Slides - Physics Seminar - Desy

m- 

Heavy flavour 

physics with 

the CMS experiment 

m+ 

Vincenzo Chiochia 

Physik-Institut - University of Zurich 

19-20 November 2013 

DESY 

V. Chiochia (Zürich University) – Heavy flavour physics with the CMS experiment - 19-20/11/2013 !1

Why heavy flavor physics? 

Recent 

results! 

Physics of beauty and charm quarks in p-p collisions 

Research area with rich of phenomenology: 

Heavy flavor production measurements 

• Tests of QCD (hard scattering, fragmentation, NRQCD, etc.) 

Spectroscopy and particle properties 

• Heavy baryon spectroscopy 

• Spectrum of standard and exotic quarkonium states 

• Particle lifetimes, masses, etc. 

Rare beauty decays 

• Complementary to direct searches: can access multi-TeV 

energy scales through loop contributions 

CMS published 24 journal articles in the heavy flavor domain 

https://twiki.cern.ch/twiki/bin/view/CMSPublic/PhysicsResultsBPH 

V. Chiochia (Zürich University) – Heavy flavour physics with the CMS experiment - 19-20/11/2013 

!2

The Large Hadron Collider 

pp, B-Physics,  

Our “toys”. 

CP Violation 

General 

Purpose, 

pp, heavy ions 

Heavy ions, pp 

G. Tonelli, CERN/INFN/UNIPI CHIPP_Grindelwald January 21-25 2012 3 


!3

LHC performance - pp run 

LHC : Performance Limitations 

Parameter/Effects Limitations Now 

Beam energy 

limited by maximum dipole field. Industrially 

available technology. 

Bunch and total beam intensity 

beam-beam effect (tune spread), small allowed 

space in Q-space, collimators (impedance, 

collective instabilities), electron cloud, radiation 

Legend: 

7 TeV 3.5 -> 4 TeV 

N ~ 1.5 10 11 

0.6 m 

σ ~ 20 µm 

N < 1.7 10 11 

N nom = 1.15 10 11 

I < 0.85 A 

Normalized emittance 

Limited by injectors and main dipole aperture ε n

Datasets and luminosity 

CMS Integrated Luminosity, pp 

Total Integrated Luminosity (fb ¡1 ) 

25 

20 

15 

10 

5 

0 

1 Apr 

Data included from 2010-03-30 11:21 to 2012-12-16 20:49 UTC 

1 May 

1 Jun 

2010, 7 TeV, 44.2 pb ¡1 

2011, 7 TeV, 6.1 fb ¡1 

2012, 8 TeV, 23.3 fb ¡1 £ 100 

5 

0 

1 Jul 

1 Aug 

1 Sep 

Date (UTC) 

1 Oct 

1 Nov 

1 Dec 

25 

20 

15 

10 

CMS Peak Luminosity Per Day, pp 

2010: ~40/pb at Linst~10 32 cm -2 s -1 

2011: ~6/fb at Linst

The CMS detector 

CMS DETECTOR 

Total weight 

Overall diameter 

Overall length 

Magnetic field 

: 14,000 tonnes 

: 15.0 m 

: 28.7 m 

: 3.8 T 

COMPACT 

STEEL RETURN YOKE 

12,500 tonnes 

SILICON TRACKERS 

Pixel (100x150 μm) ~16m2 ~66M channels 

Microstrips (80x180 μm) ~200m2 ~9.6M channels 

SUPERCONDUCTING SOLENOID 

Niobium titanium coil carrying ~18,000A 

MUON CHAMBERS 

Barrel: 250 Drift Tube, 480 Resistive Plate Chambers 

Endcaps: 468 Cathode Strip, 432 Resistive Plate Chambers 

PRESHOWER 

Silicon strips ~16m2 ~137,000 channels 

FORWARD CALORIMETER 

Steel + Quartz fibres ~2,000 Channels 

CRYSTAL 

ELECTROMAGNETIC 

CALORIMETER (ECAL) 

~76,000 scintillating PbWO4 crystals 

HADRON CALORIMETER (HCAL) 

Brass + Plastic scintillator ~7,000 channels 

MULTI-PURPOSE! 

SOLENOID 


!6

CMS: Tracker 

Insertion of the CMS 

pixel detector 

Tracker inner barrel during 

final integration 

CMS is equipped with a full-silicon tracking detector 

Three layers and two disks of pixel sensors (~66M channels) 

Ten barrel layers and 3+9 endcap wheels of strip sensors (~10M channels) 

Pseudorapidity coverage up to 2.4. Transverse momentum resolution 2-3%. 


!7

to the primary vertex; and (d) longitudinal impact parameter, d z , with respect to the primary 

vertex. 

CMS: Pixel detector 

3.2 Track Impact Parameter Resolution 

The analysis described in this section is based on the 7 TeV data collected by CMS up to the 

27th of May 2010 and corresponding to 10.9 nb 1 . In addition ~98% to the operational general selection during detailed data taking 

in Section 1.1, the events used for the measurement of the IP resolutions are required also to 

pass the uncorrected 6 GeV jet trigger. The usage of a common Hit efficiency trigger ensures >99% that the tracks 

used in both data and simulation are comparable in terms of track multiplicity and distribution 

of particle kinematic variables. The measurement of the Excellent impact parameter understanding resolution starts of detector resolution: 

from the selection of high quality tracks that have a high probability 

Hit position, 

of havingimpact been produced 

parameter, vertices 

promptly in the pp collision: a track must have its p T greater than 0.3 GeV/c and valid measurements 

on at least 7 consecutive layers of the tracker, including a measurement on the innermost 

Hit resolution vs. cluster size 

Transverse ~ 10 mm 

pixel layers (either the barrel or one of the endcap disks). Simulation studies predict that this 

Longitudinal ~ 20 mm 

simple selection is expected to reduce the fraction of fake tracks to the per mil level. For transverse 

momenta smaller than 4 GeV/c (20 GeV/c), the 

4 

fraction of non-prompt tracks that are 

2 Pr 

selected is less than 2% (10%) of the total. 

CMS preliminary 2010 

110 

Transv.Impact Parameter Resolution (µm) 

100 

90 

80 

70 

60 

50 

40 

30 

20 

10 

Data track | η | < 0.4 

Simulation track | η | < 0.4 

Excellent modeling 

of pixel hit resolution, 

multiple scattering, 

alignment 

s = 7 TeV 

2 4 6 8 10 12 14 16 18 20 

Track p (GeV/c) 

T 

CMS preliminary 2010 

130 

Longit.Impact Parameter Resolution (µm) 

120 

110 

100 

90 

80 

70 

60 

50 

40 

30 

Data track | η | < 0.4 

Simulation track | η | < 0.4 

s = 7 TeV 

Vertex resolution vs. 

number of tracks and 

average pT 

2 4 6 8 10 12 14 16 18 20 

Track p (GeV/c) 

T 

CMS-PAS-TRK-10-005 

Eur.Phys.J. C70 (2010) p.1165 

V. Chiochia (Zürich University) (a) – Heavy flavour physics with the CMS experiment (b) - 19-20/11/2013 (a) 

!8

CMS: Muon system 

Tracks: 

Excellent pT resolution ≈ 2-3% 

Efficiency above 99% for central muons 

Impact parameter resolution ~15 mm 

Muon candidates: 

Match between muon segments and a silicon 

track 

Large pseudorapidity coverage: |η| < 2.4 

Muon efficiencies evaluated with 

MC methods 

Data-driven methods: Tag & Probe 

Muon misidentification rates from data: 

D*→D 0 p, D 0 →Kp 

Ks→pp 

L→pp 

CMS-PAS-MUO-10-002 

DT coverage 


!9

quark production at the LHC 

LHC RUN 1 

~ 2 MHz at 

7.5x10 33 cm -2 s -1 

Enormous b quark production 

rate at the LHC Run 1 

Expected to more than 

double in Run 2 

High rates implies very 

selective requirements at 

trigger level to store 

interesting b decays 


!10

Triggers for heavy flavor physics 

Trigger selections based on: 

• pT and |η| of dimuons 

• dimuon invariant mass 

• secondary vertex probability 

• impact parameter 

• flight length significance 

• pointing angle 

Trigger requirements tightened following the increase in instantaneous luminosity. 

About 10% of CMS bandwidth assigned to heavy flavor physics 

Single muon trigger efficiencies measured from data (tag&probe), dimuon correlations from MC 


!11

Understanding the rates: 

Production 

measurements 


dy (pb/GeV) 

σ/dp 

2 

b-jet d 

T 

9 

10 

8 

10 

7 

10 

10 

6 

5 

10 

4 

10 

3 

10 

2 

10 

10 

Inclusive B production 

-1 

CMS L = 34 pb 

MC@NLO 

Exp. uncertainty 

Anti-k T 

R=0.5 

s = 7 TeV 

|y| < 0.5 (×625) 

0.5 < |y| < 1 (×125) 

1 < |y| < 1.5 (×25) 

1.5 < |y| < 2 (×5) 

2 < |y| < 2.2 

1 

20 30 40 50 100 200 

b-jet p (GeV) 

T 

(pb/GeV) 

b-jet dσ/dp 

T 

6 

10 |y | < 2.2 

b-jet 

5 

10 

4 

10 

3 

10 

2 

10 

-1 

CMS L = 3-34 pb 

Jet based 

Muon based 

MC@NLO 

ATLAS jet 

ATLAS muon 

s = 7 TeV 

20 40 60 80 100 120 140 160 180 200 

b-jet p (GeV) 

T 

1 

0 0.05 0.1 0.15 0.2 

2 

0 

d xy [cm] 

-2 

pp ! bX ! µX 6 1 |⌘| < 2.1 1.32 ± 0.01 ± 0.30 ± 0.15 0.95 +0.41 

0.21 

[1] 

pp ! b¯bX ! µµX 4 1 |⌘| < 2.1 (26.18 ± 0.14 ± 2.82 ± 1.05) ⇥ 10 3 (19.95 +4.68 

4.33 ) ⇥ 10 3 [4] 

pp ! b jet X 30 1 |y| < 2.4 2.14 ± 0.01 ± 0.41 ± 0.09 1.83 +0.64 

0.42 

[5] 

pp ! B + X 5 1 |y| < 2.4 28.3 ± 2.4 ± 2.0 ± 1.1 25.5 +8.8 

5.4 

[9] 

pp ! B 0 X 5 1 |y| < 2.2 33.2 ± 2.5 ± 3.1 ± 1.3 25.2 +9.6 

6.2 

[10] 

pp ! B s X ! J/ X 8 50 |y| < 2.4 (6.9 ± 0.6 ± 0.5 ± 0.3) ⇥ 10 3 (4.9 +1.9 

1.7 ) ⇥ 10 3 [11] 

Source: http://arxiv.org/abs/1201.6677 

JHEP 04 (2012) 084 

Table 1. Summary of measured b-quark cross sections in inclusive (top) and exclusive (bottom) channels and comparison with NLO JHEP 06 (2012) 110 

QCD expectations. The experimental uncertainties indicate (from left to right) the statistical, systematic and luminosity. Theoretical 

V. Chiochia uncertainties (Zürich University) are obtained– from Heavy scaleflavour variations. 4physics with the CMS experiment - 19-20/11/2013 

!13 

6 

Number / 0.004 cm 

Pull 

5 

10 CMS s = 7 TeV, L = 27.9/pb Data 

int 

BB 

µ 

µ 

p > 4 GeV, |η | < 2.1 CC 

4 

T 

10 

DD 

BD 

BC 

3 

10 

CD 

PP 

2 

10 

10 

0 0.05 0.1 0.15 0.2 

d xy [cm] 

Inclusive B cross sections using tagged jets and events with two semileptonic muons 

NLO in agreement with data but systematically below 

Further measurements needed to pin down high-pt and rapidity regions 

• MC@NLO and POWHEG EPJgiving Web ofsomewhat Conferences different predictions 

Process p T range (Pseudo)-rapidity Cross section NLO QCD Ref. 

