a.n.skachkova, n.b.skachkov (jinr, dubna)

QCD at intermediate energies:

Multimuon production and double parton interactions at PANDA

A.N.Skachkova,N.B.Skachkov

(JINR, Dubna)

2-nd PANDA 2-nd PANDA RussiaRussia

workshop workshop 26-28 April 2010, 26-28 April 2010,

IHEPIHEP

2

I. Antiproton beam with Ebeam < 15 GeV may provide an interesting information about quark dynamics inside the hadron and proton structure in the energy region where the perturbative QCD comes in interplay with a reach resonance physics.

II. Different to electron beams, used for measurements of

proton structure functions in the region of negative values of the square of transversed momentum (q^2 < 0, “space-like” region), antiproton-proton collisions allow to make measurements in the region of positive values of the square of the transverse momentum (q^2 > 0, “time-like”, region, less

studied ! ).

Anna Skachkova: : “Lepton pair production at intermediate energies and the main backgrounds ”, “Lepton pair production at intermediate energies and the main backgrounds ”, 22.11.0822.11.08

Quark-antiquark annihilation and gluon-quark scattering subprocess in antiproton-proton collision, in which the energetic

electromagentic particles (e, muon, photon) are produced, planned to be the subject of our study at PANDA experiment.

The results of Monte Carlo simulation of processes with lepton pair and prompt photon production, done by use of PYTHIA generator, are presented here. They allowed to work out the cuts needed for background separation and to make reasonable estimations of expected selection rates for these processes. In this talk we base on the results obtained for a case of Ebeam = 14 GeV (i.e., Ecm = 5.3 GeV).

The corrections for the final estimations are planned to be done in the nearest years on the basis of the existing detector simulation tools.


4

New

measuremens

and combination

with old (97-98)

results

at

low Q**2

at H1 and

ZEUS

5

Nowdays

there are

new

measuremens

at

low Q**2

at JLab

GeVQ 0.81.0 2

6

Production of lepton pair (continum Minv (l+l-) case).

The process of lepton pair production q qbar / Z˚ l+l-

is of big physical interest because:

A. The spectrum of final state leptons (e and muons) obviously

depends on the form of parton distributions inside colliding protons.

B. The transverse momentum of lepton pair PT ( l+ l-) provides the

information about intrinsic transverse momentum < kT> of

quark inside the proton ( Fermi motion).

C. Double lepton pairs production rate gives the information about

multiple parton interactions (MPI), which rate depends on the

space distribution of partons, i.e. on space structure of proton.

7

V.A. Matveev, R.M. Muradian, A.N. Tavkhelidze (MMT)( V.A. Matveev, R.M. Muradian, A.N Tavkhelidze, JINR P2-4543, JINR, Dubna, 1969; SLAC-TRANS-0098, JINR R2-4543, Jun 1069; 27p. )

process, called also as Drell-Yan( S.D. Drell, T.M. Yan, SLAC-PUB-0755, Jun 1970,12p.; Phys.Rev.Lett. 25(1970)316-320, 1970 )

The dominant mechanism

of the l+l- production is

the perturbative QED/QCD

partonic 2 2 subprocess

qiqi / Z˚ l+l-

σ = 5.9 E- 06 mbPYTHIA 6 simulation for the E beam = 14 GeV ( i.e., Ecm =5.3 GeV).

without detector effects (“ideal detector” --> all particles are detected)allows a proper account of the relativistic kinematics during the simulation


8

Signal l± :

0 ≤ El ≤ 10 GeV,

<El> = 2.6 GeV,

Epeak = 0.4 GeV

0 ≤ PTl ≤ 2 GeV,

<PTl> = 0.7 GeV <Θl> = 27.3 °

some Θl > 90°!!!


9

In each signal l± event: El ”fast” > El ”slow”

Left column

0 < El ”slow” < 5 GeV

0 < Θl ”slow” < 180° (mostly)

But !!!

