Data-driven model-independent searches for long-lived particles at the LHC

Neutral long-lived particles (LLPs) are highly motivated by many beyond the Standard Model scenarios, such as theories of supersymmetry, baryogenesis, and neutral naturalness, and present both tremendous discovery opportunities and experimental challenges for the LHC. A major bottleneck for current LLP searches is the prediction of Standard Model backgrounds, which are often impossible to simulate accurately. In this paper, we propose a general strategy for obtaining differential, data-driven background estimates in LLP searches, thereby notably extending the range of LLP masses and lifetimes that can be discovered at the LHC. We focus on LLPs decaying in the ATLAS muon system, where triggers providing both signal and control samples are available at LHC run 2. While many existing searches require two displaced decays, a detailed knowledge of backgrounds will allow for very inclusive searches that require just one detected LLP decay. As we demonstrate for the $h\to XX$ signal model of LLP pair production in exotic Higgs decays, this results in dramatic sensitivity improvements for proper lifetimes $\gtrsim 10\mathrm{m}$. In theories of neutral naturalness, this extends reach to glueball masses far below the $\overline{b}b$ threshold. Our strategy readily generalizes to other signal models and other detector subsystems. This framework therefore lends itself to the development of a systematic, model-independent LLP search program, in analogy to the highly successful simplified-model framework of prompt searches.

Figure 1

Schematic representation of signal and background events that pass the signal and orthogonal triggers. Dashed lines indicate invisible or undetected particles. All regular detector objects (prompt leptons, jets, etc.) that are produced in association with the LLP(s) are referred to as associated objects.

Figure 2

Schematic representation of the four regions A, B, C, D into which the iso-region and the non-iso-region events are divided. By construction, region A is significantly enriched with BSM signal events compared to region C.

Figure 3

Simulated limits on $\mathrm{Br}(h\to XX)$ as a function of $X$ lifetime for $m_X=10$, 25, 40 GeV at the 8 and 13 TeV LHC from a search that requires two DVs in the ATLAS muon spectrometer, in analogy to [33]. This result serves as a baseline against which to compare our projections for a data-driven search requiring just a single DV.

Figure 4

Distributions of events with one DV in iso region before defining a ${\mathrm{SR}}_Y/{\mathrm{CR}}_Y$ split, with $\sqrt{s}=13\mathrm{TeV}$ and $m_X=25\mathrm{GeV}$. $p_T(J_1)$ refers to the highest jet $p_T$ in the event. First column: BSM events with short lifetime (56 cm). Second column: BSM events with long lifetime (56 m). Third column: QCD background events under pessimistic assumption ($p_T^{\min }=0\mathrm{GeV}$). Fourth column: QCD background events under optimistic assumption ($p_T^{\min }=120\mathrm{GeV}$). ${\sigma }_{\mathrm{eff}}$ is the effective signal cross section of events with at least one detected DV in the muon system, after geometric, trigger and detection efficiencies are taken into account [setting $\mathrm{Br}(h\to XX)=1$]; see Sec. 3. ${\sigma }_{\mathrm{QCD}}$ is the cross section of SM QCD background in the shown kinematic region as estimated using the methods outlined in Sec. 3.

Figure 5

Limit projections for the data-driven search for a single DV in the ATLAS muon spectrometer (solid lines) compared to an assumed background-free search for two DVs (dashed lines). For comparison with existing limits [33], the top row shows limits that may have been achieved by performing this search at LHC run 1. The left (right) column corresponds to the pessimistic (optimistic) choice of QCD background, both normalized to give 3000 background events at LHC run 1.

Figure 6

$H_T^{\prime }$ distributions, at the 13 TeV LHC with $30{\mathrm{fb}}^{-1}$ of luminosity, in signal region A. For the pessimistic QCD case with $p_T^{\min }=0\mathrm{GeV}$ and an additional $H_T^{\prime }>80\mathrm{GeV}$ cut (top) and for the optimistic QCD case with $p_T^{\min }=120\mathrm{GeV}$ (bottom). The mass of the LLP is $m_X=25\mathrm{GeV}$, with a short lifetime of $c\tau =56\mathrm{cm}$ on the left and a long lifetime of $c\tau =56\mathrm{m}$ on the right. Green line, $h\to XX$ signal; red line, QCD background prediction from ${\mathrm{CR}}_Y$, including BSM contamination of region C. Dashed red line, QCD background prediction if there were no BSM contamination in region C. Purple dotted line, truth-level QCD in region A. Gray shading indicates the $2\sigma$ uncertainty in the SM prediction from limited region C statistics, which is significant in the lowest $H_T^{\prime }$ bin in the optimistic QCD case.

Figure 7

Partial phase space of FTH model with $y_b^{\prime }=y_b^{\mathrm{SM}}$. Blue background shading: YEGP phase of the FTH model. Brown shading, and other areas with $m_0>40\mathrm{GeV}$: More complicated exotic Higgs decay scenarios that will be explored in [64]. Blue contours: Proper lifetime ${\log }_{10}(c\tau /\text{meter})$ of $0^{++}$ glueballs. Purple contours: ${\log }_{10}$ of the perturbatively calculated exotic Higgs branching ratio to mirror gluons via intermediate mirror bottoms. Below the horizontal black line the intermediate state can be multiple mirror bottomonia, while above the black line it can be an excited quirky bound state of two mirror bottoms which annihilates to glueballs.

Figure 8

Colored contours: Reach of various proposed searches for LLPs in the YEGP phase of the FTH model (blue background shading, see Fig. 7) at the LHC with 300 (left) and $3000{\mathrm{fb}}^{-1}$ (right) of luminosity, using results from [28] and the limits for searches in the MS from this paper. Solid (dashed) contours indicate optimistic (pessimistic) assumptions for the number of $0^{++}$ glueballs produced; see the text. The magenta line shows the reach of Higgs coupling constraints for that luminosity. Projections above the black dot-dashed line should be treated with caution due to possible nonperturbative suppressions of the exotic Higgs decay branching ratio. Similar suppressions could occur for some glueball masses above 40 GeV.