La scuola di Atene Raffaello Sanzio Highlights from

La scuola di Atene, Raffaello Sanzio Highlights from LHCb Diego Martínez Santos (Universidade de Santiago de Compostela) On behalf of LHCb Diego Martinez Santos, ICNFP 2018 1

Introduction • Flavour Physics is a powerful tool to access NP beyond the reach of direct production at accelerators (G. Isidori, KAON’ 16) • Most notably in searches for sources of FV/CPV other than CKM • LHCb • Forward detector at the LHC • Very precise tracking and PID • Physics goals • Study of CP violation and rare decays of ‘light’ particles (b, c, s, τ. . . ) • But also: direct searches for BSM, EW physics, soft QCD… FPCP 2017 2

Introduction • Flavour Physics is a powerful tool to access NP beyond the reach of direct production at accelerators (G. Isidori, KAON’ 16) • Most notably in searches for sources of FV/CPV other than CKM • LHCb will upgrade to continue improving our knowledge of Flavour Physics • We speak about two upgrade phases: • Phase-I (2021 -2029): Approved, TDR’s ready, in construction FPCP 2017 CERN-LHCC-2014 -016 CERN-LHCC-2014 -001 CERN-LHCC-2013 -022 CERN-LHCC-2013 -021 3

Introduction • Flavour Physics is a powerful tool to access NP beyond the reach of direct production at accelerators (G. Isidori, KAON’ 16) • Most notably in searches for sources of FV/CPV other than CKM • LHCb will upgrade to continue improving our knowledge of Flavour Physics • We speak about two upgrade phases: • Phase-I (2021 -2029): Approved, TDR’s ready, in construction • Phase-II (2031 ->2035): Expression of Interest published FPCP 2017 4

Timescale Run-I : • 2011: 1 fb-1 @ 7 Te. V • 2012: 2 fb-1 @ 8 Te. V Run-II: • 2015: 0. 3 fb-1 @13 Te. V • 2016: 1. 7 @ 13 Te. V • 2017: 1. 7 @ 13 Te. V • 2018: ~0. 8 FPCP 2017 5

Bs μμ • Bs μμ is (still) a golden modes in NP searches w/ 4. 4 fb-1 : Δχ2 LHCb, ar. Xiv: 1703. 05747 PRL 118, 191801 (2017) SM C. Bobeth et al. , ar. Xiv: 1311. 0903. LHC + LEP + g-2 + Ωh 2 +… E. Bagnaschi et al. ar. Xiv: 1612. 05210 • Deviations from SM allowed despite LHC bounds on NP • Expected precision σ(ΒR(Bs μμ)), 50 fb-1 ~ 2. 3 x 10 - LHC + LEP + g-2 + Ωh 2 +… 10 σ(ΒR(Bs μμ)), 300 fb-1 ~ 1. 6 x 1010 • The effective Bs lifetime in Bs μμ is also sensitive to NP for precisions of percent Δχ2 50 fb-1 300 fb-1 M. Lucio σ(τ), 50 fb-1 ~ 5% σ(τ), 300 fb-1 ~ 2% FPCP 2017 6

Bs μμ/ Bd μμ B. Dutta and Y. Mimura ar. Xiv: 1501. 02044 Current • The ratio of BR’s Bd/Bs is considered a golden mode for testing new sources of flavour violation • LHCb upgrade can reach <30% precision (stat. limited) Susy unified models Consistent with kaon oscillation future 1 F. Teubert, Rn. SII 0. 9 σ(R)/R • Phase-II could achieve ~10% (…and still stat limited) 0. 8 LHCb CMS 0. 7 0. 6 0. 5 0. 4 0. 3 0. 2 0. 1 0 2015 2020 2025 2030 2035 2040 Year FPCP 2017 7

Bs mixing phase, Φs Thanks to O. Leroy CPV in interference mixing-decay: Which in b cc decays corresponds to the mechanical phase of the Bs mixing. In SM: Diego Martinez Santos, ICNFP 2018 8

