CMS-PAS-EXO-23-008

CMS-PAS-EXO-23-008
Search for dark photon bremsstrahlung by muons in proton-proton collisions at $ \sqrt{s} = $ 13 TeV
CMS Collaboration
2026-03-14

Abstract: A search for dark bremsstrahlung interactions between muons from Z boson decay and the CMS endcap calorimeters is presented. The search is performed with the CMS detector at the CERN LHC using a proton-proton collision data sample at a center-of-mass energy of 13 TeV corresponding to an integrated luminosity of 59.6 fb$ ^{-1} $. The search utilizes Z boson decays to a muon-antimuon pair. One of the muons is required to have no anomalous interactions and serves as a trigger for the event. Secondary interactions are then searched for between the second muon and the calorimeters, such that the response in tracker volume of the detector can be used to cleanly identify the muon through the invariant mass of the trigger muon and the tracker track of the second muon. The search is performed with two mutually-exclusive categories of events, one where the trajectory of the second muon is severely altered by a dark bremsstrahlung interaction and there are no nearby muon chamber hits to the projected tracker track, and one where the trajectory of the second muon is slightly altered and there is a nearby muon that is reconstructed purely by hits in the muon chambers. Limits are placed in the two dimensional plane of the square of the dark photon mixing parameter $ \epsilon^{2} $ and the dark photon mass for A' masses in the range 10 MeV to 1 GeV. This is the first search of this type at an LHC experiment.
Links: CDS record (PDF) ; CADI line (restricted) ;

Figures & Tables	Summary	References	CMS Publications

Figures
png pdf	Figure 1: Feynman diagram of the dark bremsstrahlung process
png pdf	Figure 2: The ratio of the outgoing muon energy after dark bremsstrahlung to the incident muon energy prior to dark bremsstrahlung (left). The locations of simulated dark bremsstrahlung interactions at a representative $ {A} ^{\prime} $ mass of 0.4 GeV (right). The process is enabled only within the volume of the ECAL and HCAL endcap calorimeters. The higher density of the ECAL produces more interactions per unit volume of detector, while the larger size of the HCAL results in roughly twice as many total interactions occurring within it.
png pdf	Figure 2-a: The ratio of the outgoing muon energy after dark bremsstrahlung to the incident muon energy prior to dark bremsstrahlung (left). The locations of simulated dark bremsstrahlung interactions at a representative $ {A} ^{\prime} $ mass of 0.4 GeV (right). The process is enabled only within the volume of the ECAL and HCAL endcap calorimeters. The higher density of the ECAL produces more interactions per unit volume of detector, while the larger size of the HCAL results in roughly twice as many total interactions occurring within it.
png pdf	Figure 2-b: The ratio of the outgoing muon energy after dark bremsstrahlung to the incident muon energy prior to dark bremsstrahlung (left). The locations of simulated dark bremsstrahlung interactions at a representative $ {A} ^{\prime} $ mass of 0.4 GeV (right). The process is enabled only within the volume of the ECAL and HCAL endcap calorimeters. The higher density of the ECAL produces more interactions per unit volume of detector, while the larger size of the HCAL results in roughly twice as many total interactions occurring within it.
png pdf	Figure 3: The BDT distribution in the partial disappearance region for the DY MC (yellow), signal (dashed lines) and data samples before the fit. The non-DY background is not shown. The hatched uncertainty bands on the simulated background histogram include statistical and systematic components. The systematic uncertainty of the BDT shape is primarily driven by the pileup modeling, muon identification, and muon isolation uncertainties, with an additional systematic applied based on differences observed in validation regions.
png pdf	Figure 4: Fits of the non-DY-muon background in the off-peak control regions for complete (left) and partial (right) disappearance categories. The region of invariant mass between 70 and 110 GeV is excluded from the control region as well as the fit.
png pdf	Figure 4-a: Fits of the non-DY-muon background in the off-peak control regions for complete (left) and partial (right) disappearance categories. The region of invariant mass between 70 and 110 GeV is excluded from the control region as well as the fit.
png pdf	Figure 4-b: Fits of the non-DY-muon background in the off-peak control regions for complete (left) and partial (right) disappearance categories. The region of invariant mass between 70 and 110 GeV is excluded from the control region as well as the fit.
png pdf	Figure 5: The invariant mass distributions of the tag and probe muon pair in the complete disappearance signal region (left), and the distributions of the $ \Delta\phi $ between the probe track and the nearest CSC segment multiplied by the track charge in the partial disappearance signal region (right). Statistical and systematic uncertainties in the expected background yields are represented by the hatched band. Background distributions are shown after the fit, and the DY and non-DY backgrounds are stacked with the uncertainty for the full background distribution shown as a hatched area centered on the sum of the background components. The lower panels show the ratio between the observed data and the total background expectation. The expected signal distributions for $ m({A} ^{\prime})= $ 0.2 GeV and $ m({A} ^{\prime})= $ 1.0 GeV are normalized to their cross sections with $ \epsilon^{2} = $ 1.
png pdf	Figure 6: The observed (solid) and expected (dashed) 95% CL upper limits on the kinetic mixing $ \epsilon^{2} $ are shown for the combined analysis of both search categories. The 68 and 95% confidence intervals on the expected limit are shown in the green and yellow bands, respectively.

