Dark Matter Searches with Mono-Photon Signature at Future E+E− Colliders

Dark matter searches with mono-photon signature at future e+e− colliders

Jan Kalinowskia, Wojciech Kotlarskib, Krzysztof Mekalaa, Pawel Sopickia, Aleksander Filip Zarnecki˙ a

a Faculty of Physics, University of Warsaw b Institut f¨urKern- und Teilchenphysik, TU Dresden

Research supported by

XXVIII International Workshop on Deep-Inelastic Scattering and Related Subjects Electroweak Physics and Beyond the Standard Model session April 14, 2021 A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 1 / 16 Outline

1 Introduction

2 Analysis framework

3 Results

4 Conclusions

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 2 / 16 Introduction

Dark Matter production The mono-photon signature is considered to be the most general way to look for DM particle production in future e+e− colliders.

DM can be pair produced in the e+e− collisions via exchange of a new mediator particle, which couples to both electrons (SM) and DM states This process can be detected, if additional hard photon radiation from the initial state is observed in the detector... Considered in this contribution is radiative DM production at high energy e+e− colliders: 500 GeV ILC and 3 TeV CLIC.

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 3 / 16 Introduction International Linear Collider

Technical Design (TDR) completed in 2013 arXiv:1306.6328 superconducting accelerating cavities 250 – 500 GeV c.m.s. energy (baseline), 1 TeV upgrade possible footprint 31 km polarisation for both e− and e+ (80%/30%)

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 4 / 16 Introduction

Compact LInear Collider

Conceptual Design (CDR) presented in 2012 CERN-2012-007 high gradient, two-beam acceleration scheme staged implementation plan with energy from 380 GeV to 3 TeV footprint of 11 to 50 km e− polarisation (80%)

For details refer to arXiv:1812.07987

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 5 / 16 Introduction

Running scenarios

ILC Total of 4000 fb−1 assumed at 500 GeV (H-20 scenario) 2×1600 fb−1 for LR and RL beam polarisation combinations 2×400 fb−1 for RR and LL beam polarisation combinations assuming polarisation of ±80% for electrons and ±30% for positrons arXiv:1903.01629 CLIC Total of 5000 fb−1 assumed at 3 TeV 4000 fb−1 for negative electron beam polarisation 1000 fb−1 for positive electron beam polarisation assuming polarisation of ±80% for electrons arXiv:1812.06018

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 6 / 16 Introduction Unique analysis approach

4000 Most of the studies performed so far Vector, 500GeV ILD focused on heavy mediator exchange [GeV]

95 3000

Λ expected WIMP (EFT limit) and coupling values O(1) exclusion region ⇒ extracted were limits on DM 2000 Only statistical uncertainties _ or mediator masses 4ab-1, P(e ,e+) = (+80%,-30%) 1000 4ab-1, polarisation mix _ 1.6ab-1, P(e ,e+) = (+80%,-30%) 4ab-1, unpolarised In our study: EFT not valid 50 100 150 200 250 focus on light mediator Mχ [GeV] exchange (DM even lighter) ILD study: arXiv:2001.03011 consider very small mediator Phys. Rev. D 101, 075053 (2020) couplings to SM,Γ Γ SM tot CLIC study: arXiv:2103.06006

“Experimental” approach (model-independent) ⇒ focus on cross section limits as a function of mediator mass and width A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 7 / 16 Analysis framework

Simulating mono-photon events Dedicated simulation procedure for Whizard, with all “detectable” photons generated on Matrix Element level, matched with soft ISR.

For details: J. Kalinowski et al., Eur. Phys. J. C 80 (2020) 634, arXiv:2004.14486 √ Two variables, calculated separately Detector acceptance s = 500 GeV for each emitted photon:

102 θγ [GeV]

p +

q = 4E E · sin , q − 0 γ 2 θ p γ 10 q = 4E E · cos , γ + 0 γ BeamCal (E >30 GeV) 2 γ LumiCal (E >10 GeV) γ Tracking (E >5 GeV) are used to separate “soft ISR” ME photon cut 1 Hard photon cut γ p =0.1, 1, 10, 100 GeV emission region from the region T described by ME calculations. 1 10 102 q- [GeV] A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 8 / 16 Analysis framework