Some of less energetic slow leptons have Θl ”slow” > 90°

Right column

0.4 < El ”fast” < 10 GeV

0 < Θl ”fast” < 60°


10

Fake electron distributions

The number of events which include fake electrons is about 1 -

2% of events

The most of electron pairs do appear as Dalitz pairs (πºe+e-)

produced in decays of neutral mesons, which appear from η, ω

or more heavy mesons or barions, produced in their turn as the resonance states according to the LUND fragmentation model.

The life time of these resonance states is rather short the electron

pairs are produced close to the interaction point the vertex position information will not be

efficient for the Signal / Background separation..

In approximation when particles are allowed to decay in cylinder volume R=2500 x

L=8000 mm


11

Fake muon distributions

1.The part of signal events which include fake muons is about 16%.

2.Up to 4 fake muons in the final state.

3. Fake muons production

vertices are distributed

within detector volume

Vertex position information will be useful

for Signal / Background separationAnna Skachkova: : “Lepton pair production at intermediate energies and the main backgrounds ”, “Lepton pair production at intermediate energies and the main backgrounds ”,

22.11.0822.11.08

12

Parents of fake leptons

The most probable parents of fake

electrons are neutral pions (Dalits decay)

muons are charged pions

The most probable grandparents of fake

electrons are string (Lund model),

ρ+,η, ω, Δ0 ,Δ+ ,Λ0

muons are string (Lund model),

ρ0, ρ+, ω, Δ+ ,Δ++ ,Λ0


13

Global variable for

q qbar l +l - process - Minv l+l-

Minv l+l- = √(Pl+ + Pl-)² = q2

Minv l +l- min = Minv¯qq = 1 GeV – originates from the internal

PYTHIA restriction

Minv¯qq = √(Pq + P¯q)² = m_hat


14

Background QCD processes for the q q l +l -

one

The generation was done with the use of more than 20 QCD subprocesses existed in PYTHIA (including the signal one ¯q q l +l - ).

The main contributions come from the following partonic subprocesses:

q + g q + g (gives 50% of events with the σ = 4.88 mb);

g + g g + g (gives 30% of events with the σ = 2.96 mb);

q + q’ q + q’ (gives 18% of events with the σ = 1,75 mb);

q + q bar g + g (gives 0.6% of events with the σ = 5.89 E- 02 mb);

¯q + q l+ + l- (has 0.00005% of events with the σ = 5.02 E- 06 mb);

So, initially we have 1 signal event among 2.000.000 of QCD background S/B ≃ 5.5 * 10-6

The simulation was done within approximation when particles are allowed to decay

in cylinder volume R=2500 x L=8000 mmAnna Skachkova: : “Lepton pair production at intermediate energies and the main backgrounds ”, “Lepton pair production at intermediate energies and the main backgrounds ”,

22.11.0822.11.08

15

QCD Muon background

The shape of muon distributions, produced in these background events do not differ from those of fake “decay” muons, produced in the signal process

a rather high probability of appearing the muon pair with the different signs of their charges in QCD events (which are other than the signal one)

fake pretty good the signal events


16

QCD background electrons

The shape of QCD background electron’s distributions are identical to the

ones, surrounding the signal process.


The electron pairs are produced close to the

interaction point the vertex position information

will not be efficient for the Signal / Background separation..


17

Parents of QCD background leptons

The most probable parents of fakeelectrons are neutral pions (πºe+e-)

muons are charged pions

The most probable grandparents of fakeelectrons are strings (Lund model),

ρ+,η, ω, Δ0 ,Δ+

muons are strings (Lund model),ρ0, ρ+, ω


18

The main source of background for the signal q qbar l +l – are the Minimum-Bias processes:

1. Low - PT scattering (gives 68% of events with the σ = 34.25 mb);

2. Single diffractive (gives 6% of events with the σ = 3.32 mb); and other

3. ¯q + q l+ + l- (has 0.000012% of events with the σ = 5.9 E- 06 mb);

So, we have 1 signal event among 8.333.333 of Mini-bias bkgd

S/B ≃ 10-7

This source of background is 5 times harder !!! than QCD background


19

Minimun-bias background electrons

The shape of QCD background electron’s distributions are identical to the

ones, surrounding the signal process.


