

UPL

ING

𝜀

MASS OF THE DARK PHOTON

4 x 1010 electrons on target

8

0 20 40 60 80 100, GeVECALE

0

20

40

60

80

100

120

, GeV

HC

ALE

1

10

210

310

410

510

III

I

II

0 20 40 60 80 100, GeVECALE

0

20

40

60

80

100

120

, GeV

HC

ALE

1

10

210

310

0 20 40 60 80 100, GeVECALE

0

20

40

60

80

100

120

, GeV

HC

ALE

1

10

B

C

A

0 20 40 60 80 100, GeVECALE

0

20

40

60

80

100

120

, GeV

HC

ALE

1

10

210

310

410

510

III

I

II

0 20 40 60 80 100, GeVECALE

0

20

40

60

80

100

120

, GeV

HC

ALE

1

10

210

310

0 20 40 60 80 100, GeVECALE

0

20

40

60

80

100

120

, GeV

HC

ALE

1

10

B

C

A

FIG. 7: Event distribution in the (EECAL;EHCAL) plane from the runs II(top row) and III (bottom row) data. The left panelsshow the measured distribution of events at the earlier phase of the analysis. Plots in the middle show the same distribution afterapplying all selection criteria, but the cut against upstream interactions. The right plots present the final event distributionsafter all cuts applied. The dashed area is the signal box region which is open. The side bands A and C are the one used for thebackground estimate inside the signal box. For illustration purposes the size of the signal box along EHCAL-axis is increasedby a factor five.

hadron electroproduction in the ECAL which satisfy theenergy conservation EECAL+EHCAL ' 100 GeV withinthe detector energy resolution. The leak of these eventsto the signal region mainly due to the HCAL energy reso-lution was found to be negligible. The fraction of eventsfrom the region III was due to pileup of e� and beamhadrons. It was beam rate dependent with a typical valuefrom about a few % up to ' 20%.

V. DIMUON EVENTS FROM THE REACTIONe�Z ! e�Zµ+µ�

To evaluate the performance of the setup, a cross-checkbetween a clean sample of & 104 observed and MC simu-lated µ+µ� events was made. The process (16) was usedas a benchmark allowing to verify the reliability of theMC simulation and to estimate the corrections to thesignal reconstruction e�ciency and possible additionaluncertainties in the A0 yield calculations. Let us firstbriefly review the description of the gamma conversioninto a muon-antimuon pair implemented in Geant4. Thedimuon production was also used as a reference for theprediction of background, see Sec. VIII.

A. Simulation of dimuon events

The dimuon production has been simulated withGeant4 [50] and a code developed by NA64 used alsofor simulation of dark photon production [46]. Here, wereport our comparison with data based mostly on Geant4simulation for decays and propagation of muons throughthe detectors. However, we anticipate that comparisonof dimuon results with the NA64 code will follow in thefuture, as it will be an important additional cross-checkof the A0 yield calculations reported in this work and inRef.[54].The gamma conversion into a muon-antimuon pair

�Z ! µ+µ�Z (17)

on nuclei is a well known reaction in particle physics(Bethe-Heitler process). The simulation of this reactionin Geant4 is based on the di↵erential cross section forelectromagnetic creation of muon pairs on nuclei (A,Z)in terms of the energy fraction of muons [64, 65]:

d�

dx+= 4↵Z2r2µ

⇣1� 4

3x+x�

⌘log(W ) , (18)

No event in signal box


Improvement of setup for 2018 run

!15

and a massive, hermetic hadronic calorimeter (HCAL) of ' 30 nuclear interactionlengths in total. The modules HCAL1�3 served as an e�cient veto to detect muons or

hadronic secondaries produced in the e�A interactions in the target, while the zero-

degree calorimeter HCAL0 was used to reject events accompanied by a hard neutrals

from the upstream e� interactions. The new straw-tube tracker is manufactured with

Figure 1. Schematic illustration of the setup to search for A0 ! invisible decays of thebremsstrahlung A0s produced in the reaction eZ ! eZA0 of 100 GeV e� incident on theactive ECAL target.

6 mm in diameter straw tubes covered by Cu layer. Each coordinate plane consist

of two layers of straw shifted each other to the size of the radius. Two mother board

for di↵erent layers are currently used. New 6 Amplifier AST-1-1 (32channels) for

each coordinate which has to operate with 3 TDC units, are located directly on the

each chamber stand. Cables are 17 twisted pairs, each of 60 cm in length. Since the

2017 run, the reconstruction of events in the straw tube tracker has been significantly

improved.

