DISPLACED PHYSICS AT THE LHC Eric Kuflik Cornell University with Csaba Csaki (Cornell) Salvator Lombardo (Cornell) Oren Slone (Tel Aviv) Tomer Volansky (Tel Aviv)
OUTLINE • Motivation ✴ Dynamical R-Parity Violation — displaced LSP ✴ Twin Higgs models — displaced Higgs decays • Bounds on models
DYNAMICAL R-PARITY VIOLATION C. Csaki, EK, T. Volansky, Phys.Rev.Lett. 112 (2014) 131801 [arXiv:1309.5957] C. Csaki, EK, O. Slone, T. Volansky, JHEP 1506 (2015) 045 [arXiv:1502.03096]
NO SUPERSYMMETRY? There must be supersymmetry! Naturalness Grand Unification Dark Matter Is supersymmetry natural? Bounds typically assume large missing-energy electron muon � R-parity violation to jet the rescue!? jet MET
R-PARITY CHEAT SHEET • Renormalizable baryon and lepton number violating Proton Decay operators allowed: e + LQ ¯ u ¯ d ¯ W = LL ¯ d + ¯ d Q Q π 0 • Small and hierarchical in order to not generate K 0 p ¯ Q proton decay, di-nucleon decay, flavor violation, etc… d ˜ • Impose R-Parity d u ¯ L u, ¯ u, ¯ ( Q, ¯ d, L, ¯ e ) → − ( Q, ¯ d, L, ¯ e ) � ( H u , H d ) → ( H u , H d ) • SM particles are even / superpartners are odd ˜ f ˜ λ • Every vertex contains an even number of super partners • LSP cannot decay - every event has missing f energy
R-PARITY VIOLATION Proton Decay R-Parity conservation is not required Q Q π 0 • If broken, the breaking must be small K 0 p ¯ Q d Perhaps… ˜ d u ¯ L • Baryon or lepton number is exact - why? • Couplings are hierarchal - some large, FCNC some small ¯ d α d β † j i ˜ L k • Approximate symmetry of the MSSM, but it may be strongly broken elsewhere Q α Q β † i j
RPV OPERATORS Which RPV terms are largest in the visible sector? • Standard RPV operators? e + λ 0 LQ ¯ d + λ 00 ¯ u ¯ d ¯ O hRPV = λ LL ¯ d O hBL = µ 0 LH u If we know R-parity violating operators are small, why do we only consider renormalizable ones?
RPV OPERATORS Which RPV terms are largest in the visible sector? • Standard RPV operators? e + λ 0 LQ ¯ d + λ 00 ¯ u ¯ d ¯ O hRPV = λ LL ¯ d � O hBL = µ 0 LH u � • Higher order holomorphic operators? hRPV = κ 00 LH u H d H u O d5(1) O d5(2) hRPV = ρ H d QQQ + ρ 0 H d Q ¯ u ¯ e � • Non-holomorphic operators? d † + η 0 Q ¯ uL † + η 00 QQ ¯ d † + κ ¯ eH d H † e ¯ O nhRPV = η ¯ u ¯ u O nhBL = κ 0 L † H d