Bs mixing phase, Φs • Neutral meson mixing (ΔF = 2 transitions) are extremely sensitive to new source of flavour violation. eg, for O(1) couplings bounds on ΛNP at 100 – 105 Te. V • eg, in SUSY the B mesons can mix via gluino box Phase I upgrade, 50 fb-1 < 0. 005 rad @ 300 fb-1 Phase II upgrade, 300 fb-1 Diego Martinez Santos, ICNFP 2018 9

LHCb-CONF-2018 -002 The CKM angle γ γ 10

Bd K*μμ angular observables JHEP 02 (2016) 104 +… P 5’ is one of the coefficients of the differential decay rate of the helicity angles, designed to have reduced TH uncertainties 2 bins show a ~3σ deviation (each) w. r. t SM prediction(s) But this significance can be diluted by hadronic effects Diego Martinez Santos, ICNFP 2018 11

Lepton Universality Violation in b sll JHEP 08 (2017) 055 LHCb observed some interesting deviations in the ratios Full Run-I dataset ar. Xiv: 1705. 05802 2. 3σ 2. 2σ, 2. 4σ Can be “explained” by similar BSM than P 5’ Diego Martinez Santos, ICNFP 2018 12

LUV in semileptonic b decays LHCb, 3 fb-1 Full Run-I dataset R(D*) leptonic tau: 0. 336� 0. 027 � 0. 030 PRL 115 (2015) 111803 R(D*) hadronic: 0. 291� 0. 019 � 0. 026 � 0. 013 PRL 120 (2018) 171802 Overall (WA) ~4σ deviation w. r. t SM prediction in the R(D) vs R(D*) plane, towards larger values Check it with larger stats, as well as in other channels Recent measurement on Bc decays Diego Martinez Santos, ICNFP 2018 13

LUV in semileptonic b decays • • PRL 120, 121801 (2018) LHCb analysis using leptonic τ decays 3 D template fit: missing mass, estimated lifetime with kinematic constraints, and a categorical variable Z (bins of missing transverse momentum and q 2) Full Run-I dataset Consistent with expectations from R(D*) Within less than 2σ from SM (0. 25 -0. 28) Also first evidence for Bc J/ψτν Diego Martinez Santos, ICNFP 2018 14

Σ→pμμ full Run-I analysis • PRL 120, 221803 (2018) Motivation: Hyper. CP experiment found evidence for Σ pμμ decays, all with the dimuon mass of ~ 214 Me. V, which suggested the possibility of a light dimuon resonance LHCb can check this • X→ μμ : Found no evidence for the 214 Me. V particle. Upper limit for the resonant channel: 1. 4 x 10 -8 @95% CL Consistent with SM as well as with Hyper. CP 15

LFV: Β(s) eμ JHEP 1803 (2018) 078 • Run-I dataset • Used BDT to discriminate from background • Invariant mass fit in different bremsstrahlung categories • Without photon recovery • With photon recovery Diego Martinez Santos, ICNFP 2018 16

LFV: Β(s) eμ • • JHEP 1803 (2018) 078 Run-I dataset Used BDT to discriminate from background Invariant mass fit in different bremsstrahlung categories Normalization of the signal yield to B channels with known BR: B+ J/ψK+, Bd Kπ Diego Martinez Santos, ICNFP 2018 17

LFV: Β(s) eμ JHEP 1803 (2018) 078 • • Run-I dataset Used BDT to discriminate from background Invariant mass fit in different bremsstrahlung categories Normalization of the signal yield to B channels with known BR: B+ J/ψK+, Bd Kπ • Upper limit set with CLs World best upper limits: Diego Martinez Santos, ICNFP 2018 18

LFV: D eμ, τ μμμ JHEP 02 (2015) 121 PLB 754(2016)167 Other results in LFV • ΒR(τ 3μ) < 4. 6 (5. 6) x 10 -8 @ 90 (95)% CLJHEP 02 (2015) 121 • Competitive with Belle (2. 1 x 10 -8 @ 90%CL) PLB 687 (2010) 139 • BR(D eμ) < 1. 3 (1. 6) x 10 -8 @90 (95)% CL (world best)PLB 754(2016)167 ΒR(τ 3μ) < 4. 6 (5. 6) x 108 @ 90 (95)% CL BR(D eμ) < 1. 3 (1. 6) x 10 -8 @90 (95)% CL Full Run-I dataset Diego Martinez Santos, ICNFP 2018 19