2.1 Data sample from 2016-2018 runs

The events were collected with the hardware trigger requiring an in-time cluster

in the ECAL with the energy EECAL . 80 � 85 GeV. The results reported herecame mostly from a set of data in which nEOT = 4.3 ⇥ 1010 of electrons on target(EOT) were collected with the beam intensity ranging from ' 1.5 ⇥ 106/spill to' 5 ⇥ 106/spill e� per spill in the year 2016, about nEOT = 5.4 ⇥ 1010 collectedwith intensity ' (5 � 6) ⇥ 106 e�/spill in the year 2017, and nEOT = 1.87 ⇥ 1011collected with intensity up to ' (8 � 9) ⇥ 106 e�/spill in the year 2018. The totalstatistics thus corresponds to nEOT = 2.84⇥1011 EOT. Data of these three runs wereanalyzed with similar selection criteria and finally summed up, taking into account

the corresponding normalization factors. The combined analysis is described below.

– 3 –

HCAL0: Rejection of events with hard neutral from upstream e- interactions

ST1,2: New straw-tube trackers: VETO against hadron electro-production in the beam material upstream the ECAL.


hadronic interactions in the downstream part of the beam line accompanied by the

large transverse fluctuations of hadronic secondaries, which is di�cult to simulate

reliably. Table I summarizes the conservatively estimated background inside the

signal region, which is expected to be 0.53 ± 0.17 events. After determining all theselection criteria and estimating background levels, we examined the events in the

signal box and found no candidates, as shown in Fig. 2.

Figure 3. The NA64 90% C.L. exclusion region in the (mA0 , ✏) plane. Constraints from

the E787, E949 [1], BaBar [13] experiments, recent NA62 [14] results, as well as the muon

↵µ favored area are also shown. Here, ↵µ =gµ�22 . For more limits obtained from indirect

searches and planned measurements see e.g. Ref. [1, 2].

3.3 Results and calculation of limits

This allows us to obtain the mA0-dependent upper limits on the mixing ✏ which were

calculated as follows. In the final combined statistical analysis the three runs I-III

were analysed simultaneously using the multi-bin limit setting technique [11] based

on the RooStats package [16]. First, the above obtained estimates of background,

e�ciencies, and their corrections and uncertainties were used to optimize better the

main cut defining the signal box by comparing sensitivities, defined as an average

expected limit calculated using the profile likelihood method. The calculations were

done with uncertainties used as nuisance parameters, assuming their log-normal dis-

tributions [17]. For this optimization, the most important inputs were the expected

values from the background extrapolation into the signal region from the data sam-

ples of the runs I,II,III. The errors for background prediction were estimated from

the variation of the extrapolation functions. It was found that the optimal cut value

depends very weakly on the A0 mass choice and can be safely set to EECAL < 50

GeV for the whole mass range.

– 8 –

Combined results (2016-2018)

!16

Signal efficiency: ∼50% slightly mass dependent.

TOT: 2.84 x 1011 electrons on target

CO

UPL

ING

𝜀


NA64 collaboration, Phys. Rev. Lett. 123, 121801 (2019)


NEW constraints on sub-GeV DM parameter space (2016-2018)

!17

First time NA64 constraints on light

thermal DM exceeding sensitivity of beam

dump exp. (suppressed by 𝝐2𝛼D)

NA64 collaboration, Phys. Rev. Lett. 123, 121801 (2019) Figure 4. The top raw shows the NA64 limits in the (y;m�) plane obtained for ↵D = 0.5

(left panel) and ↵D = 0.1 (right panel) from the full 2016-18 data set. The bottom

raw shows the NA64 constraints in the (↵D;mA0) plane on the pseudo-Dirac (left panel)

and Majorana (right panel) DM. The limits are shown in comparison with bounds from

the results of the LSND and E137 [1], MiniBooNE [15], BaBar [13], and direct detection

experiments . The favoured parameters to account for the observed relic DM density for

the scalar, pseudo-Dirac and Majorana type of light DM are shown as the lowest solid line

in top plots.

3.4 Constraints on sub-GeV dark matter

Using constraints on the cross section of the DM annihilation freeze out (resulting in

Eq.(3.2)), and obtained limits on mixing strength of Fig. 3, one can derive constraints

on the LDM models, which are shown in the (y;m�) and (↵D;m�) planes in Fig. 4

for the masses m� . 1 GeV. The y limits are shown together with the favouredparameters for scalar, pseudo-Dirac (with a small splitting) and Majorana scenario

of LDM taking into account the observed relic DM density [1]. The limits on the

variable y are calculated by using Eq.(3.3) under the convention ↵D = 0.1 and =0.5,

and mA0 = 3m� [1, 2]. This choice of the ↵D region is compatible with the bounds

derived based on the running of the dark gauge coupling arguments. The plot shows

also the comparison of our results with bounds from other experiments. It should

be noted that for smaller values of ↵D NA64 limits will be stronger, due to the fact

– 10 –

mA’= 3mX 𝛼D = 0.5

mA’= 3mX 𝛼D = 0.1

𝜒

𝜒e

e

�

e

e

ee

�

ee

e

e

e

e



3 W Diagrams

⌫e

µ

W

⌫e

e

W�

W+

�

3

A0

𝛾

NN

e 𝜀


Current bounds on thermal relic DM & projected NA64 sensitivity

!18

CO

UPL

ING

S.N. Gninenko – NA64++ – PBC Workshop, CERN, January 16–17, 2019

5.