The electron pairs are produced close to the

interaction point the vertex position information

will not be efficient for the Signal / Background separation..


20

The plots show the distributions over summarized energy of the final state

particles in the cones of radius R isolation = √ η2+φ2 respect to the

(η – pseudorapidity)

upper plot signal eventssignal events

bottom plot Mini-bias background

Isolation criteria (R R isolationisolation = 0.2 ) = 0.2 )E (of particles) = 0.5 GeV

allows to separate 100% of Mini-bias bkg leptons

with the loss of 8% of signal events

(after applied 3 cuts discussed above + cut M inv (l+,l-) > 0.9)

Lepton (µ) isolation criteria


21

Cuts used for separation of single lepton pairs from Minimum bias and QCD background

1. selection of events with the only 2 leptons, having;

2. these 2 leptons must have opposite charges;

3. the vertex of lepton origin lies within the R< 15mm from the interaction point;

4. Minv ( 2 leptons) ≥ 0.9 GeV;

5. leptons do satisfy the isolation criteria: the summed energy ( Esum ) of all particles within the cone of

R(in eta-phi space) =0.2 is less than 0.5 GeV, i.e., Esum < 0.5 GeV.

22

Efficiency of cuts and S/B ratio

The total loss of signal events after application of all five cuts is expected to be about 20%.

So, one can expect to gain a huge sample of about

7x106 signal dilepton events per 1 year (10^7 sec.)

Direct photon production in 4pp collisions at PANDA energies

(E beam = 14 GeV)

A.N.Skachkova

(JINR, Dubna)

PANDA EM group workshop, Ferrara 2007

First (pp -> X) experiments were done at CERN:

at ISR ( pp - collider, Ecm = 31, 45, 53 GeV ): R412 experiment (Darriulat et. al., 1976); R107 (Amaldi et. al., 1978);

WA98 (Aggarwal et. al., 2000); and many others at higher E.

Also at Fermilab and BNL:

E557 (Baltrusaitis et. al. 1979, pBe, Ecm = 19.4, 23.7 GeV)

E629 (McLaughlin et. al. 1983, pC, Ecm = 19.4 GeV)

N. Skachkov: EM group. ITEP PANDA workshop, 22 October 2007

Previous experiments

The proton-antiproton cross section ( ) in quark-parton model is expressed as follows:

.

pp

tdˆ/

dHere f( x, Q**2) are the structure functions, is the 22 parton level ( ) cross section ;

= is the square of transfered momentum, is the energy fraction carried by a quark in a proton.

In our case, ,( or ) and

213 )(ˆ ppt 2Q

4321 pppp

2,1x

22 )( TpQ gqqd qgqdd ˆ

Nikolay Skachkov: ISHEPP XIX, Dubna, Sept.29-Oct.4, 2008

There is also another diagram that describes fragmentation into a photon.

Its contribution is supressed by photon isolation criteria

and drops with -growth tp

Physical goal: New information about proton structure functions (gluon-distrib.) in new x-Q**2 kinematical region.

Our Physical goal: To estimate the possibility of getting the information about proton structure functions

f(x, Q**2) .

Main interest: To estimate the size of the x-Q**2 kinematical region in which the structure functions can be measured.