Event selection Delphes framework used for detector simulation and event reconstruction. On generator level: Require: 1, 2 or 3 ME photons single photon with nonradiative events for signal γ p > 3 GeV & |ηγ| < 2.8 (ILC) only (for normalisation) T γ γ pT > 10 GeV & |η | < 2.6 (CLIC) all ME photons with q > 1 GeV & E γ > 1 GeV no other activity in the detector ± other reconstructed objects rejected are events with no electrons γ q± > 1 GeV & E > 1 GeV no LumiCal photons for any of the ISR photons no BeamCal photons no jets at least one ME photon with γ ◦ γ ◦ pT > 2 GeV & 5 < θ < 175 (ILC 500 GeV) γ ◦ γ ◦ pT > 5 GeV & 7 < θ < 173 (CLIC 3 TeV)

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 9 / 16 Analysis framework Background distributions For mono-photon events, two SM backgrounds are relevant: (radiative) Bhabha scattering and (radiative) neutrino pair production Two variables fully describe mono-photon event kinematics γ ⇒ use 2D distribution of (pT , ηγ) to constrain DM production γ γ η η 2 2

1 104 1 105

0 0

3 −1 −1 10

− 2 −2 104 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 γ γ log p /pmin log p /pmin ILC 500 GeV X T T CLIC 3 TeV X T T (-80%/+30%) 1600 fb−1 (+80%) 1000 fb−1 A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 10 / 16 Analysis framework Signal distributions Simpliﬁed DM model used to simulate DM pair-production (see backup)

Expected mono-photon distribution for fermion DM withM χ = 50 GeV and vector mediator of 400 GeV @ ILC 2.4 TeV @ CLICΓ /M = 0.03 γ γ

η 10 η 2 2 1 1 1 1 0 0 10−1 −1 − −1 10 1 − 2 −2 − 10 2 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 γ γ log p /pmin log p /pmin X T T X T T Signal normalised to unpolarised DM pair-production cross section of 1 fb A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 11 / 16 Results

Cross section limits for radiative events (with tagged photon) Vector mediator with Γ/M = 3%

ILC @ 500 GeV CLIC @ 3 TeV

-1 LR 1600 fb neg 4000 fb-1 -1 [fb] RL 1600 fb [fb] γ 10 γ pos 1000 fb-1 χ -1 χ

χ RR 400 fb χ

→ -1 → - LL 400 fb - combined e e + + e 95%CL e 95%CL σ σ 1 1

102 103 102 103 104 MY [GeV] MY [GeV]

Limits calculated with CLs approach using RooFit v3.60 A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 12 / 16 Results

Cross section limits for total DM production cross section Corrected for probability of hard photon tagging! see backup slides Combined limits for mediator with Γ/M = 3% ILC @ 500 GeV CLIC @ 3 TeV

Γ/m = 0.5 Γ/m = 0.5 2 [fb] Γ/m = 0.1 [fb] 10 Γ/m = 0.1 χ χ χ 2 Γ/m = 0.03 χ Γ/m = 0.03

→ 10 → - -

e Γ e Γ

+ /m = 0.01 + /m = 0.01 e 95%CL e 95%CL σ σ

10 10

102 103 102 103 104 MY [GeV] MY [GeV] √ Radiation suppressed for narrow mediator with MY ∼ s ⇒ weaker limits A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 13 / 16 Results Impact of systematic uncertainties following ILD study: Phys. Rev. D 101, 075053 (2020), arXiv:2001.03011 Cross section limits for mediator with Γ/M = 3% ILC @ 500 GeV CLIC @ 3 TeV

RL (stat) pos (stat) 2 [fb] 2 RL (stat+sys) [fb] 10 pos (stat+sys) χ 10 χ χ χ H-20 (stat) combined (stat) → → - - e e

+ H-20 (stat+sys) + combined (stat+sys) e 95%CL e 95%CL σ σ

10 10

102 103 102 103 104 MY [GeV] MY [GeV] √ Inﬂuence of systematic eﬀects reduced for light mediators, MY < s A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 14 / 16 Results

Coupling limits with systematic uncertainties Combined coupling limits for assumed mass and width of the mediator. ILC @ 500 GeV CLIC @ 3 TeV

Γ/m = 0.5 Vector Γ/m = 0.1 Scalar 10−1 Γ/m = 0.03 1 Pseudoscalar Γ/m = 0.01 Axial-vector (95%CL)

eeY −1 g −2 10 10 (95%CL limit) eeY g −2 − 10 10 3

102 103 102 103 104 MY [GeV] MY [GeV]

Almost uniform sensitivity to geeY up to kinematic limit. Coupling limits weakly dependent on the assumed coupling structure!