New results and projection for sub-GeV thermal DM (I) αD = 0.1, mA = 3mχ

•  Sensitivity of a beam-dump ~ε4αD, NA64~ε2 •  Bounds from LSND, SLAC, MiniBooNE for ~1022, 1019, 1020 POT•  NA64 can cover significant area with ~ a few 1012 EOT

Favored (y, mχ) for observed

relic DM density for the scalar,

pseudo-Dirac and Majorana type.

NA64 2018NA64 2016preliminary


𝛼D = 0.1 and mA’=3m𝜒

Setup upgrade required

New VHCAL: to improve detector hermiticity and reject high-pt hadronic secondaries from beam interactions upstream the ECAL dump. Search expected to be BKG free up to ~ 1013 EOT


2) The NA64 search for X/A’ → e+e-

!19

e

e

�

e

e

ee

�

ee

e

e

e

e



3 W Diagrams

⌫e

µ

W

⌫e

e

W�

W+

�

3

VISIBLE DECAY MODE

A0

SM

SM

Pair production of SM particles

N

𝛾

N

e e𝜀

𝜀


8Be anomaly and X boson

!20

Could be explained by new  ‘protophobic’ gauge boson X with mass around 17 MeV

,

− → → ′ → − +

− +

ECAL1 MicroMegas ECAL2

e-

e+

X-bosone-

30X0

∼

A. J. Krasznahorkay et al. Phys. Rev. Lett.116, 042501 (2015)  and new results for 4He arXiv:1910.10459

J. L. Feng et al. Phys. Rev. D95, 035017 (2017)


The NA64 search for X→ e+e- - experimental setup

!21

2

e−, 100 GeV

S1V1 T1

T2

Vacuum vesselSRD

γ

e−Magnet2

Magnet1

e−

e+WCAL X

V3

MU4

MU1

MU2

MU3

HCAL1

HCAL2

HCAL3

HCAL4

T3T4

S4

ECAL

S3

S2V2

FIG. 1: Schematic illustration of the setup to search for A0, X ! e+e� decays of the bremsstrahlung A0, X produced in thereaction eZ ! eZA0(X) of 100 GeV e� incident on the active WCAL target.

various phenomenological aspects of light vector bosonsweakly coupled to quarks and lepton, see e.g. [6–11]

Another strong motivation to the search for a new lightboson decaying into e+e� pair is provided by the DarkMatter puzzle. An intriguing possibility is that in ad-dition to gravity a new force between the dark sectorand visible matter, transmitted by a new vector boson,A0 (dark photon), might exist [12, 13]. Such A0 couldhave a mass mA0 . 1 GeV, associated with a sponta-neously broken gauged U(1)D symmetry, and would cou-ple to the Standard Model (SM) through kinetic mixingwith the ordinary photon, � 12✏Fµ⌫A

0µ⌫ , parametrized bythe mixing strength ✏ ⌧ 1 [14–16], for a review see, e.g.[4, 17, 18]. A number of previous experiments, such asbeam dump [19–33], fixed target [34–36], collider [37–39] and rare particle decay [40–51] experiments have al-ready put stringent constraints on the mass mA0 and ✏ ofsuch dark photons excluding, in particular, the parame-ter space region favored by the gµ�2 anomaly. However,a large range of mixing strengths 10�4 . ✏ . 10�3 cor-responding to a short-lived A0 still remains unexplored.These values of ✏ could naturally be obtained from theloop e↵ects of particles charged under both the darkand SM U(1) interactions with a typical 1-loop value✏ = egD/16⇡2 [16], where gD is the coupling constantof the U(1)D gauge interactions. In this Letter we reportthe first results from the NA64 experiment specificallydesigned for a direct search of the e+e� decays of newshort-lived particles at the CERN SPS in the sub-GeVmass range [52–55].

