Selected results on heavy flavour physics at LHCb Matthew CHARLES - - PowerPoint PPT Presentation

SLIDE 1

Selected results on heavy
 flavour physics at LHCb

Matthew CHARLES (UPMC/LPNHE)

1

!

SLIDE 2

Plan

• Quick intro to heavy flavour physics & LHCb
• Searches for CP violation in charm decays
• Baryon spectroscopy
• Onward and upward

2

SLIDE 3

Introduction

3

SLIDE 4

Particle Physics today

• Our strawman theory, the Standard Model,

works embarrassingly well.

• We know that it's incomplete, that there is

New Physics to discover.

• Matter-antimatter asymmetry; dark matter; dark

energy; neutrino masses; gravity; etc

• Our collective goal today: discover that NP

.

• Our goal for tomorrow: understand it.

4

SLIDE 5

The Standard Model

The fundamental particles...

5

u c t d s b e μ τ νe νμ ντ γ W± Z g H

quarks leptons bosons

... and how they interact

[... plus their antiparticles]

fermions

SLIDE 6

Chirality

• In the Standard Model, chirality is a big deal.
• The SU(2) -- i.e. weak -- interaction only talks to left-handed

fermions.

• So in the SM, each quark generation is represented as
• a doublet of left-handed particles with SU(2) interactions
• two singlet right-handed particles
• These weak flavour eigenstates are different from the mass

eigenstates, but can be expressed as superpositions of them.

• Phase convention: up-type quarks are aligned. Then:
• { u, c, t } are also mass eigenstates
• {d', s', b' } are linear combinations of mass eigenstates...

6

• u

d′

• L

,

• c

s′

• L

,

• t

b′

• L

weak isospin −1/2 weak isospin +1/2

SLIDE 7

The CKM Matrix

• Write linear relation between mass eigenstates (d,s,b) and

flavour eigenstates (d',s',b') as a matrix VCKM:

• Notation:
• The complex elements of this matrix are free parameters

in the SM and have to be determined experimentally.

• Unitarity constraints & removal of unphysical phases => 4 actual free params

7

  d0 s0 b0   = VCKM   d s b  

VCKM =

⎛ ⎜ ⎝

Vud Vus Vub Vcd Vcs Vcb Vtd Vts Vtb

⎞ ⎟ ⎠

SLIDE 8

The CKM Matrix

8

VCKM =

⎛ ⎜ ⎝

Vud Vus Vub Vcd Vcs Vcb Vtd Vts Vtb

⎞ ⎟ ⎠

Image by Anna Phan, Quantum Diaries.
 http://www.quantumdiaries.org/2012/05/10/needle-in-a-haystack/

± VCKM = ⎛ ⎝ 0.97427 ± 0.00014 0.22536 ± 0.00061 0.00355 ± 0.00015 0.22522 ± 0.00061 0.97343 ± 0.00015 0.0414 ± 0.0012 0.00886+0.00033

−0.00032

0.0405+0.0011

−0.0012

0.99914 ± 0.00005 ⎞ ⎠ Current best-fit magnitudes, from PDG 2014:

These two elements have non-tiny complex phases

SLIDE 9

CP violation

• C = charge conjugation
• P = parity
• Weak interaction violates C and P

.

• It can also violate CP

. This occurs when a process and its CP conjugate have different rates, for example:
 Γ(D → f) ≠ Γ(D → f)

• How can this happen?
• It's always, always an interference effect. For example...

9

SLIDE 10

CPV

10

¯ u c u ¯ u s ¯ s ¯ u c s ¯ u u ¯ s

W +

SM tree SM penguin T = |T|eiθteiφt | | P = |P|eiθpeiφp ≡ r|T| ei(θt+∆θ)ei(φt+∆φ) Total amplitudes: aD = T + P = |T|eiθteiφt 1 + rei∆θei∆φ aD = T + P = |T|eiθte−iφt 1 + rei∆θe−i∆φ Difference in the rates prop. to: |aD|2 − |aD|2 = −4r sin ∆θ sin ∆φ Asymmetry:

A = |aD|2 |aD|2 |aD|2 + |aD|2 = 4r sin ∆θ sin ∆φ 2 + O(r) ' 2r sin ∆θ sin ∆φ if r ⌧ 1

CPV requires strong and weak phase differences.

SLIDE 11

CPV

11

¯ u c u ¯ u s ¯ s ¯ u c s ¯ u u ¯ s

W +

SM tree SM penguin

¯ u c s ¯ u u ¯ s ˜ e

NP penguin ... and NP could contribute, changing the amount of CPV. Thus, test for NP: Does CP asymmetry match SM expectation? Asymmetry:

A = |aD|2 |aD|2 |aD|2 + |aD|2 = 4r sin ∆θ sin ∆φ 2 + O(r) ' 2r sin ∆θ sin ∆φ if r ⌧ 1

SLIDE 12

CPV in charm

• There's a lot more to say about CPV...
• ... but for today we'll focus on one specific type:


direct CP violation in SCS decays of charm mesons.

• Of interest because expected CP asymmetries in

the SM are small, but can often be enhanced by NP .

• How small? Generically up to O(10−3) in direct CPV.


Much theory progress on this in recent years.

• Thus: a sensitive probe of NP

.

• Very well suited to LHCb.

12 Grossman, Kagan & Nir, PRD 75, 036008 (2007) Bianco, Fabbri, Benson & Bigi, Riv. Nuovo. Cim 26N7 (2003) Bigi, arXiv:0907.2950 Bobrowski, Lenz, Riedl & Rorhwild, JHEP 03 009 (2010) Bigi, Blanke, Buras & Recksiegel, JHEP 0907 097 (2009) Brod, Kagan & Zupan, Phys.Rev. D86 014023 (2012) Gedalia, Kamenik, Ligeti & Perez, PLB 714 55 (2012) Giudice, Isidori & Paradisi, JHEP 1204 060 (2012) Hiller, Hochberg & Nir, Phys.Rev. D85 116008 (2012) etc etc etc

SLIDE 13

LHCb

13

SLIDE 14

LHC

14

SLIDE 15

LHCb

15

M1 M3 M2 M4 M5 RICH2 HCAL ECAL SPD/PS Magnet z 5m y 5m 10m 15m 20m TT T1 T2 T3 Vertex Locator

SLIDE 16

LHCb data-taking in Run 1...