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 15 / 16 Results

Coupling limits with systematic uncertainties Combined coupling limits for assumed mass and width of the mediator. ILC @ 500 GeV CLIC @ 3 TeV

Γ/m = 0.5 Vector Γ/m = 0.1 Scalar 10−1 Γ/m = 0.03 1 Pseudoscalar Γ/m = 0.01 Axial-vector (95%CL) Y χ χ −1 −2 g 10 10 (95%CL limit) Y χ eeY χ g g −2 −3 10 10 eeY g

102 103 102 103 104 MY [GeV] MY [GeV]

Almost uniform sensitivity to geeY up to kinematic limit. Coupling limits weakly dependent on the assumed coupling structure!

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 15 / 16 Conclusions

New framework for mono-photon analysis developed focus on light mediator exchange and very small mediator couplings to SM Mono-photon production at e+e− colliders sensitive to wide range of DM pair-production scenarios + − O(1 fb) limits on the radiative production e e → χχγtag O(10 fb) limits on the DM pair-production e+e− → χχ(γ) √ except for the resonance region MY ∼ s O(10−3 − 10−2) limits on the mediator coupling to electrons √ up to the kinematic limit MY ≤ s Limits largely independent on the mediator type/coupling For for heavy mediators, limits from EFT analysis can be reproduced For light mediators limits more stringent than those expected from direct resonance search in SM decay channels

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 16 / 16 Thank you!

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 16 / 16 Backup slides

Simpliﬁed DM model Simpliﬁed model covering most popular scenarios of DM pair-production

Possible DM candidates: Cross section for e+e− → χχ for real or complex scalar Mχ = 50 GeV and MY = 300 GeV Majorana or Dirac fermion Γ 4 Y = 3 GeV [fb] 10

σ Γ real vector Y = 30 GeV 3 Γ = 150 GeV 10 Y Possible mediators: 102 scalar 10 pseudo-scalar 1 vector 10−1 2×102 103 2×103 axial-vector s [GeV]

(mixed couplings, eg. V-A or V+A, also possible)

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 16 / 16 Backup slides

Simpliﬁed DM model

Dark matter particles, Xi , couple to the SM particles via an mediator, Yj . Each simpliﬁed scenario is characterized by one dark matter candidate and one mediator from the set listed below: particle mass spin charge self-conjugate type

XR mXR 0 0 yes real scalar

XC mXC 0 0 no complex scalar 1 XM mXM 2 0 yes Majorana fermion DM 1 XD mXD 2 0 no Dirac fermion

XV mXV 1 0 yes real vector

YR mYR 0 0 yes real scalar

YV mYC 1 0 yes real vector

mediator TC mTC 0 1 no charged scalar

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 16 / 16 Backup slides

Simpliﬁed DM model Lagrangian describing mediator coupling to electrons given by

L 3 e¯(g 1 + ıγ5g 5 )eY +e ¯γ (g 1 + γ5g 5 )eY µ eeY eYR eYR R µ eYV eYV V

The interaction of mediators with dark matter is described by

2 ∗ ∗ µ LXXY 3 gXR YR XR YR + ıgXC YV (XC (∂µXC ) − (∂µXC )XC )YV + X¯ (g 1 + ıγ5g 5 )X Y + X¯ γ (g 1 + γ5g )X Y µ D XD YR XD YR D R D µ XD YV XD YV D V X¯ (g 1 + ıγ5g 5 )X Y + g 5 ψ¯ γ γ5ψ Y µ M XM YR XM YR M R XM YV M µ M V

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 16 / 16 Backup slides

Detector simulation for ILC running at 500 GeV ILCgen model for Delphes includes proper modelling of forward detectors BeamCal Reconstruction eﬃciency for e+e− → e+e−

150

100 Y [mm] 50

−50

−100

−150 −150 −100 −50 0 50 100 150 X [mm] Included in the oﬃcial Delphes repository as delphes card ILCgen.tcl

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 16 / 16 Backup slides

Detector simulation for CLIC running at 3 TeV CLICdet model for Delphes also modiﬁed to include forward calorimeters LumiCal+ BeamCal Reconstruction eﬃciency for e+e− → e+e−

Included in the repository as delphes card CLICdet Stage3 fcal.tcl

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 16 / 16 Backup slides

ILC vs CLIC comparison of simulation and analysis setup ILCgen CLICdet @ 500 GeV @ 3 TeV Generator level cuts γ pT min. 2 GeV 5 GeV Θγ min. 5◦ 7◦ Detector acceptance (Delphes model) tracking |η| < 3 |η| < 2.54 ECAL |η| < 3 |η| < 3 LumiCal 3 < |η| < 4 3 < |η| < 4 BeamCal 4 < |η| < 5.8 4 < |η| < 5.3 Detector level cuts γ pT min. 3 GeV 10 GeV |ηγ| max. 2.8 2.6 (7◦) (8.5◦)