[GeV/c] range [µb] [µb]

6.6 ± 0.4, and n(L b )=7.6 ± 0.4. 

recent 

The larger 

result [42] 

n value 

from the 

for 

LHCb 

L b indicates 

Collaboration 

a more 

measures 

steeply 

a strong 

falling 

p 

p T 

distribution than observed for the mesons, also suggesting that the production of L hadron production 

T dependence of the ratio of 

L b production to B-meson production, f Lb /( f u + f d ), with f Lbb ⌘baryons, 

B(b ! L b ) and f q ⌘ B(b ! 

relative to B mesons, varies as a function B q ). Largerof f Lb pvalues T , with areaobserved larger at Llower b /B ratio p T , which at lower suggests transverse that the discrepancy observed 

between the LEP and 

momentum. The right plot of Fig. 3 shows the p L Tevatron b 

T spectrum data may shape be due compared to the lower to p B+ T of the L 

and B b 0 baryons produced at 

, 

the Tevatron. 

where the distributions are normalized to the common bin with p T = 10 13 GeV. 

(pp→ b-hadron X) (µb/GeV) 

dσ/dp 

T 

-1 

10 

-2 

10 

10 

10 

1 

-3 

(10-13 GeV)) 

)/(dσ/dp 

(dσ/dp 

T 

T 

-2 

10 

10 20 30 40 50 

10 20 30 40 50 

b-hadron p (GeV) 

b-hadron p (GeV) 

T 

T 

Figure CMS 3: Preliminary, Comparison s=7 of TeV production rates Spring for 2012 B + [6], B 0 [7], B 0 s [9], and L b versus p T . The left 

4.8 

plot shows the absolute comparison, where the inner error bars correspond to the total bin-tobin 

uncertainties, while the outer error bars represent the total bin-to-bin and normalization 

value ± stat. ± syst. ± lum. error 

(luminosity) 

NLO predictions compatible with data 

pp → Λ b 

X → J/ψ Λ X 

11.6 ± 0.6 ± 1.2 ± 2.0 

uncertainties P T>10 GeV, |y|5 GeV, |y|

Spectroscopy: 

A new heavy baryon 

and other peaks 


On the media 

First particle discovery at CMS! 

May 1st, 2012 


!16

Ξ 

- 

b 

B 

Y 

The Jb baryon family 

B=-1 

Σ - 

bbb 

J=3/2 + 

Σ 

*- ddb 

udb uub Σ 

*+ 

b b 

dsb usb 

π" 

ssb 

dsb 

usb 

Δ - ddd 

uuu Δ ++ 

ssb 

dds 

uus 

uds 

dss 

Σ + 

n 

$ 

- 

b p 

uss 

Ξ - sss 

Ξ o 

5791 γ" ddu 

duu 

" 

Ω - ! - dds 

J=1/2 + 

uds ! o uus 

Y 

dss 

uss 

dbb 

ubb 

# 

Ξ’ o 

b I 3 

sbb 

! 

dsb 

usb • Charm sector cousin: c(2645) 

ssb 

! 

0 ! 

n 

Ω 

- 

b p 

ddu 

duu 

Λ 

dds uds Σ o uus Σ + 

Width < 1MeV 

dss 

uss 

Ξ - Ξ o 

I 3 

B=-2 


Ξ b 

* 

Σ 

- 

ddb udb Λ b 

uub 

b Σ 

+ 

b 

• B-baryon multiplets: 

B 

Mass [MeV] 

Analog to charmed baryon spectrum 

Q value: M(Jb *0 )-M(Jb - )-M(p + )≈11-29 MeV 

“Heavy neutron” with u-s-b quarks 

• Given the predicted Q, Γ(Ξb* 0 ) < 1 MeV. 

Introduction 

J=1/2 + 

# 

- dbb 

ubb 

b 

#’ b 

sbb 

! 

- 

ddb udb " b 

uub 

b ! 

+ 

b 

? 

? 

" - dd 

(known) 

# b 

*- 

# - + c ⇡ 

Phys. Rev. Let 

Phys. Rev. D 66, 014502 (2002) 

• Expected Q(Ξb* 0 ) = M(Ξb* 0 ) − M(Ξb − ) − M(π + ) ~ 11-29 M 

! + 

Phys. Rev. D 77, 034012 (2008) 

Phys. Rev. D 79, 014502 (2009) 

Phys. Rev. D 84, 014025 (2011) 

arXiv:1203.3378v1 [hep-lat] 

arXiv:1203.3378v1 !17

Jb* 0 decay chain 

Introduction 

• Ξb *0 decay chain: 

p + 

Secondary 

vertices 

⇡ ⇤ 

c ( 0 )=7.89 cm 

c ( 

)=4.91 cm 

⇤ 0 

Can decay beyond pixel volume 

Requires reconstruction of very 

displaced tracks and vertices 

⇡ ⌅ 

c ( b 

) = 470 µm 

⌅ 

PV 

⌅ ⇤0 

b 

⌅ b 

⇡ + PV 

J/ 

µ 

µ + 

Displaced J/c decay vertex 

within pixel volume 

& c.c. 

Ernest Aguiló (HEPHY) 6 

June 26th 2012 


!18

candidates per 20 MeV 

- 

Ξ b 

70 

60 

50 

40 

30 

20 

10 

Observation of the Jb* 0 baryon 

CMS 

pp, s = 7 TeV 

-1 

L = 5.3 fb 

Jb→JJ/c 

- 

Ξ b 

data 

Opposite-sign pπ Ξ 

Same-sign pπ Ξ 

Signal+background fit 

Background fit 

108 ± 14 candidates 

M=5795.0 ± 3.1 MeV 

s=23.7 ± 3.2 MeV 

0 

5.6 5.7 5.8 5.9 6 6.1 

- 

M(J/ψΞ ) [GeV] 

Additional 

prompt pion 

candidates per 2 MeV 

*0 

Ξ b 

16 

14 

12 

10 

8 

6 

4 

2 

21 events observed, 3.0±1.4 

background expected 

Resolution from MC: sMC=1.9±0.1 MeV 

CMS 

pp, s = 7 TeV 

-1 

L = 5.3 fb 

(b) 

Opposite-sign data 

Signal+background fit 

Background 

0 

0 10 20 30 40 50 

- + 

- 

M(J/ψΞ π ) - M(J/ψΞ ) - M(π) [MeV] 

Q= 

Jb *0 →JJ/cp 

Q = 14.8 ± 0.7 ± 0.3 MeV 

M(Jb) = 5945.0 ± 0.7(stat.) ± 0.3(syst.) ± 2.7(PDG) MeV 

GBW=2.1±1.7 MeV (Theory: ! 0.93 MeV) 

Significance from ln(Ls+b/Lb)=6.9 


Phys. Rev. Lett. 108, 252002 (2012) 

!19

Possible Speculations 

Exotic mesons 

Many new charmonium(–like) do not fit into 

quark model spectrum easily. Theoretical 

speculations include: 

• Molecular states: loosely bound states composed of a 

pair of mesons, probably bound by the long-range colorsinglet 

pion exchange 

• Tetraquarks: bound states of four quarks, bound by 

colored-force between quarks, decay through 

rearrangement, some are charged or carry strangeness, there 

are many states within the same multiplet 

• Hybrid charmonium: bound states composed of a 

pair of quarks and one excited gluon 

• Conventional charmonium: quark model spectrum 

could be distorted by the coupled-channel effects 

9 

Courtesy: S.Zhu 


!20

Exotic mesons 

TABLE 9: As in Table 4, but for new unconventional states in the c¯c and b¯b regions, ordered by mass. For X(3872), the values 

given are based only upon decays to ⇡ + ⇡ J/ . X(3945) and Y (3940) have been subsumed under X(3915) due to compatible 

properties. The state known as Z(3930) appears as the c2(2P )inTable4. See also the reviews in [81–84] 

Well established 

at CMS 

JHEP 04 (2013) 154 

Seen also at D0 

arXiv:1309.6580 

Seen also at 

D0, Belle 

arXiv:1010.5827 

State m (MeV) (MeV) J PC Process (mode) Experiment (# ) Year Status 

X(3872) 3871.52±0.20 1.3±0.6 1 ++ /2 + B ! K(⇡ + ⇡ J/ ) Belle [85, 86] (12.8),BABAR [87] (8.6) 2003 OK 

(

Search for X(4140) 

First evidence (3.8s) for a near-threshold 

narrow peak in the J/cf system, reported by 

CDF in B + →J/c(mm)f(KK)K + decays in 2009, 

based on 2.7 fb -1 . 

Result updated in 2011 with 6 fb -1 , significance 

over 5s. Assuming relativistic BW: 

M1=4143.0 +2.9 -3.0(stat)±0.6(syst) MeV 

G1=15.3 +10.4 -6.1(stat)±2.5(syst) MeV 

Evidence (3.1s) for a second structure: 

0 

0 

5.2 5.3 5.4 

1 1.01 1.02 

+ 

2 

M2=4274.4 +8.4 -6.7(stat)±1.9(syst) MeV 

m(J/ Kφψ ) [GeV/c 

] 5 m K 

G2=32.3+21.9(stat)±7.6(syst) MeV 

PRL 102, 242002 (2009) 

Could be cc bound state but well above opencharm 

threshold (3740 MeV). Some models: line is a fit to the data with a Gaussian signal fun 

FIG. 1: (a) The mass distribution of J/ψφK + arXiv:1101.6058 ;the 

40 

10 

10 

6 

CDF 2011 

a) 

Molecular linear background function. (b) The K + K − mass 

(Ds Ds) state b) 

a) 

19±7 b) 

L=6 fb c) 

-1 

8 candidates 

8 

4 

30 

tions inside the B mass window (black solid) and 

Hybrid particle (qqg) 

2 

sidebands (red dotted). 