16

8 TeV: 2 fb−1 7 TeV: 1 fb−1 7 TeV: 0.035 pb−1

SLIDE 17

As of June 8

Photo (c) Zefram

... and Run 2

17

13 TeV 13 TeV

SLIDE 18

Why LHCb?

• Very nice detector (for charged tracks)
• Precision vertexing & tracking
• Hadron & muon ID across wide momentum range
• Huge statistics! Cross-sections in acceptance:
• σbb(7 TeV) = 49 ± 7 μb => 20kHz of bb
• σcc(7 TeV) = 1419 ± 134 μb => 600 kHz of cc
• σbb(13 TeV) = 101 ± 10 μb => 40kHz of bb
• σcc(13 TeV) = 2940 ± 240 μb => 1.2 MHz of cc

18

Eur.Phys.J. C71 (2011) 1645 ; Nucl.Phys.B 871, 1 ; JHEP 1510 (2015) 172 ; JHEP 1603 (2016) 159

SLIDE 19

Overflowing with charm & beauty

• In Run 1 we got 600 kHz of charm, and in Run 2

this rises to 1.2 MHz!

• We can't keep all of this (or even read it all out)
• Use trigger system to select the events that are

cleanest and most interesting for physics.

• Hardware (L0): fast but crude decision, 1 MHz output
• Software (HLT): inclusive, then exclusive reconstruction
• Big challenge & goal for the upgrade: making the

trigger smarter (and the output leaner) so that we can save keep the physics efficiency up.

19

SLIDE 20

Search for CPV in D+ → K− K+ π+

• Phys. Rev. D 84, 112008 (2011) -- 35 pb−1

20

SLIDE 21

D+ → K− K+ π+: Why? How?

• SCS charm decay, CP asymmetry sensitive to NP
• Could measure asymmetry integrated across PHSP

...

• ... but we can be smarter.
• These are 3-body decays, so there are many

interfering contributions

• e.g. D+ → 𝜚π+, K*0K+, a0(1430)0π+, ...
• NP might show up in some but not others.
• Strong phase varies across the phase space.

21

A = B(D+ → K−K+π+) − B(D− → K+K−π−) B(D+ → K−K+π+) + B(D− → K+K−π−).

SLIDE 22

D+ → K− K+ π+: Dalitz plot

22

)

4

/c

2

(GeV

2

+

π

• K

m

0.5 1 1.5 2

)

4

/c

2

(GeV

2

+

K

• K

m

1 1.5 2 2.5 3

1 10

2

10

3

10

LHCb

PHSP density is uniform, so all structure* is due to |A|2.

* Neglecting efficiency variation, background, etc.

SLIDE 23

D+ → K− K+ π+: Asymmetries

• Idea: divide D+ and D− Dalitz plots into bins, count

yield in each, check for differences in distribution.

• Any significant difference => CP violation!
• Work with normalised yields so that overall effects,

e.g. σ(D+) ≠ σ(D−), cancel out.

• To test significance, calculate figure of merit:



 
 
 with NDF = Nbins − 1.

• Careful choice of binning necessary for sensitivity.


Studied with toy MC.

23

χ2 =

Nbins

X

i

✓difference in normalised yields in bin i uncertainty on the difference ◆2

The per-bin significance, SCP

,i

SLIDE 24

D+ → K− K+ π+: Binnings

24

One binning, and (jumping ahead) the per-bin significances:

SLIDE 25

D+ → K− K+ π+: Validation with MC

• Generated toy MC (isobar model from CLEO-c)
• Baseline: no CP asymmetries generated
• Verified that we don't produce false positives
• Effect of K+/K− detector efficiency asymmetry included
• Then tested sensitivity by introducing CPV:

25

CPV Adaptive I Adaptive II p(3σ) hSi p(3σ) hSi no CPV 0% 0.84σ 1% 0.84σ 2 in φ(1020) phase 5% 1.6σ 2% 1.2σ 3 in φ(1020) phase 38% 2.8σ 12% 1.9σ 4 in φ(1020) phase 76% 3.8σ 41% 2.7σ 5 in φ(1020) phase 97% 5.5σ 79% 3.8σ 6 in φ(1020) phase 99% 7.0σ 98% 5.2σ 6.3% in κ(800) magnitude 16% 1.9σ 24% 2.2σ 11% in κ(800) magnitude 83% 4.2σ 95% 5.6σ

SLIDE 26

D+ → K− K+ π+: The data

• Key control sample: Ds+ → K− π+ π+
• Also used D+ → K− π+ π+ ; sidebands of both modes

26

)

2

(MeV/c

+

π

+

π

• K

m 1800 1850 1900 )

2

Events / ( 0.28 MeV/c 20000 40000

lower upper

+

D

LHCb (a)

)

2

(MeV/c

+

π

+

K

• K

m 1800 1850 1900 1950 2000 )

2

Events / ( 0.48 MeV/c 5000 10000 15000

lower middle upper

+

D

+ s

D

LHCb (b)

CF control mode
 3760k D+ → K− π+ π+ Signal mode
 370k D+ → K− π+ π+ CF control mode
 515k Ds+ → K− π+ π+

2010 data only (0.035 fb−1 at 7 TeV)

SLIDE 27

D+ → K− K+ π+: The data

• Validated that we do not see CPV in control samples
• Various checks for detector effects, not discussed here
• Results from D+ → K− K+ π+ signal mode:

27

)

4

/c

2

(GeV

2

+

π

• K

m

0.5 1 1.5 2

)

4

/c

2

(GeV

2

+

K

• K

m

1 1.5 2 2.5 3

CP

S

• 3
• 2
• 1

1 2 3

LHCb (a)

)

4

/c

2

(GeV

2

+

π

• K

m

0.5 1 1.5 2

)

4

/c

2

(GeV

2

+

K

• K

m

1 1.5 2 2.5 3

CP

S

• 3
• 2
• 1

1 2 3

LHCb (b)

