A
T

L
-P

H
Y

S-
SL

ID
E

-2
01

0-
18

1
19

Ju
ly

20
10

 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

1

The charmonium and beauty 
physics programme in ATLAS

Maria Smizanska
Lancaster University, UK 



 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

Layout of the talk

2

• ATLAS J/ψ selection strategy for early beam 
conditions

• Mass determination, method, results
• Kinematic properties of J/ψ with early selections
• First performance results with J/ψ 
• B-physics program

– two examples of early measurements under preparation
– two examples future high sensitivity B-measurements



 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

Early J/ψ: event selections

3

•     p-p collision data at 7 TeV, taken between March 30th and May 17th 2010
• Integrated luminosity of data used for this study:   6.4 ± 1.3nb-1

• Strategy: 
collect largest possible statistics; determine mass, resolution and J/ψ 
properties, understand backgrounds

• Trigger requirements:
• Minimum Bias  Trigger Scintillators (MBTS) mounted at each end of the  detector in 

front of the Liquid Argon Endcap-Calorimeter cryostats at z= ± 3.56m. The MBTS 
trigger  - requires at least two hits from either sides of the detector.

• L1 minimum bias trigger was  not prescaled for runs with luminosity < 1028  cm−2 s−1. 
• A dedicated muon software  trigger commissioning chain at the Event Filter level 

initiated by the MBTS L1 trigger searches for muon track in the entire Muon 
Spectrometer

•  Analysing data in MBTS stream we requested at least one  muon must pass 
the EF  muon-commissioning chain with a muon of any pT reconstructed in the 
Muon System

• To ensure collision events are selected, at least 3 tracks form a primary vertex. 



 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

µµ and  J/ψ selections

4

Types of muons used: 
• Combined muon:  

• statistical combination of track parameters and the covariance matrices of Muon 
System(MS)  track and Inner detector (ID) track; 

• the tracks with tight matching criteria selected to  create a combined muon track 
traversing the ID and MS 

• Tagged muon:  
• muon segments matched to  ID tracks extrapolated to  MS. Reconstructed muon adopts 

parameters of ID track.

• Pairs of muons with at least one Combined muon were retained

Cosmic ray background: 
• may come from a pair formed by  a cosmic muon and a muon from the  collision. The 

probability is very small ( < 10−4) from the 900 GeV data analysis 
• A cosmic muon mimicking a J/ψ decaying back-to-back is excluded - muons detected 

in the MS can only have momentum higher than 3 GeV.



 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

µµ and  J/ψ selections, cont

5

•    ID selections, Vertexing:
• >=  1 hit in the pixels and 6 hits in silicon strip layers  
• pT > 0.5 GeV on each track 
• Tracks fitted to a common vertex using vertexing tools based on 

Kalman filter. 
• No constraints on mass or pointing to the primary vertex, and a 

very high vertex fit χ2 upper limit is applied (χ2 < 200).

• Only ID track parameters of muons used for this J/psi study 
•    Same sign pairs retained for cross-checking. 
•   Cuts not optimized to reject backgrounds, since the aim of this study is to 
understand the shape of the low pT combinatorial background



 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

Early J/ψ signal in ATLAS 

6

• J/ψ mass and number of signal events from  
unbinned maximum-likelihood fit

An unbinned maximum-likelihood fit is used to extract the J/!mass and the number of J/! signal

candidates from the data. The likelihood function is defined by:

L=

N
!

i=1

"

fsignal(m
i
µµ)+ fbkg(m

i
µµ)

#

(1)

whereN is the total numberofpairsofoppositely chargedmuons in the invariantmass range2<mµµ < 4

GeV.The fsignal and fbkg areprobability density functions thatmodel the J/! signal andbackgroundmass

shapes in this range anda0 is the fraction of pairs originating from J/!decay. For the signal, themass is

modeledwith aGaussian distribution:

fsignal(mµµ,"mµµ)! a0
1

"
2# S"mµµ

e

#(mµµ#mJ/!)
2

2(S"mµµ)2 (2)

whose mean value mJ/! is the J/! mass and its width is a product S"mµµ, where the scale factor S

is a parameter of the fit and "mµµ is the measured mass error calculated for each muon pair from the

covariance matrix of the vertex finding. For the fits made before vertexing the per-candidate error is

calculated from the covariance matrices of the original track parameters. For the background, themass

distribution is assumed to follow aflat function:

fbkg(mµµ)! (1#a0) (3)

