TL;DR
This paper proposes a new model for dark matter that connects it to the puzzling B-physics anomalies observed in particle physics experiments. The model predicts a colored "partner" particle for dark matter that could be discovered at the Large Hadron Collider (LHC), potentially opening a new window into dark matter detection.
What Problem Does This Paper Address?
The Standard Model of particle physics, while incredibly successful, has two major shortcomings: it doesn't explain the existence of dark matter, and it struggles to account for certain anomalies observed in the decays of B mesons (known as B-physics anomalies). These anomalies hint at new particles and forces beyond the Standard Model, possibly involving a TeV-scale leptoquark that interacts with third-generation quarks and leptons. This paper aims to bridge these two mysteries by proposing a dark matter model that is naturally linked to such a leptoquark, and then exploring how this model could be tested at the LHC.
Methods
The authors construct a "simplified model" that incorporates a Majorana fermion as the dark matter candidate and a colored fermion as its coannihilating partner. This dark sector is connected to the Standard Model via a U₁ vector leptoquark. They first calculate the "dark matter relic surface," which describes the range of parameters (like particle masses) that would produce the observed abundance of dark matter in the early universe through a process called thermal freeze-out, heavily influenced by the coannihilation partner.
The core of their experimental investigation lies in analyzing the collider signatures of the colored coannihilation partner. They predict that this partner will hadronize into "psionium" and "open-psi" bound states, similar to quarkonia in the Standard Model. They then "recast" existing LHC searches (dilepton and monojet searches) to constrain the model's parameter space. Crucially, they propose a novel, dedicated search strategy for the pair production of the colored partner, focusing on specific decay channels that are expected to be sensitive to the compressed mass spectrum predicted by their model. This involves developing a multivariate analysis (MVA) technique using a Boosted Decision Tree (BDT) classifier.
Datasets
The paper does not introduce new datasets. Instead, it utilizes:
- Simulated event samples: Generated using Monte Carlo event generators (MadGraph5, Pythia8) and detector simulation (Delphes3) for both signal processes (based on their model) and background processes (primarily tt production).
- Existing experimental data: Recasting of results from published ATLAS and CMS searches for leptoquarks, dilepton resonances, and monojets.
Key Insights and Ideas
- Unified explanation for dark matter and B-physics anomalies: The model elegantly connects two seemingly disparate puzzles in particle physics by proposing a dark sector that is intrinsically linked to the leptoquark responsible for the B-physics anomalies.
- Hadronization of colored dark sector particles: The prediction that colored dark matter partners hadronize into observable bound states (like psionium) is a novel and potentially powerful signature for dark matter searches.
- Compressed spectrum as a challenge and opportunity: The predicted near-degeneracy in mass between dark matter and its colored partner makes direct dark matter detection difficult but opens up unique collider signatures.
- RT ratio as a powerful discriminator: The ratio of missing transverse energy to visible transverse momentum (RT) is identified as a key observable for distinguishing signal events from background in scenarios with compressed spectra and invisible particles.
- Multivariate analysis significantly outperforms cut-based methods: The BDT classifier, trained on a combination of low-level and high-level observables, demonstrates a substantial improvement in signal-background discrimination compared to simpler cut-based analyses.
Key Results
- Exclusion of parameter space: Recasting existing dilepton and monojet searches excludes coannihilation partner masses below 280 GeV and 400 GeV, respectively.
- Reach of proposed search: The proposed dedicated search strategy is expected to probe dark matter masses up to around 600 GeV with current LHC luminosity.
- BDT performance: The multivariate analysis using a BDT classifier significantly outperforms cut-based strategies, achieving a factor of two improvement in background rejection.
- Relic abundance constraints: The model can reproduce the observed dark matter relic abundance for a wide range of leptoquark and dark matter masses, provided a specific mass splitting between the dark matter and its colored partner is maintained.
Limitations and Future Work
- Model dependence: The results are based on a simplified model. A full UV completion would be needed for a more complete picture.
- Background uncertainties: The analysis relies on MC simulations for background estimation, and acknowledges that reducible backgrounds (e.g., fake taus) and uncertainties in mistagging efficiencies require more detailed study.
- Trigger requirements: For processes with soft final state objects, the trigger requirements at the LHC need careful consideration, and the possibility of triggering on the proposed observable R_T could be investigated.
- Experimental validation: The authors emphasize that a more detailed analysis by LHC collaborations would provide a more robust estimation of the limits.
- Future searches: Including other signal categories (e.g., two hadronic taus) and exploring the impact of improved background estimations are suggested future directions.
Summary updated