)

4

/c

2

(GeV

2

+

π

• K

m

0.5 1 1.5 2

)

4

/c

2

(GeV

2

+

K

• K

m

1 1.5 2 2.5 3

CP

S

• 3
• 2
• 1

1 2 3

LHCb (c)

)

4

/c

2

(GeV

2

+

π

• K

m

0.5 1 1.5 2

)

4

/c

2

(GeV

2

+

K

• K

m

1 1.5 2 2.5 3

CP

S

• 3
• 2
• 1

1 2 3

LHCb (d)

Per-bin asymmetry significance

Binning scheme χ2/ndf p-value (%) 25 adaptive 32.0/24 12.7 106 adaptive 126.1/105 7.9 199 uniform 191.3/198 82.1 530 uniform 519.5/529 60.5

(a) (b) (c) (d)

Sadly, consistent with no CPV.

SLIDE 28

D+ → K− K+ π+: Afterword

• Result just shown was the first CPV analysis from LHCb.
• Similar method was used later (not by me) for


B+ → h+ h− h+ modes.

28

N raw

A

• 1
• 0.8
• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 1

]

4

c /

2

[GeV

low

)

−

K

+

(K

2

m

5 10 15

]

4

c /

2

[GeV

high

)

−

K

+

(K

2

m

5 10 15 20 25

N raw

A

• 1
• 0.8
• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 1

LHCb

(a)

N raw

A

• 1
• 0.8
• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 1

]

4

c /

2

) [GeV

−

π

+

π (

2

m

5 10 15 20

]

4

c /

2

) [GeV

−

π

+

(K

2

m

5 10 15 20 25

N raw

A

• 1
• 0.8
• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 1

LHCb

(b)

N raw

A

• 1
• 0.8
• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 1

]

4

c /

2

[GeV

low

)

−

π

+

π (

2

m

5 10 15

]

4

c /

2

[GeV

high

)

−

π

+

π (

2

m

5 10 15 20 25

N raw

A

• 1
• 0.8
• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 1

LHCb

(c)

N raw

A

• 1
• 0.8
• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 1

]

4

c /

2

) [GeV

−

K

+

(K

2

m

10 20

]

4

c /

2

) [GeV

−

π

+

(K

2

m

5 10 15 20 25

N raw

A

• 1
• 0.8
• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 1

LHCb

(d)

Figure 3: (colour online) Measured AN

raw in Dalitz plot bins of background-subtracted and

acceptance-corrected events for (a) B± → K±K+K−, (b) B± → K±π+π−, (c) B± → π±π+π− and (d) B± → π±K+K− decays.

• Very large local

asymmetries seen!

• Demonstrates that

method is sensitive.

• Begs the question: why

don't we see similar CPV in D+ decays?

• Weak phase difference

must be small across the whole D+ PHSP .

• Phys. Rev. D 90, 112004 (2014)

SLIDE 29

Search for CPV in D0 → K− K+, π− π+

LHCb-CONF-2011-023 -- 35 pb−1

• Phys. Rev. Lett. 108 (2012) 111602 -- 0.6 fb−1

LHCb-CONF-2013-003 -- 1 fb−1

• Phys. Rev. Lett. 116, 191601 (2016) -- 3 fb−1

29

SLIDE 30

How many different papers?!

• Same basic method, four progressively larger data

sets (from 0.035 fb−1 to 3 fb−1)

• Some improvements over time as we learned more. Will

mostly elide them here for simplicity.

• Report focused on 1 fb−1, since:
• It was most recent public result at the time of writing
• Other people did most of the heavy lifting for 3 fb−1
• Today will use 1 fb−1 to present the method, then

give a round-up of all four results.

• Will also discuss complementary measurements

with D0 from B decays (muon-tagged).

30

SLIDE 31

ΔACP: The observable

• We want to study the CP asymmetries



 
 for f = K− K+, π− π+

• Because f = f, we need to tag the initial flavour
• Two methods, D*+ tag & muon tag:

31

Af = Γ(D → f) − Γ(D → ¯ f) Γ(D → f) + Γ(D → ¯ f)

π+ K+ K− D0 D*+ μ− K+ K− D0 ν̄, ...

pp collision pp collision

B̄

Charge of the pion or muon tags the flavour of the D0.

D∗+ → D0π+ B → D0µ−¯ νµ(X)

SLIDE 32

ΔACP: The observable

• What we measure is the raw (yield) asymmetry:





• ... which includes the physics asymmetry ACP but also

nuisance asymmetries (production, efficiency)

• Provided these are small, we can Taylor expand:
• ... and then cancel them in the difference:





• Method very robust against systematic effects.

32

Araw(f) ' ACP (f) + Aprod + Aeff(f) + Aeff(tag) Araw(f) = N(D → f) − N(D → f) N(D → f) + N(D → f)

Araw(f) Araw(f 0) ' ACP (f) ACP (f 0) ⌘ ∆ACP

cancels (f=f)

SLIDE 33

ΔACP: Subtleties

• Sketch on previous slide covers the key points.
• Full formalism (skipped!) handles some higher-order

effects and gets to the same place in the end:

• D0 mixing cancels; time-dependent CPV mostly cancels
• Second-order effects from kinematic correlations between

nuisance asymmetries handled by binning or reweighting in kinematic variables

• Exclude edges of detector with large local asymmetry
• Will touch on these again later.

