Flavor Physics and CP Violation Conference, Victoria BC, 2019 1

Production rates and branching fractionsof heavy hadrons & quarkonia at LHC experiments

P. Ronchese

on behalf of the ATLAS, CMS and LHCb collaborationsUniversity and INFN Padova

Measurements of production cross-sections of inclusive b-hadrons pairs, bottom mesons andbaryons, and quarkonia at LHC will be shown. Recent measurements of branching fractions ofbottom baryons, bottom mesons in the final state with baryons, and a new result about a searchfor intermediate states in meson decay will also be shown.

I. INTRODUCTION

Measurements of heavy hadron and quarkonia crosssections at LHC allow probing QCD processes; theyare also reference or ingredient for searches and mea-surements of rarer or new processes, as well as thebaseline for associated production of heavy flavourand other objects.

The study of decay properties and branching frac-tions does allow a test of form-factor models as wellas the search for new and exotic states that can beproduced in the decay.

Results from ATLAS [1], CMS [2] and LHCb [3] willbe shown in the following.

II. PRODUCTION CROSS-SECTIONS

A. Inclusive b-hadrons

An inclusive b-hadron pair production cross sectionmeasurement was obtained by ATLAS [4] at an en-ergy

√s = 8 TeV; final states were selected looking

for a J/ψ coming from the first hadron and decayingto µ+µ−, and a muon coming from the second hadron.The fiducial volume was defined requiring the twomuons from the J/ψ to have

∣∣ηµ,J/ψ∣∣ < 2.3 and the

third muon to have |ηµ| < 2.5; a minimum transversemomentum was also required: pT,µ > 6GeV. Thetotal cross section in the fiducial volume was found:

σ(B(→ J/ψ[→ µ+µ−] +X)B(→ µ+X))) =

(17.7± 0.1(stat)± 2.0(syst)) nb .

B. Bottom mesons and baryons

1. B+ production

The differential pp → B+ + X cross-section ver-sus transverse momentum or rapidity was measuredby CMS [5] at an energy

√s = 13 TeV in the region

|yB | < 1.45 or |yB | < 2.1 and 10 GeV < pT,B <100 GeV or 17 GeV < pT,B < 100 GeV; the ratiowith the corresponding cross-section at

√s = 7 TeV

was measured and compared with FONLL [7–9] andPYTHIA [10] predictions. Results are shown inFig. 1.

[GeV]BT

p5 10 15 20 25 30 35

(7 T

eV)

σ(1

3 T

eV)

/ σ

0

1

2

3

4

5

6 |<1.45)BData (|y

|<2.1)BData (|y

FONLL

PYTHIA

CMS (7 TeV)-1 (13 TeV), 5.8 pb-148.1 pb

(a)

|B|y0 0.5 1 1.5 2 2.5

(7 T

eV)

σ(1

3 T

eV)

/ σ

0

1

2

3

4

5

6 >10 GeV)BT

Data (p

>17 GeV)BT

Data (p

FONLL

PYTHIA

CMS (7 TeV)-1 (13 TeV), 5.8 pb-148.1 pb

(b)

FIG. 1: Ratios of B+ differential production cross sectionsat

√s = 13 TeV and

√s = 7 TeV measured by CMS [5].

An analogous study was done by LHCb [11], thatmeasured the double differential cross-section ver-sus transverse momentum and rapidity in the region2.0 < |yB | < 4.5 , pT,B < 40 GeV. Again the ratiowith the corresponding cross-section at

√s = 7 TeV

was measured and compared with FONLL [12] predic-tions. Differential cross-sections are shown in Fig. 2;the integrated cross sections were found:

σ(pp→ B±X (√s = 7 TeV) =

(43.0± 0.2(stat)± 2.5(syst)± 1.7(b.r.)) µb

σ(pp→ B±X (√s = 13 TeV) =

(86.6± 0.5(stat)± 5.4(syst)± 3.4(b.r.)) µb

where the last uncertainty comes from theB± → J/ψK± branching fraction.

WedB1130

arX

iv:1

906.

