[Paper Review] Searches for long-lived and highly-ionizing particles at the CMS and ATLAS experiments

Teresa Lenz|arXiv (Cornell University)|Sep 27, 2016

Particle physics theoretical and experimental studies6 references3 citations

TL;DR

This paper presents searches for long-lived and highly-ionizing particles at the CMS and ATLAS experiments using energy loss per unit path length ($dE/dx$) as a key discriminant. By exploiting high $dE/dx$ values from multi-charged or magnetic monopole-like particles, the studies set upper limits on cross sections and exclude models up to specific masses, including double-charged particles up to 650 GeV and magnetic monopoles with $g_D = 1.0$ up to 1340 GeV.

ABSTRACT

Long-lived particles are contained in a variety of beyond Standard Model theories, including supersymmetric models, universal extra dimensions, or technicolor theories. If the lifetime of such a particle is long enough, the particle can enter - or even pass through - the detector before it decays. Therefore, searches for long-lived particles require a very different search strategy compared to conventional searches for particles beyond the Standard Model. If the new particle is not only weakly interacting, the particle can be reconstructed itself and not only via its decay products. A very specific characteristic of such new heavy charged particles is their large ionization losses when traveling through the detector. This article summarizes searches for long-lived particles at the CMS and ATLAS experiments that exploit the potentially high ionization losses per path length (dE/dx) of the new particle. The presented searches are performed on 8 and/or 13 TeV data. Additionally, an overview of the methodology of dE/dx measurements at the CMS and ATLAS experiments is given.

Motivation & Objective

• To search for long-lived particles beyond the Standard Model that exhibit high ionization losses due to large charge or low velocity.
• To develop and apply $dE/dx$ measurement techniques in CMS and ATLAS tracking systems to identify highly-ionizing particles.
• To set model-dependent and model-independent cross-section upper limits for long-lived particles such as multi-charged particles and magnetic monopoles.
• To improve background estimation using data-driven methods in regions with low expected signal, leveraging $dE/dx$ and cluster energy dispersion.
• To extend sensitivity to new physics scenarios not accessible by conventional searches, particularly those with suppressed couplings or conserved quantum numbers.

Proposed method

• Utilized the harmonic-2 estimator $I_h = \left( \frac{1}{N} \sum_{i=1}^{N} (\Delta E_i / \Delta x_i)^{-2} \right)^{-1/2}$ to compute the most probable $dE/dx$ from silicon tracker measurements in CMS.
• Applied the Asymmetric Smirnov-Cramér-von Mises discriminator to assess compatibility of $dE/dx$ measurements with minimally ionizing particles (MIPs), enabling suppression of SM backgrounds.
• In ATLAS, used $f^{\text{HT}}_{\text{TRT}}$ and $N^{\text{HT}}_{\text{TRT}}$ — high-threshold $dE/dx$ and hit counts in the transition radiation tracker — to identify highly ionizing tracks.
• Combined $dE/dx$ with energy dispersion $w$ in electromagnetic calorimeters to enhance discrimination between signal and background events.
• Employed control regions with defined $w$ and $f^{\text{HT}}_{\text{TRT}}$ bounds to predict background yields in signal regions using data-based estimation.
• Used Drell-Yan production models with modified electromagnetic couplings to interpret results and set upper limits on cross sections for various particle types and charges.

Experimental results

Research questions

• RQ1 Can long-lived, highly-ionizing particles with multiple charges be detected via their enhanced $dE/dx$ in the CMS and ATLAS tracking systems?
• RQ2 How effectively can $dE/dx$ measurements discriminate between Standard Model particles and hypothetical long-lived BSM particles with high ionization losses?
• RQ3 What are the sensitivity limits for double-charged particles and magnetic monopoles in $dE/dx$-based searches at 8 and 13 TeV?
• RQ4 How do data-driven control regions using $w$ and $f^{\text{HT}}_{\text{TRT}}$ improve background estimation in low-background, high-signature regions?
• RQ5 To what extent do $dE/dx$-based searches complement conventional BSM searches at the LHC?

Key findings

• Double-charged particles are excluded up to a mass of 650 GeV in the CMS search for long-lived charged particles.
• The ATLAS search for magnetic monopoles with $g_D = 1.0$ excludes masses up to 1340 GeV for spin-1/2 particles, based on 8 TeV data.
• A model-independent fiducial cross-section upper limit of 0.5 fb is set for magnetic monopoles, valid in regions of nearly constant selection efficiency.
• The $dE/dx$-based search strategy successfully predicts background yields in signal regions with $0.013 \pm 0.002$ and $0.026 \pm 0.003$ events for $z=2$ and $z \geq 3$ selections, respectively, with zero observed events.
• The harmonic-2 estimator and likelihood discriminator in CMS provide robust, calibrated $dE/dx$ measurements that suppress tail bias from Landau-distributed energy deposits.
• The combination of $w$ and $f^{\text{HT}}_{\text{TRT}}$ in ATLAS enables effective discrimination between highly ionizing signals and SM backgrounds, especially in low-momentum regimes.