Studio di decadimenti rari a LHCb · rari a LHCb Matteo Palutan INFN-LNF 100° Congresso SIF, Pisa...
Transcript of Studio di decadimenti rari a LHCb · rari a LHCb Matteo Palutan INFN-LNF 100° Congresso SIF, Pisa...
Studio di decadimenti rari a LHCb
Matteo PalutanINFN-LNF
100° Congresso SIF, Pisa 2014
Matteo Palutan, INFN-LNF
Indirect searches for New Physics
2
- Photon polarization in b→sγ - Branching fractions and angular distributions of b→sμ+μ− - The very rare decays B0(s) → μ+μ−
• Recent LHCb results on FCNC b decays:
• Lepton Flavour Violating decays
OUTLINE
• High energy“real” new particles can be produced and discovered via their decays
Direct and indirect approach complement each other
• High precision“virtual” new particles can be discovered in loop processes, such as Flavor Changing Neutral Current (FCNC)
Matteo Palutan, INFN-LNF
FCNC decays
3
• Contribution from New Physics as correction to the SM
• In the Standard Model FCNC decays are forbidden at tree level, and occur at loop level only → strongly suppressed!
• SM diagrams for b→sμ+μ−: “ electroweak penguins”
What is the scale of ΛNP?What is its coupling cNP? loop tree
Matteo Palutan, INFN-LNF
Three impersonations of the EW penguin
4
αQED suppression
helicitysuppression
“Photon penguin” “Electroweak penguin” “Higgs penguin”
BR~3x10-4 BR~10-6 BR~3.6x10-9
angular distributions BR
BR,γ polarization
Matteo Palutan, INFN-LNF
FCNC processes in effective field theory
5
• Effective Hamiltonian for b→s FCNC transistions
• Wilson coefficients Ci encode short-distance physics and possible NP effects, computed perturbatively
• Local operators Oi with different Lorentz structure absorbe long distance effects
• O´i helicity flipped operators, ms/mb suppressed in SM
photon polarization in b→sγ
Matteo Palutan, INFN-LNF
Photon polarization in b→sγ decays
7
• The SM predicts that the photon emitted in b→sγ decays is predominantly left-handed (up to order ms2/mb2), since the recoil s-quark that couples to a W-boson is left-handed
• Several models beyond the SM predict the photon to acquire a significant right-handed component due to the exchange of heavy fermions in the electroweak penguin loop
[Atwood et al PRL 79 (1997) 185]
[Gronau and Pirjol PRD 66 (2002) 054008]
[Yu et al. JHEP12 (2013) 102]
Matteo Palutan, INFN-LNF
Photon polarization in B+ → K+π−π+γ
8
• Can infer the photon polarization from the up-down asymmetry of the photon direction in the K+π−π+ rest-frame. Unpolarized photons would have no asymmetry (conceptually similar to Mme. Wu experiment)
The up-down asymmetry is defined as:
Where sij = (pi+pj)2 and the ai contain the information about the K+π−π+ resonances and their interference: the odd terms is cosθ carry the photon polarization: if λγ≠0 the photon is polarized. In SM λγ ≈-1 (+1) for radiative B (anti-B) decays, up to O(ms2/mb2) corrections
[Gronau and Pirjol PR
D 66 (2002) 054008]
Matteo Palutan, INFN-LNF9
• There are a large number of overlapping resonances in the m(K+π−π+) mass spectrum. No attempt is made to separate these in the analysis, we simply bin in 4 bins of m(K+π−π+)
Photon polarization in B+ → K+π−π+γ
• At LHCb we look at B+ → K+π−π+γ decays using calorimeter photons, observe ~13k signal candidates in 3 fb-1
Matteo Palutan, INFN-LNF
Up-down asymmetry
10
• Combining the four m(K+π−π+) bins, the up-down asymmetry Aud is different from zero at 5.2σ
• To extract the polarization value from the up-down asymmetry theory input is needed about the amplitude composition of the hadronic system (ongoing)
• This is the first observation of photon polarization in b→sγ decays
[PRL 112 (2014) 161801]
• Alternative approach is being finalized at LHCb: extraction of λγ
from mixing-induced CP violation in B0s → φγ
b→s μ+μ− decays
Matteo Palutan, INFN-LNF
Experimental data on b→sμ+μ−
12
great interest, big effort...