0899

2v2

[he

p-ex

] 1

1 N

ov 2

019

2 Flavor Physics and CP Violation Conference, Victoria BC, 2019

0 10 20 30 401

1.5

2

2.5

3

3.5

4

4.5

R(1

3Te

V/7

TeV

)

pT [GeV/c]

LHCb

2.0< y< 4.5

DataFONLL

2 3 41

1.5

2

2.5

3

3.5

4

4.5

R(1

3Te

V/7

TeV

)

y

LHCb

0< pT < 40GeV/c

DataFONLL

FIG. 2: Ratios of B+ differential production cross sectionsat

√s = 13 TeV and

√s = 7 TeV measured by LHCb [11].

2. Ξ−b production

Recently LHCb measured the ratio fΞ−b/fΛ0

bof the

fragmentation fractions to Ξ−b and Λ0b [13]. The decay

of the Ξ−b baryon has been studied since some time anda measurement of a branching fraction was done [14],but its absolute determination requires knowing thefragmentation ratio.

The quantity that is directly accessible is the ratioRof the number of reconstructed decays in the channelΞ−b → J/ψΞ−, and a normalization one, with Λ0

b →J/ψΛ0, but the ratio R can be expressed also as theratio of the products of fragmentation and branchingfractions; the latter can be expressed as the productsof the ratios of partial widths and lifetimes:

R =fΞ−

b

fΛ0b

· Γ(Ξ−b → J/ψΞ−)

Γ(Λ0b → J/ψΛ0)

·τΞ−

b

τΛ0b

=N(Ξ−b → J/ψΞ−)

N(Λ0b → J/ψΛ0)

·εΛ0

b

εΞ−b

.

The ratio of widths can be assumed to be 2/3 fromSU(3) flavor symmetry [15–17] and the ratio of life-times can be taken from PDG [18] so that the fragmen-tation ratio can be obtained. The decay has been re-constructed pairing a J/ψ → µ+µ− with a Λ0 → pπ−

or a Ξ− → Λ0π−. For the reconstruction of the Ξ−

or Λ0 tracks have been classified as “long” or “down-stream”, depending on the track originating before orafter the vertex detector. For Λ0 downstream trackshave been used, due to the long lifetime, while a long

track was used as candidate for the pion coming fromthe Ξ− decay. The signal yields were estimated fromfits to the mass distributions, as shown in Fig. 3.

]2c mass [MeV/Λψ/J5500 5550 5600 5650 5700 5750

)2 cC

andi

date

s / (

2 M

eV/

500

1000

1500LHCb

=7,8 TeVsData

Full PDF

Λψ/J →0bΛ

]2c mass [MeV/−Ξψ/J

5700 5800 5900

)2 cC

andi

date

s / (

5 M

eV/

20

40

60

LHCb=13 TeVs

Data

Full PDF−Ξψ/J →−

bΞ)L

−πΛ→−Ξ(

FIG. 3: Mass distributions for Ξ−b and Λ0

b reconstructedby LHCb [13].

Using the mass difference mΞ−b− mΛ0

bas free pa-

rameter in the fit and using the Λ0b mass from the

PDG [18] the most precise measurement of Ξ−b masswas obtained:

mΞ−b

=

(5796.70± 0.39(stat)± 0.15(syst)± 0.17(mΛ0b)) MeV

where the last uncertainty comes from the Λ0b mass.

With the signal yields from the fits and the efficien-cies from simulation the ratio R was obtained, and inthe end the fragmentation fraction was extracted:

fΞ−b

fΛ0b

= (6.7± 0.5(stat)± 0.5(syst)± 2.0(f.s.))× 10−2

[√s = 7, 8 TeV]

fΞ−b

fΛ0b

= (8.2± 0.7(stat)± 0.6(syst)± 2.5(f.s.))× 10−2

[√s = 13 TeV]

WedB1130

Flavor Physics and CP Violation Conference, Victoria BC, 2019 3

where the last uncertainty is due to the flavor symme-try assumption and taken to be 30%.