33

SLIDE 34

ΔACP: The data (1 fb−1)

34

)

2

c ) (MeV/

+

K

• m(K

1820 1840 1860 1880 1900

)

2

c Candidates / (0.2 MeV/

5000 10000 15000 20000 25000

LHCb

Preliminary )

2

c ) (MeV/

+

π

• π

m(

1820 1840 1860 1880 1900

)

2

c Candidates / (0.2 MeV/

1000 2000 3000 4000 5000 6000 7000

LHCb

Preliminary

K+ K− (2240k in 1 fb−1) π+ π− (690k in 1 fb−1)

Tagging with D*+ → D0 π+. Define δm ≡ mcand(D∗+) − mcand(D0) − m(π+),

5 10

• 5
• 4
• 3
• 2
• 1

1 2 3 4 5

)

2

m (MeV/c δ

1 5

)

2

Events / ( 0.1 MeV/c

10000 20000 30000 40000 50000

/ndof = 1.24)

2

χ Fit ( Background Data

Preliminary

LHCb

δ

χ

δ

χ

δ

×

χ

δ

χ

δ

χ

δ

χ

δ

χ

δ

χ

δ

χ

δ

χ

δ

×

χ

5 10

• 5
• 4
• 3
• 2
• 1

1 2 3 4 5

)

2

m (MeV/c δ

1 5

)

2

Events / ( 0.1 MeV/c

2000 4000 6000 8000 10000 12000 14000 16000

Preliminary

LHCb

/ndof = 1.00)

2

χ Fit ( Background Data

δ

χ

δ

χ

δ

χ

Example plot (MagUp TOS) Example plot (MagUp TOS)

SLIDE 35

ΔACP: Checks & systematics

35

Slow pion |p| (MeV/c)

5000 10000 15000

(MeV/c)

x

Slow pion p

• 1000

1000

• 1
• 0.8
• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 1

LHCb

Slow pion |p| (MeV/c)

5000 10000 15000

(MeV/c)

x

Slow pion p

• 1000

1000

• 1
• 0.8
• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 1

LHCb

Exclude detector edges where the local asymmetry is large. Dipole B-field up Dipole B-field down

Slow pion |p| (MeV/c)

5000 10000 15000

(MeV/c)

x

Slow pion p

• 1000

1000

• 1
• 0.8
• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 1

LHCb

Slow pion |p| (MeV/c)

5000 10000 15000

(MeV/c)

x

Slow pion p

• 1000

1000

• 1
• 0.8
• 0.6
• 0.4
• 0.2

0.2 0.4 0.6 0.8 1

LHCb

Away from beampipe Close to beampipe

SLIDE 36

ΔACP: Checks & systematics

• Test key steps & assumptions in the method:
• Vary fiducial cuts (prev. slide)
• Kinematic weighting (check vs. unweighted)
• Peaking background from misID/misrec
• Fit procedure
• Multiple candidates
• For 1/fb: soft pion IPCHI2 (blip)

36

SLIDE 37

ΔACP: Results

37

35 pb−1 : (−0.28 ± 0.70 ± 0.25)% − ± ± 1.0 fb−1 : (−0.34 ± 0.15 ± 0.10)% − ± ± 0.6 fb−1 : (−0.82 ± 0.21 ± 0.11)% − ± ± 3.0 fb−1 : (−0.10 ± 0.08 ± 0.03)%

New dataset Finer kinematic binning Reject edge regions Reprocess existing data Kinematic weighting Add PV constraint to fit Reprocess existing data Better kinematic weighting

D*+ tagged 1.0 fb−1 : (+0.49 ± 0.30 ± 0.14)% Muon tagged ± ± 3.0 fb−1 : (+0.14 ± 0.16 ± 0.08)%

Reprocess existing data

• Phys. Lett. B 723 (2013), 33-43

JHEP 07 (2014) 041

SLIDE 38

ΔACP & AΓ combination

38

LHCb

no CPV AΓ SL K−Κ+ and π−π+ AΓ prompt K−Κ+ AΓ prompt π−π+ ∆ACP SL ∆ACP prompt

10 5

• 5
• 10
• 10
• 5
• 5

10

∆aCP

dir

aCP

ind

×10-3 ×10-3

Consistent with no CPV (p=0.32)

SLIDE 39

Some other mixing/CPV things

39

Nuclear Physics, Section B 871 (2013), 1

Prompt charm production in pp collisions at √s = 7 TeV

] c [GeV/

T

p 1 2 3 4 5 6 7 8 )] c b/(GeV/ µ [

T

p /d σ d ×

• m

10

• 5

10

• 4

10

• 3

10

• 2

10

• 1

10 1 10

2

10

= 7 TeV s

LHCb

GMVFNS FONLL LHCb data 2.0 <y< 2.5, m=0 2.5 <y< 3.0, m=1 3.0 <y< 3.5, m=2 3.5 <y< 4.0, m=3 4.0 <y< 4.5, m=4

(d)

• Phys. Lett. B. 718 (2013) 902

Measurement of the D± production asymmetry in 7 TeV pp collisions

] c [GeV/

T

p

5 10 15

Production asymmetry

• 0.015
• 0.01
• 0.005

0.005

LHCb (a)

η

2.5 3 3.5 4 4.5

Production asymmetry

• 0.02
• 0.015
• 0.01
• 0.005

LHCb (b)

JHEP 06 (2013) 112

Search for CP violation in D+ → φπ+ and D+

s → K0

Sπ+ decays

]

2

c mass [MeV/

+

π φ

1850 1900 1950 2000

)

2

c Candidates / (MeV/

2 10 3 10 4 10 5 10

LHCb

(a) +

D

+ s

D

]

2

c mass [MeV/

• π

φ

1850 1900 1950 2000

)

2

c Candidates / (MeV/

2 10 3 10 4 10 5 10

LHCb

(b)

• D
• s

D

]

2

c mass [MeV/

+

π

S

K

1800 1850 1900 1950 2000

)

2

c Candidates / (MeV/

2 10 3 10 4 10

LHCb

(c) +

D

+ s

D

]

2

c mass [MeV/

• π

S

K

1800 1850 1900 1950 2000

)

2

c Candidates / (MeV/

2 10 3 10 4 10

LHCb

(d)

• D
• s

D

JHEP 04 (2016) 033

[ps]

D

t 1 2 3 4 5 Candidates per 0.047 ps

2000 4000 6000 8000 10000

LHCb

Model-independent measurement of mixing parameters in D0 → K0

Sπ+π− decays

SLIDE 40

40

SLIDE 41

Heavy baryon spectroscopy Search for Ξcc

JHEP 1312 (2013) 090 -- 0.65 fb−1

41

• Eur. Phys. J. C74 (2014) 3026 (chapter 19.4)