Matteo Palutan, INFN-LNF
Angular analysis of B0→ K*0[→K+π− ]μ+μ−
13
• B0→ K*μ+μ− is the golden mode to test new vector
(-axial) couplings in b→s transitions: sensitivity to O7 , O9 and O10 and their primed counterparts.
• K*→Kπ is self tagged, hence angular analysis ideal to test helicity structure
• Decay described by 3 helicity anges and q2 = m(μ+μ− )
folding technique (φ→φ+π for φ<0 ) applied to reduce the number of fit parameters
FL: fraction of K* longitudinal polarization
AFB: Forward-Backward lepton asymmetry
S3: Asymmetry in K* transverse polariz.
AIm: T-odd CP asymmetry
Matteo Palutan, INFN-LNF
B0→ K*0μ+μ− angular observables (1fb-1)
14
[JHEP 08 (2013) 131]
• SM prediction from [C. Bobeth et al. JHEP 07 (2011) 067]
• Zero crossing point of AFB free from leading FF uncertainties
we measure
• Results based on 900 signal events from 1fb-1 at 7 TeV (angular analysis on 3 fb-1 being finalized)
to be compared with
Matteo Palutan, INFN-LNF
B0→ K*0μ+μ− angular observables (1fb-1)
15
• And fortunately also ATLAS and CMS with ~0.4k candidates in 5 fb-1 start to contribute to this analysis. They particularly competitive at large q2
Matteo Palutan, INFN-LNF
B0→ K*0μ+μ− angular observables (1fb-1)
16
• SM prediction from [Decotes-Genon et al. JHEP 05 (2013) 137]
• Can also apply different angular foldings to access different angular terms: focus on observables where leading form-factor uncertainties cancel, e.g.
• In 1 fb-1, LHCb observes a local (4.3<q2<8.7 GeV2) discrepancy of 3.7σ in P´5 (the probability that at least one bin varies by this much is 0.5%)
[PRL 111 (2013) 191801]
eagerly await for LHCb 3 fb-1 analysis!
Matteo Palutan, INFN-LNF
Understanding the P´5 anomaly?
17
• Several global fits to the available b→sγ and b→sl+l−
have been attempted, which include the observed anomaly => discrepancy with SM is 3-4.5 σ. Fits favour C9NP ~ -1.5 (non-SM vector current), and can also be explained by introducing a flavour-changing Z’ boson ~1-7 TeV
[Decotes-Genon, Matias, Virto ’13Altmannshofer, Straub ’13Gaul, Goertz, Haisch ’14
Beaujean, Bobeth, van Dyk ’13
Jaeger, Camalich ’13Lyon, Zwicky ’14Hofer, Matias ‘14
• By introducing form-factor uncertainties as nuisance parameters in the global fit, the discrepancy wrt SM is reduced to 2σ
• Form-factor uncertainties are being rediscussed, especially in the large recoil region, with appreciable impact on the comparison with data, and different opinions, of course!
(4.3<q2<8.7 GeV2)
Matteo Palutan, INFN-LNF
Understanding the P´5 anomaly?
18
• Several global fits to the available b→sγ and b→sl+l−
have been attempted, which include the observed anomaly => discrepancy with SM is 3-4.5 σ. Fits favour C9NP ~ -1.5 (non-SM vector current), and can also be explained by introducing a flavour-changing Z’ boson ~1-7 TeV
[Decotes-Genon, Matias, Virto ’13Altmannshofer, Straub ’13Gaul, Goertz, Haisch ’14
Beaujean, Bobeth, van Dyk ’13
Jaeger, Camalich ’13Lyon, Zwicky ’14Hofer, Matias ‘14
• By introducing form-factor uncertainties as nuisance parameters in the global fit, the discrepancy wrt SM is reduced to 2σ
• Form-factor uncertainties are being rediscussed, especially in the large recoil region, with appreciable impact on the comparison with data, and different opinions, of course!