C. Quarkonia

Several measurements of the production cross-section for quarkonia have been done at LHC exper-iments; a special interest can be found in the pro-duction of quarkonia pairs. Quarkonia pairs can beproduced in single parton scattering (SPS), that’s as-sumed to dominate and lead to strongly correlatedpairs with small rapidity differences, but, in the highparton densities in proton-proton collisions, also dou-ble parton scattering (DPS) can occur producing mul-tiple heavy flavour particles with large ∆y [19, 20].

A measurement of the DPS contribution in doubleJ/ψ production was done by ATLAS [21] at

√s =

8 TeV. In the analysis J/ψ pairs coming from dif-ferent pp interactions were removed with a cut onthe distance along the beam direction between thereconstructed vertices, while the residual pile-up con-tamination was estimated looking at the kinematicvariables distributions in sidebands. In double partonscattering J/ψ candidates are assumed to be producedindependently, so a template ∆y∆φ distribution hasbeen built with J/ψ pairs from different events andhas been normalized to data at large rapidity differ-ence. Then event weights in each ∆y∆φ bin have beencomputed from the ratio of the normalized templateand full data; this weight can be used as an estimate ofthe DPS fraction, that can be compared to predictionfrom NLO versus rapidity difference and transversemomentum as shown in Fig. 4.

The total cross-section, in two fiducial regions∣∣yJ/ψ∣∣ < 1.05 , 1.05 <

∣∣yJ/ψ∣∣ < 2.1 with pT,J/ψ >

8.5 GeV, pT,µ > 2.5 GeV and |ηµ| < 2.3, were foundto be:

σFid(∣∣yJ/ψ

∣∣ < 1.05) =

(15.6± 1.3(stat)± 1.2(syst)± 0.2(b.r.)± 0.3(lum)) pb

σFid(1.05 <∣∣yJ/ψ

∣∣ < 2.1 ) =

(13.5± 1.3(stat)± 1.1(syst)± 0.2(b.r.)± 0.3(lum)) pb

with a DPS fraction

fDPS = (9.2± 2.1(stat)± 0.5(syst))% .

A similar measurement at√s = 13 TeV has been

done from LHCb [25, 26]; the measured total cross-section for the production in the region 2.0 < |y| <4.5 , pT < 10 GeV was:

σ = (15.2± 1.0(stat)± 0.9(syst)) nb .

The DPS component prediction was obtained froma large number of pseudoexperiments, where two un-correlated J/ψ mesons were produced according to

FIG. 4: The DPS and total differential cross-sections as afunction of difference in rapidity and the transverse mo-mentum of the di-J/ψ measured by ATLAS [21], comparedwith LO DPS [22] and NLO* SPS [23, 24] predictions.

the measured differential cross-sections, and SPS pre-dictions from theoretical calculations using severalapproaches (LO, NLO, color singlet or color octet).Several data distributions were fitted with a two-component model to obtain the DPS fraction givingresults in the range:

fDPS = ((42± 25)÷ (86± 55))% .

As an example, the comparison between the mea-sured and predicted differential cross-section vs.pT,J/ψJ/ψ is shown in Fig. 5.

III. BRANCHING FRACTIONS

A. Bottom baryon decay

Ratios of branching fractions of b-hadrons with aJ/ψ or a ψ(2S) in the final state allow testing thefactorization of amplitudes; some recent result of suchratios involve baryons. A measurement of the ratio

WedB1130

4 Flavor Physics and CP Violation Conference, Victoria BC, 2019

0 5 100

1

2

3

4

5

6

7

dσ(J/ψJ/ψ)

dp T

(J/ψJ/ψ)

[nb

GeV/c

]

pT(J/ψJ/ψ ) [GeV/c]

LHCb 13TeV

DPSSPS: LO kT

SPS: NLO∗CS′

SPS: NLO∗CS′′〈kT〉=0.5GeV/c

SPS: NLO∗CS′′〈kT〉=2GeV/c

SPS: LO CO〈kT〉=0.5GeV/c

SPS: LO CO〈kT〉=2GeV/c

×××××SPS: NLO CS

FIG. 5: Comparisons between the measured and predicteddifferential cross-section vs. pT,J/ψJ/ψ from LHCb [25, 26].

of branching fractions of Λ0b → ψ(2S)Λ0 and Λ0

b →J/ψΛ0 was done by ATLAS [27] a few years ago givinga result that shows a discrepancy from covariant quarkmodel prediction [28, 29]. Another measurement hasjust been done by LHCb [30] that reconstructed ψfrom non-prompt muons, a Λ0 with two tracks of thesame type, “long” or “downstream”, built a commonvertex and applied a constrained fit with the massesof the ψ and the Λ0.