The fit returns values of the free parameters a0,mJ/! and S and a covariancematrix of the fit. They

are used to calculate other characteristics of the data. The number of J/! signal decays Nsig and its

uncertainty is calculated froma0 and itsuncertainty, and the total numberofpairsN. Themass resolution

$m is calculated by summing the result of the function fsignal(mµµ,"mµµ) over all candidates;$m defines

the region of the distribution for which the integral of sum over fsignal(mµµ,"mµµ) retains 68.27% of

Nsig. The uncertainty on$m is calculated using the covariancematrices of the fitted vertices. Number of

background events Nbck in themass interval mJ/! ±3$m and its error are calculated from a0,N and the
uncertainties ofa0,mJ/! and$m. The samefit procedure is applied to the prompt J/!MCevents, in this

case fbkg(mµµ)! 0.
InFigure 1 the invariantmass for all oppositely chargedmuonpairs passing vertexing is shown. The

fit results from data and MC, both with and without the application of vertexing on the di-muons are

summarised inTable 2. No systematic errors are considered in the current study. The J/!mass returned

by the fit is 3.095GeVwith an error of 0.004GeV.The number of J/! signal decays isNsig = 612±34,
themass resolution of J/! signal is$m = 82±7MeVand the number of background pairs in themass
range corresponding tomJ/! ±3$m isNbck = 332±9. In the same figure, the fit function to the prompt
J/! MC samples, normalised to the number of signal events observed, is also shown. In Figure 2, the

same data points are shown and the shape of the distribution is compared to the complete MC sample

(prompt J/! andminimumbias) as described in Section 3.

InTable2 it is seen that the vertexing has no statistically significant e!ect on themass resolution and

mean position, while the background is reduced slightly (about 6%). Larger background rejection from

vertexing is expected if there is a significant contribution from leptons originating from semileptonic

decays of heavy flavour, B andDmesons [11].

Figure 3 shows the invariant mass distribution of the opposite sign pairs after vertexing, with the

same sign superimposed. The same sign pairs contribution in the J/! signal region is lower than the

background fromopposite signpairs, as expected. The small excess in thebackground contribution from

opposite signpairs is in agreementwith the expected contribution ofmuons fromB andDmesondecays

in the current data.

5

An unbinned maximum-likelihood fit is used to extract the J/!mass and the number of J/! signal

candidates from the data. The likelihood function is defined by:

L=

N
!

i=1

"

fsignal(m
i
µµ)+ fbkg(m

i
µµ)

#

(1)

whereN is the total numberofpairsofoppositely chargedmuons in the invariantmass range2<mµµ < 4

GeV.The fsignal and fbkg areprobability density functions thatmodel the J/! signal andbackgroundmass

shapes in this range anda0 is the fraction of pairs originating from J/!decay. For the signal, themass is

modeledwith aGaussian distribution:

fsignal(mµµ,"mµµ)! a0
1

"
2# S"mµµ

e

#(mµµ#mJ/!)
2

2(S"mµµ)2 (2)

whose mean value mJ/! is the J/! mass and its width is a product S"mµµ, where the scale factor S

is a parameter of the fit and "mµµ is the measured mass error calculated for each muon pair from the

covariance matrix of the vertex finding. For the fits made before vertexing the per-candidate error is

calculated from the covariance matrices of the original track parameters. For the background, themass

distribution is assumed to follow aflat function:

fbkg(mµµ)! (1#a0) (3)

The fit returns values of the free parameters a0,mJ/! and S and a covariancematrix of the fit. They

are used to calculate other characteristics of the data. The number of J/! signal decays Nsig and its

uncertainty is calculated froma0 and itsuncertainty, and the total numberofpairsN. Themass resolution

$m is calculated by summing the result of the function fsignal(mµµ,"mµµ) over all candidates;$m defines

the region of the distribution for which the integral of sum over fsignal(mµµ,"mµµ) retains 68.27% of