(4.3<q2<8.7 GeV2)
looks interesting but we need more data, and better understanding of theoretical uncertainties...
B0s,d → μ+μ−
Matteo Palutan, INFN-LNF
B0(s) → μ+μ− decays
20
• B0(s) → μ+μ− decays are a particular interesting case of EW penguin. The helicity suppression of the vector(-axial) terms, makes these decays very sensitive to new (pseudo-)scalar interactions → Higgs penguin!
e.g. in MSSM, branching fraction scales as
• Predicted precisely in the SM:
BR(B0s → μ+μ−) error budget
• The B0/B0s ratio is also powerful to discriminate among NP models: a deviation from the value predicted by SM, ~(Vtd/Vts)2, would indeed also imply the breaking of the Minimal Flavour Violation hypothesis
[Bobeth et al. PRL 112 (2014) 101801]
Matteo Palutan, INFN-LNF
It’s a long history...with great contributors!
21
Matteo Palutan, INFN-LNF
CMS vs LHCb
22
• Good trigger and muon ID• No hadron PID• Excellent silicon tracking to resolve signal decays in the high pile-up environment• Di-muon mass resolution 32-75 MeV/c2
• CMS: 5+20 fb-1 at 7 and 8 TeV
• Efficient muon trigger• Good muon and hadron PID• Track impact parameter resolution ≤20μm• Luminosity levelling at 4x1032 cm-2s-1
• Di-muon mass resolution 25 MeV/c2
• LHCb: 1+2 fb-1 at 7 and 8 TeV
~1 fb-1 at LHCb is equivalent to ~10 fb-1 at CMS
Matteo Palutan, INFN-LNF
B0(s) → μ+μ− results at EPS2013
23
• During EPS 2013 CMS and LHCb collaborations presented their results based on 25 fb-1 and 3fb-1, respectively
Matteo Palutan, INFN-LNF
Combined CMS and LHCb results at CKM2014
24
• Full simultaneous fit of CMS and LHCb data
• Statistical significance (Wilks’ theorem ):
likelihood profile scan for Bs
6 best S/B bins
Matteo Palutan, INFN-LNF
Comparison with SM
25
• 1D LL scan of the B0/B0s ratio: • 2D LL scan performed for the signal strength BR/BRSM :
theoretical errors included in the fits
• Compatibility with SM:
1.2σ for B0s 2.2σ for B0
• Compatibility with SM (and MFV):
2.3σ for B0/B0s ratio
B0/B0s ratioSignal strength
Matteo Palutan, INFN-LNF
Implications
26
• Latest results on B0(s) → μ+μ− strongly constrain the parameter space for many NP models, complementing direct searches from ATLAS/CMS: in particular, large tanβ with light pseudo-scalar Higgs in CMSSM is strongly disfavored
• Start to explore less constrained (and more “natural”) SUSY models: “split-family” SUSY, pMSSM
[Mamhoudi arXiv:1310.2556]
Barbieri et al. ’14Althmanshofer et al. ’13Mamhoudi ’13
Matteo Palutan, INFN-LNF
More implications
27
B0/B0s ratio fixed by MFV to SM value: it is therefore very relevant to clarify the experimental picture on B0
• Test the MFV hypothesis: flavour breaking in NP model described by CKM matrix → naturally small effects in FCNC observables
Chivukula, Georgi ’87D’Ambrosio, Giudice, Isidori, Strumia ’02
• Model independent constraints:the precision achieved now is such that B0(s) → μ+μ− sensitivity to (Z,γ) penguin cannot longer be considered sub-leading, and starts to compete with the golden mode B0→ K*μ+μ−
Althmanshofer et al. 1308.1501
Charles et al. 1309.2293
Lepton Flavour Violationand exotic searches
Matteo Palutan, INFN-LNF
The quest for lepton flavour violation
29
• The discovery of neutrino oscillations implies charged-LFV at some level. Many extensions of the SM introduce large cLFV effects.