Events have been weighted with the inverse of effi-ciency; the latter was estimated in the simulation aswell as the background from decays of B0 → ψ(K0

S →π+π−) or Ξ−b → ψ(Ξ− → Λ0π−): the invariant massdistributions are shown in Fig. 6.

]2c [MeV/Λ(2S)ψm5400 5500 5600 5700

)2 cC

andi

date

s / (

8 M

eV/

210

310

410

510 long trackLHCb (a)

]2c [MeV/ΛψJ/m5400 5500 5600 5700

)2 cC

andi

date

s / (

8 M

eV/

210

310

410

510

610 long trackLHCb (b)

]2c [MeV/Λ(2S)ψm5400 5500 5600 5700

)2 cC

andi

date

s / (

4 M

eV/

210

310

410

510 downstream trackLHCb (c)

]2c [MeV/ΛψJ/m5400 5500 5600 5700

)2 cC

andi

date

s / (

4 M

eV/

210

310

410

510

610 downstream trackLHCb (d)

FIG. 6: Fits to the invariant-mass distributions forΛ0b → ψ(2S)Λ0 (a,c) and Λ0

b → J/ψΛ0 (b,d) obtained byLHCb [30]. The signal (blue, dashed), the combinatorialbackground (green, dotted), the B0 → ψK0

S background(cyan, long-dash-dotted) and the Ξ−

b → ψΞ− background(violet, dash-triple-dotted) are indicated.

Taking the signal yields from the mass distributionsfit and branching fractions of the ψ from PDG [18] thebranching fractions ratio was obtained:

B(Λ0b → ψ(2S)Λ0)

B(Λ0b → J/ψΛ0)

=

0.513± 0.023(stat)± 0.016(syst)± 0.011(b.r.)

where the last uncertainty comes from the ψbranching fraction.

B. Baryon production in meson decays

1. B0d,s decay

Some special interest related to baryons can befound in heavy hadron decays not only when baryonthemselves are decaying, but also when they arepresent in the final state. Their presence can be usedto look for possible pentaquark intermediate states; anevidence was claimed by LHCb [31, 32] in the decay ofΛ0b → J/ψpK−. The presence of a baryon and an an-

tibaryon can also test possible glueball states [33, 34].LHCb then studied the decays B0

d,s → J/ψpp [35];

both the decays are suppressed: the B0d decay in this

channel is suppressed by Cabibbo while the B0s decay

in the same channel is suppressed by OZI. A branch-ing fraction at the level of 10−9 would be expected,with some enhancement via a resonant contributionfrom fJ(2220)→ pp.

The branching fraction is measured by a compari-son with a normalization channel, so that the ratio ofbranching fractions is measured, using the well knownB0s → J/ψφ decay as reference. The branching frac-

tion of the studied channel is given by the ratio of re-constructed decays, multiplied by the branching frac-tions of B0

s → J/ψφ, φ → K+K− and, only for B0d,

the ratio of fragmentation fractions:

B(B0d → J/ψpp) =

NB0d→J/ψpp

NB0s→J/ψφ

· B(B0s → J/ψφ) · B(φ→ K+K−) · fs

fd

B(B0s → J/ψpp) =

NB0s→J/ψpp

NB0s→J/ψφ

· B(B0s → J/ψφ) · B(φ→ K+K−) .

The number of events was obtained by an extendedmaximum likelihood fit to the mass distributions, asshown in Fig. 7; the product B(B0

s → J/ψφ) · B(φ →K+K−) · fs/fd was measured [36] at

√s = 7 TeV

as well as the fragmentation ratio fs/fd [37, 38] andscaled to

√s = 13 TeV [39].