Four-quark combination (ccss) 

6 

6 22±8 

20 

candidates 

0 

No significant first structure from Belle 1 1.1 1.2 in 1.3 1.4 1.5 1.6 1.7 

6 

4 

4 

exclusive B decays. 3.2s evidence for second 

m(µ + µ − K + K − ) − m(µ + µ − 4 

) for events in 

structure 10 at 4350 MeV in gg→J/cf 

2 

2 

2 

mass window. Events from reference [12] and f 

Exclusion limits on partial width disfavor 

0 

data are shown in (a) top and bottom. In the 

5.4 

1 molecular 1.01 1.02 scenarios 1.03 with 1.040 ++ , 2 ++ 0 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 0 

0 

1 1.1 1.2 1.3 1.4 1.5 

1 1.1 1.2 1.3 2 

2 

eV/c 

] 

m [GeV/c ] 

signal region (∆M < 1.07GeV/c 2 1.4 1.5 

- + - 

- 

2 

), the new da 

+ - 

K K 

m(µ + µ K K )-m(µ + - + - 

- 

2 

µ ) GeV/c m(µ + )-m(µ + - + - 

- 

2 

µ K K µ ) [GeV/c ] m(µ + µ K K )-m(µ + µ ) [GeV/c ] 

within 1σ of the expectation (6 events compar 

expected). Over the entire examined region 

stribution V. Chiochia of (Zürich J/ψφK University) + ;thesolidblue 

– Heavy flavour physics FIG. with 2: (a) the The CMS experiment mass difference, 

data sets 

- 19-20/11/2013 ∆M, betweenµ 

are consistent at the + µ − K 

7% + K 

probability − 

!22 l 

2 

Candidates/2 MeV/c 

2 

Events/10 MeV/c 

+ − + 

2 

Candidates per 5 MeV/c 

60 

50 

40 

30 

20 

10 

2 


115±12 B + 

candidates 

CDF 2011 

L=6 fb -1 

a) 

2 


2 

Candidates/2 MeV/c 

40 

30 

B + X(4140) 

→ K + → J/cf K 

X(4270) 

+ 

? 

20 

10

Search for X(4140) 

3 

5 GeV 

8 GeV 

round 

4 5.45 

+ 

) [GeV] 

) / 20 MeV 

+ 

Candidates / 5 MeV 

N(B 

CMS, 

400 

100 

350 

300 

250 

200 

150 

100 

50 

−1 

s = 7 TeV, L = 5.2 fb 

arXiv:1309.6920 

− 

1.0125 < m(K 

+ 

K ) < 1.0265 GeV 

Δm < 1.568 GeV 

Data 

Fit 

Background 

5.15 0 

5.2 5.25 5.3 5.35 + − 

5.4 + 

5.45 5 

m(J/ψK K K ) [GeV] 

with the standard event selection (left) and the tighter 

-1 

CMS, s = 7 TeV, L=5.2 fb 

300 

how the result of fitting these distributions Data to a Gausomial 

background 250 while the dashed Three-body curves PS (global showfit) 

the 

Global fit 

±1σ uncertainty band 

+ 

200 

Event-mixing (J/ψ, φ, K ) 

+ 

Event-mixing (J/ψ, φ K ) 

1D fit 

150 

ibed below. With the exception of this cross-check, all 

tive criteria. 

Search based on 5.2 fb -1 in B + →J/cfK + decays: 

Largest sample so far: 2480±160 B + candidates 

Fitted Dm distribution corrected for detector 

efficiency 

Efficiency fairly uniform over mass spectrum (

Searches for Xb 

Search for the X(3872) counterpart in the 

bottomonium sector - called here Xb 

Decay channel Y(1S)p + p - 

Mass predicted close to the BB or BB* 

threshold. Search scans 10-11 GeV mass 

range 

Search in two rapidity regions due to 

different mass resolution 

Dipion mass distribution assumed to be as 

for Y(2S) and similar to X(3872) 

“R” ratio of observed Xb and Y(2S) 

candidates corrected for detector efficiency 

R=6.56% motivated by X(3872) case 

would yield >5s observation over the full 

mass range 

No excess observed. 95% CL upper bounds 

on R within (0.9 - 5.4)%. 

First upper limit on possible Xb state at 

hadron collider 


numbers of U(2S) ! U(1S)p + p events are 7100 ± 150 an 

cap regions, where the uncertainties are statistical. The 

the U(2S) resonance are shown in Fig. 2 for the barrel and 

the fits superimposed. 

6 5 Summary 

regions. The systematic uncertainties mentioned above are implemented as nuisance parame 

ters in the fit, assuming log-normal or flat priors. The expected discovery potential is estimated 

by injecting various amounts of signal events into the fits and evaluating the resulting p-values 

The expected signal significance for the assumption R = 6.56%, motivated by the ratio of pro 

duction cross sections times branching fractions for X(3872) and y(2S) reported in Ref. [8], i 

larger than five standard deviations Υ (2S)(s) across Υ(3S) 

the explored X b CMS mass barrel range, as shown by th 

dashed curve in Fig. 8000 3. The observed p-values displayed in Fig. s = 38 by TeVthe solid line show no 

-1 

indication of an X L = 20.7 fb 

b signal. The smallest local p-value is 0.004 at 10.46 GeV, corresponding 4000 to 

p > 13.5 GeV 

a statistical significance of 2.6s, which is reduced to 0.8s when taking T into account the “look 

6000 

|y| < 1.2 

elsewhere effect” [29]. The expected and observed 95% confidence level upper limits on R 

derived using a modified frequentist approach (CL S ) [30, 31], are shown in Fig. 4 as a function 

of the X b mass. The4000 

observed upper limits on R are in the range 0.9–5.4% at 95% confidenc 

2000 

level. The expected upper limits, which are derived for a pure background hypothesis, are les 

stringent than those obtained from the p-value calculations. This is because the p-value calcu 

lations are only concerned 

2000 

with the probability of the background fluctuating to a signal-lik 

peak in the invariant-mass distribution, while the upper limits on R also include the system 

atic uncertainties in the signal normalization from the signal decay model and X b polarization 

assumptions. 9.8 10 10.2 10.4 10.6 10.8 11 

9.8 

[GeV] 

p-value 


M + Υ(1S)π 

- 

π 

Figure 1 1: The reconstructed invariant-mass distributions 

and endcap (right) regions. Peaks corresponding to U( 

U(1S)p + 2σ 

-2 

10 

CMS 

p decays are indicated with the arrows. 3σ 

-4 

10 

4 10 -6 

Results 

-8 

10 

s = 8 TeV 

-1 

L = 20.7 fb 

The search for the X b is performed in the mass 

6σ 

regions 1 

-10 

10 

Observed 

cluding the mass intervals around the U(2S) and U(3S) r 

Expected for R=6.56% 

-12 

10 

7σ 

10 10.2 10.4 10.6 10.8 11 

CMS 

M Xb 

[GeV] 

± 1σ Expected 

-1 

Figure 3: Observed 10% 

(solid s = curve) 8 TeVand expected for R = 6.56% (dotted ± 2σ Expected curve) local p-values, a 

a function of the8% 

assumed L X b mass. -1 

= 20.7 fb 

6.56% Median expected 

6% 

Observed 

Upper limit on R 

4% 

5 Summary 

A search for an exotic bottomonium state in the decay channel X b ! U(1S)p + p , followed by 

U(1S) ! µ + 2% 

µ , in pp collisions at p s = 8 TeV at the LHC has been presented. This analysi 

was performed using data collected by the CMS experiment, corresponding to an integrated 

luminosity of 20.7 fb 1 -2 

. Candidates were reconstructed from two identified muons and two 

1% 10 

additional charged tracks assumed to be pions. The search was conducted in the kinemati 

region p T (U(1S)p + 10 p ) > 10.13.510.2 GeV and 10.3 |y(U(1S)p 10.4 10.5 + p 10.6 )| < 10.7 2.0. The 10.8U(2S) 10.9! 11 U(1S)p + p 

process was used as a normalization channel, canceling many of the systematic M Xb 

[GeV] uncertainties 

Excluding the known U(2S) and U(3S) resonances, no significant excess above the!24 

background 

Figure 4: Upper limits at the 95% confidence level on R, the production cross section for the X 

arXiv:1309.0250 

4σ 

5σ 

Candidates / 6 MeV

Looking outside the box: 

Rare B decays 


Why searching for Bs,d→m + m - ? 

otivation: search for new physics 

otivation: Decays highly suppressed search in SM for new physics 

Forbidden at tree level 

Decays b →s(d) highly FCNC suppressed transitions only through in Standard Penguin or Model 

DecaysBox highly diagrams suppressed in Standard Model 

⇥ effective FCNC, helicity suppression 

⇥ effective Cabibbo FCNC, (|Vtd| |V td |) 

⇥ Cabibbo-enhancement of B 0 (|V ts | > |V td |) 

s ⇥ µ + µ over B 0 ⇥ µ + µ 

Sensitivity to new physics, e.g. extended 

of only 

Higgs 

Bs 0 in ⇥ MFV 

sector 

µ + µ models 

and 

over 

SUSY 

Bparticles: 

0 ⇥ µ + µ 

only in MFV models 

Indirect 2HDM sensitivity branching ~(tanb) to new 4 and physics m(H + ) 

MSSM branching ~(tanb) 6 

Indirect sensitivity to new physics 

⇥ 2HDM: Leptoquarks B ⇤ (tan ) 4 ,m H +; MSSM: B ⇤ (tan ) 6 

4th generation top 

⌅ sensitivity to extended Higgs boson sectors 

⌅ sensitivity constraints toon extended parameter Higgs regions boson sectors 

⌅ constraints on parameter regions 


b 

b 

B s 

B s 

b 

b 

s 

s 

s(d) 

s(d) 

⇥ 2HDM: B ⇤ (tan ) 4 ,m H +; MSSM: B ⇤ (tan ) 6 

B 0 s ⇥ µ + µ 

considered as golden channel 

t 

+ t 

Z 0 0 0 

+ 

, H ,h , ... 

W , H 

+ 

Z 0 0 0 

+ 

, H ,h , ... 

W , H 

t 

t 

+ χ ~ 0 

W 

+ χ ~ , 

W , 

0 

~ ~ t,c,u d l ν 

~ ~ t,c,u d l ν 

− 

W , χ 

~ 0 

− 

W , χ 

~ 0 

µ − + 

µ 

µ 

l 

+ 

l 

l 

− 

l 

µ − + 

[1] JHEP 1307 (2013) 77 

[2] PRL 109, 041801 (2012), arXiv:1208.0934 

[3] Phys. Rev. D85: 033005, 2011 

+ 

− 

!26

A 30-years quest! 