SLIDE 42

Charm baryon ground states

42

n p

• Σ

Σ / Λ

Σ / Λ

+

Σ

+ c

Σ /

+ c

Λ

+ c

Σ /

+ c

Λ

'0 c

Ξ /

c

Ξ

'0 c

Ξ /

c

Ξ

'+ c

Ξ /

+ c

Ξ

'+ c

Ξ /

+ c

Ξ

+ cc

Ξ

++ cc

Ξ

+ cc

Ω

c

Ω

c

Σ

++ c

Σ

• Ξ

Ξ C=0 C=1 C=2

+

2 1 =

P

J

• ∆

∆

+

∆

++

∆

*-

Σ

*0

Σ

*+

Σ

*-

Ξ

*0

Ξ

*0 c

Σ

*+ c

Σ

*++ c

Σ

*0 c

Ξ

*+ c

Ξ

*+ cc

Ξ

*++ cc

Ξ

• Ω

*0 c

Ω

*+ cc

Ω

++ ccc

Ω

3

I S C C=0 C=1 C=2 C=3

+

2 3 =

P

J

Baryons = 3 quarks (qqq) States on this slide have no


• rbital or radial excitations.

Pattern follows from
 symmetry rules. Wavefunction must be antisymmetric under quark exchange. More possibilities when
 there are three quark
 flavours in the game (csq).

SLIDE 43

Charm baryon ground states

43

n p

• Σ

Σ / Λ

Σ / Λ

+

Σ

+ c

Σ /

+ c

Λ

+ c

Σ /

+ c

Λ

'0 c

Ξ /

c

Ξ

'0 c

Ξ /

c

Ξ

'+ c

Ξ /

+ c

Ξ

'+ c

Ξ /

+ c

Ξ

+ cc

Ξ

++ cc

Ξ

+ cc

Ω

c

Ω

c

Σ

++ c

Σ

• Ξ

Ξ C=0 C=1 C=2

+

2 1 =

P

J

• ∆

∆

+

∆

++

∆

*-

Σ

*0

Σ

*+

Σ

*-

Ξ

*0

Ξ

*0 c

Σ

*+ c

Σ

*++ c

Σ

*0 c

Ξ

*+ c

Ξ

*+ cc

Ξ

*++ cc

Ξ

• Ω

*0 c

Ω

*+ cc

Ω

++ ccc

Ω

3

I S C C=0 C=1 C=2 C=3

+

2 3 =

P

J

Masses of these states can be (roughly) modelled in terms of: Constituent quark masses + spin-spin (hyperfine) couplings Origin of sum rules for light baryons: Same approach also works
 fairly well for baryons with heavy quarks.

(mN + mΞ)/2 = (3mΛ + mΣ)/4, mΣ∗ − m∆ = mΞ∗ − mΣ∗ = mΩ − mΞ∗, mΣ∗ − mΣ = mΞ∗ − mΞ,

SLIDE 44

Charm baryon ground states

44

n p

• Σ

Σ / Λ

Σ / Λ

+

Σ

+ c

Σ /

+ c

Λ

+ c

Σ /

+ c

Λ

'0 c

Ξ /

c

Ξ

'0 c

Ξ /

c

Ξ

'+ c

Ξ /

+ c

Ξ

'+ c

Ξ /

+ c

Ξ

+ cc

Ξ

++ cc

Ξ

+ cc

Ω

c

Ω

c

Σ

++ c

Σ

• Ξ

Ξ C=0 C=1 C=2

+

2 1 =

P

J

• ∆

∆

+

∆

++

∆

*-

Σ

*0

Σ

*+

Σ

*-

Ξ

*0

Ξ

*0 c

Σ

*+ c

Σ

*++ c

Σ

*0 c

Ξ

*+ c

Ξ

*+ cc

Ξ

*++ cc

Ξ

• Ω

*0 c

Ω

*+ cc

Ω

++ ccc

Ω

3

I S C C=0 C=1 C=2 C=3

+

2 3 =

P

J

We'll look for this state (ccd)

SLIDE 45

Ξcc: theory & experimental status

• Quark model: must exist and decay weakly
• Theory expectations for properties:
• Mass: most say 3500-3700 MeV (but some outliers)
• Lifetime: 100-250 fs
• Experimental situation unclear
• Ξcc+ → Λc+ K− π+ claim by SELEX at 3519 MeV with very

short lifetime (<33 fs) and very high cross-section

• Estimated 20% of Λc+ come from Ξcc+ alone
• Not reproduced by FOCUS, BABAR, Belle
• But nigh impossible to rule out, because production

cross-section can depend on collision environment

45

SLIDE 46

Ξcc: LHCb measurement

• Start with inclusive Λc+ → p K− π+ selection
• Use inclusive Λc+ as normalisation
• Add tracks to form Ξcc+ → Λc+ K− π+ candidates
• Will search for a Ξcc+ signal peak
• Will measure or put a limit on production ratio:
• Require reconstructed Λc+ → p K− π+ in trigger
• Introduced during 2011 run => only 0.65 fb−1 useful.
• Central preselection included some unhelpful cuts.

46

R ≡ σ(Ξ+

cc) B(Ξ+ cc → Λ+ c K−π+)

σ(Λ+

c )

= Nsig Nnorm εnorm εsig

SLIDE 47

Ξcc: LHCb measurement

47

]

2

c ) [MeV/

+

π

−

K p m(

2260 2280 2300 2320

)

2

c Entries / ( 0.8 MeV/

500 1000 1500 2000 2500 3000

LHCb

5% of data

Λc+ yield in full 0.65 fb−1: (818 ± 7) x 103

δm = mcand(Λ+

c K−π+) − mcand(Λ+ c ) − m(K−) − m(π+),

Define signal mass difference δm: Scan 380 < δm < 880 MeV/c2, approx 3300 < m(Ξcc) < 3800 MeV/c2

]

2

c m [MeV/ δ

560 580 600

)

2

c Entries / ( 0.6 MeV/

20 40 60 80 100 120 140

LHCb

simulation

Λc+ normalisation sample (data) Ξcc+ resolution (signal MC)