The B0d,s → J/ψpp decay branching ratios have fi-

nally been extracted:

WedB1130

Flavor Physics and CP Violation Conference, Victoria BC, 2019 5

FIG. 7: Fit to invariant mass distribution of B0d,s → J/ψpp

from LHCb [35].

B(B0d → J/ψpp) =

(4.51± 0.40(stat)± 0.44(syst))× 10−7

B(B0s → J/ψpp) =

(3.58± 0.19(stat)± 0.33(syst))× 10−6 .

Due to the very low phase space available the mo-mentum uncertainty is negligible; that does allow asa side results the most precise single measurements ofB0d and B0

s masses:

mB0d

= (5279.74± 0.30(stat)± 0.10(syst)) MeV

mB0s

= (5366.85± 0.19(stat)± 0.13(syst)) MeV .

2. B+ decay

Another study including the look for intermedi-ate states has been done by CMS about the decayB+ → J/ψΛ0p [40]: that decay was first seen at B-factories [41, 42]; in this new study new exotic stateswere searched in the J/ψΛ0 or J/ψp systems.

As in the previous study of B0d,s decay branching

fraction has been measured as ratio with the normal-ization channel B+ → J/ψK∗+ (K∗+ → K0

Sπ+ ,

K0S → π+π−).The ratio and the absolute vaule of branching frac-

tions were measured as:

B(B+ → J/ψΛ0p)

B(B+ → J/ψK∗+)=

1.054± 0.057(stat)± 0.028(syst)± 0.011(b.r.)

B(B+ → J/ψΛ0p) =

(15.07± 0.81(stat)± 0.40(syst)± 0.86(b.r.))× 10−6

where the last uncertainty comes from the involvedcascade decays branching fractions.

The distributions of invariant masses of the J/ψΛ0

and J/ψp systems have then been studied and com-pared with expectations, from pure phase space or

phase space corrected for reflections from K∗+ → Λ0presonances. To do that the event sample has been di-vided in M(Λ0p) invariant mass bins and in each binthe first 8 Legendre polynomials and momenta havebeen computed using the Λ0p helicity angle to describethe angular distributon. Simulated events have thenbe reweighted using the Λ0p mass distribution ratioas reference, or the weights given by Legendre poly-nomials and moments. The distributions obtained inthis way have been fitted to data.

The J/ψp and J/ψΛ0 invariant mass distributionsare shown in Fig. 8, compared with the simulation us-ing pure phase space, the simulation reweighted withthe Legendre polynomials or a function fitted to thecos θK∗ distribution in data.

p) [GeV]ψM(J/

4.04 4.06 4.08 4.1 4.12 4.14 4.16

Yie

ld /

5 M

eV0

2000

4000

6000

8000

10000 Data

Pure phase space

= 8max> with ljU<P

fit*K

θcos

(8 TeV)-119.6 fbCMS

) [GeV]ΛψM(J/4.22 4.24 4.26 4.28 4.3 4.32 4.34

Yie

ld /

5 M

eV

0

2000

4000

6000

8000

10000

Data

Pure phase space

= 8max> with ljU<P

fit*K

θcos

(8 TeV)-119.6 fbCMS Preliminary

FIG. 8: Invariant mass distributions of J/ψp and J/ψΛ0

obtaind by CMS [40] in the decay B+ → J/ψΛ0p, com-pared to the simulation using pure phase space (black),phase space corrected by the Legendre polynomials (red,solid) and a fit to the cos θK∗ distribution (red, dashed).

The quality of the data description from the differ-ent hypotheses has been estimated generating a largenumber of pseudoexperiments according to the PDFfor the pure phase space or the reweighted angular dis-tributions; a log-likelihood ratio has then been com-puted, to extract a compatibility, or incompatibility,significance. The significance of the incompatibility ofdata with the pure phase space was found to be muchlarger than the incompatibility with the phase spacecorrected by the Legendre polynomials:

J/ψp J/ψΛ0 Λ0p

pure phase-space 5.5÷ 7.4 6.1÷ 8.0 3.4÷ 4.6

reweighted phase-space 1.3÷ 2.8 1.3÷ 2.7 −

WedB1130

6 Flavor Physics and CP Violation Conference, Victoria BC, 2019

IV. CONCLUSIONS

ATLAS, CMS and LHCb have produced many mea-surements of heavy hadron production cross-sectionsand decay branching fractions:

• cross-sections have been compared to predic-tions and simulations and are input for othermeasurements,

• branching fractions allow test model predictions,

• in the study of decays the possible presence ofintermediate exotic states has been investigated.