The experiment employs the optimized 100 GeV elec-tron beam from the H4 beam line in the North Area(NA) of the CERN SPS. The beam delivers ' 5⇥106 e�per SPS spill of 4.8 s produced by the primary 400 GeVproton beam with an intensity of a few 1012 protons ontarget. The NA64 setup designed for the searches of Xbosons and A0 is schematically shown in Fig. 1. Twoscintillation counters, S1 and S2 were used for the beamdefinition, while the other two, S3 and S4, were used todetect the e+e� pairs. The detector is equipped with amagnetic spectrometer consisting of two MPBL magnets

and a low material budget tracker. The tracker was aset of four upstream Micromegas (MM) chambers (T1,T2) for the incoming e� angle selection and two setsof downstream MM, GEM stations and scintillator ho-doscopes (T3, T4) allowing the measurement of the out-going tracks [56, 57]. To enhance the electron identifica-tion the synchrotron radiation (SR) emitted by electronswas used for their e�cient tagging and for additional sup-pression of the initial hadron contamination in the beam⇡/e� ' 10�2 down to the level ' 10�6 [55, 58]. The useof SR detectors (SRD) is a key point for the hadron back-ground suppression and improvement of the sensitivitycompared to the previous electron beam dump searches[23, 24]. The dump is a compact electromagnetic (e-m)calorimeter WCALmade as short as possible to maximizethe sensitivity to short lifetimes while keeping the leakageof particles at a small level. It is followed by another e-mcalorimeter (ECAL), which is a matrix of 6⇥ 6 shashlik-type modules [55]. The WCAL(ECAL) was assembledfrom the tungsten(lead) and plastic scintillator plateswith wave lengths shifting fiber read-out. The ECAL has' 40 radiation lengths (X0) and is located at a distance' 3.5 m from the WCAL. Downstream the ECAL the de-tector was equipped with a high-e�ciency veto counter,V3, and a massive, hermetic hadron calorimeter (HCAL)[55] used as a hadron veto and muon identificator.The method of the search for A0 ! e+e� decays is

described in [52, 53]. The application of all further con-siderations to the case of theX ! e+e� decay is straight-forward. If the A0 exists, it could be produced via thecoupling to electrons wherein high-energy electrons scat-ter o↵ a nuclei of the active WCAL dump target, followedby the decay into e+e� pairs:

e� + Z ! e� + Z +A0(X); A0(X) ! e+e� (1)

The reaction (1) typically occurs within the first fewradiation lengths (X0) of the WCAL. The downstreampart of the WCAL served as a dump to absorb completelythe e-m shower tail. The bremsstrahlung A0 would pene-trate the rest of the dump and the veto counter V2 with-

Addition of W calorimeter Short in length to allow X to escape

Zooming in (next slide)


The NA64 search for X→ e+e- - experimental signature

!22

WCAL: 30-40 X0 Sandwich W-Sc

2

e−, 100 GeV

S1V1 T1

T2

Vacuum vesselSRD

γ

e−Magnet2

Magnet1

e−

e+WCAL X

V3

MU4

MU1

MU2

MU3

HCAL1

HCAL2

HCAL3

HCAL4

T3T4

S4

ECAL

S3

S2V2

FIG. 1: Schematic illustration of the setup to search for A0, X ! e+e� decays of the bremsstrahlung A0, X produced in thereaction eZ ! eZA0(X) of 100 GeV e� incident on the active WCAL target.

various phenomenological aspects of light vector bosonsweakly coupled to quarks and lepton, see e.g. [6–11]

Another strong motivation to the search for a new lightboson decaying into e+e� pair is provided by the DarkMatter puzzle. An intriguing possibility is that in ad-dition to gravity a new force between the dark sectorand visible matter, transmitted by a new vector boson,A0 (dark photon), might exist [12, 13]. Such A0 couldhave a mass mA0 . 1 GeV, associated with a sponta-neously broken gauged U(1)D symmetry, and would cou-ple to the Standard Model (SM) through kinetic mixingwith the ordinary photon, � 12✏Fµ⌫A

0µ⌫ , parametrized bythe mixing strength ✏ ⌧ 1 [14–16], for a review see, e.g.[4, 17, 18]. A number of previous experiments, such asbeam dump [19–33], fixed target [34–36], collider [37–39] and rare particle decay [40–51] experiments have al-ready put stringent constraints on the mass mA0 and ✏ ofsuch dark photons excluding, in particular, the parame-ter space region favored by the gµ�2 anomaly. However,a large range of mixing strengths 10�4 . ✏ . 10�3 cor-responding to a short-lived A0 still remains unexplored.These values of ✏ could naturally be obtained from theloop e↵ects of particles charged under both the darkand SM U(1) interactions with a typical 1-loop value✏ = egD/16⇡2 [16], where gD is the coupling constantof the U(1)D gauge interactions. In this Letter we reportthe first results from the NA64 experiment specificallydesigned for a direct search of the e+e� decays of newshort-lived particles at the CERN SPS in the sub-GeVmass range [52–55].