Tables
png pdf	Table 1: Summary of the systematic and statistical uncertainties and the relative effect on the total yields of events. The uncertainties are shown for both the complete (comp.) and partial (part.) disappearance region. The uncertainty on the signal yield is calculated for the $ \mathrm{m}_{{A} ^{\prime}}= $ 0.2 GeV signal point. The other signal points have matching uncertainties within the statistical uncertainty on the simulation samples. The set of systematic uncertainties which are considered correlated between categories are indicated; all correlations are for signal, none for background.
png pdf	Table 2: The number of background events, as determined during the fit, compared to the observed data events.

Summary

A novel search for dark matter production via dark bremsstrahlung from muons interacting with calorimeter material in the CMS detector is described. Events are selected to isolate candidate muon pairs arising from decays of Z bosons by finding probe tracks that match with fully reconstructed muons to produce an invariant mass near the Z boson mass. The probe tracks are then extrapolated into the calorimeters and muon chambers to search for missing transverse momentum. Because of the strong dependence on the hadron calorimeter (HCAL) resolution for rejecting backgrounds, tracks are required to pass through the upgraded HCAL endcap and selected from a sample of proton-proton collisions at $ \sqrt{s}= $ 13 TeV corresponding to an integrated luminosity of 59.6 fb$ ^{-1} $ collected in 2018. Two categories of event are considered, one where the muon is not observed in the CMS muon systems or the last layers of the HCAL endcap (``complete disappearance'') and a second where the muon is observed with a large decrease in momentum (``partial disappearance''). The Drell--Yan (DY) background in the complete disappearance category is estimated using a dedicated control region where energy is required in the last layer of the HCAL endcap traversed by the probe track. The DY background in the partial disappearance category is estimated using simulation, with corrections derived from control regions in data that provide information about the HCAL energy deposits and individual muon chamber hits of a typical muon. The non-DY backgrounds are evaluated from data by fitting the non-peaking invariant mass distribution in regions away from the Z peak. No excess is observed, so the results are interpreted as limits on the mixing parameter for a minimal light dark matter model, using the effective fixed target luminosity of the CMS calorimeter volumes and the number of selected tracks. The upper limits placed on the mixing parameter $ \epsilon^2 $ range from 0.45--1.0 as the mass of the dark photon ranges from 0.01--1.0 GeV. These results are the first time a search for a dark photon has been performed through interactions between muons and a fixed target using a collider as a source of muons.