Nsig. The uncertainty on$m is calculated using the covariancematrices of the fitted vertices. Number of

background events Nbck in themass interval mJ/! ±3$m and its error are calculated from a0,N and the
uncertainties ofa0,mJ/! and$m. The samefit procedure is applied to the prompt J/!MCevents, in this

case fbkg(mµµ)! 0.
InFigure 1 the invariantmass for all oppositely chargedmuonpairs passing vertexing is shown. The

fit results from data and MC, both with and without the application of vertexing on the di-muons are

summarised inTable 2. No systematic errors are considered in the current study. The J/!mass returned

by the fit is 3.095GeVwith an error of 0.004GeV.The number of J/! signal decays isNsig = 612±34,
themass resolution of J/! signal is$m = 82±7MeVand the number of background pairs in themass
range corresponding tomJ/! ±3$m isNbck = 332±9. In the same figure, the fit function to the prompt
J/! MC samples, normalised to the number of signal events observed, is also shown. In Figure 2, the

same data points are shown and the shape of the distribution is compared to the complete MC sample

(prompt J/! andminimumbias) as described in Section 3.

InTable2 it is seen that the vertexing has no statistically significant e!ect on themass resolution and

mean position, while the background is reduced slightly (about 6%). Larger background rejection from

vertexing is expected if there is a significant contribution from leptons originating from semileptonic

decays of heavy flavour, B andDmesons [11].

Figure 3 shows the invariant mass distribution of the opposite sign pairs after vertexing, with the

same sign superimposed. The same sign pairs contribution in the J/! signal region is lower than the

background fromopposite signpairs, as expected. The small excess in thebackground contribution from

opposite signpairs is in agreementwith the expected contribution ofmuons fromB andDmesondecays

in the current data.

5

An unbinned maximum-likelihood fit is used to extract the J/!mass and the number of J/! signal

candidates from the data. The likelihood function is defined by:

L=

N
!

i=1

"

fsignal(m
i
µµ)+ fbkg(m

i
µµ)

#

(1)

whereN is the total numberofpairsofoppositely chargedmuons in the invariantmass range2<mµµ < 4

GeV.The fsignal and fbkg areprobability density functions thatmodel the J/! signal andbackgroundmass

shapes in this range anda0 is the fraction of pairs originating from J/!decay. For the signal, themass is

modeledwith aGaussian distribution:

fsignal(mµµ,"mµµ)! a0
1

"
2# S"mµµ

e

#(mµµ#mJ/!)
2

2(S"mµµ)2 (2)

whose mean value mJ/! is the J/! mass and its width is a product S"mµµ, where the scale factor S

is a parameter of the fit and "mµµ is the measured mass error calculated for each muon pair from the

covariance matrix of the vertex finding. For the fits made before vertexing the per-candidate error is

calculated from the covariance matrices of the original track parameters. For the background, themass

distribution is assumed to follow aflat function:

fbkg(mµµ)! (1#a0) (3)

The fit returns values of the free parameters a0,mJ/! and S and a covariancematrix of the fit. They

are used to calculate other characteristics of the data. The number of J/! signal decays Nsig and its

uncertainty is calculated froma0 and itsuncertainty, and the total numberofpairsN. Themass resolution

$m is calculated by summing the result of the function fsignal(mµµ,"mµµ) over all candidates;$m defines

the region of the distribution for which the integral of sum over fsignal(mµµ,"mµµ) retains 68.27% of

Nsig. The uncertainty on$m is calculated using the covariancematrices of the fitted vertices. Number of

background events Nbck in themass interval mJ/! ±3$m and its error are calculated from a0,N and the
uncertainties ofa0,mJ/! and$m. The samefit procedure is applied to the prompt J/!MCevents, in this

case fbkg(mµµ)! 0.
InFigure 1 the invariantmass for all oppositely chargedmuonpairs passing vertexing is shown. The

fit results from data and MC, both with and without the application of vertexing on the di-muons are

summarised inTable 2. No systematic errors are considered in the current study. The J/!mass returned

by the fit is 3.095GeVwith an error of 0.004GeV.The number of J/! signal decays isNsig = 612±34,
themass resolution of J/! signal is$m = 82±7MeVand the number of background pairs in themass
range corresponding tomJ/! ±3$m isNbck = 332±9. In the same figure, the fit function to the prompt
J/! MC samples, normalised to the number of signal events observed, is also shown. In Figure 2, the

same data points are shown and the shape of the distribution is compared to the complete MC sample

(prompt J/! andminimumbias) as described in Section 3.