PHYSICAL REVIEW D VOLUME 30, NUMBER 11 1 DECEMBER 1984 

Two-body 

decays of B mesons 

R. Giles, J. Hassard, M. Hernpstead, K. Kinoshita, W. W. MacKay, F. M. Pipkin, and Richard Wilson 

Harvard University, Cambridge, Massachusetts 02138 

P. Haas, T. Jensen, H. Kagan, and R. Kass 

Ohio State University, Columbus, Ohio, 43210 

S. Behrends, K. Chadwick, J. Chauveau, T. Gentile, Jan M-. Guida, Joan A. Guida, A. C. Melissinos, 

S. L. Olsen, G. Parkhurst, D. Peterson, R. Poling, C. Rosenfeld, E. H. Thorndike, and P. Tipton 

University ofRochester, Rochester, Xetv York 14627 

PRL 111, 101804 (2013) PHYSICAL REVIEW LETTERS 

week e 

6 SEPTEMB 

D. Besson, J. Green, R. Namjoshi, F. -Sannes, P. Skubic, f A. Snyder, ~ and R. Stone 

Rutgers University, New Brunswick, New Jersey 08854 

A. Chen, M. Goldberg, M. E. Hejazifar, N. Horwitz, A. Jawahery, P. Lipari, G. C. Moneti, 

C. G. Trahern, and H. van Hecke 

Syracuse University, Syracuse, New York 13210 

M. S. Alam, S. E. Csorna, L. Garren, M. D. Mestayer, R. S. Panvini, and Xia Yi 

Vanderb&lt University, Nashville, Tennessee 37235 

P. Avery, C. Bebek, K. Berkelman, D. G. Cassel, J. W. DeWire, R. Ehrlich, T. Ferguson, R. Galik, 

M. G. D. Gilchriese, B. Gittelman, M. Hailing, D. L. Hartill, S. Holzner, M. Ito, J. Kandaswamy, 

D. L. Kreinick, Y. Kubota, N. B. Mistry, F. Morrow, E. Nordberg, M. Ogg, K. Read, A. Silverman, 

P. C. Stein, S. Stone, R. Wilcke, ~ and Xu Kezun 

Cornell University, Ithaca, New York 14853 

A. J. Sadoff 

Ithaca College, Ithaca, New York 14850 

(Received 8 June 1984; revised manuscript received 10 September 1984) 

Various exclusive and inclusive decays of B mesons have been studied using data taken with the 

CLEO detector at the Cornell Electron Storage Ring. The exclusive modes examined are mostly decays 

into two hadrons. The branching ratio for a B meson to decay into a charmed meson and a DOI: 10.1103/PhysRevLett.111.101804 

charged pion is found to be about 2%. Upper limits are quoted for other final states PE, m.+m 

pox, p+p, e+e, and p e+. +— We also give an upper limit on inclusive g production and improved 

charged multiplicity measurements. 

In the standard model (SM) of particle physics, treelevel 

diagrams do not contribute to flavor-changing 

neutral-current (FCNC) decays. However, FCNC decays 

I. INTRODUCTION 

may proceed through higher-order loop diagrams, and this 

&4/o at 90% confidence level, a charmed particle should 

be present in almost all 8-meson 

opens up the possibility for contributions from non-SM 

decays. 

Even if we assume that the spectator model (Fig. 1) de- 

Weak decays of 8 mesons provide a unique testing 

ground for both electroweak theory and quantum chromodynamics 

(QCD). 8 þ possible have wayssmall to branching fractions of BðB 0 s ! 

scribes 8 decay, there are still many 

mesons are composed of a b antiquark 

and a light quark, either a d (8 ) or a u (8+). In final-state primary partons interact þ withÞ¼ð3:57 each other 0:30Þ10 9 , corresponding to the 

describe how hadrons might form. In one model, the four 

the standard electroweak model' the b quark can decay without regard to color or chargedecay-time and produce integrated hadron branching fraction, and BðB 0 ! 

into the c quark or the u quark by emitting a virtual intermediate 

vector boson. The possible decay schemes of native model is based on the fact that QCD requires had- 

states proportional to the available phase space. An alter- 

þ Þ¼ð1:07 0:10Þ10 10 [1,2]. Charge conjugation 

which is implied we call throughout the this Letter. Several extensions of 

V. B and 

Chiochia B are shown in 

(Zürich 

Fig. 1University) for both the b~e and rons 

– Heavy 

to be color singlets. In this model, 

flavour physics with the CMS experiment - 19-20/11/2013 

b~u cases. In the "spectator" diagrams the B meson de- 

"spectator model without color mixing, " the quarks from 

8' 

Measurement of the B 0 s ! þ Branching Fraction and Search 

for B 0 ! þ with the CMS Experiment 

S. Chatrchyan et al.* 

(CMS Collaboration) 

(Received 18 July 2013; published 5 September 2013) 

Results are presented from a search for the rare decays B 0 s ! þ and B 0 ! þ in pp collisions 

pffiffi 

at s ¼ 7 and 8 TeV, with data samples corresponding to integrated luminosities of 5 and 20 fb 1 , 

respectively, collected by the CMS experiment at the LHC. An unbinned maximum-likelihood fit to the 

dimuon invariant mass distribution gives a branching fraction BðB 0 s ! þ Þ¼ð3:0 þ1:0 

0:9 Þ10 9 , 

where the uncertainty includes both statistical and systematic contributions. An excess of B 0 s ! þ 

events with respect to background is observed with a significance of 4.3 standard deviations. For the decay 

B 0 ! þ an upper limit of BðB 0 ! þ Þ < 1:1 10 9 at the 95% confidence level is determined. 

Both results are in agreement with the expectations from the standard model. 

particles. In the SM, the rare FCNC decays B 0 s ðB 0 Þ ! 

the SM, such as supersymmetric models with nonuniversal 

PACS numbers: 13.20.He, 12.15.Mm 

p 

corresponding 

ffiffi 

to integrated luminosities of 5 

s ¼ 7 TeV and 20 fb 1 at 8 TeV collected 

Compact Muon Solenoid (CMS) experiment at th 

Hadron Collider (LHC). For these data, the peak l 

ity varied from 3:5 10 30 to 7:7 10 33 cm 2 s 

average number 

pffiffi 

of interactions per bunch crossing 

was 9 (21) at s ¼ 7ð8Þ TeV. 

The search for the B ! þ signal, w 

denotes B 0 s or B 0 ,isperformedinthedimuoni 

mass regions around the B 0 s and B 0 masses. T 

possible biases, the signal region 5:20

Event characteristics 

Key ingredients: 

Good dimuon vertex, correct B mass assignment, 

isolation, • Signal B s 0 momentum ⇥ µ + pointing to interaction point 

µ 

Analysis overview 

µ 

_ µ 

B 

⇤ 

two 

two 

muons 

muons 

from 

from 

one 

one 

decay 

decay vertex 

vertex 

Signal characteristics: mass around m 

mass around B 0 0 s s 

Two muons from long-lived a well reconstructed B 

long-lived decay vertex 

well reconstructed 

Mass compatible with Bs (or B 0 secondary vertex 

well reconstructed) 

secondary vertex 

momentum aligned with flight direction 

momentum aligned with flight direction B 

Dimuon momentum aligned with B flight ! 

• Background 

Background 

⇤ two semileptonic (B) decays (gluon splitting) 

⇤ two semileptonic (B) decays (gluon splitting) 

Background sources: 

⇤ one semileptonic (B) decay and one misidentified hadron 

⇤ one semileptonic (B) decay and one misidentified hadron 

Two semi-leptonic ⇤ rare 

⇤ rare B single 

single decays B decays 

decays (e.g. from gluon splitting) 

peaking (B 

One semi-leptonic peaking B decay (Bs 0 s 0 ⇥ ⇥ + K+ K ) 

K+ misidentified K ) hadron _ 

non-peaking (B Hadronic B decays 

non-peaking (Bs 0 s 0 ⇥ K µ+ ⇥) 

⇥ K µ+ ⇥) 

B 

⇥ mass resolution 

mass resolution 

• Peaking: ⇥ notBs→K well-reconstructed - K + 

secondary vertex 

– 

not 

More 

well-reconstructed 

problematic within the B 

secondary 0 mass window 

vertex B 

⇥ pointing angle 

• Rare 

⇥ 

semileptonic: 

pointing angle 

Bs→K - m + n, Lb→pmn 

⇤ High signal efficiency and high background reduction 

High signal efficiency and high background reduction 

⇤ 

µ 

µ 

µ 

Urs Langenegger Search for B 

Urs Langenegger Search for B 

s(d) 0 s(d) 0 ⇥ µ+ µ with the CMS experiment (2011/07/22) 5 

⇥ µ+ µ with the CMS experiment (2011/07/22) 5 


!28

Discriminating variables 

Boosted decision tree (BDT) selection 

-1 

4 

Trained on signal MC sample and dimuon data 10 sidebands 

Data (sideband) 

Same BDT for normalization (B + →J/cK) and control (Bs→J/cf) channels 

3 

0 - 

" µ + µ (MC) 

Candidates 

10 

10 

2 

10 

! Mass windows: 5.2-5.3 GeV for B and 5.3-5.4 

! Split into barrel (both |" ! | < 1.4) and endcap 

! l 3D > 15 or 20 , # 3D < 0.050 or 0.025 

! p T! > 4.5 or 4.0 GeV, p TB > 6.5 GeV, 

! B fit $ 2 /dof < 1.6, dca min > 0.15 mm (endca 

12 input variables: kinematic, Tracker only, Muon only,Tracker+Muon variables 

Robustness studies 

CMS, 1.14 fb s = 7 TeV 

Insensitive to invariant mass using MC signal events with shifted mass 

Output independent on high- or low-mass sideband 

Insensitive to multiple collisions (pileup) 

B s 

PV 

l 3D 

µ 

Secondary 

vertex 

µ 

! 3D 

6000 Dimuon 

data sidebands 

- 

B → µ + 

s µ 

5000 

4000 

3000 

2000 

1000 

CMS L = 20 fb ( s = 8 TeV) 

-1 

500 

400 

300 

200 

100 

CMS L = 20 fb ( s = 8 TeV) 

-1 

Dimuon 


B → µ + 

s µ 

1 

Examples of background separation variables 

-1 

10 

-1 

CMS L = 20 fb ( s = 8 TeV) 

0 20 40 60 80 100 5000 

1000 

Dimuon l 3D 

/!(l ) 

3D 

" Select best primary vertex based on 

consistency with B candidate 

momentum 

-1 

direction 

L = 20 fb ( s 

" Average of 5.5 primary vertices per 

- 

- 

B → µ µ 

9/1/11800 

Keith Ulmer -- University of Colorad 

+ 

s 

4000 

- 

B → µ + 

s µ 

600 

400 

200 


3000 

2000 

1000 

EPS-HEP 2013 

CMS = 8 TeV) 

Dimuon 


0 50 100 

l 3D 

/σ(l ) 