The experiment employs the optimized 100 GeV elec-tron beam from the H4 beam line in the North Area(NA) of the CERN SPS. The beam delivers ' 5⇥106 e�per SPS spill of 4.8 s produced by the primary 400 GeVproton beam with an intensity of a few 1012 protons ontarget. The NA64 setup designed for the searches of Xbosons and A0 is schematically shown in Fig. 1. Twoscintillation counters, S1 and S2 were used for the beamdefinition, while the other two, S3 and S4, were used todetect the e+e� pairs. The detector is equipped with amagnetic spectrometer consisting of two MPBL magnets

and a low material budget tracker. The tracker was aset of four upstream Micromegas (MM) chambers (T1,T2) for the incoming e� angle selection and two setsof downstream MM, GEM stations and scintillator ho-doscopes (T3, T4) allowing the measurement of the out-going tracks [56, 57]. To enhance the electron identifica-tion the synchrotron radiation (SR) emitted by electronswas used for their e�cient tagging and for additional sup-pression of the initial hadron contamination in the beam⇡/e� ' 10�2 down to the level ' 10�6 [55, 58]. The useof SR detectors (SRD) is a key point for the hadron back-ground suppression and improvement of the sensitivitycompared to the previous electron beam dump searches[23, 24]. The dump is a compact electromagnetic (e-m)calorimeter WCALmade as short as possible to maximizethe sensitivity to short lifetimes while keeping the leakageof particles at a small level. It is followed by another e-mcalorimeter (ECAL), which is a matrix of 6⇥ 6 shashlik-type modules [55]. The WCAL(ECAL) was assembledfrom the tungsten(lead) and plastic scintillator plateswith wave lengths shifting fiber read-out. The ECAL has' 40 radiation lengths (X0) and is located at a distance' 3.5 m from the WCAL. Downstream the ECAL the de-tector was equipped with a high-e�ciency veto counter,V3, and a massive, hermetic hadron calorimeter (HCAL)[55] used as a hadron veto and muon identificator.The method of the search for A0 ! e+e� decays is

described in [52, 53]. The application of all further con-siderations to the case of theX ! e+e� decay is straight-forward. If the A0 exists, it could be produced via thecoupling to electrons wherein high-energy electrons scat-ter o↵ a nuclei of the active WCAL dump target, followedby the decay into e+e� pairs:

e� + Z ! e� + Z +A0(X); A0(X) ! e+e� (1)

The reaction (1) typically occurs within the first fewradiation lengths (X0) of the WCAL. The downstreampart of the WCAL served as a dump to absorb completelythe e-m shower tail. The bremsstrahlung A0 would pene-trate the rest of the dump and the veto counter V2 with-

Signature:  1) EWCAL+EECAL = 100 GeV 2) No activity in V2,3 and   HCAL  3) Signal in S3, S4 4) e-m shower in ECAL


The NA64 search for X→ e+e- - results (2017-2018)

!23

No signal- like event in  signal box

NA64 collaboration, PRL 120, 231802 (2018), PRD 107, 071101 (R) 2020

CO

UPL

ING

𝜀

~ 8x1010 EOT


The NA64 search for X→ e+e- - prospects (2021)

!24

Feasibility under study

NA64 collaboration, PRL 120, 231802 (2018), PRD 107, 071101 (R) 2020

CO

UPL

ING

𝜀


The NA64 search for ALP

!25

2

FIG. 1: The left panel illustrates schematic view of the setup to search for the a ! �� decays of the a’s produced in the reactionchain e�Z ! e�Z�; �Z ! aZ induced by 100 GeV e�’s in the active ECAL dump. The right panel shows an example of thea ! �� decay in the HCAL2 module.

the energy-momentum tensor[20], and its two-photon in-57teraction is given by Eq.(1) with the replacement F̃µ⌫ !58F

µ⌫ . Usually it is assumed that gs�� = O(M�1Pl ) and59

that the dilaton mass ms = O(MPl), where MPl is the60Planck mass. However, in some models, see, e.g., [21]),61the dilaton could be rather light. Since there are no firm62predictions for the coupling gs�� the searches for such63particles become interesting.64

Experimental bounds on ga�� for light a’s in the eV-65MeV mass range, can be obtained from laser experi-66ments [22, 23], from experiments studying J/ and ⌥67particles [24], from the NOMAD experiment by using68a photon-regeneration method at the CERN SPS neu-69trino beam [25], and from orthopositronium decays [26].70Limits on ALPs in the sub-GeV mass range have been71typically placed by the beam-dump experiments or from72searches at e+e� colliders [6], leaving the large area7310�2 . ga�� . 10�5 GeV�1 of the (ga�� ;ma) -parameter74space still unprobed. Additionally, since the theory pre-75dictions for the coupling, mass scale, and decay modes76of ALPs are still quite uncertain it is crucial to perform77independent laboratory tests on the existence of such par-78ticles in the mass and coupling strength range discussed79above. One possible way to answer these questions is to80search for ALPs in a beam dump experiment. However,81for the coupling lying in the range 10�2 . ga�� . 10�482GeV�1 this method is not very promising, because for83the masses in sub-GeV region the a is expected to be a84relatively short-lived particle.85