InTable2 it is seen that the vertexing has no statistically significant e!ect on themass resolution and

mean position, while the background is reduced slightly (about 6%). Larger background rejection from

vertexing is expected if there is a significant contribution from leptons originating from semileptonic

decays of heavy flavour, B andDmesons [11].

Figure 3 shows the invariant mass distribution of the opposite sign pairs after vertexing, with the

same sign superimposed. The same sign pairs contribution in the J/! signal region is lower than the

background fromopposite signpairs, as expected. The small excess in thebackground contribution from

opposite signpairs is in agreementwith the expected contribution ofmuons fromB andDmesondecays

in the current data.

5

• δmμμ - measured mass uncertainty of each pair 
of muon tracks

• S - scale factor to cover for unaccounted 
uncertainties on track parameters (e.g. non-
gaussian tails)

•  The measured mass agrees with PDG within statistical precision of first data
•  mass resolution agrees with that expected from MC 

Table 2: Summary of fit results tomass distributions of J/! ! µ+µ" candidates. The number of back-
groundevents is given in the rangemJ/!±3"m. The samefit is applied toprompt J/!MCdata, assuming
fbkg # 0 in the formula 1. Results for data before vertexing are shown for comparison.

mJ/!, GeV "m,MeV Nsig Nbck S

all data 3.095±0.004 82±7 612±34 332±9 1.21±0.07
MC 3.098±0.001 74±0.4 1.09±0.01
data n/v 3.096±0.004 82±7 612±34 351±10 1.20±0.07

BB data 3.097±0.005 36±6 69±9 8±1 1.12±0.14
MC 3.098±0.001 37±0.7 1.10±0.02
data n/v 3.099±0.005 38±7 69±9 8±1 1.14±0.15

EB data 3.089±0.008 66±12 88±11 34±3 1.32±0.16
MC 3.097±0.001 53±0.8 1.08±0.01
data n/v 3.089±0.009 66±12 87±11 36±3 1.30±0.17

EE data 3.095±0.006 88±9 437±31 324±10 1.17±0.09
MC 3.098±0.001 82±0.5 1.09±0.01
data n/v 3.096±0.006 88±9 437±31 344±10 1.16±0.09

Table 3: The fraction ofmuon pairs in the reconstructed J/! in data andMC

combined combined combined tagged

Data (26.4±1.4)% (73.6±1.4)%
MC (31±0.3)% (69±0.3)%

9



 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

Properties of early J/ψ in ATLAS

7

• Data agree with MC predictions of 
resolution and PDG mass 

• Also good agreement between data and 
MC on kinematic properties of J/psi 

• Essential conclusions derived from 
these first  J/ψ signal studies.



 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

Properties of muons from early J/ψ

8

• Our early analysis can access very low pT J/ψ  
producing soft pT muons, see left top

•  Muons with enough energy to cross the 
calorimeters reach the MS mainly in the 
forward region 

• This is a consequence of the muon acceptance 
of the ATLAS detector without any threshold 
requirement on the muon trigger, see the 
muon efficiency (left bottom, black) 
determined from MC



 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

Performance of early J/ψ in ATLAS

9

• J/ψ mass resolution varies with the pseudorapidity 
of  muons  accordingly to MC expectations 

– endcap  2.5 > | η| > 1.05, barrel |η| < 1.05  

• no statistically significant mass shifts from the PDG 
value observed in any of the pseudorapidity regions

 both 
muons 
barrel BB

 both 
muons in 
endcap EE

one muon  
barrel one 
endcap (EB)

Table 2: Summary of fit results tomass distributions of J/! ! µ+µ" candidates. The number of back-
groundevents is given in the rangemJ/!±3"m. The samefit is applied toprompt J/!MCdata, assuming
fbkg # 0 in the formula 1. Results for data before vertexing are shown for comparison.