Dark photons PRL 120 (2018) no. 6, 061801 • Search for dark photons (A’), that mix to offshell photons shared production mechanism normalization to DY • Uses dedicated “turbo” trigger: writes on tape only the relevant part of the event Turbo trigger: Write on tape the relevant information of the event as measured online: + Advantage: Little event size afford higher output rate and hence efficiency + Disadvantage: Give up offline reconstruction quality, analysis reoptimizations Diego Martinez Santos, ICNFP 2018 20

Dark photons PRL 120 (2018) no. 6, 061801 • Search for dark photons (A’), that mix to offshell photons shared production mechanism normalization to DY • Uses dedicated “turbo” trigger: writes on tape only the relevant part of the event • 1. 6 fb-1 of Run-II (13 Te. V) data • Prompt search up to m. Z • Displaced search up to 350 Me. V Diego Martinez Santos, ICNFP 2018 21

BSM bosons from gg μμ LHCb-PAPER-2018 -008 ar. Xiv: 1805. 09820 • Search for a prompt dimuon resonance in the full Run-I data (supersedes preliminary result of Moriond EW) • Statistically independent from the Dark Photon search can be combined by recast searches) • Search inspired in NMSSM, but done in bins of η, pt to easy interpretations in different models • Good mass resolution of LHCb allows to explore closely the Y region • No signal found, upper limits in σ x BR set (CLs). ~ 5 pb level Diego Martinez Santos, ICNFP 2018 22

Conclusions • Presented results on beauty, charm, strange and tau decays • Also BSM direct searches • The Hyper. CP anomaly seems to vanish • On the other hand, interesting deviations in LFU tests Diego Martinez Santos, ICNFP 2018 23

Conclusions Overall the Standard Model may look impressive, with millions of successful measurements on its side… Diego Martinez Santos, ICNFP 2018 24

Conclusions Overall the Standard Model may look impressive, with millions of successful measurements on its side… … but we are not afraid Diego Martinez Santos, ICNFP 2018 25

BACKUP Diego Martinez Santos, ICNFP 2018 26

Lepton Flavour Violation • LFV forbidden in SM, but allowed in BSM positive signal is compelling NP discovery F. Teubert, Rn. SII 25 fb-1 50 ab-1 10 BR(τ 3μ)x 10 -9 50 fb-1 50 ab-1 LHCb Belle-II 300 fb-1 50 ab-1 nil bkg • LHCb searches for LFV LS 2 1 2015 2020 LS 3 2025 LS 4 2030 2035 2040 Year Several world best results in hadron decays τ 3μ Phase-II competitive w/ Belle II B Keμ expected O(10 -9) for Phase - II Diego Martinez Santos, ICNFP 2018 27

LFV: τ μμμ • • • JHEP 02 (2015) 121 Run-I dataset Used BDT to discriminate from background Classification of events in bins of 2 BDT’s and μμμ mass Normalization of the signal yield to Ds � ( μμ)π Background estimate from sidebands Limit set with CLs ΒR(τ 3μ) < 4. 6 (5. 6) x 10 -8 @ 90 (95)% CL Competitive with Belle (2. 1 x 10 -8 @ PLB 687 (2010) 139 90%CL) LHCb Upgrade (s) will explore the 10 -9 region Diego Martinez Santos, ICNFP 2018 28

Hidden sector bosons from flavour decays PRL 115 161802 (2015) PRD 95, 071101(R) (2017 • Searches for BSM bosons (χ) in b sμμ transitions performed in full Run-I data • Provide sensitivity in the region ~2 mμ < mχ < ~4. 7 Ge. V • Lifetimes up to 1000 ps • Upper limits for B Kχ(μμ) and B K*χ(μμ) are in the 1010 – 10 -7 range, depending on the dimuon mass and the χ lifetime (analysis reoptimized) J/ψ Ψ(2 S), Ψ(3770) • See talk by Preema Pais • Similar searches possible in s dμμ, eg, Σ pμμ KS ππ misid � Diego Martinez Santos, ICNFP 2018 Ψ(4160) 29