In this Letter we propose and describe a direct86search for ALPs with the coupling to two photons from87the (ga�� ;ma) -parameter space uncovered by previous88searches, which might be produced in the NA64 experi-89ment at the CERN SPS . The application of the obtained90results to the s ! �� decay case is straightforward.91

The NA64 detector located at the CERN SPS H4 elec-92tron beam [27] is schematically shown in Fig. 1. It con-93sists of a set of beam defining scintillator counters S1�494and veto V1,2, a magnetic spectrometer consisting of two95dipole magnets (MBPL1,2) and a low-material-budget96tracker, composed of two upstream Micromegas cham-97bers MM1,2, and four downstream MM3�6 stations, two98

straw-tube ST1,2 and GEM1,2 chambers [28]. A syn-99chrotron radiation detector (SRD) is used for the iden-100tification of incoming e�’s [29, 30] and suppression of101the hadron contamination in the beam down to the level102⇡/e

� . 10�5. An active dump, consisting of a preshower103detector (PRS) and an electromagnetic (e-m) calorimeter104(ECAL), made of a matrix of 6 ⇥ 6 Shashlik-type mod-105ules, is assembled from Pb and Sc plates of ' 40 radia-106tion lengths (X0). A large high-e�ciency veto counter107VETO, and a massive, hermetic hadronic calorimeter108(HCAL) composed of three modules HCAL1-3 completes109the setup. Each module is a 3⇥3 cell matrix with a thick-110ness of' 7.5 nuclear interaction lengths. The events from111e� interactions in the PRS and ECAL were collected with112the trigger provided by the S1�4 requiring also an in-time113cluster in the ECAL with the energy EECAL . 80 GeV.114The detector is described in more detail in Ref. [31].

γ

Z

γ

e−e− e−

γ

Z

γ

γ a a

γ

FIG. 2: Diagramatic illustration the a production in the re-action of Eq.(2).

115

If ALPs exist, one would expect a flux of such highenergy particles from the dump because both scalars andpseudoscalars could be produced in the forward direc-tion through the Primako↵ e↵ect in interactions of highenergy bremsstrahlung photons, generated by 100 GeVelectrons in the target, with virtual photons from theelectrostatic field of the target nuclei:

e�Z ! e�Z�; �Z ! aZ; a ! �� (2)

as illustrated in Fig.2. If the ALP is a relatively long-116lived particle, it would penetrate the first downstream117HCAL1 module serving as shielding and would be ob-118served in the NA64 detector with two distinctive signa-119tures, either 1) via its decay into 2� inside the HCAL2120or HCAL3 modules (denoted further as HCAL2,3), or121

2










γ

Z

γ

e−e− e−

γ

Z

γ

γ a a

γ


115


e�Z ! e�Z�; �Z ! aZ; a ! �� (2)


2










γ

Z

γ

e−e− e−

γ

Z

γ

γ a a

γ


115


e�Z ! e�Z�; �Z ! aZ; a ! �� (2)


Production via Primakoff effect

Closing the gap between beam  dump and colliders

5

number of the a decays for the signature i. The a yield

FIG. 5: The NA64 90% C.L. exclusion region in the (ma; ga��)plane as a function of the (pseudo)scalar mass ma derivedfrom the present analysis. The yellow band represents theparameter space for the benchmark DFSZ [3] and KSVZ [4]models extended with a broader range of E/N values [7, 8].Constraints from the BABAR [51], E137 [52], E141 [53, 54],LEP [55], and PrimEx [56] experiments, as well as limits fromCHARM [57] and NuCal [58], updated in Ref.[59] are alsoshown. For more limits from indirect searches and proposedmeasurements; see, e.g., Refs. [13–15].

from the reaction chain (3) was obtained with the cal-culations described in Ref.[47] assigning . 10% system-atic uncertainty due to di↵erent form-factor parametriza-tions [48, 49]. An additional uncertainty of ' 10% wasaccounted for the data-MC di↵erence for the dimuonyield [34, 50]. The signal detection e�ciency for eachsignature in (4) was evaluated by using signal MC andwere found slightly ma dependent. For instance, forthe signature 1 and ma ' 10 MeV, the "1a and itssystematic error was determined from the product ofe�ciencies accounting for the geometrical acceptance(0.97 ± 0.02), the primary track (' 0.83 ± 0.04), SRD(& 0.95±0.03), ECAL(0.95±0.03), VETO ( 0.94±0.04),HCAL1 (0.94 ± 0.04), and HCAL2,3 (0.97 ± 0.02) sig-