SLIDE 48

Ξcc: Scan for signal

• Scan across dm in 1 MeV/c2 steps. At each δm:
• Measure yield (for cross-section ratio)
• Estimate local significance (yield / stat error)
• Three components:
• Signal peaks in δm and m(Λc+)
• Λc+ background peaks in m(Λc+) but not in δm
• Combinatorial background peaks in neither
• Two methods used:
• Baseline: 2D sideband subtraction in δm, m(Λc+)
• Crosscheck: Cut in m(Λc+), 1D fit to δm

48

SLIDE 49

Ξcc: Scan for signal

49

)

2

candidate (MeV/c

+ c

Λ Mass of 2200 2250 2300 2350 2400 )

2

candidate (MeV/c

+ cc

Ξ m for δ 450 500 550 600 650 700

a b c d α β γ δ

Baseline 2D sideband subtraction: simple, analytic interpolation from sidebands to estimate expected background in signal box.

)

2

m (MeV/c δ 200 400 600 800 1000 1200 1400 Entries (arbitrary norm.) 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45

2

lower sideband: <2248 MeV/c

c

Λ

2

upper sideband: >2328 MeV/c

c

Λ

2

signal box: 2273-2303 MeV/c

c

Λ

Crosscheck: fit smooth function across range excluding sig box, interpolate to estimate expected background in sig box.

)

2

m (MeV/c δ 200 400 600 800 1000 1200 1400 Entries per 100 MeV bin 20 40 60 80 100

/NDF = 15.3/11 (17.1%)

2

χ

m(Λc+) sidebands m(Λc+) and δm sidebands

SLIDE 50

Ξcc: Yield, significance

50

]

2

c m [MeV/ δ

400 600 800

Signal yield

• 15
• 10
• 5

5 10 15

LHCb

]

2

c m [MeV/ δ

400 600 800

Signal yield

• 15
• 10
• 5

5 10 15

LHCb

]

2

c m [MeV/ δ

400 600 800

Signal yield

• 15
• 10
• 5

5 10 15

Baseline method Crosscheck method

LHCb

)

2

m (MeV/c δ 400 500 600 700 800 900 Local significance

• 2
• 1

1 2

25 Tiles 1D Fit & Count

Yield at each δm step with the two methods Local significance

Best per-test p-value in toy 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 p-value corrected for LEE 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Best per-test p-value in toy 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 p-value corrected for LEE 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Correct for LEE with toys: Global p-values: Baseline: 99% Crosscheck: 53%

SLIDE 51

Ξcc: Upper limits

• To set upper limits on R, also need efficiencies.
• Ξcc+ efficiency depends mildly on δm, but very

strongly on mean lifetime.

• Due to cuts requiring displaced daughter tracks; and
• Λc+ lifetime is short and trigger requires displaced vertex
• Quote upper limits for five illustrative lifetimes:

51

R, largest 95% CL UL in range ×103 δm (MeV/c2) 100 fs 150 fs 250 fs 333 fs 400 fs 380–429 12.6 2.7 0.73 0.43 0.33 430–479 11.2 2.4 0.65 0.39 0.29 480–529 14.8 3.2 0.85 0.51 0.39 530–579 10.7 2.3 0.63 0.38 0.29 580–629 10.9 2.3 0.63 0.38 0.29 630–679 14.2 3.0 0.81 0.49 0.37 680–729 9.5 2.0 0.56 0.33 0.25 730–779 10.8 2.3 0.63 0.37 0.28 780–829 12.8 2.8 0.74 0.45 0.34 830–880 12.2 2.6 0.70 0.42 0.32 380–880 14.8 3.2 0.85 0.51 0.39 ]

2

c m [MeV/ δ

400 600 800

at 95% CL R Upper limit on

• 4

10

• 3

10

• 2

10

• 1

10

100fs 150fs 250fs 333fs 400fs

LHCb

R ≡ σ(Ξ+

cc) B(Ξ+ cc → Λ+ c K−π+)

σ(Λ+

c )

= Nsig Nnorm εnorm εsig ,

SLIDE 52

Ξcc: In context

• Theory => expect R ~ 10−4 to 10−5 at LHCb
• SELEX equivalent: R = 9%
• Our UL: O(10−2) for τ=100fs, to O(10−4) at τ=400fs
• Disfavour large production à la SELEX
• Follow-on analysis with 3/fb and more modes has

started, promises to be much more sensitive.

• Especially for short lifetimes

52

SLIDE 53

Search for Ξb resonances

• Phys. Rev. Lett. 114, 062004 (2015) -- 3 fb−1 (Ξb0 π−)

JHEP 1605 (2016) 161 -- 3 fb−1 (Ξb− π+)

53

SLIDE 54

Charm baryon ground states

54

n p

• Σ

Σ / Λ

Σ / Λ

+

Σ

+ c

Σ /

+ c

Λ

+ c

Σ /

+ c

Λ

'0 c

Ξ /

c

Ξ

'0 c

Ξ /

c

Ξ

'+ c

Ξ /

+ c

Ξ

'+ c

Ξ /

+ c

Ξ

+ cc

Ξ

++ cc

Ξ

+ cc

Ω

c

Ω

c

Σ

++ c

Σ

• Ξ

Ξ C=0 C=1 C=2

+

2 1 =

P

J

• ∆

∆

+

∆

++

∆

*-

Σ

*0

Σ

*+

Σ

*-

Ξ

*0

Ξ

*0 c

Σ

*+ c

Σ

*++ c

Σ

*0 c

Ξ

*+ c

Ξ

*+ cc

Ξ

*++ cc

Ξ

• Ω

*0 c

Ω

*+ cc

Ω

++ ccc

Ω

3

I S C C=0 C=1 C=2 C=3

+

2 3 =

P

J

We'll look for the beauty versions of these states (bsu, bsd)