DYNAMICAL R-PARITY VIOLATION R SM messengers h S i M • R-Parity is broken in (a hidden sector) by field S • In the low energy EFT, S is the spurion of R-Parity breaking • Charge of S under U(1) B-L and U(1) R will determine leading operators
LEADING OPERATOR? U(1) B − L U(1) R O hRPV , O hBL − 1 3 / 2 O d5(1) 1 5 / 2 hRPV O d5(2) − 1 7 / 2 hRPV O nhRPV , O nhBL 1 1 / 2 e + λ 0 LQ ¯ d + λ 00 ¯ u ¯ d ¯ O hRPV = λ LL ¯ d O hBL = µ 0 LH u hRPV = κ 00 LH u H d H u O d5(1) O d5(2) hRPV = ρ H d QQQ + ρ 0 H d Q ¯ u ¯ e d † + η 0 Q ¯ uL † + η 00 QQ ¯ d † + κ ¯ eH d H † e ¯ O nhRPV = η ¯ u ¯ u O nhBL = κ 0 L † H d
LEADING OPERATOR? U(1) B − L U(1) R U(1) B − L U(1) R S 1 1 / 2 O hRPV , O hBL − 1 3 / 2 O d5(1) 1 5 / 2 hRPV O d5(2) − 1 7 / 2 hRPV O nhRPV , O nhBL 1 1 / 2 W dRP V = S K dRP V = S ∗ M 2 O nhRPV + S ∗ M O hRPV + S O hBL , M O nhBL Leading operators: e + λ 0 LQ ¯ d + λ 00 ¯ u ¯ d ¯ O hRPV = λ LL ¯ d O hBL = µ 0 LH u
LEADING OPERATOR? U(1) B − L U(1) R U(1) B − L U(1) R S − 1 − 1 / 2 O hRPV , O hBL − 1 3 / 2 O d5(1) 1 5 / 2 hRPV O d5(2) − 1 7 / 2 hRPV O nhRPV , O nhBL 1 1 / 2 S M 2 O nhRPV + S S M 2 O d 5(1) W dRP V = hRPV , K dRP V = M O nhBL Leading operators: O d5(1) hRPV = ρ H d QQQ + ρ 0 H d Q ¯ u ¯ e d † + η 0 Q ¯ uL † + η 00 QQ ¯ d † + κ ¯ eH d H † e ¯ O nhRPV = η ¯ u ¯ u O nhBL = κ 0 L † H d
LEADING OPERATOR? U(1) B − L U(1) R U(1) B − L U(1) R S − 1 − 1 / 2 O hRPV , O hBL − 1 3 / 2 O d5(1) 1 5 / 2 hRPV O d5(2) − 1 7 / 2 hRPV O nhRPV , O nhBL 1 1 / 2 S M 2 O nhRPV + S S M 2 O d 5(1) W dRP V = hRPV , K dRP V = M O nhBL Leading operators: Not the standard RPV terms
THE KAHLER OPERATORS ✓ ◆ d 4 θ S ⇤ Z k + 1 d † u j L † d † e j ¯ ijk Q i Q j ¯ k + η 0 2 η 00 η ijk ¯ u i ¯ ijk Q i ¯ k M 2 Different helicity and flavor structure compared to standard operators e + LQ ¯ u ¯ d ¯ LL ¯ d + ¯ d
THE KAHLER OPERATORS ✓ ◆ d 4 θ S ⇤ Z k + 1 d † u j L † d † e j ¯ ijk Q i Q j ¯ k + η 0 2 η 00 η ijk ¯ u i ¯ ijk Q i ¯ k M 2 Different helicity and flavor structure compared to standard operators e + LQ ¯ u ¯ d ¯ LL ¯ d + ¯ d Different helicity structure All superfields can be the same flavor A neutrino-top-stop interaction
THE KAHLER OPERATORS ✓ ◆ d 4 θ S ⇤ Z k + 1 d † u j L † d † e j ¯ ijk Q i Q j ¯ k + η 0 2 η 00 η ijk ¯ u i ¯ ijk Q i ¯ k M 2 All operators (with just one scalar) are chirally suppressed or suppressed by SUSY breaking If SUSY breaking effects are absent, operators are automatically hierarchal and suppressed for light flavors!