mJ/!, GeV "m,MeV Nsig Nbck S

all data 3.095±0.004 82±7 612±34 332±9 1.21±0.07
MC 3.098±0.001 74±0.4 1.09±0.01
data n/v 3.096±0.004 82±7 612±34 351±10 1.20±0.07

BB data 3.097±0.005 36±6 69±9 8±1 1.12±0.14
MC 3.098±0.001 37±0.7 1.10±0.02
data n/v 3.099±0.005 38±7 69±9 8±1 1.14±0.15

EB data 3.089±0.008 66±12 88±11 34±3 1.32±0.16
MC 3.097±0.001 53±0.8 1.08±0.01
data n/v 3.089±0.009 66±12 87±11 36±3 1.30±0.17

EE data 3.095±0.006 88±9 437±31 324±10 1.17±0.09
MC 3.098±0.001 82±0.5 1.09±0.01
data n/v 3.096±0.006 88±9 437±31 344±10 1.16±0.09

Table 3: The fraction ofmuon pairs in the reconstructed J/! in data andMC

combined combined combined tagged

Data (26.4±1.4)% (73.6±1.4)%
MC (31±0.3)% (69±0.3)%

9



 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

Comparison with like sign pairs  

10

Early di-muon pairs selected at lowest pT have 
specific features visible when comparing like sign 
pairs with J/ψ candidates
• like sign pairs almost match the level  of the 

J/ψ background (unlike pairs)  in the side 
bands 

• source of both dominated by muons 
from K/pi decays

• very little b/c content  in tails
 

  Di-muon pairs of opposite sign in the J/ψ region have evidently different kinematic properties from the like 
sign pairs
  Di-muon pairs of opposite sign in the J/ψ region have evidently different kinematic properties from the like 
sign pairs



 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

ATLAS B-physics program 

11

• ATLAS B-physics program is realised in following sub-projects
HF quarkonia measurements 
B → J/ψ (inclusive, exclusive) channels
Rare B-decays Bsd →μμ,  b →s μμ,  b →d μμ 
Production properties of B and D-mesons decaying into hadrons

• Each sub-project has tasks/measurements for early, medium 
and advanced periods
• First measurements, in addition to physics results, serve to 

improveunderstanding of detector performance to allow  later high 
precision measurements 

• Selected examples of MC based studies are given further  for 
the early and for advanced periods
• Complete B-physics program arXiv:0901.0512 ; CERN-OPEN-2008-020, 

Chapter 11.



 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

Early measurements with exclusive B → J/ψ 

12

 Mass (MeV)
4900 5000 5100 5200 5300 5400 5500 5600 5700
0

50

100

150

200

250

300

350

400

 Mass (MeV)
4900 5000 5100 5200 5300 5400 5500 5600 5700
0

50

100

150

200

250

300

350

400

 Decay time (ps)
-2 0 2 4 6 8 10

1

10

210

310  X J/bb 

 X J/pp 

0* K J/ dB

Parameter Simulated value Fit result with statitical error
!, ps!1 0.651 0.73 ± 0.07
m(B), GeV 5.279 5.284 ± 0.006