• Important advantage wrt the quark sector: the reach for NP energy scale is not much affected by QCD uncertainties in the SM predictions
• The MEG collaboration at PSI using 3.6x1014 stopped muons reached an amazing sensitivity to μ→ eγ :
[PRL 110 (2014) 201801]
• In principle, τ are more sensitive per event than μ since mass typically decreases GIM suppression, >500;however production rate is much lower.
Matteo Palutan, INFN-LNF
Search for lepton flavour violation in τ→μμμ
30
• Possible as penguin with neutrino oscillation; SM prediction ~10-40, beyond experimental reach
• (some) NP predictions: SUSY ~10-10, mSUGRA+seesaw ~10-9, non universal Z´ ~10-8
• With ~1.4x109 τ at the B-factories the current limits are:
arXiv:1001.3221
• At the LHC τ are copiously produced (mainly from charm decays, Ds→ τν): ~1011 τ/fb-1 (~5x1014 at HL-LHC!).
BR(τ→μμμ)<4.6x10-8 at 90% CL
Belle: BR(τ→μμμ) <2.1x10-8 at 90%CL
BaBar: BR(τ→μμμ) <3.3x10-8 at 90%CL arXiv:1002.4550
Few days ago, LHCb presented at TAU2014 the search based on 3 fb-1
Matteo Palutan, INFN-LNF
Search for Majorana neutrinos
31
• Limit on BR(B- → π+μ−μ−) from 3fb-1 can be translated (with a model-dependent assumption on the decay width) to an upper limit on the coupling between muon and fourth generation neutrino
• In LHCb, heavy Majorana neutrinos can be sought in B- → π+μ−μ− decay, which is forbidden in SM but can proceed via production of on-shell massive neutrinos
B- → π+μ−μ−
• Observation of neutrino oscillations is a strong theoretical motivation for Majorana neutrinos to exist
[PRL 112 (2014) 131802]
decay width fromAtre et al. JHEP 05 (2009) 030+ S. Stone, Z. Xing ’13
Matteo Palutan, INFN-LNF
Search for Majorana neutrinos - implications
32
• Limit on BR can be translated to a model-dependent upper limit on the coupling between muon and fourth generation neutrino
[JHEP 05 (2009) 030]
• With this result LHC join the search for Heavy Neutral Leptons performed all around the world, both at colliders and fixed target experiments.
Conclusions
Matteo Palutan, INFN-LNF
Conclusions
34
• Flavor-changing transitions represent a unique window on physics beyond the SM: there is still a lot to learn and explore
• There are few interesting anomalies, but in general the agreement with the SM is excellent: large NP contributions, O(SM), ruled out in many cases
LHC (and LHCb) is acting as a fantastic flavour-factory
need combined th+exp precision at the few % level
• Interplay between low energy precision measurements and direct searches as strong as ever
• Ambitious program for upgrading integrated luminosity in the next years at LHC: LHCb aims at ~50fb-1
Matteo Palutan, INFN-LNF
Conclusions
35
• Flavor-changing transitions represent a unique window on physics beyond the SM: there is still a lot to learn and explore
• There are few interesting anomalies, but in general the agreement with the SM is excellent: large NP contributions, O(SM), ruled out in many cases
- LHC & LHCb are acting as a fantastic flavour-factory
- need combined th+exp precision at the few % level
• Interplay between low energy precision measurements and direct searches as strong as ever
• Ambitious program for upgrading integrated luminosity in the next years at LHC: LHCb aims at ~50fb-1
We don’t know yet what is the scale of NP: cast a wide net!