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 16 / 16 Backup slides

Background distributions For considered SM backgrounds before and after detector level selection (radiative) Bhabha scattering and (radiative) neutrino pair production

8 107 10 e+e-γ (gen) e+e-γ (gen) ννγ 6 ννγ 7 (gen) 10 (gen)

Events 10 Events e+e-γ (det) e+e-γ (det) ννγ 5 ννγ 106 (det) 10 (det) 104 105 103 104 102 103 10 102 0 50 100 150 200 250 0 500 1000 1500 γ γ p [GeV] p [GeV] T T ILC 500 GeV CLIC 3 TeV (-80%/+30%) 1600 fb−1 (+80%) 1000 fb−1

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 16 / 16 Backup slides

ISR rejection eﬃciency Fraction of events generated by Whizard removed by ISR rejection procedure (ISR photons emitted in the phase-space region covered by ME) ILC @ 500 GeV CLIC @ 3 TeV

50 50 [%] [%] 40 40 ISR rej. ISR rej. f f 30 30

20 20 Γ/m = 0.50 Γ/m = 0.50 Γ/m = 0.10 Γ/m = 0.10 10 10 Γ/m = 0.03 Γ/m = 0.03 Γ/m = 0.01 Γ/m = 0.01 0 0 102 103 102 103 104 MY [GeV] MY [GeV]

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 16 / 16 Backup slides Tagging eﬃciency Detectable hard photon emitted only in a fraction of signal event + − + − σ e e → χ χ γtag = fmono-photon · σ e e → χ χ (γ)

ILC @ 500 GeV CLIC @ 3 TeV

Γ/m = 0.50 Γ/m = 0.50

[%] Γ/m = 0.10 [%] Γ/m = 0.10 15 Γ/m = 0.03 15 Γ/m = 0.03 Γ/m = 0.01 Γ/m = 0.01 mono-photon mono-photon f 10 f 10

5 5

102 103 102 103 104 M [GeV] M [GeV] Y √ Y Emission strongly suppressed for narrow mediator with MY ∼ s A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 16 / 16 Backup slides

Systematic uncertainties

following ILD study: Phys. Rev. D 101, 075053 (2020), arXiv:2001.03011 Considered sources of uncertainties: Integrated luminosity uncertainty of 0.26% uncorrelated between polarisations Luminosity spectra shape uncertainty correlated between polarisations Uncertainty in neutrino background normalisation of 0.2% (th+exp) correlated between polarisations Uncertainty in Bhabha background normalisation of 1% (th+exp) correlated between polarisations ⇒ nuisance parameters in the model ﬁt (7 for ILC, 5 for CLIC)

A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 16 / 16 Backup slides M2 Eﬀective mass scale limits Λ2 = Y |geeY gχχY | √ For MY s, limits on the eﬀective mass scale of new interactions no longer depend on the assumed mediator mass or width ⇒ EFT approximation can be used ILC @ 500 GeVΛ lim ∼ 3 TeV CLIC @ 3TeVΛ lim ∼ 7 TeV

Vector Vector

Scalar Scalar 10 Pseudoscalar Pseudoscalar [TeV] [TeV]

Y Axial-vector Y Axial-vector χ χ 10 χ χ g g eeY eeY g g / / Y Y M M

1 1

102 103 102 103 104 MY [GeV] MY [GeV] A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 16 / 16 Backup slides M2 Eﬀective mass scale limits Λ2 = Y |geeY gχχY | √ For MY s, limits on the eﬀective mass scale of new interactions no longer depend on the assumed mediator mass or width ⇒ EFT approximation can be used ILC @ 500 GeVΛ lim ∼ 3 TeV CLIC @ 3TeVΛ lim ∼ 7 TeV

RL (stat) pos (stat)

RL (stat+sys) pos (stat+sys) 10 H-20 (stat) combined (stat) [TeV] [TeV]

Y H-20 (stat+sys) Y combined (stat+sys)

χ χ 10 χ χ g g eeY eeY g g / / Y Y M M

1 1

102 103 102 103 104 MY [GeV] MY [GeV] A.F.Zarnecki˙ (University of Warsaw) DM searches with mono-photons April 14, 2021 16 / 16