3D 

Flight length 

significance 

0 0.05 0.1 

α 3D 

Pointing 

angle 

0 1 2 3 4 

δ 3D 

/σ(δ ) 

3D 

Impact parameter 

significance 

0 1 2 3 4 

χ 2 /dof 

B-vertex reduced 

chi 2 


!29

Isolation variables 

Relative isolation of muon pairs 

8000 

7000 

6000 

5000 

4000 

3000 

2000 

1000 

Cone with DR=0.7 around di-muon momentum 

Include all tracks with pT>0.9 GeV from same PV or 

dCA

Signal normalization 

Branching ratios calculated w.r.t. normalization channel B + →J/c(m + m - )K + 

Many systematic uncertainties cancel in ratio 

No need for absolute luminosity and b-quark cross section 

Large B + yield and well known branching ratio to J/cK + (3% uncertainty) 

Ratio of b-quark fragmentation fractions to Bs/B + : fs/fu=(256±20)×10 -3 [JHEP 04 (2013) 001] 

From LHCb 

Br(B s ! µ + µ )= N(B s ! µ + µ ) f u ✏ B+ 

tot 

N(B + ! J/ K + ) f s 

✏ B s 

tot 

Br(B + ! J/ K + ) 

Candidates / 0.010 GeV 

CMS L = 20 fb ( s = 8 TeV) 

40000 

35000 

30000 

25000 

20000 

15000 

10000 

5000 

-1 

Barrel 

~82’700 

candidates 

Candidates / 0.010 GeV 

6000 

5000 

4000 

3000 

2000 

1000 

CMS L = 20 fb ( s = 8 TeV) 

-1 

Endcap 

2011-12 

B + Candidates: 

380×10 3 barrel 

90×10 3 endcap 

5 5.2 5.4 5.6 5.8 

[GeV] 

m µµK 

5 5.2 5.4 5.6 5.8 

[GeV] 

m µµK 


!31

Branching ratio measurement 

BDT output divided into 4 (2) bins for 2012 (2011) data and barrel/endcap categories 

Simultaneous UML fit of Bs and B 0 candidates: 

B s and B 0 decays signal 

Peaking backgrounds (e.g. B 0 →Kp, Bs→KK) 

Rare s-l backgrounds (e.g. Lb→pmn) 

Combinatorial background 

Event-per-event mass resolution included 

BR(B S 

→ µµ) = ( 3.0 +0.9 −0.8 

(stat) +0.6 −0.4 

(syst))×10 −9 

BR(B d 

→ µµ) = ( 3.5 +2.1 −1.8 

(stat+syst))×10 −10 

Significances 

Bs→µµ: 4.3 σ (exp. median 4.8 σ) 

B 0 →µµ: 2.0 σ 

CMS+LHCb combination 

(~24% precision) 

CMS-PAS-BPH-13-007 

−1 

LHCb 3fb 

SM 

−1 

CMS 25fb 

CMS+LHCb 

preliminary 

(2.9±0.7)×10 -9 

0 1 2 3 4 5 6 7 

0 − −9 

B(Bs→ µ + µ ) [10 ] 

CMS: Phys.Rev.Lett. 111 (2013) 101804 

LHCb: Phys. Rev. Lett. 08 (2013) 117 


!32

Exclusion limits on B 0 →m + m - !33 

No significant excess observed in the B0 mass window 

Upper limit on BR computed using CLs method 

BR(B d 

→ µµ)

4 References 

Future sensitivity 

Table 1: Number of expected events for B 0 s ! µ + µ and B 0 ! µ + µ at different integrated 

Table luminosity 1: Number values. of We expected also report events the expected for B 0 uncertainty in the branching fraction measurement 

for the B 0 s ! µ + µ and B 0 ! µ + s ! µ + µ and B 0 ! µ + µ at different integrated 

luminosity values. We also report the expected µ , the range uncertainty of significance in theof branching B 0 ! µ + µ fraction (the range measurement 

indicates for the the B±1s 0 of the distribution of significance), and the relative uncertainty on the B 0 to 

B 0 s ! µ + µ and B 0 ! µ + µ , the range of significance of B 0 ! µ + µ (the range 

s branching fractions. 

indicates the ±1s of the distribution of significance), and the relative uncertainty on the B 0 to 

B 0 s branching fractions. 

CMS PAS FTR-13-022 

L (fb 1 ) No. of B 0 s No. of B 0 dB/B(B 0 s ! µ + µ ) dB/B(B 0 ! µ + µ ) B 0 sign. d B(B0 !µ + µ ) 

B(B 0 s !µ + µ ) 

20 16.5 2.0 35% >100% 0.0–1.5 s >100% 

L (fb 1 ) No. of B 0 s No. of B 0 dB/B(B 0 s ! µ + µ ) dB/B(B 0 ! µ + µ ) B 0 sign. d B(B0 !µ + µ ) 

100 144 18 15% 66% 0.5–2.4 s 71% B(B 0 s !µ + µ ) 

20 300 16.5 433 54 2.0 12% 35% 45% >100% 1.3–3.3 0.0–1.5 s s 47% >100% 

100 3000 2096 144 256 18 12% 15% 18% 66% 5.4–7.6 0.5–2.4 s s 21% 71% 

300 433 54 12% 45% 1.3–3.3 s 47% 

3000 2096 256 12% 18% 5.4–7.6 s 21% 

S/(S+B) weighted events / ( 0.02 GeV) 


120 

100 

120 

80 

100 

60 

80 

40 

60 

40 

CMS Simulation - Scaled to L = 300 fb 

-1 


20 

-1 

data 

full PDF 

- 

B s →µ + µ 

- 

B d 

→µ + µ 

data 

combinatorial bkg 

full PDF 

semileptonic - bkg 

B 

peaking s →µ + µ 

bkg - 

B d 

→µ + µ 


semileptonic bkg 

peaking bkg 

0 

0 

4.9 5 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 

4.9 100 5 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 

20 

m µµ (GeV) 

m µµ (GeV) 

50 

Figure 1: Fit results of the invariant mass distribution for 300 fb 1 and 3000 fb 1 . The improvement in 

0 

the mass resolution for the 3000 fb 1 0 

4.9 5 5.1 5.2 5.3 5.4 5.5 5.6projection 5.7 5.8 is 5.9expected from 4.9 an 5 improved 5.1 5.2 5.3 inner 5.4tracker 5.5 5.6 system 5.7 and 5.8 5.9 

removing endcap candidates. 

(GeV) 

(GeV) 

Figure 

5 Summary 

1: Fit results of the invariant mass distribution for 300 fb 1 and 3000 fb 1 . The improvement in 

the mass resolution for the 3000 fb 1 projection is expected from an improved inner tracker system and 

removing In the coming endcapyears, candidates. the LHC accelerator and the CMS detector will undergo a series of upgrades 

in two major steps. The first will result in a data sample corresponding to 300 fb 1 

5of integrated Summary luminosity and the second to 3000 fb 

V. Chiochia (Zürich University) – Heavy flavour physics with the CMS experiment 1 . With the increased data sample sizes it 

- 19-20/11/2013 

will be possible to reduce both systematic and statistical errors leading to high precision mea- 

m µµ 




450 

data 

full -1PDF 

400 CMS Simulation - Scaled to L = 3000 - 

B 

fb s →µ + µ 

- 

450 

B d 

→µ + µ 

350 

data 


full PDF 

400 

semileptonic bkg - 

300 

B 

peaking s →µ 

bkg 

+ µ 

- 

350 

250 

200300 

150250 

100200 

50150 

-1 

B d 

→µ + µ 


semileptonic bkg 

peaking bkg 

m µµ 

!34

H eff = 

New physics in b→s transitions? 

4 Analysis method 

Described by effective hamiltonian in operator product expansion 

b→s transitions sensitive to O (‘) 7,9,104 Analysis method 

4G F 

p 

2 

V tb V ⇤ 

ts 

2 

4 

i=1 

Decay B 0 →K*l - l + provides 

samples sufficiently large to 

measure observables sensitive to 

effective coefficients. E.g: 

One very famous variable: 

A FB ∝−Re[(2C eff 

7 + q2 

C eff 

m 2 9 )C 10 ] 

b 

requirement are greater thanselections 99%. and haveGeneral a reconstructed description B 0 compatible of Hamiltonian with the in operator generatedproduct B 0 in theexp 

ev 

to the number of events that pass the acceptance criteria. The compatibility of gen 

reconstructed p particles is enforced by requiring the reconstructed K + , p , µ + , and 

(Dh)2 +(Dj) 2 < 0.3 for hadrons and 0.004 for muons, where Dh and Dj are the 

The analysis measures A in h and j between the reconstructed and generated particles, and j is the azimuth 

FB , F L , and dB/dq 2 of the decay B 0 ! K ⇤0 µ + µ as a function of q 2 . 

Figure 1 shows the relevantthe angular planeobservables perpendicular needed to thetobeam define direction. the decay: Theq efficiency and purity of this co 

K is the angle 

requirement are greater than 99%. 

between the kaon momentum and the direction opposite to the B 

b → s transitions 0 B 0 in the K 

are sensitive ⇤0 K 

to O (′ ⇤0 ) 

7 , rest ) 

O(′ 9 , ) 

frame, q O(′ l is the angle between the positive (negative) muon momentum and the direction op- 

10 

posite to the B 0 B 0 in the dimuon rest frame, and f is the angle between the plane containing 

the two muons and the plane containing the kaon and pion. Since the extracted 

The analysis measures 3 i=1,2 Tree A FB , F L , and dB/dq 2 angular parameters 

A 

of the decay B 0 ! K ⇤0 µ + µ as a fun 

FB and F L and the acceptance times efficiency i-3-6,8 Gluon dopenguin 

not depend on f, f is integrated out. 