SLIDE 55

Charm baryon ground states

55

n p

• Σ

Σ / Λ

Σ / Λ

+

Σ

+ c

Σ /

+ c

Λ

+ c

Σ /

+ c

Λ

'0 c

Ξ /

c

Ξ

'0 c

Ξ /

c

Ξ

'+ c

Ξ /

+ c

Ξ

'+ c

Ξ /

+ c

Ξ

+ cc

Ξ

++ cc

Ξ

+ cc

Ω

c

Ω

c

Σ

++ c

Σ

• Ξ

Ξ C=0 C=1 C=2

+

2 1 =

P

J

• ∆

∆

+

∆

++

∆

*-

Σ

*0

Σ

*+

Σ

*-

Ξ

*0

Ξ

*0 c

Σ

*+ c

Σ

*++ c

Σ

*0 c

Ξ

*+ c

Ξ

*+ cc

Ξ

*++ cc

Ξ

• Ω

*0 c

Ω

*+ cc

Ω

++ ccc

Ω

3

I S C C=0 C=1 C=2 C=3

+

2 3 =

P

J

Σ0

c

Σ+

c

Σ++

c

Ξ0

c

Ξ+

c

Ω0

c

Ξ0

c

Ξ+

c

Λ+

c

j = 0, JP = 1

2 +

j = 1, JP = 1

2 +

Σ∗0

c

Σ∗+

c

Σ∗++

c

Ξ∗0

c

Ξ∗+

c

Ω∗0

c

j = 1, JP = 3

2 +

Decay weakly

We'll look for the beauty versions of these states (bsu, bsd)

SLIDE 56

Charmed baryon mass (GeV)

1/2+ 1/2+ 1/2+ 1/2+ 1/2+ 3/2+ 3/2+ 1/2– 3/2– 3/2–

Λc Σc Ξc Ωc

2.3 2.5 2.7 0.0 0.2 0.4 π γ ππ π π π

1/2–

π π

3/2+

γ ?

Λcππ

pD π

Δ Δ

∇ ∇

Charmed baryon spectrum from the PDG (L,R>0 states removed)

Ξc Ξʹc Ξ*c

Same pattern expected for b-baryons, tho

(cud) (cqq) (csq) (css)

Ξb: Three isodoublets

56

Caution: simplified wavefunctions!

sd diquark : (↑↓) bsd : ↑(↑↓) j = 0, JP = 1/2+ Ξb

18

sd diquark: (↑↑) bsd : ↑(↑↑) bsd : ↓(↑↑) j = 1, JP = 1/2+ j = 1, JP = 3/2+ Ξb′ Ξb*

SLIDE 57

Ξb: Overview of analyses

• Reconstruct Ξb ground state, cut on m(Ξb)
• Fit δm = m(Ξb π) − m(Ξb) − m(π)
• ... using signal MC to fix the resolution shape
• Lineshape = resolution ⊗ P-wave rel. Breit-Wigner
• Measure properties of peak(s)
• Measure production relative to Ξb
• Do these for both:
• Ξb0 π−, using Ξb0 → Ξc+ π−, Ξc+ → p K− π+
• Ξb− π+, using Ξb− → Ξc0 π−, Ξc0 → p K− K− π+

57

SLIDE 58

Ξb0 π−: Data & fits

58

]

2

c ) [MeV/

b

Ξ (

cand

m

5600 5700 5800 5900 6000

2

c Entries per 10 MeV/

20 40 60 80 100 120 140

LHCb

]

2

c ) [MeV/

b

Ξ (

cand

m

5600 5700 5800 5900 6000

2

c Entries per 10 MeV/

20 40 60 80 100 120

inset: in δm sig window

m(Ξb0)

δm(Ξ0

b )

= 3.653 ± 0.018 ± 0.006 MeV /c2, δm(Ξ⇤

b )

= 23.96 ± 0.12 ± 0.06 MeV /c2, Γ(Ξ⇤

b )

= 1.65 ± 0.31 ± 0.10 MeV, Γ(Ξ0

b )

< 0.08 MeV at 95% CL.

Results of fit: ← narrow!

]

2

c m [MeV/ δ

10 20 30 40

2

c Entries per 0.45 MeV/

20 40 60 80 100 120 140

LHCb

−

π

b

Ξ

+

π

b

Ξ

]

2

c m [MeV/ δ

2 3 4 5

2

c Entries per 0.1 MeV/

5 10 15 20 25 30 35

Ξb′− Ξb*− Ξb′− z

• m

δ Ξ π Ξ π

SLIDE 59

Ξb− π+: Data & fits

59

Inclusive m(Ξb−) Results of fit:

]

2

c ) [MeV/

− b

Ξ (

cand

m

5700 5800 5900 6000

2

c Entries per 4 MeV/

20 40 60 80 100 120 140 160 180 200

LHCb

]

2

c m [MeV/ δ

10 20 30 40

2

c Entries per 0.45 MeV/

10 20 30 40 50 60

LHCb

m(⌅∗0

b ) − m(⌅− b ) − m(⇡+)

= 15.727 ± 0.068 ± 0.023 MeV /c2, Γ(⌅∗0

b )

= 0.90 ± 0.16 ± 0.08 MeV,

SLIDE 60

Ξb: Systematics

• Won't go through full list, but key points:
• Use control sample D*+ → D0 π+ to validate both mass

scale (good to few keV) and resolution scale (to ~10%)

• Also use narrow Ξb′− peak to validate resolution model
• Statistical uncertainties dominate.

60

Effect m Γ Fit bias correction 0.016 Simulated sample size 0.007 0.034 Multiple candidates 0.009 0.007 Resolution model 0.001 0.072 Background description 0.002 0.001 Momentum scale 0.009 0.001 RBW shape 0.017 0.011 Sum in quadrature 0.023 0.082 Statistical uncertainty 0.068 0.162

Source δm(Ξ0

b)

δm(Ξ⇤

b )

Γ(Ξ⇤

b )

Simulated sample size 0.002 0.005 Multiple candidates 0.004 0.048 0.055 Resolution model 0.002 0.003 0.070 Background description 0.001 0.003 0.019 Momentum scale 0.003 0.014 0.003 RBW spin and radial parameter 0.000 0.023 0.028 Sum in quadrature 0.006 0.055 0.095 Statistical uncertainty 0.018 0.119 0.311

Sys errors in units of MeV/c2 (for δm), MeV (for Γ)

Ξb0 π− Ξb− π+

SLIDE 61

Relative production rates

• Can also compare yield of resonances to that of

ground states (with additional trigger requirement).