nal event detection. The signal e�ciency loss . 7%due to pileup was taken into account using reconstructeddimuon events [34]. The VETO and HCAL1 e�cien-cies were defined as a fraction of events below the cor-responding energy thresholds with the main uncertaintyestimated to be . 4% for the signal events, which iscaused by the pileup e↵ect from penetrating hadrons.The trigger e�ciency was found to be 0.95 with a smalluncertainty of 2%. The total signal e�ciency ✏a variedfrom 0.51±0.09 to 0.48±0.08 for the amass range of 10-50MeV. The total systematic uncertainty on na calculatedby adding all errors in quadrature did not exceed 20%for both signatures. The attenuation of the a flux dueto interactions in the HCAL1 was found to be negligible.The combined signal region excluded in the (ma; ga��)plane at 90 % C.L. is shown in Fig. 5 together with theresults of other experiments. Our limits are valid for bothscalar and pseudoscalar cases and exclude the region inthe coupling range 2 ⇥ 10�4 . ga�� . 5 ⇥ 10�2 GeV�1for masses ma . 55 MeV.This work serves as a memorial to Danila Tlisov and

his contributions to the NA64 collaboration. We grate-fully acknowledge the support of the CERN manage-ment and sta↵ and the technical sta↵s of the partic-ipating institutions for their vital contributions. Wewould like to thank M.W. Krasny for providing uswith reference [53] and useful comments, B. Döbrichfor providing information on the E141, and updatedCHARM and NuCal exclusion curves, and G. Lan-franchi for valuable discussions. This work was sup-ported by the Helmholtz-Institut für Strahlen-und Kern-physik (HISKP), University of Bonn, the Carl Zeiss Foun-dation 0653-2.8/581/2, and Verbundprojekt-05A17VTA-CRESST-XENON (Germany), Joint Institute for Nu-clear Research (JINR) (Dubna), the Ministry of Scienceand Higher Education (MSHE) and RAS (Russia), ETHZurich and SNSF Grant No. 169133 and No.186158(Switzerland), and FONDECYT Grants No.1191103, No.190845, and No. 3170852, UTFSM PI M 18 13, ANIDPIA/APOYO AFB180002 (Chile).

[1] S. Weinberg, Phys. Rev. Lett. 40, 223 (1978);F. Wilczek, Phys. Rev. Lett. 40, 279 (1978).

[2] R.D. Peccei and H. Quinn, Phys. Rev. Lett. 38 (1977)1440.

[3] M. Dine, W. Fischler, and M. Srednicki, Phys. Lett.B104, 199 (1981);A. Zhitnitski, Sov. J. Nucl. Phys. 31, 260 (1980) [Yad.Fiz. 31, 497 (1980).

[4] J. E. Kim, Phys. Rev. Lett. 43, 103 (1979);M. Schifman, A. Vainstein and V. Zakharov, Nucl. Phys.166, 493 (1981).

[5] N. V. Krasnikov, V. A. Matveev and A.N. Tavkhelidze,Sov. J. Part. Nucl. 12, 38 (1981);J. E. Kim, Phys. Rep. 150, 1 (1987);H. Y. Cheng, Phys. Rep. 158, 1 (1988);

G. Ra↵elt, Phys. Rep. 198, 1 (1990).[6] G. G. di Cortona et al., JHEP 1601, 034 (2016).[7] J.E. Kim, Phys. Rev. D 58, 055006 (1998); L. Di Luzio,

F. Mescia and E. Nardi, Phys. Rev. Lett. 118, 031801(2017).

[8] M. Tanabashi et al. (Particle Data Group), Phys. Rev.D 98, 030001 (2018).

[9] R. H. Parker, C. Yu. W. Zhong, B. Estey, and H. Müller,Science 360, 191 (2018).

[10] G. W. Bennett et al. (Muon g-2 Collaboration), Phys.Rev. D 73, 072003 (2006).

[11] H. Davoudiasl and W. J. Marciano, Phys. Rev. D 98,075011 (2018); C.-Yi Chen, H. Davoudiasl, W. J. Mar-ciano, and C. Zhang, Phys. Rev. D 93, 036006 (2016);W. J. Marciano, A. Masiero, P. Paradisi, and M. Passera,

Search expected to be BKG free up to ~ 5 x1012 EOT

NA64 collaboration, CERN-EP-2020-068 arXiv:2005.02710

~ 2.84x1011 EOT

https://arxiv.org/abs/2005.02710


The NA64 physics prospectsNA64 program: submitted as input to the European Strategy Group in the context of the PBC

Table 1. The NA64 research program: Projections for searches for Dark Sector physics and other rare

processes with e, µ,⇡,K beams.