6X 

Although 

X 

the K 

C(µ)O(µ)+ 

+ p invariant 

Figure 

mass 

1 shows the relevant angular observables needed to define the decay: q 

(C(µ)O(µ)+C 0 must 

(µ)O 0 (µ)) 5bei=7 consistent Photon penguin with a K ⇤0 , there can be contributions 

K i 

from a spinless (S-wave) K between the kaon i=9,10 momentum Electroweak and penguin the direction opposite to the B 0 B 0 in the K ⇤0 

i=7,. . . ,10,P,S 

i=S Higgs (scalar) penguin 

LH LH + p combination. This is parametrized with two terms related to 

the S-wave fraction, F S , andframe, theRH 

interference q l is the angle between the positive (negative) muon momentum and the di 

i=P amplitude Pseudoscalar between penguin the S-wave and P-wave decays, 

A posite to B 0 B 0 S . Including this component, the angular distribution in the dimuon of rest B 0 ! frame, K ⇤0 µ and + µ fcan is the be written angle between the plane 

as [18]: 

the two muons and the plane containing the kaon and pion. Since the extracted angu 

eters A FB and F L and the acceptance times efficiency do not depend on f, f is inte 

1 d 3 G 

Although the K + p invariant mass must be consistent with a K ⇤0 , there can be co 

G dcos q K dcos 

from 

q l dq 

a 2 = 9 ⇢apple 2 

spinless 

16 

(S-wave) 

3 F S + 4 3 A K + S cos q K 1 cos 2 q l 

h p combination. This is parametrized with two term 

the S-wave + (1 fraction, F S ) 2F 

B 0 F S L , cos and 2 qthe → K ∗ K 

l + 1interference l − cos 2 amplitude SM between theBSM 

q l 

S-wave and 

cays, A S . Including this component, the is angular the most distribution prominent of(large B 0 ! statistic K ⇤0 µ and + µ flavou can 

as [18]: + 1 (1) 

2 (1 F f-integrated Studies L) 1 cos 2 q 

statistical K angular 1 + cos 

limited 2 distribution 

q l 

B s → φµ + µ − , Λ b → Λµ + µ 

1 d 3 G 

G dcos q K dcos q l dq 2 = 9 ⇢apple 2 

+ 4 i 

16 3 F S + 4 3 A S cos q K 1 cos 2 q 

3 A FB 1 cos 2 q K cos q l . 

l 

h 

+ (1 F S ) 2F L cos 2 q K 1 cos 2 q l 

Angular ϕanalysis 

forward 

μ − 

θl 

B 0 

backward 

θK 

+ 1 2 (1 F L) 1 cos 2 q K 1 + cos 2 q l 

+ 4 i 

3 A FB 1 cos 2 q K cos q l . 

q 2 = dimuon invariant mass 

ϕ 

μ + K + π − 

Introduce 3 relative angles to describe angular distribution of final state particles. 

Ci eff are linear combinations of C(m) 

Folding φ → φ + π if φ < 0 increase Figure sensitivity 1: Sketchfor showing some coefficients. 

the definition of the angular observables for the decay B 0 ! 

K ⇤0 (K + p )µ + µ . 

B 0 

V. Chiochia (Zürich University) – Heavy flavour physics with the e.g. CMS A FB 

experiment = 3 - 19-20/11/2013 

4 (1 − F L)A Re θl 

T 

μ − K + θK 

!35

B 0 →K*mm angular analysis 

K* longitudinal polarization Muon F-B asymmetry Differential branching fraction 

CMS data: arXiv:1308.3409 

SM: Phys. Rev. D 87 (2013) 034016 

Angular fit of the Kpmm system in bins of dimuon invariant mass q 2 

Signal yields ranging from 23 to 103 events in each bin 

Vertical shaded bands correspond to J/c and c(2S) resonances 

Results consistent with SM predictions and previous 

measurements 

World’s most precise measurement of AFB and FL at high q 2 !36 

Next: analyze 2012 samples and measure more angular variables 

see the localized excesses observed by LHCb [arXiv:1308.1707] and interpretations 

[e.g.: arXiv:1308.1501, arXiv:1309.2466, ...] 

V. Chiochia (Zürich University) – Heavy flavour physics with the CMS experiment - 19-20/11/2013

Summary 

The excellent performance of the CMS detector and LHC 

have allowed key contributions to the field heavy flavor 

physics 

Wide range of topics covered: Standard Model physics 

and the great beyond 

Several aspects of heavy flavor production are not yet 

well understood and spectroscopy is not a closed chapter! 

No sign of physics beyond the Standard Model yet, but we 

keep on searching! 

Analysis of the large datasets collected in 2012 is in 

progress. Expect new results (and perhaps surprises?) at 

the forthcoming conferences! 


!37

BACKUP SLIDES 


!38

Future upgrades 

The HL-LHC scenario 

40 fb -1 /year 

13/14 TeV 

>50 fb -1 /year 

13/14 TeV 

End of Phase 1 

300-500 fb -1 

PHASE 1 PHASE 2 

HL-LHC 

14 TeV 

>250 fb-1/year 

LHCb Upgrade 

ATLAS/CMS Upgrades phase 1 

ATLAS/CMS 

Upgrades phase 2 

At HL-LHC ~140 events/bx spread over ±15 cm (3 σ) 

~3000 fb -1 expected in 10 years running 

~3000 fb -1 expected in 10 years running 

Excellent opportunity to search for (rare) and new phenomena 


!39

in Table 2.3. Comparing the efficiencies at zero pileup with and without the dynamic ROC data 

Figure 2.1 shows a conceptual layout for the Phase 1 upgrade pixel detector. The current 3-layer 

loss expected for 2 ⇥ 10 

barrel (BPIX), 

Phase 

2-disk endcap (FPIX) 

1: 

system 

CMS 

is replaced with 

Pixel 

a 4-layer 34 cm 

barrel, 2 s 1 (25 ns crossing time), the dynamic data losses cause a only 

upgrade 

3-disk endcap 

3.5% (4.0%) loss of tracking efficiency for muons (t¯t) with the current pixel detector. However 

system for four hit coverage. Moreover the addition of the fourth barrel layer at a radius of 

at the pileup conditions for 2 ⇥ 10 

16 cm provides a safety margin in case the first silicon strip layer of the Tracker 34 cm 

Inner 2 s 

Barrel 

1 (25 ns crossing time) simulations show that the 

expected loss in efficiency due to the dynamic data loss increases to 8.6% (5.2%) for muons (t¯t). 

2.1. 

(TIB) 

Simulation 

degrades more 

Setup 

rapidly 

and Reconstruction 

than expected, but its main role is in providing redundancy in 

With much lower dynamic data loss for the upgrade 17 pixel detector, the resultant loss in tracking 

pattern recognition and reducing fake rates with high pile-up. 

efficiency is less than 0.5% in both pileup conditions. The track fake rate is hardly affected by 

the dynamic data loss. 

Table 2.2: Total material weight for the pixel barrel and forward pixel detectors, and for the carbon 

fiber tube outside of the pixel barrel thatTo is isolate neededthe foreffects Upgrade 

the TIB ofand highfor pileup, beamwe pipe canbakeout. 

compare the performance at zero pileup with those 

Upgrade 

Outer rings at high pileup without any ROC dynamic data loss simulated. The results in Table 2.3 for which 

=0 =0.5 =1.0 

=1.5 

=2.0 

Four barrel layers and three endcap 

Volume 

28 the ROC data loss was not implemented Chaptershow 2. Expected that thePerformance pileup conditions & Physics for 2Capabilities 

⇥ 10 

Mass (g) 

34 cm 2 s 1 

=2.5 (25 ns crossing time) by itself would cause a 7.3% disks (4.7%) on each loss of side tracking efficiency for muons 

Inner rings 

Present (t¯t) in Detector the current Phase pixel 1 detector. UpgradeThe Detector extra pixel layer in the upgrade detector adds information 

that reduces this loss in efficiency by more thanCarbon half to 3.2% 

fibre 

(1.3%) 

structure, 

for muons 

CO2 (t¯t), and decreases 

50.0 cm 

BPIX |h| < 2.16 16801 6686 

the fake rate by about a factor of two. evaporative cooling, first layer closer 

FPIX |h| < 2.50 Current 8582 7040 to interaction point (2.9 cm) 

=2.5 With both 3 barrel the layerseffects of dynamic data loss and the high pileup expected for 2 ⇥ 10 34 cm 2 s 1 

Current 

Barrel Outer Tube |h| < 2.16 9474 9474 Analog readout with on-chip 

=2.0 (25 ns crossing time), the loss in tracking efficiency for the current detector is 15.9% (9.9%) for 

=0 

=0.5 

=1.0 

=1.5 

muons (t¯t), while for the upgraded detector itdigitization 

is reduced by more than a factor of 4 (6) to 3.7% 

Figure 2.1: Left: Conceptual layout comparing(1.5%) the different for muons layers (t¯t). andAlthough disks in the thiscurrent not catastrophic, and the degradation is worse than linear 

New (Old): 79 (48) million pixel cells 

black 

upgrade 

points. 

pixel 

Note 

detectors. 

that the 

Right: 

“barrel 

Transverse-oblique 

outer tube” 

and 

mentioned 

is expected view comparing 

above 

to become 

and 

the 

in 

unacceptable pixel 

Table 

barrel 

2.2 is included 

layers at moderately in 

in 

higher pileup as illustrated in Figure 2.6. 

the two detectors. 

distributed over 1184 (768) detector 

the material budget for both present and upgrade 

The track 

pixel 

fake 

detectors 

rate also 

for 

rapidly 

the comparisons 

increases with 

shown 

pileup is about a factor of two lower for the 

upgrade pixel detector. 

modules 

in Figure 2.2. 

Pixels 

Pixels 

Current Pixel Detector 

Upgrade Pixel Detector 

100 50 

45 

0.7 

Current Pixel Detector δR, TTbar Upgrade Events Pixel Detector 

(a) 

90 0.18 40 

Current Detector No Pileup 

35 

Upgrade Detector No Pileup 

0.6 

0.16 

like the electronic boards and connections, out of the30 

80 tracking volume. 

0.5 

0.14 25 

20 

70 0.12 

15 

0.4 

detector and of the Phase 1 upgrade pixel detector. 0.1 10 

60 

5 

0.3 

0.08 

30 

50 0 10 20 30 40 50 60 70 80 90 100 

0.06 

for a0.2 

limited range in h that covers most of the tracking 2.5 region. Current Pixel Detector 

40 

0.04 2 Upgrade Pixel Detector 

radlen 

Transverse Resolution nuclen [um] 

4 barrel layers 

Since the extra pixel layer could easily increase the material of the pixel detector, the upgrade 

50 

detector, support, and services are redesigned to be lighter than the present system, using an45 

40 

ultra-lightweight support with CO 2 cooling, and by relocating much of the passive material, 

Average Tracking Efficiency (%) 

20 

Table 2.2 shows a comparison of the total material mass in the simulation of the present pixel 

15 

Since significant mass reduction was10 

5 

achieved by moving material further out in z from the interaction point, the masses are given 

Ratio 

0.1 

Also shown in Table 2.2 is the mass of the carbon 0.02 fiber 1.5 tube that sits outside of the pixel detector0and 

is needed by the Tracker Inner Barrel (TIB) 0 0 and 1020 20for3040 bakeout 40 5060 60 of the 7080 80 beampipe. 90100 

100 

1.5 

30 1 

0 1 

By0 10 0 20 30 20 40 50 40 60 70 60 80 90 80 100 100 

-3 -2 -1 0 1 2 3 

-3 -2 -1 Number 0of Tracks 1 Average 2Pileup 

3 

Number of Tracks 

Average Pileup 

eta 

eta 

convention, 

Radiation 

the material 

lengths 

for this tube is usually included as part of the pixel system “material 

Figure budget”; 2.2: this The tube amount is expected of material to remain the unchanged pixel 