References
1	V. C. Rubin, N. Thonnard, and W. K. Ford, Jr.	Rotational properties of 21 SC galaxies with a large range of luminosities and radii, from NGC 4605 (R = 4kpc) to UGC 2885 (R = 122 kpc)	Astrophys. J. 238 (1980) 471
2	D. Clowe, A. Gonzalez, and M. Markevitch	Weak lensing mass reconstruction of the interacting cluster 1E0657-558: Direct evidence for the existence of dark matter	Astrophys. J. 604 (2004) 596--603	astro-ph/0312273
3	Planck Collaboration	Planck 2015 results. XIII. Cosmological parameters	Astron. Astrophys. 594 (2016) A13	1502.01589
4	J. Billard et al.	Direct detection of dark matter\textemdashAPPEC committee report	Rept. Prog. Phys. 85 (2022) 056201	2104.07634
5	D. S. Akerib et al.	Snowmass2021 Cosmic Frontier Dark Matter Direct Detection to the Neutrino Fog	in. 3, 2022 Snowmass 202 (2022) 1	2203.08084
6	LZ Collaboration	Dark Matter Search Results from 4.2 Tonne-Years of Exposure of the LUX-ZEPLIN (LZ) Experiment		2410.17036
7	G. Arcadi et al.	The waning of the WIMP? A review of models, searches, and constraints	EPJC 78 (2018) 203	1703.07364
8	M. Battaglieri et al.	US Cosmic Visions: New Ideas in Dark Matter 2017: Community Report	in U.S. Cosmic Visions: New Ideas in Dark Matter. 7, 2017	1707.04591
9	B. W. Lee and S. Weinberg	Cosmological Lower Bound on Heavy Neutrino Masses	PRL 39 (1977) 165
10	C. Boehm, T. A. Ensslin, and J. Silk	Can Annihilating dark matter be lighter than a few GeVs?	JPG 30 (2004) 279	astro-ph/0208458
11	M. Pospelov, A. Ritz, and M. B. Voloshin	Secluded WIMP Dark Matter	PLB 662 (2008) 53	0711.4866
12	B. Batell, A. Freitas, A. Ismail, and D. Mckeen	Flavor-specific scalar mediators	PRD 98 (2018) 055026	1712.10022
13	E. Izaguirre, G. Krnjaic, P. Schuster, and N. Toro	Analyzing the Discovery Potential for Light Dark Matter	PRL 115 (2015) 251301	1505.00011
14	NA64 Collaboration	Search for Light Dark Matter with NA64 at CERN	PRL 131 (2023) 161801	2307.02404
15	Y. Kahn, G. Krnjaic, N. Tran, and A. Whitbeck	M3: a new muon missing momentum experiment to probe $ (g - 2)\mu $ and dark matter at Fermilab	Journal of High Energy Physics 2018 (2018)
16	CMS Collaboration	The CMS experiment at the CERN LHC	JINST 3 (2008) S08004
17	CMS Collaboration	Performance of the CMS Level-1 trigger in proton-proton collisions at $ \sqrt{s} = $ 13 TeV	JINST 15 (2020) P10017	CMS-TRG-17-001 2006.10165
18	CMS Collaboration	The CMS trigger system	JINST 12 (2017) P01020	CMS-TRG-12-001 1609.02366
19	CMS Collaboration	Electron and photon reconstruction and identification with the CMS experiment at the CERN LHC	JINST 16 (2021) P05014	CMS-EGM-17-001 2012.06888
20	CMS Collaboration	Performance of the CMS muon detector and muon reconstruction with proton-proton collisions at $ \sqrt{s}= $ 13 TeV	JINST 13 (2018) P06015	CMS-MUO-16-001 1804.04528
21	CMS Collaboration	Description and performance of track and primary-vertex reconstruction with the CMS tracker	JINST 9 (2014) P10009	CMS-TRK-11-001 1405.6569
22	CMS Collaboration	Particle-flow reconstruction and global event description with the CMS detector	JINST 12 (2017) P10003	CMS-PRF-14-001 1706.04965
23	CMS Collaboration	CMS Technical Design Report for the Phase 1 Upgrade of the Hadron Calorimeter	link
24	CMS Tracker Group Collaboration	The CMS phase-1 pixel detector upgrade	JINST 16 (2021) P02027	2012.14304
25	CMS Collaboration	Track impact parameter resolution for the full pseudo rapidity coverage in the 2017 dataset with the CMS phase-1 pixel detector	cms detector performance summary, 2020 CDS
26	CMS Collaboration	Local reconstruction algorithms in the cathode strip chambers of CMS	EPJ Web Conf. 214 (2019) 02014
27	J. Alwall et al.	The automated computation of tree-level and next-to-leading order differential cross sections, and their matching to parton shower simulations	JHEP 07 (2014) 079	1405.0301
28	J. Alwall et al.	Comparative study of various algorithms for the merging of parton showers and matrix elements in hadronic collisions	EPJC 53 (2008) 473	0706.2569
29	T. Sjostrand et al.	An introduction to PYTHIA 8.2	Comput. Phys. Commun. 191 (2015) 159	1410.3012
30	CMS Collaboration	Extraction and validation of a new set of CMS PYTHIA8 tunes from underlying-event measurements	EPJC 80 (2020) 4	CMS-GEN-17-001 1903.12179
31	NNPDF Collaboration	Parton distributions from high-precision collider data	EPJC 77 (2017) 663	1706.00428
32	GEANT4 Collaboration	GEANT 4---a simulation toolkit	NIM A 506 (2003) 250
33	T. Eichlersmith et al.	Simulation of dark bremsstrahlung in Geant4	Comput. Phys. Commun. 287 (2023) 108690	2211.03873
34	J. D. Bjorken, R. Essig, P. Schuster, and N. Toro	New fixed-target experiments to search for dark gauge forces	PRD 8 (2009) 80 075018
35	CMS Collaboration	Technical proposal for the Phase-II upgrade of the Compact Muon Solenoid	CMS Technical Proposal CERN-LHCC-2015-010, CMS-TDR-15-02, 2015 CDS
36	A. L. Read	Presentation of search results: the $ \text{CL}_\text{s} $ technique	JPG 28 (2002) 2693
37	T. Junk	Confidence level computation for combining searches with small statistics	NIM A 434 (1999) 435	hep-ex/9902006
38	G. Cowan, K. Cranmer, E. Gross, and O. Vitells	Asymptotic formulae for likelihood-based tests of new physics	EPJC 71 (2011) 1554	1007.1727
39	CMS Collaboration	The CMS Statistical Analysis and Combination Tool: Combine	Comput. Softw. Big Sci. 8 (2024) 19	CMS-CAT-23-001 2404.06614