Process New Physics Comments, Projections for limits

e�

beam Required number of EOT: 5⇥ 1012

A0 ! e+e�, and Dark photon 10�5 < ✏ < 10�2, 1 . mA0 . 100 MeVA

0 ! invisible 2⇥ 10�6 < ✏ < 10�3, 10�3 . mA0 . 1 GeVA

0 ! �� sub-GeV Dark Matter (�) Scalar, Majorana, pseudo-Dirac DM↵S,M

D. 1, ↵p�D

D. 0.1, for m� . 100 MeV

X ! e+e� new gauge X- boson 8Be* anomaly, ✏upe

< 10�5; ✏lowe

> 2⇥ 10�3milliQ particles Dark Sector, charge quantisation 10�4 < mQ < 0.1 e, 10�3 < mmQ < 1 GeVa ! ��, invisible Axion-like particles ginv

a��. 2⇥ 10�5, ma . 200 MeV

µ�

beam Required number of MOT: 1011 � 5⇥ 1013

Zµ ! ⌫⌫ gauge Zµ-boson of Lµ � L⌧ , < 2mµ (g-2)µ anomaly; gVµ . 10�4, with . 1011 MOTZµ ! �� Lµ � L⌧ charged Dark Matter (�) y . 10�12 for m� . 300 MeV with ' 1012 MOTmilliQ Dark Sector, charge quantisation 10�4 < mQ < 0.1 e, 10�3 < mmQ < 2.5 GeV

aµ ! invisible non-universal ALP coupling gY . 10�2, maµ . 1 GeVµ� ⌧ conversion Lepton Flavour Violation �(µ� ⌧)/�(µ ! all) . 10�11

⇡�, K

�beams Current limits, PDG’2018 Required number of POT(KOT):5⇥ 1012(5⇥ 1011)

⇡0 ! invisible Br(⇡0 ! invisible) < 2.7⇥ 10�7 Br(⇡0 ! invisible) . 10�9⌘ ! invisible Br(⌘ ! invisible) < 1.0⇥ 10�4 Br(⌘ ! invisible) . 10�8⌘0 ! invisible Br(⌘0 ! invisible) < 5⇥ 10�4 Br(⌘ ! invisible) . 10�7

K0S! invisible no limits Br(K0

S! invisible) . 10�9

K0L! invisible no limits Br(K0

L! invisible) . 10�7

complementary to K� ! ⇡⌫⌫

beams. The recently concluded NA64 runs in 2016-2018 consisted of physics programs whichaddress the two most important issues currently accessible with electron beam: a high sensitivitysearch for dark photon A0 mediator of sub-GeV Dark Matter production in invisible decay modesand search for visible decays of dark photon A0 ! e+e� and of a new 17 MeV gauge X-boson,X ! e+e�, which can resolve the anomaly observed in the excited 8Be nuclei transitions. Theincoming SPS runs 2021-23, combined with the 2016-18 runs, provides us with the opportunityto meet and perhaps exceed our original goals for the program with electron beam, and to starton a new physics program summarised in Table 1. Therefore, the NA64 Collaboration proposesto carry out further searches for Dark Sector particles and others rare processes in missing energyevents from i high-energy electron interactions at H4 beam, and extend them to the M2 muon andhadron beams at the CERN SPS. Six months of running time at H4 line in 2021-23 will allows usto accumulate at least a factor 10 more statistics, (3� 5)⇥ 1012 EOT, and to explore most of thesub-GeV Dark Matter parameter space, either to observe or completely rule out the 17 MeV gaugeX-boson explanation of the 8Be anomaly and put stringent bounds on the visible decays A0 ! e+e�of dark photons. For the M2 muon beam we propose to focus on the unique opportunity to discoverya new state, e.g. the Zµ, weakly coupled predominantly to muon that could resolve the longstandingmuon (g-2)µ anomaly. Two months of running at M2 line will allow us to collect enough muons inorder to get a conclusive result. We also propose to explore Dark Sector states in invisible decays⇡0, ⌘, ⌘

0,K

0S,K

0L! invisible of neutral mesons with ⇡,K beams.

– 10 –

!26

CERN-PBC-REPORT-2018-007

Could provide an explanation of (g-2)µ anomaly


Summary and Outlook

!27

▪ NA64: Active beam dump + missing-energy approach is very powerful probe for Dark  

Sector physics.

▪ Experiment exceeded sensitivity of previous beam dump exps. to thermal light  

dark matter.

▪ To fully exploit NA64 potential probing most of the remaining parameter space  

predicted by the DM relic density accumulate 5x1012 EOT for A’ → 𝛘�̅� after LS2

▪ Exploration of the remaining parameter space X→ e+e-

▪ New permanent location being prepared with active participation of NA64.

▪ Proposed searches in NA64 with leptonic and hadronic beams: unique sensitivities  

highly competitive/complementary to similar projects.

searching for hidden/dark sectors with the ...10physics department, university of patras, 265 04...

Documents