Figure Tracking 

detector for 

2.6: 

the shown 

Average efficiency 

Phase in1 tracking 

units upgrade. of radiation 

efficiency 

length 

(a) and 

(left), 

average Primary trackvertex fake rate resolution 

50 

50 

(b) for the t¯t sample as 

45 

δZ, TTbar Events 

45 

δZ, TTbar Events 

and in units of nuclear interaction length (right) 

a function 

40 

as a function 

of the average 

of h; this 

Current 

pileup. 

is 

Detector 

given 

No 

Results 

Pileup 

for the 

were 

40 

current 

determined usingCurrent theDetector expected 

50 PileupROC data loss 

V. Another 

pixel 

Chiochia comparison 

detector 

(Zürich 

(green 

University) of the “material 

histogram), 

– 

and 

Heavy budget” 

the 

flavour for 

Phase 

expected 

1 

physics the 35 current 

upgrade 

for with each and 

detector 

the given Phase Upgrade 

CMS 1 Detector 

(black 

average experiment pixelNo detectors Pileup 

points). 

pileup, 

The 

- and 19-20/11/2013 

was 

shaded 

for 

35 

Upgrade Detector 50 Pileup 

the current pixel detector (blue squares) !40 and 

30 

30 

solution [um] 

Transverse Resolution [um] 

Ratio 

solution [um] 

35 

30 

25 

2.5 

Average Track Fake Rate (%) 

2 

25 

20 

15 

10 

5 

δR, TTbar Events 

Current Pixel Current Detector Detector 50 Pileup 

Upgrade Detector 50 Pileup 

Upgrade Pixel Detector 

03 0 10 20 30 40 50 60 70 80 90 100 

(b)

Bs(mm) BDT validation 

Use differences between data and MC for systematics 

B ± →J/ψ K ± 3% ; B s →J/ψ ϕ 9.5% (2011) and 3.5% (2012) 

4000 

3000 

2000 

1000 

400 

350 

300 

250 

200 

150 

100 

50 

CMS L = 5 fb ( s = 7 TeV) 

5000 

+ 

+ 

B → J/ψ K 

data 

MC simulation 

Barrel 

-1 -0.5 0 0.5 1 

BDT 

CMS L = 5 fb ( s = 7 TeV) 

450 

→ J/ψ φ B s 

data 

MC simulation 

-1 

-1 

1600 

1400 

1200 

1000 

800 

600 

400 

200 

CMS L = 5 fb ( s = 7 TeV) 

200 

180 

160 

140 

120 

100 

80 

60 

40 

20 

CMS L = 5 fb ( s = 7 TeV) 

1800 

+ 

+ 

B → J/ψ K 

data 

MC simulation 

Endcap 

-1 -0.5 0 0.5 1 

BDT 


data 

MC simulation 

-1 

-1 

22000 

20000 

18000 

16000 

14000 

12000 

10000 

8000 

6000 

4000 

2000 

1600 B s → J/ψ φ 

1400 

1200 

1000 

800 

600 

400 

200 

CMS L = 20 fb ( s = 8 TeV) 

+ 

+ 

B → J/ψ K 

data 

MC simulation 

Barrel 

-1 -0.5 0 0.5 1 

BDT 

CMS L = 20 fb ( s = 8 TeV) 

data 

MC simulation 

-1 

-1 

+ 

4000 B 

+ → J/ψ K 

3500 

3000 

2500 

2000 

1500 

1000 

500 

-1 -0.5 0 0.5 1 

BDT 

CMS 

400 

-1 

L = 20 fb ( s = 8 TeV) 

350 

300 

250 

200 

150 

100 

50 

CMS L = 20 fb ( s = 8 TeV) 

data 

MC simulation 

Endcap 


data 

MC simulation 

-1 

-1 -0.5 0 0.5 1 

BDT 

-1 -0.5 0 0.5 1 

BDT 

-1 -0.5 0 0.5 1 

BDT 

-1 -0.5 0 0.5 1 

BDT 


!41

Lifetime measurements 


Lb lifetime 

Events / ( 0.006 GeV ) 

-n fit )/ 

250 

200 

150 

100 

obs 

(n 

[ps] 

τ Λb 

50 

b 

CMS 

s = 7 TeV L = 5 fb 

Fit function 

Signal 

Prompt background 

45.4 5.5 5.6 5.7 5.8 5.9 6 

3 

Invariant mass J/ [GeV] 

2 

1 

0 

-1 

-2 

-1 

Nonprompt background 

5.4 5.5 5.6 5.7 5.8 5.9 6 

Invariant mass J/ [GeV] 

1.6 

1.4 

1.2 

Lb→J/cL 

Events / ( 0.16 ps ) 

-n fit )/ 

obs 

(n 

3 

10 

b 

10 

2 

10 

1 

4 

3 

2 

1 

0 

-1 

-2 

-3 

CMS 

s = 7 TeV L = 5 fb 

Fit function 

Signal 

Prompt background 

0 2 4 6 8 10 12 14 

Proper decay time [ps] 

-1 

Nonprompt background 

0 2 4 6 8 10 12 14 

Proper decay time [ps] 

0.7 

1.0 

PDG 2012 average 

Not used in PDG 2012 average 

0.6 

0.8 

Used in PDG 2012 average 0.5 

CMS - this measurement 

0.6 

0.4 

1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 

Year of publication 

1.1 

1.0 

0.9 

0.8 

,PDG2012 

0 

B 

/τ 

τ Λb 

Lb baryon lifetime in J/cL decays 

Theoretical predictions in tension with earlier 

measurements 

Use dimuon trigger without lifetime 

significance cut. 

Lifetime from combined maximum-likelihood 

mass and proper decay time fit 

About 1000 signal events 

t(Lb)=1.503 ± 0.052(stat.) ± 0.031(syst.) ps 

World average: 1.425 ± 0.032 ps 

B 0 lifetime determined in J/cKs decays as 

cross check compatible with PDG 

Dominant systematic uncertainty from proper 

decay time efficiency 

Compatible with ATLAS determination and 

world average 

JHEP 07 (2013) 163 


!43

Bs lifetime difference 

-2 

-1012345 

-3 

-4 

-5 

pull 

0.05 0.1 0.15 0.2 0.25 0.3 

proper decay length [cm] 

B s 

troduction 

Bs lifetime difference DGs=GH-GL 

Figure 3: Proper decay length projection of the five-dimensional maximum CMS PAS likelihood BPH-11-006 fit to 

the data. The points are the data distribution, the solid blue line shows the full fit, the green 

Events / ( 0.0045 GeV ) 

2200 CMS preliminary, 5 fb -1 

-1 

3 

CMS preliminary, 5 fb 

1 

10 

2000 s = 7 TeV 

Data 

s = 7 TeV 

1800 

Signal 

Data 

dash-dotted line is the signal model component, with the CP-even (odd) component Signal shown 

Background 

Background 

1600 

in blue short dashed line (magenta 

~15k 

dash-dot-dotted Fit line), and the red long dashed CP line even is the 

1400 

CP odd 

2 

10 

Fit 

SM expectation DGs/Gs= 0.12±0.06 

signal 

¯B s). The resulting mass eigenstates background from mixing component. The pull between the proper decay length distribution and the fit is 

Selecting decays to J/cf candidates, 

candidates Bs→J/cf 

with J/c→mm and f→KK. 

cay of a B s meson is characterized by the possibility that it may go through the mixeen 

its two flavour eigenstates (B s 

ected to have sizeable mass and decay width difference. Since the CP-violating mixing 

s expected to be small in the Standard Model, the two mass eigenstates are approxiequal 

to CP eigenstates. The light mass eigenstate is expected to be a CP-even state 

shorter lifetime than the heavy mass eigenstate which is a CP-odd state [1]. The decay 

ifference of the two B s eigenstates is found to be several percent of the mean decay 

. In the decay B s ! J/yf with J/y ! µ + µ and f ! K + K , the final state is an admixthe 

CP-even and CP-odd eigenstates. A measurement of the lifetime difference (DG s) of 

mass eigenstates can be performed using the proper decay time distribution of B s me- 

]. The additional information required to separate the light and heavy mass eigenstates 

btained from the angular distribution of the final decay products. 

1200 

1000 

shown in the histogram below. 

800 0.05 0.1 0.15 0.2 0.25 0.3 

10 

600 

uncertainty is evaluated by 400 generating pseudo-data with the modified scale factor and fitting 

5D fit of mass, proper decay 200 

with the time nominal and value. 3 

0 

angular variables to separate CP states 

To evaluate the effect of assuming f s to be zero in the fitting model, pseudo-experiments have 

-2 

0.048 ± 0.024 (stat.) ± 0.003 (syst.) ps -1 

-1012345 

-2 

-3 

-3 

been generated with f -1012345 

s = 0.04 -4 rad, which is the Standard Model -4 expected value. Similarly, 

-5 

-5 

5.25 5.3 5.35 5.4 5.45 

0.05 0.1 0.15 0.2 0.25 0.3 

The two CP 

+ - 

Invariant mass J/ψ K K [GeV] 

B proper decay length [cm] 

Compatible with world average 

the systematic uncertainties due to non-inclusion of S-wave components in the signal 

s 

PDF is 

estimated using pseudo-experiments by adding the model components of the S-wave by setting 

the |A s | 2 to 0.03 and the phase equal to previous measurements. In both cases, the pull mean 

e B s is a pseudo-scalar meson while the J/y and f are vector mesons, the orbital angular 

tum of the two decay products can have the values L = 0, 1, and 2. 

ents can be statistically disentangled by measuring the angular distributions of the 

cay products. A set of three angles Q =(q T , y T , j T ) in the transversity basis [4], as 

ted in Fig. 1, is used to describe the decay topology. Here, q T and j T are the polar and 

hal angles of the µ + in the rest frame of the J/y respectively, where the x-axis is defined 

irection of the B s and the xy-plane by the decay plane of f ! K + K . The helicity angle 

e angle of the K + in the f rest frame with respect to the negative B s flight direction. 

T 

0 

-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 

cos(ψ ) 

T 

1: Definition of the three angles, (q T , y T and j T ), used for the description of the decay 

y. 

Events / ( 0.1 ) 

1600 

1400 

1200 

1000 

800 

600 

400 

200 

-1 


pull 

s = 7 TeV 

1200 

5.25 5.3 5.35 5.4 5.45 

0 

-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 

cos(θ T ) 

1200 0.05 0.1 0.15 0.2 0.25 0.3 

Figure 4: The angular projection of the five-dimensional maximum likelihood fit to the data. 