!, ps 0.132 ± 0.004

 Mass (MeV)
4900 5000 5100 5200 5300 5400 5500 5600 5700
0

50

100

150

200

250

300

350

400

 Mass (MeV)
4900 5000 5100 5200 5300 5400 5500 5600 5700
0

50

100

150

200

250

300

350

400

 Decay time (ps)
-2 0 2 4 6 8 10

1

10

210

310  X J/bb 

 X J/pp 

0* K J/ dB

!s, ps!1 0.683 0.743 ± 0.051
m(B), GeV 5.343 5.359 ± 0.006

!, ps 0.152 ± 0.001

 Mass (MeV)
4900 5000 5100 5200 5300 5400 5500 5600 5700
0

50

100

150

200

250

300

350

400

 Mass (MeV)
4900 5000 5100 5200 5300 5400 5500 5600 5700
0

50

100

150

200

250

300

350

400

 Decay time (ps)
-2 0 2 4 6 8 10

1

10

210

310  X J/bb 

 X J/pp 

0* K J/ dB

 Mass (MeV)
5000 5100 5200 5300 5400 5500 5600 5700
0

200

400

600

800

1000

1200

1400

1600

1800

2000

2200

 Mass (MeV)
5000 5100 5200 5300 5400 5500 5600 5700
0

200

400

600

800

1000

1200

1400

1600

1800

2000

2200

 Decay time (ps)
-2 0 2 4 6 8 10

1

10

210

310

410
 X J/bb 

 X J/pp 

  J/ sB

 
Applying simultaneous mass -      
lifetime likelihood  fit to events 
•  B →  J/ψ K0*  (10 pb-1) 
•  Bs → J/ψ φ   (150 pb-1) 
  Lifetimes measured with   
sensitivity  better than 10%. Early 
lifetime measurements test the 
calibrations and alignments 
necessary for precise CPV studies

 
Applying simultaneous mass -      
lifetime likelihood  fit to events 
•  B →  J/ψ K0*  (10 pb-1) 
•  Bs → J/ψ φ   (150 pb-1) 
  Lifetimes measured with   
sensitivity  better than 10%. Early 
lifetime measurements test the 
calibrations and alignments 
necessary for precise CPV studies

 
Applying simultaneous mass -      
lifetime likelihood  fit to events 
•  B →  J/ψ K0*  (10 pb-1) 
•  Bs → J/ψ φ   (150 pb-1) 
  Lifetimes measured with   
sensitivity  better than 10%. Early 
lifetime measurements test the 
calibrations and alignments 
necessary for precise CPV studies



 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

Differential cross section B+ →J/ψ K +  

13

The B+ → J/ψK+  total and 
differential production cross-
sections  
• With 10 pb−1 the total cross-

section can be measured with 
a statistical precision better 
than 5% 

•  The differential cross-section 
with precision of the order of 
10%.

pT range [GeV] pT ! [10,18] pT ! [18,26] pT ! [26,34] pT ! [34,42]
A [%] 20.1±1.0 37.3±1.7 45.0±3.1 51.6±4.7

!(B+) [ MeV] 38.5±2.0 42.3±2.1 46.1±3.2 46.6±4.0

Table 3: Efficiency A and B+ masswidth !(B+) for the various pT bins.

B+ mass (MeV)
5200 5300 5400 5500 5600 5700 5800

E
ve

nt
s 

/ (
 6

.5
 M

eV
 )

0

10

20

30

40

50

60

70 nbkg =  310 +/- 29
nsig =  710 +/- 36
sigma =  38.5 +/- 2.0 Mev

B+ mass (MeV)
5200 5300 5400 5500 5600 5700 5800

E
ve

nt
s 

/ (
 6

.5
 M

eV
 )

0

10

20

30

40

50

60

70

(a) 10! pT < 18GeV

B+ mass (MeV)
5200 5300 5400 5500 5600 5700 5800

E
ve

nt
s 

/ (
 6

.5
 M

eV
 )

0

10

20

30

40

50

60

70 nbkg =  183 +/- 26
nsig =  779 +/- 36
sigma =  42.3 +/- 2.1 Mev

B+ mass (MeV)
5200 5300 5400 5500 5600 5700 5800

E
ve

nt
s 

/ (
 6

.5
 M

eV
 )

0

10

20

30

40

50

60

70

(b) 18! pT < 26GeV

B+ mass (MeV)
5200 5300 5400 5500 5600 5700 5800

E
ve

nt
s 

/ (
 6

.5
 M

eV
 )

0

5

10

15

20

25

30 nbkg =  51 +/- 16
nsig =  335 +/- 23
sigma =  46.1 +/- 3.2 Mev

B+ mass (MeV)
5200 5300 5400 5500 5600 5700 5800

E
ve

nt
s 

/ (
 6

.5
 M

eV
 )

0

5

10

15

20

25

30

(c) 26! pT < 34GeV

B+ mass (MeV)
5200 5300 5400 5500 5600 5700 5800

E
ve

nt
s 

/ (
 6

.5
 M

eV
 )

0

2

4

6

8

10

12

14

16 nbkg =  27.2 +/- 9.8
nsig =  165 +/- 15
sigma =  46.6 +/- 4.0 Mev

B+ mass (MeV)
5200 5300 5400 5500 5600 5700 5800

E
ve

nt
s 

/ (
 6

.5
 M

eV
 )

0

2

4

6

8

10

12

14

16

(d) 34! pT < 42GeV

Figure 4: Fit of the B+ mass in various pT ranges: pT ! [10,18] GeV (a), pT ! [18,26] GeV (b),
pT ! [26,34] GeV(c), pT ! [34,42] GeV (d).

total cross-section
A [%] 29.8±0.8
!(B+) [ MeV] 42.2±1.3

Table 4: Overall efficiency and B+ masswidth for all B+ with pT > 10GeV.