HV pions decaying to di-jets EPJC 77 (2017) 812 • Signature: search for a displaced vertex with two bjets (low efficiency to search for all 4) • This signature is possible in Hidden Valley pions from SM Higgs decay • Di-jet mass fit in bins of transverse displacement • 2 fb-1 of Run-I data, @7 and 8 Te. V. Diego Martinez Santos, ICNFP 2018 30

Trigger system: status and prospects L 0 (Hardware) HLT 1 (Software) HLT 2 (Software) Main bottleneck for K. Can’t be changed LHCb Upgrade HLT (Software) Not designed for K, but flexible. K triggers being implemented ε(2011 -2012) ~ 1 -2% ε(Run-II) improved HLT 1, 2 ~ 18% Maximum possible ~30% (L 0 won’t allow more) (Note: This logo may not be official) ε(Upgrade) ~ 80 -100%? Simulation studies show that rate would be under control V. Chobanova et al, CERN-LHCb-PUB-2016 -017 31

Strangeness transitions MFV paradigm (small) CCKM X ( Non-MFV paradigm SM + BSM ) (small) CCKM X SM + C? X BSM In a scenario in with BSM > few Te. V, s-d transitions are central (SM dynamics has the strongest suppression of all quark transitions) CCKM X SM + C? X BSM Diego Martinez Santos, ICNFP 2018 eg, for c~ 1, bounds from kaon physics go up to a scale of 105 Te. V 32

Bd K*μμ angular observables V. Chobanova et al. ar. Xiv: 1702. 02234 T. Blake Now (yes, the red points have error bars) q 2 Example ~300 fb-1 Moreover, the q 2 dependence will be known with high precision very valuable input to the discussion on NP vs hadronic power corrections to SM Diego Martinez Santos, ICNFP 2018 V. Chobanova et al. ar. Xiv: 1702. 02234 33

Bd K*μμ angular observables Moreover, the q 2 dependence will be known with high precision very valuable input to the discussion on NP vs hadronic corrections to SM CERN-LHCC-2017 -003 … and larger datasets would allow to make finer q 2 binning Diego Martinez Santos, ICNFP 2018 34

Lepton Universality Violation in b sll T. Blake Systematics could be ultimately reduced to ~ 1 -2% T. Hurth et at. , ar. Xiv: 1705. 06274 Diego Martinez Santos, ICNFP 2018 35

LFV: D 0 eμ PLB 754(2016)167 • • Run-I dataset Used BDT to discriminate from background D 0’s are tagged from D*� D 0π�decays Invariant mass fit in different BDT categories, two mass variables: help discriminating against ~irreducible background from D 0 ππ • Normalization of the signal yield to D 0 Kπ • Limit set with CLs World best upper limit: BR(D eμ) < 1. 3 (1. 6) x 10 -8 @90 (95)% CL Diego Martinez Santos, HQL 2018 36

LFV: τ μμμ • • • JHEP 02 (2015) 121 Run-I dataset Used BDT to discriminate from background Classification of events in bins of 2 BDT’s and μμμ mass Normalization of the signal yield to Ds � ( μμ)π Background estimate from sidebands τ μμμ Ds � ( μμ)π Diego Martinez Santos, HQL 2018 37

LFV: τ μμμ • • • JHEP 02 (2015) 121 Run-I dataset Used BDT to discriminate from background Classification of events in bins of 2 BDT’s and μμμ mass Normalization of the signal yield to Ds � ( μμ)π Background estimate from sidebands Limit set with CLs ΒR(τ 3μ) < 4. 6 (5. 6) x 10 -8 @ 90 (95)% CL Competitive with Belle (2. 1 x 10 -8 @ PLB 687 (2010) 139 90%CL) LHCb Upgrade (s) will explore the 10 -9 region Diego Martinez Santos, HQL 2018 38