V. Chiochia (Zürich University) – Heavy Theflavour plots show physics thewith projections the CMS onexperiment the angular- variables 19-20/11/2013 cos(y T ), cos(q T ) and the angle f. The !44 

ay of vector mesons is further described by the time evolution of three different ampliith 

different angular dependencies. The amplitudes at time t = 0 are defined using the 

dinal component A 0 (0) for L = 0, which is CP-even, and the transverse components 

T 

T 

Events / ( 0.1 ) 

1000 

800 

600 

400 

200 

-1 


s = 7 TeV 

Events / ( 0.0034 cm ) 

pull 

Events / ( π/10) 

1400 

1000 

800 

600 

400 

200 

-1 


s = 7 TeV 

0 

-3 -2 -1 0 1 2 3 

φ [rad] 

T

Summary of CPV in Bs decays 

Rather consistent picture from Tevatron and LHC experiments 

Closing on the SM value also in this case 

CMS currently working on the determination of the CP violating phase fs 

0.25 

0.20 

0.15 

LHCb 1.0 fb –1 + CDF 9.6 fb –1 + D 8 fb – 1 

+ ATLAS 4.9 fb – 1 

D 

HFAG 

Fall 2012 

68% CL contours 

( ) 

LHCb 

0.10 

Combined 

CMS 

0.05 

CDF 

SM 

ATLAS 

0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 


!45

Baryon and meson 

spectroscopy 


Heavy baryons with beauty 

In heavy quark effective theory an heavy baryon is made of 

Static color field from heavy quark 

B baryon spectrum 

Cloud corresponding to the light di-quark system 

“Hydrogen atom of QCD” 

This talk 

Q 

q 

q 

Thomas Kuhr, EKP, Uni KA Lattice QCD Workshop, Dec. 11 2007 

page 4 

Antisymmetric flavor configuration [q1,q2]: L-type baryons 

Symmetric flavor configuration {q1,q2}: S-type baryons 

Lb 0 =|bdu> 

Sb=|buu> 

qi is a strange quark: Jb family - both qi are strange quarks: Vb - 


!47

Jb - selection 

Ξb − selection 

• Full 2011 dataset (5.3 fb −1 ) 

• Ξb − reconstruction: 

• Displaced + prompt J/ψ trigger. 

Muons matched to trigger objects. 

p + 

⇡ ⇤ 

• Two tracks forming a very displaced 

vertex. Proton: highest |p| track. 

• Mass constrained kinematic fit 

• J/ψ mass constr. kinematic fit 

• Same track 

not used twice 

• Choose closest PV, in 3D, 

to Ξb − trajectory. 

⌅ 

µ + 

⇤ 0 

20 MeV Ks mass veto 

J/ 

PV 

µ 

PDG: 5790.5 ± 2.7 MeV 

V. Chiochia Ernest (Zürich Aguiló University) (HEPHY) – Heavy flavour physics with the CMS 8 experiment - 19-20/11/2013 June 26th 2012!48







p + 

⇡ ⇤ 



• Mass constrained kinematic fits 






⌅ 

µ + 

⇤ 0 

⇡ ⌅ 


20 MeV Ω − mass veto 

To save CPU time, before the 

kinematic fit: 

PV 

J/ 

µ 

• |Minv(Λπ) −MPDG(Ξ)| < 50 MeV 

• 5.5 GeV < Minv(J/ψΞ) < 6.2 GeV 

V. Chiochia Ernest (Zürich Aguiló University) (HEPHY) – Heavy flavour physics with the CMS 9 experiment - 19-20/11/2013 June 26th 2012 !49







p + 

⇡ ⇤ 



• Mass constrained kinematic fits 






⌅ 

µ + 

⇤ 0 


20 MeV Ω − mass veto 

⇡ ⌅ 

To save CPU time, before the 

kinematic fit: 

PV 

⌅ b 

J/ 

µ 

• |Minv(Λπ) −MPDG(Ξ)| < 50 MeV 

• 5.5 GeV < Minv(J/ψΞ) < 6.2 GeV 

PDG: 5790.5 ± 2.7 MeV 

V. Chiochia Ernest (Zürich Aguiló University) (HEPHY) – Heavy flavour physics with the 10 CMS experiment - 19-20/11/2013 June 26th 2012 !50


100 

80 

60 

3 

×10 

X(3872) production 

1 

10 

T 

|y| < 1.2 

Candidates / 3.125 MeV 

3 

×10 

16 

14 

12 

CMS 

s = 7 TeV 

-1 

L = 4.8 fb 

data 

total fit 

background 

signal 

Β [nb/GeV] 

⋅ 

/dp 

T 

prompt 

X(3872) 

dσ 

R fiducial 

10 

10 

Β [nb/GeV] 

-1 

⋅ 

/dp 

-2 

0.12 

0.11 

0.1 

0.09 

10 

0.08 

T 

prompt 

X(3872) 

LO NRQCD CMS s = 7 TeV 

LO NRQCD uncertainty -1 

L = 4.8 fb 

CMS CMS s = s7 = TeV 7 TeV 

|y| < 1.2 

-1 -1 

L = L 4.8 = 4.8 fb fb 

|y| < LO |y| 1.2 < NRQCD 1.2 

LO NRQCD uncertainty 

LO NRQCD 

LO NRQCD uncertainty 

40 

0.07 

0.07 

10 

-2 

3.75 3.8 3.85 3.9 3.95 4 10 10 15 20 

m(J/ψ π + π - ) [GeV] 

- 

0.06 

(J/ψ π + 25 30 

p π ) [GeV] 

0.06 

-2 

T 

20 

10 

Figure 

11910 

6: Measured 

candidates 

differential cross 0.05 section for prompt X(3872) production times 0.05branching 

10 15 20 

- 

(J/ψ π 

0 

+ 25 30 

fraction B(X(3872) ! J/yp + p ) as a function of p T . The inner p error bars π ) [GeV] indicate the statistical 

0.04 

T 

0.04 

3.6 3.7 uncertainty 3.8 and3.9 the outer error 4 bars represent 10 10 20 the 15total30 uncertainty. 20 

- 

p (J/ψ π + 40 50 10 20 30 

- 

) [GeV] 

p (J/ψ π + 40 50 

- 

m(J/ψ π + - 

π ) [GeV] 

(J/ψ π + 25Predictions 30 from a NRQCD 

model [11] Figure are shown 6: Measured by the solid differential line, withcross the dotted sectionp 

lines for prompt representing π ) X(3872) [GeV] theproduction uncertainty. times Thebranching 

π ) [GeV] 

T T 

T 

data pointsfraction are placed B(X(3872) where the ! J/yp value + p of the ) astheoretical a function of prediction p T . The inner is equal error tobars its mean indicate value the statistical 

uncertainty and Figure outer3: error Ratios barsof represent the X(3872) the total anduncertainty. y(2S) crossPredictions sections times from branching a NRQCD fractions, without 

Important measurement 

over each bin, according 

to better 

to the 

(R fiducial 

constraint 

prescription in 

, left) andthe [28]. 

with particle (R, right) nature 

model [11] are shown by the solid line, with the dotted acceptance lines representing corrections thefor uncertainty. the muonThe 

and pion pairs, as a 

function of p T . The inner error bars indicate the statistical uncertainty and the outer error bars 

Measured at central data rapidities points are using placed where J/c(mm)pp the valuedecays of the theoretical and assuming prediction Jequal PC =1 to ++ its mean value 

these results, over theeach measured bin, according integrated represent tocross the prescription total section uncertainty. for prompt [28]. The X(3872) data points production are placed timesat branching 

section fractionlargely is: dominated by prompt production 

the centre of each p T bin. 

Total cross (~75%) 

dσ 

Β [nb/GeV] 

⋅ 

/dp 

1 

T 

-1 

prompt 

X(3872) 

dσ 

10 

1 

-1 

CMS s = 7 TeV 

-1 

L = 4.8 fb 

|y| < 1.2 

Ratio to c(2S) cross these section results, and the measured Table non-prompt 4: integrated Relativefraction variations, cross section independent for percent, prompt ofX(3872) the on integrated pT production cross times section branching 

! X(3872)+anything) 

ratio R for different 

X(3872) and y(2S) polarization hypotheses: transversely (longitudinally) polarized J/y are 

NRQCD underestimates s prompt (pp 

fraction is: 

the measured denoted 

·B(X(3872) 

as cross CST (CSL) section ! J/yp 

in the + p 

Collins–Soper 

)=1.06 ± 0.11 

frame 

(stat.) 

and 

± 0.15 

HXT 

(syst.) 

(HXL) 

nb. 

in the helicity frame. Unpolarized 

scenarios (labelled unpol) are also included. 

This result assumes 

s prompt that the X(3872) and y(2S) states are unpolarized. 

(pp ! X(3872)+anything) ·B(X(3872) Polarization ! J/yp + The NRQCD prediction 

Prompt cross section Relative Polarization Relative 

for the prompt = X(3872) cross section times branching 

X(3872) 

fraction 

y(2S) 

in the p kinematic )=1.06 ± 

shifts (%) X(3872) 

region 0.11 (stat.) 

y(2S) 

of this ± 0.15 (syst.) nb. 

shifts (%) 

NRQCD analysis= 

is 4.01 ± 0.88 nb [11],(10 significantly < pT < 30 GeV, above CST |y| the < 1.2) measured CSL (>3s value. larger 28than data) CST unpol 8 

Non-prompt fraction = This 0.263±0.028 result assumes that the X(3872) and CSL y(2S) CST states are unpolarized. +31 CSLThe NRQCD unpol prediction +22 

for the prompt X(3872) cross sectionHXT times branching HXL fraction +86 in the HXTkinematic unpol region +28 of this 

7 Measurement analysis is 4.01 of ± 0.88 thenb p [11], + psignificantly invariant-mass HXL above HXT the measured distribution 

49 value. HXL unpol 31 

CST CST 1 unpol CST +8 

The decay properties of the X(3872) are further investigated with a measurement of the p 

V. Chiochia (Zürich University) – Heavy flavour physics with the CMS CSL experiment CSL - 19-20/11/2013 

5 unpol CSL + p 

22 

invariant-mass distribution from X(3872) decays !51 

HXT 

to J/yp + HXT 

p . Here, the 

6 

same event 

unpol 

selection 

HXT 

as 

27 

R 

0.12 

0.11 

0.1 

0.09 

0.08 

7 Measurement of the p + p invariant-mass distribution 

13 

JHEP 04 (2013) 154 

CMS s = 7 TeV 

-1 

L = 4.8 fb 

|y| < 1.2 

9

Summary of X(4140) searches 

M1 (MeV) G1 (MeV) M2 (MeV) G2 (MeV) 

Belle - - 4350 +4.6 -5.1±0.7 13 +18 -9±4 

CDF 

4143.0 +2.9 -3.0±0.6 15.3 +10.4 -6.1±2.5 4274.4 +8.4 -6.7±1.9 32.3+21.9±7.6 

CMS 4148.0±2.4±6.3 28 +15 -11±19 4313.8±5.3±7.3 38 +30 -16±16 

D0 4159.0±4.3±6.6 19.9±12.6 +1.0 -8.0 4328.5±12.0 30 (constrained) 

LHCb - - - - 

>5s >3s N.A. 


!52

Slides - Physics Seminar - Desy

Create successful ePaper yourself

Delete template?

Save as template?