4.4 Lifetime measurement

Themeasurement of the lifetime " of the selected B+ candidates is a sensitive tool to confirm the beauty
contents in a sample, in particular the number of the reconstructed B+ " J/#K+ decays obtained in the
bb̄ " J/#X dataset. Theproper decay time is defined as t = $/c. For this analysis, no cut on the proper
decay length $ (Equation 2) of the J/# candidate or the B+ candidate should be applied.
Theproperdecay timedistribution in the signal regionB+ " J/#K+ canbeparametrised asaconvo-

lution of an exponential functionwith aGaussian resolution function, while the background distribution

7

B-PHYSICS – PRODUCTION CROSS-SECTION MEASUREMENTS AND STUDY OF THE . . .

79

1117

Fit of the B+ mass in  four 
pT! ranges

Signal lifetime ! [ps] 1.637±0.036
BG lifetime !1 [ps] 1.320±0.24
BG lifetime !2 [ps] 0.370±0.067

Table 5: Results for the lifetime fit.

5 Statistical and Systematic Uncertainties

From the analysis presented above, the expected number of reconstructed B+ candidates amounts to 160
perpb!1 . This implies that sufficient statistics canbecollected for a reliable cross-sectionmeasurement,
after just a fewmonthsofdata takingat the initial low luminosityphaseof theLHC.This scenario isvalid
for a luminosity less than L = 1032 cm!2s!1, since the analysis was performed without pileup events
and contains no special trigger requirements or prescaling other than a singlemuonwith pµT > 6 GeVat
level-1.
For the measurements presented in this note the main sources of systematic uncertainties are the

same. The uncertainty from the luminosity in the initial phase is estimated to be 10 % and will be
reduced to about 6.5 % after 0.3 fb!1 of data. The uncertainty from the PDF’s is estimated to be 3 %,
while the scale uncertainty of the NLO calculations is about 5 %. Finally, the uncertainty originating
from the muon identification is about 3 %. Assuming Gaussian distributions for the above mentioned
uncertainties, the total systematic uncertainty of the signal varies from 9.2% to 12%and is dominated
by the uncertainty in the luminosity.
Given that a statistical precision of O(1%) will be reachedwith an integrated luminosity of 0.1 fb!1,

the contribution of the systematics will dominate the uncertainties of the first measurements. This is
the case even for the differential cross-section measurement. Although the statistics in each pT bin
is limited, the total uncertainty is dominated by the systematic uncertainties in the branching ratio of
the B+ ! J/"K+ and in the luminosity, which are of the same order. For the exclusive cross-section
measurement in the B+!J/"K+ channel, the relative uncertainties of the differential and total cross-
sections are given in Table 6. Therein, the first row of the table contains the quadratic sum of the
statistical uncertainty corresponding to an integrated luminosity of 0.01 fb!1 and the uncertainty in the
efficiency. The latter is based on the statistics of the Monte Carlo dataset used. The second row is
calculated by adding in quadrature the above uncertainty to the systematic uncertainty of the luminosity
and the branching ratio for every pT bin.
For the high statistics pT bins as well as for the total cross-section, the total relative uncertainty

is dominated by systematic errors, originating mainly from the uncertainty in the luminosity, which is
assumed tobe10%for the initial phase, and the10%uncertainty in thebranching ratioofB+ ! J/"K+.
The effect of the assumed background shape on the measurements is estimated to be less than 1 %.
Finally, the precision of the lifetimemeasurement, for the same integrated luminosity is 2.5%,where no
systematic effects are taken into account.

pT range [GeV] pT " [10,18] pT " [18,26] pT " [26,34] pT " [34,42] pT " [10,inf)
stat. + A [%] 7.7 6.9 10.5 13.9 4.3
total [%] 16.1 15.8 17.6 19.8 14.8

Table 6: Statistical and total uncertainties for the B+ ! J/"K+ differential and total cross-sectionmea-
surements for an integrated luminosity of 0.01 fb!1. Total uncertainties include luminosity and BR
systematic uncertainties.

