Faking ordinary photons by displaced dark photon decays

Yuhsin Tsai, Lian-Tao Wang, and Yue Zhao

Phys. Rev. D 95, 015027 – Published 30 January 2017

Abstract

A light metastable dark photon decaying into a collimated electron/positron pair can fake a photon, either converted or unconverted, at the LHC. The detailed object identification relies on the specifics of the detector and strategies for the reconstruction. We study the fake rate based on the ATLAS (CMS) detector geometry and show that it can be O(1) with a generic choice of parameters. Especially, the probability of being registered as a photon is angular dependent. Such detector effects can induce bias to measurements on certain properties of new physics. In this paper, we consider the scenario where dark photons in final states are from a heavy resonance decay. Consequently, the detector effects can dramatically affect the results when determining the spin of a resonance. Further, if the decay products from the heavy resonance are one photon and one dark photon, which has a large probability to fake a diphoton event, the resonance is allowed to be a vector. Because of the difference in detectors, the cross sections measured in ATLAS and CMS do not necessarily match. Furthermore, if the diphoton signal is given by the dark photons, the standard model $Z γ$ and $Z Z$ final states do not necessarily come with the $γ γ$ channel, which is a unique signature in our scenario. The issue studied here is relevant also for any future new physics searches with photon(s) in the final state. We discuss possible ways of distinguishing dark photon decay and a real photon in the future.

Received 27 September 2016

DOI:https://doi.org/10.1103/PhysRevD.95.015027

Physics Subject Headings (PhySH)

Extensions of gauge sector Particle decays Particle production

Hypothetical gauge bosons

Particles & Fields

Authors & Affiliations

Yuhsin Tsai¹, Lian-Tao Wang^2,3, and Yue Zhao⁴

¹Maryland Center for Fundamental Physics, Department of Physics, University of Maryland, College Park, Maryland 20742, USA
²Department of Physics, The University of Chicago, Chicago, Illinois 60637, USA
³Enrico Fermi Institute and Kavli Institute for Cosmological Physics, The University of Chicago, Chicago, Illinois 60637, USA
⁴Michigan Center for Theoretical Physics, University of Michigan, Ann Arbor, Michigan 48109, USA

Article Text (Subscription Required)

Click to Expand

References (Subscription Required)

Click to Expand

Issue

Vol. 95, Iss. 1 — 1 January 2017

Reuse & Permissions

Access Options

Article Available via CHORUS

Download Accepted Manuscript

Author publication services for translation and copyediting assistance advertisement

Images

Figure 1
Left: Probability of one dark photon decay by regions, where $r_{⊥}$ is the transverse distance of the decay vertex from the beam pipe. We take the detailed positions of the third-to-the-last layer of SCT from [19] in our analysis. The probability for each region is $η$ dependent, and we integrate the result from the central region to $| η | = 2.3$ in our estimation based on our simulation. Right: Comparison of energy ratios between the hard and soft leptons from the SM photon conversion and the dark photon decay. The plot is made by assuming $m_{γ^{'}} = 100 MeV$ , and the decay into $e^{+} e^{-}$ is from the transverse mode of $γ^{'}$ . When $m_{γ^{'}} ≪ E_{γ^{'}}$ , the distribution is insensitive to the size of $m_{γ^{'}}$ .
Reuse & Permissions
Figure 2
Ratio of the fake photon events being identified as converted and unconverted photons. The blue curve is derived from the ratio of the dark photon decay probability in the converted and unconverted regions. The light (dark) purple band includes a $2 σ$ fluctuation of the photon events as in Eq. (2) assuming 20 (200) events of both the signal and the background. According to [35], we take the ratio of converted versus unconverted photon events in the SM to be 25%.
Reuse & Permissions
Figure 3
Left: Probability of a dark photon produced from $Φ (750) \to 2 γ^{'}$ decaying into $e^{+} e^{-}$ between radial distance [371, 1500] mm (from the second layer of SCT to ECAL with $η = 0$ ) that has its two hits in the $η$ direction of the strip cells (defined as $Δ x$ ) to be within the average thickness of the cell ( $< 4.7 mm$ ). Here we assume the $γ^{'}$ moves in the radial ( $η = 0$ ) direction. Right: Probability of one dark photon fakes a real photon. The result is derived according to the decay probability with (solid curve) and without (dash-dotted curve) the upper bound on the angular separation in the ECAL for the fake normal photon signals. See Sec. 2 for more discussions. Here we estimate the acceptance of this cut by assuming dark photons to be produced in the transverse mode. After boosting the open angle to the lab frame, we calculate the acceptance as a function of the decay position.
Reuse & Permissions
Figure 4
The required cross section times branching ratio of the scalar resonance $Φ (750) \to 2 γ^{'}$ that generates a 5 fb diphoton signal at the 13 TeV ATLAS search, assuming $Br (γ^{'} \to e^{+} e^{-}) = 1$ . Different color schemes represent various dark photon masses $m_{γ^{'}}$ , and the shaded regions give the 90% C.L. exclusion bounds from the 8 TeV ATLAS search of a $photon + MET$ [28] (exclusion regions on the left) and prompt lepton jets [25] (exclusion regions on the right). Here we carry out the analysis conservatively by assuming the highly collimated $e^{+} e^{-}$ from the light $γ^{'}$ decays cannot be reconstructed if the decay happens after the first layer of the ECAL ( $≲ 1590 mm$ ). Since the MET in [28] is defined as the imbalance of the total $p_{T}$ of the constructed objects, we treat such displaced decay as missing energy. When the decay happens before the pixel ( $≲ 34 mm$ ), the highly collimated lepton pairs are identified as lepton jets. The rescaling of the resonance production between $13 / 8 TeV$ is based on the $g g \to Φ (750)$ .
Reuse & Permissions
Figure 5
Here we take the events that pass a mild diphoton $p_{T}$ cut, as discussed in context. We show the angular distribution of the decay products in the rest frame of the heavy resonance, for scalar (left) and vector (right). Different colors correspond to various choices of $ε$ . We take $m_{γ^{'}} = 100 MeV$ as our benchmark.
Reuse & Permissions
Figure 6
The angular distribution of the decay products in the rest frame of the heavy resonance, assuming $m_{γ^{'}} = 100 MeV$ . Different colors specify various categories of photon identification, and the different angular distributions of the signal may provide a handle to distinguish the real and fake photon events. When generating the $V \to γ γ^{'}$ events, we assume the process is dominated by the transverse component of the massive vectors $V$ and $γ^{'}$ . Note, the scalar decay is a flat distribution in the rest frame before taking into consideration the angular dependent acceptance from the detector. The spin of the vector resonance has a large probability to point along to the beam direction. Thus the vector bosons in the final states are more concentrated in the central region in the rest frame, i.e., smaller $\cos (θ_{CM}^{γ^{'}})$ .
Reuse & Permissions

Physical Review D

covering particles, fields, gravitation, and cosmology