Instead of only considering � e + LQ ¯ u ¯ d ¯ � W = λ LL ¯ d + ¯ d Let’s also consider K = 1 d † + Q ¯ uL † + QQ ¯ e ¯ d † � � u ¯ ¯ M
LSP DECAYS STOP EXAMPLE • Stop can decay via the operator K = 1 M QQ ¯ d † � • Will be chirally suppressed ( ∝ m d,s,b ) • Can decay into displaced anti-bottoms ◆ 2 � 2 ◆ ✓ M 2 / h S i � ✓ 300 GeV 1 � � c τ ˜ b ' 10 cm t ! ¯ b ¯ � � 10 9 GeV m ˜ η 00 � � t 333
SOME MORE… ˜ t ¯ ˜ ¯ b b t d ∗ : QQ ¯ ¯ b ¯ b ˜ t ¯ ¯ ˜ t ν ν : uL ∗ Q ¯ t t
SUMMARY #1 • RPV operators suppressed by messenger scale, light fermion masses, and/or SUSY breaking • Operators also break flavor symmetries, may have additional flavor suppression • LSP likes to decay to 3rd generation particles, and in much of the parameter space it is displaced τ LSP ∼ 1 mm − 1 km and sometimes prompt and sometimes collider stable…
NEUTRAL NATURALNESS
UN-COLORED TOP PARTNERS Do top partners need to be colored? y 2 16 π 2 N c Λ 2 t ∼ • To cancel quadratic divergence (at one-loop) • Need to relate the × N c × N c couplings • Need 3 colors of top partners image: Roni Harnik
THE TWIN HIGGS A full copy of the SM SM` SM Z 2
MODELS • Many UV completions, Folded SUSY, Quirky Little Higgs, Holographic Twin Higgs, Orbifold Higgs, … • Phenomenology depends strongly on the details. • “Fraternal Twin Higgs”— only 3rd generation partners N. Craig, A. Katz, M. Strassler, and R. Sundrum, JHEP 1507 (2015) 105 [arXiv:1501.05310] • Lightest partner will be glueballs of QCD’ (similar in Folded SUSY)
DISPLACED HIGGS The SM Higgs can decay into mirror glueballs Twin Mirror Glueball Higgs Mirror x Gluon SM Mirror Glueball Mirror Higgs Top Which then decay back to the SM Twin Higgs SM x Mirror Glueball Particles SM Mirror Higgs Mirror Top Gluon
DISPLACED SEARCHES AT THE LHC
CMS DISPLACED DIJET Search for long-lived neutral particles decaying to quark- antiquark pairs in proton-proton collisions at sqrt(s) = 8 TeV • Search for 2 displaced jets jet with p T > 60 GeV Secondary Vertex jet • Important cuts • H T > 325 GeV (trigger) • m DV > 4 GeV (no b’s, detector interactions) p p • N tracks > 4, 5 • At most one prompt (IP <0.5 mm) track per jet Primary Vertex • Dijet consistent with DV
ATLAS DV + muon/e /jets/MET Search for massive, long-lived particles using multitrack displaced vertices or displaced lepton pairs in pp collisions at sqrt(s) = 8 TeV with the ATLAS detector • Search for displaced vertex μ /e with • N tracks > 5 • m DV > 10 GeV jets / MET • Trigger/cut on associated 0.5 mm object • muon, p T > 55 GeV p p 2 mm • electron, p T > 125 GeV • MET > 180 GeV • Jets 4, 5 or 6, p T > 65, 60, 55 GeV
CONSTRAINTS ON RPV C. Csaki, E. Kuflik, S. Lombardo, O. Slone, T. Volansky JHEP 1508 (2015) 016 [arXiv:1505.00784]
RPV SCENARIOS • Looked at cases motivated by naturalness • Light stops, gluinos, and higgsinos • Only considered direct production of LSPs
RECAST • Feynrules → Madgraph → Pythia → Delphes • LSP displaced by writing proper lifetimes to LHE file (VTIMUP column) • Stop and gluino hadronize (R- hadrons supported by Pythia 8) • Displaced R-hadrons typically enhance DV reconstruction eff.
RECAST • Use displaced tracking efficiency parametrized by track IP and L xy • Applied cuts and vertex reconstruction procedure for ATLAS + CMS displaced vertex searches • ATLAS: vertex tracks with IP > 1.5 mm, merge truth-level vertices within 1 mm of each other • CMS: reconstruct displaced dijet using track information • Typically reproduce efficiencies for benchmark models within 20% • Recast HSCP at parton level • Prompt searches applied directly
Recommend
More recommend