9

B-PHYSICS – PRODUCTION CROSS-SECTION MEASUREMENTS AND STUDY OF THE . . .

81

1119



 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

Production polarization of Λb  with 5 fb-1

14

Polarization varies with pseudorapidity 
thus ATLAS/CMS and LHCb can perform 
complemetary measurements to map full 
range.
With 5 fb-1 the  Λb polarization in ATLAS 
can be measured with precision of 0.07 



• The ATLAS performance was analysed for 
first  J/ψ→μμ  Shown that di-muon 
performance with real data consistent with 
MC predictions

• The MC simulation of Bs →μμ potential 
(left) to  test potential with 10 fb-1 was done 
with  trigger  menus for > 1033 

• both muons required pT> 6 GeV. 
• Low  pT B-physics di-muon triggers will be 

applied at low instantaneous luminosities 
of early LHC period to maximize reach for 
first sensitivity 

• At  ~ 1034 dedicated triggers prepared to  
use full ATLAS potential for Bs →μμ 

• Bs →μμ Bd →μμ in physics program for 
ATLAS upgrade

Bs →μμ signal and backgrounds after applying all 
selection cuts - relevant at > 1033 cm -2 s -1

• The ATLAS performance was analysed for 
first  J/ψ→μμ  Shown that di-muon 
performance with real data consistent with 
MC predictions

• The MC simulation of Bs →μμ potential 
(left) to  test potential with 10 fb-1 was done 
with  trigger  menus for > 1033 

• both muons required pT> 6 GeV. 
• Low  pT B-physics di-muon triggers will be 

applied at low instantaneous luminosities 
of early LHC period to maximize reach for 
first sensitivity 

• At  ~ 1034 dedicated triggers prepared to  
use full ATLAS potential for Bs →μμ 

• Bs →μμ Bd →μμ in physics program for 
ATLAS upgrade

• The ATLAS performance was analysed for 
first  J/ψ→μμ  Shown that di-muon 
performance with real data consistent with 
MC predictions

• The MC simulation of Bs →μμ potential 
(left) to  test potential with 10 fb-1 was done 
with  trigger  menus for > 1033 

• both muons required pT> 6 GeV. 
• Low  pT B-physics di-muon triggers will be 

applied at low instantaneous luminosities 
of early LHC period to maximize reach for 
first sensitivity 

• At  ~ 1034 dedicated triggers prepared to  
use full ATLAS potential for Bs →μμ 

• Bs →μμ Bd →μμ in physics program for 
ATLAS upgrade

 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

ATLAS potential for B → µµ

15

Selection cut B0s ! µ+µ! efficiency bb̄ ! µ+µ!X efficiency
Iµµ > 0.9 0.24 (2.6±0.3) ·10!2

Lxy > 0.5mm 0.26 (1.4±0.1) ·10!2 (1.0±0.7) ·10!3
" < 0.017 rad 0.23 (8.5±0.2) ·10!3
Mass in ["#,2#] 0.76 0.079
TOTAL 0.04 0.24·10!6 (2.0±1.4) ·10!6

Events yield 5.7 14+13!10



 The Charmonium and Beauty  physics programme in ATLAS, M.Smizanska, BEACH 2010, Perugia

Summary
• Early J/ψ data taken with minimum bias trigger show 
excellent agreement with expected performance 

• Reproducing J/ψ PDG mass in all pseudorapidity regions 
- confirms that  pT scale understood at low pT range 

• J/ψ mass resolution over entire pseudorapidity regions 
of detector - consistent with MC.

• B-physics program prepared for both early and advanced 
periods.

• ATLAS will significantly contribute to B →μμ potential as 
an instantaneous LHC luminosity  will be increased to  
several times 1033 and to a nominal value 1034   Rare B 
decays for detector upgrade  being prepared.

16