• Correct for efficiency, obtain production ratios:
• Remembering isospin partner modes, implies

significant fraction of Ξb come from decays of higher- mass states (like D*+ and D0)

61

σ(pp → Ξ0

b X)B(Ξ0 b

→ Ξ0

b π)

σ(pp → Ξ0

b X)

= 0.118 ± 0.017 ± 0.007, σ(pp → Ξ⇤

b X)B(Ξ⇤ b

→ Ξ0

b π)

σ(pp → Ξ0

b X)

= 0.207 ± 0.032 ± 0.015, σ(pp → Ξ⇤

b X)B(Ξ⇤ b

→ Ξ0

b π)

σ(pp → Ξ0

b X)B(Ξ0 b

→ Ξ0

b π)

= 1.74 ± 0.30 ± 0.12,

σ(pp → Ξ⇤0

b X)B(Ξ⇤0 b

→ Ξ

b π+)

σ(pp → Ξ

b X)

= 0.28 ± 0.03 ± 0.01.

SLIDE 62

Ξb: Putting it together

62

Ξb− Ξb0

≈3 MeV? 5.9±0.6 MeV 2.3 ± 0.7 MeV 155.30 ± 0.07 MeV ≈20 MeV? Ξbʹ0 → Ξb− π+ :
 ≈134 MeV?

⇒ Forbidden?

Ξbʹ− → Ξb0 π− :
 143.21 ± 0.02 MeV

⇒ Allowed Measurements in black Theory/guesses in pink

20.31 ± 0.13 MeV Ξbʹ0 → Ξb0 π0 : ≈140 MeV ⇒ Allowed?

SLIDE 63

Ξb: Angular analysis

• Self-consistent interpretation within quark model
• But would be nice to actually measure the JP
• Tried an inclusive angular analysis for (Ξb0 π−)

63

Expect a polynomial in cosθh

• f order (2J−1)...

... but if Ξb*/ʹ is unpolarised, hit pathological case where nearly all coefficients are zero and distribution is flat. Thus, data consistent with
 J(Ξb′−)=1/2 & J(Ξb*−)=3/2 but limited information

• => Angular analysis, defining θh to be the angle between


[direction of the Ξb0 in the Ξb*− rest frame] and
 [direction of the Ξc+ in the Ξb0 rest frame]

Figures adapted from SLAC-R-868 (Ziegler, V.)

~ Ξ+

c2

~ Ξ∗−

b

= ~ ~ ⇡+

1

~ Ξ0

b1

~ Ξ+

c1

~ ⇡+

1

Ξb*− rest frame Ξb0 rest frame

~ Ξ0

b1

~ Ξ0

b2 = ~

~ ⇡−

2

θh

)

h

θ cos(

• 1
• 0.5

0.5 1

Normalized yield

0.05 0.1 0.15 0.2

) = 1/2

h

θ f(cos )]/2

h

θ (

2

a)cos − ) = [a+3(1

h

θ f(cos

LHCb

)

h

θ cos(

• 1
• 0.5

0.5 1

Normalized yield

0.05 0.1 0.15 0.2

) = 1/2

h

θ f(cos )]/2

h

θ (

2

a)cos − ) = [a+3(1

h

θ f(cos

LHCb

Ξb′− Ξb*−

SLIDE 64

What next?

64

SLIDE 65

For LHCb

• Run 2 restart: so far, so good!
• Key improvements: real-time calibration, trigger, DAQ
• Several results from Run 1 showed tension with SM:
• Angular analyses of b→sμ+μ− (B0, Bs, Λb)
• Lepton universality in B0 → D*− l+ νl) and B+ → K+ l+ l−
• Vub (exclusive vs inclusive)
• Interesting pattern -- remains to be seen if they hold up!
• After Run 2: hardware upgrade
• Boost lumi by order of magnitude
• Total overhaul of trigger, tracking, PID systems

65

SLIDE 66

Upcoming

• Follow-on work to analyses discussed earlier:
• Search for Ξcc in 3/fb
• Search for baryon number violation (Ξb0 oscillations)
• CKM angle gamma from charmless B decays
• Work ongoing (M2 stage), with Eli, Emilie
• Further studies of charmless B decays for Run 2

66

SLIDE 67

Last words

• Thanks for hanging in there with me
• Lots of good physics in LHCb's Run 1
• Searches for NP

, e.g. for CPV in (D+ → K− K+ π+),
 (D0 → K− K+, π− π+)

• Studies of SM physics, e.g. Ξcc and Ξb spectroscopy
• Still plenty more to do with the Run 1 dataset...
• ... but will soon have significant stats from Run 2 as well!
• Looking forward to Run 2 -- and the upgraded detector

for Run 3

• Busy, exciting time for flavour physics!

67

SLIDE 68

68

SLIDE 69

69

˜ t

SLIDE 70

70

)

2

m (MeV/c δ 400 500 600 700 800 900 Local significance

• 2
• 1

1 2 )

2

m (MeV/c δ 400 500 600 700 800 900 Local significance

• 2
• 1

1 2

)

2

m (MeV/c δ 400 500 600 700 800 900 Local significance

• 2
• 1

1 2

25 Tiles 1D Fit & Count

Figure 28: Local signal significance as a function of δm in the unblinded data set, for the baseline 25-Tiles method (upper left), and the crosscheck 1D Fit & Count method (upper right). The results from the two methods are compared in the lower plot.

SLIDE 71

71

]

2

c m [MeV/ δ

400 500 600 700 800

)

2

c Entries / ( 4 MeV/

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

LHCb

]

2

c m [MeV/ δ

400 500 600 700 800

)

2

c Entries / ( 4 MeV/

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

LHCb

Figure 3.17:

Mass difference spectrum requiring 2273 < mcand(Λ+

c ) < 2303 MeV

/c2. Candidates are also required to be consistent with (left) an intermediate Σc(2455)++, (right) an intermediate Σc(2520)++.