SPARES
Matteo Palutan, INFN-LNF37
Photon polarization in B+ → K+π−π+γ
Matteo Palutan, INFN-LNF38
Photon polarization in B+ → K+π−π+γ[PRL 112 (2014) 161801]
Matteo Palutan, INFN-LNF39
B0 → K*0μ−μ+ angular distribution
• in [PRL 111 (2013) 191801] fit P‘4,5,6,8=S4,5,7,8/sqrt(FL(1-FL)) with different angular foldings; e.g. for P’5: φ→-φ for φ<0; θl→ π-θl for θl> π/2- 24 bins in total
amplitudes depend on Wilson coefficient and form factors
Matteo Palutan, INFN-LNF
Understanding the P´5 anomaly?
40
• Several global fits to the available b→sγ and b→sl+l−
have been attempted, which include the observed anomaly => discrepancy with SM is 3-4.5 σ. Fits favour slightly negative C9NP (non-SM vector current), and can also be explained by introducing a flavour-changing Z’ boson ~1-7 TeV
[Decotes-Genon, Matias, Virto ’13Altmannshofer, Straub ’13Gaul, Goertz, Haisch ’14
Beaujean, Bobeth, van Dyk ’13
• By introducing form-factor uncertainties as nuisance parameters in the global fit, the discrepancy wrt SM is reduced to 2σ
G. Isidori at ICHEP 2014
(4.3<q2<8.7 GeV2)
• Form-factor uncertainties are being rediscussed, especially in the large recoil region, with appreciable impact on the comparison with data, and different opinons, of course!
Jaeger, Camalich ’13Lyon, Zwicky ’14Hofer, Matias ‘14
Matteo Palutan, INFN-LNF
Differential BRs at high q2 (>14GeV2)
41
Matteo Palutan, INFN-LNF42
B → K(*)μ−μ+ isospin asymmetry
Matteo Palutan, INFN-LNF43
B+ → K+μ−μ+ angular distributionJHEP 05 (2014) 082
Matteo Palutan, INFN-LNF44
Lepton universality in B+ → K+l+l−
Matteo Palutan, INFN-LNF
B0(s) → μ+μ− decays
45
SM
NP
Matteo Palutan, INFN-LNF
B0(s) → μ+μ− time-integrated BR
46
Time-integrated BR vs CP-averaged BR
Lifetime bias in the analysis efficiency
a residual dependence vs analysis lifetime-dependent cuts is also corrected
Matteo Palutan, INFN-LNF
LHCb
47
~1011 b-bbar couples per fb-1
~ 3 fb-1 accumulated so far
Matteo Palutan, INFN-LNF
B0(s) → μ+μ− at LHCb
48
~ B0s → μ+μ− 12 SM evts/fb-1
Matteo Palutan, INFN-LNF
Lepton flavour violation in τ→μμμ
49
Matteo Palutan, INFN-LNF
Search for Majorana neutrinos
50
• Limit on BR can be translated to a model-dependent upper limit on the coupling between muon and fourth generation neutrino
• BR upper limits as a function of mass and lifetime in [1-1000]ps
decay width fromAtre et al. JHEP 05 (2009) 030+ S. Stone, Z. Xing ’13
Matteo Palutan, INFN-LNF
Search for lepton flavour violation in B(s)→μe
51
• Decays of the type B(s)→μe are allowed in models with a local gauge symmetry between quarks and leptons, with lepto-quark linking different quark/lepton generations
[Pati, Salam PRD 10 (1974) 275]
• With 1 fb-1 LHCb has put limits x20 more stingent than the previous best limits set by CDF
[Valencia, Willenbrock arXiv:hep-ph/9409201v1]
[PRL 111 (2013) 141801]
• These limits can be translated into limits on the value of the lepto-quark mass in the framework of the Pati-Salam model:
mLQ(Bs→μe)>101 TeV/c2 at 95% CL
mLQ(B→μe) >126 TeV/c2 at 95% CL
BR(B0s→μe)< 1.1 x 10-8 at 90% CL
BR(B0→μe) < 2.8 x 10-9 at 90% CL
Matteo Palutan, INFN-LNF
Projections
52
2012: LHCb Upgrade Framework TDR http://cdsweb.cern.ch/record/1443882/files/LHCB-TDR-012.pdf