All those measurements allow important tests ofQCD.

[1] ATLAS Collaboration, J. Inst. 03, S08003 (2008).[2] CMS Collaboration, J. Inst. 03, S08004 (2008).[3] LHCb Collaboration, J. Inst. 03, S08005 (2008).[4] ATLAS Collaboration, J. High Energy Phys. 11, 062

(2017).[5] CMS Collaboration, Phys. Lett. B 771, 435 (2017).[6] CMS Collaboration, Phys. Rev. Lett. 106, 112001

(2011).[7] M. Cacciari et al., J. High Energy Phys. 05, 007

(1998).[8] M. Cacciari et al., J. High Energy Phys. 03, 006

(2001).[9] M. Cacciari et al., J. High Energy Phys. 10, 137

(2012).[10] T. Sjostrand et al., Comp. Phys. Comm. 178, 852

(2008).[11] LHCb Collaboration, J. High Energy Phys. 12, 026

(2017).[12] M. Cacciari et al., Eur. Phys. J. C 75, 610 (2015).[13] LHCb Collaboration, Phys. Rev. D 99, 052006 (2019).[14] LHCb Collaboration, Phys. Rev. Lett. 115, 241801

(2015).[15] M. J. Savage and M. B. Wise, Nucl. Phys. B 326, 15

(1989).[16] M. Voloshin, arXiv:1510.05568.[17] Y. K. Hsiao et al., Phys. Lett. B 751, 127 (2015).[18] M. Tanabashi et al.(Particle Data Group), Phys. Rev.

D 98, 030001 (2018).[19] S. Baranov et al., Phys. Lett. B 705, 116 (2011).[20] C. H. Kom et al., Phys. Rev. Lett. 107, 082002 (2011).[21] ATLAS Collaboration, Eur. Phys. J. C 77, 76 (2017).[22] C. Borschensky and A. Kulesza, arXiv:1610:00666.[23] J.-P. Lansberg and H.-S. Shao, Phys. Lett. B 751, 479

(2015).

[24] J.-P. Lansberg and H.-S. Shao, Phys. Rev. Lett. 111,122001 (2013).

[25] LHCb Collaboration, J. High Energy Phys. 06, 047(2017).

[26] LHCb Collaboration, J. High Energy Phys. 10, 068(2017).

[27] ATLAS Collaboration, Phys. Lett. B 751, 63 (2015).[28] T. Gutsche et al., Phys. Rev. D 88, 114018 (2013).[29] T. Gutsche et al., Phys. Rev. D 92, 114008 (2015).[30] LHCb Collaboration, J. High Energy Phys. 03, 126

(2019).[31] LHCb Collaboration, Phys. Rev. Lett. 115, 072001

(2015).[32] LHCb Collaboration, Phys. Rev. Lett. 117, 082002

(2016).[33] C. J. Morningstar and M. Peardon, Phys. Rev. D 60,

034509 (1999).[34] Y. Chen et al., Phys. Rev. D 73, 014516 (2006).[35] LHCb Collaboration, Phys. Rev. Lett. 122, 191804

(2019).[36] LHCb Collaboration, Phys. Rev. D 87, 072004 (2013).[37] LHCb Collaboration, J. High Energy Phys. 04, 001

(2013).[38] LHCb Collaboration, LHCb-CONF-2013-011,

https://cds.cern.ch/record/1559262.[39] LHCb Collaboration, Phys. Rev. Lett. 118, 191801

(2017).[40] CMS Collaboration, CMS-PAS-BPH-18-005,

https://cds.cern.ch/record/2668754.[41] BaBar Collaboration, Phys. Rev. Lett. 90, 231801

(2003).[42] Belle Collaboration, Phys. Rev. D 72, 051105 (2005).

WedB1130