Cross Sections

Total: = 2.9E-03 mb;

q + qbar gluon + photon (62%):

= 1.78E-03 mb;

gluon + q q + photon (38%):

= 1.10E-03 mb

N. Skachkov: EM group. ITEP PANDA workshop, 22 October 2007

The diagrams contribution

The contributions of both diagrams to the total cross section “Annihilation”

diagram

“QCD Compton”

diagram

0 ≤ 0 ≤ E E γγ ≤ 10 GeV, < ≤ 10 GeV, <E E γγ > = 2.6 GeV > = 2.6 GeV 0 ≤ 0 ≤ PT PT γγ ≤ 2 GeV, < ≤ 2 GeV, <PT PT γγ > = 0.67 GeV > = 0.67 GeV 0 ≤ 0 ≤ ΘΘ γγ ≤ 180 ≤ 180 ° , ° , <<ΘΘ γγ > = 29.6 °> = 29.6 ° some some ΘΘ γγ > 90> 90°°

Anna Skachkova: EM group workshop, Ferrara, 15-16 October 2007

Signal photons distributions

In pbar + p gamma + X process (chosing pbar beam direction as the z-axis) the role of the transfered momentum Q

plays the photon transverse momentum, i.e.

As it shall be shown below, at E E beam beam = 14 GeV= 14 GeV

we can expectwe can expect < 2 GeV, < 2 GeV, i.e,i.e,

This region is under study now at HERA and JLab but in “space- This region is under study now at HERA and JLab but in “space- like” region of the square of the momentum transferred q^2 <0.like” region of the square of the momentum transferred q^2 <0.

QTp

Tp

22 )( TpQ 24GeV

0.05 < x < 0.7


Structure functions distributions

22 )( TpQ 24GeV

0.05 < x < 0.7

The x-distribution, given by the generated structurefunctions, allows, togetherwith the photon Pt spect-rum, to estimate the range of kinematical x-Q^2 region, available for themeasurement at PANDA .

0 ≤ 0 ≤ E E γγ ≤ 7 GeV, < ≤ 7 GeV, <E E γγ > = 1.2 GeV > = 1.2 GeV 0 ≤ 0 ≤ PT PT γγ ≤ 1 GeV, < ≤ 1 GeV, <PT PT γγ > = 0.29 GeV > = 0.29 GeV 0 ≤ 0 ≤ ΘΘ γγ ≤ 180 ° , < ≤ 180 ° , <ΘΘ γγ > = 23.5 °> = 23.5 ° some some ΘΘ γγ > 90°> 90°

Up to 7 neutral photons in signal event (i.e. ≤≤ 14 photons/event).

Distribution of N_ πº / event well explains the distribut. of N_ γ / event

(due to πº 2 γ).

The most probable parents of πº areη, ρ+, strings and Δ0


Neutral pions distributions

0 ≤ 0 ≤ E E γγ ≤ 4 GeV, < ≤ 4 GeV, <E E γγ > = 0.6 GeV > = 0.6 GeV 0 ≤ 0 ≤ PT PT γγ ≤ 1 GeV, < ≤ 1 GeV, <PT PT γγ > = 0.16 GeV > = 0.16 GeV 0 ≤ 0 ≤ ΘΘ γγ ≤ 180 ° , < ≤ 180 ° , <ΘΘ γγ > = 29.6 °> = 29.6 ° some some ΘΘ γγ > 90°> 90°

Vx, Vy, Vz distributions show mostly zero value

Up to 10 background photons in a signal event (in some few events up to 14)


Fake (decay) photons distributions

The plots show the distributions over summed energy of stable particles in

the cones of radius

R = √ η2+φ2 respect to the upper plot direct photon

bottom plot fake photon

Isolation criteria E (of stable particles) =0 in the R=0.3

allows to separate 100% of fake

photons with loss of 4.2% of signal events


Gamma isolation criteria

Conclusion.

The simulation with PYTHIA ( used as the first step approximation ) has shown that with E beam = 14 GeV :E beam = 14 GeV :

1. PANDA can perform the measurement of proton structure

functions in the time-like !!! (q2 >0) x-Q2 kinematical region:

0.05 < x < 0.7;

2. This measurement will have the advantage over DIS because it shall be more sensitive to gluon distribution.

3. To separate the contribution of fake photons, which come from the decays of pions and other neutral fake mesons in the signal and mini-bias events, a lot of work with full GEANT simulation is needed in nearest years to find a way to identify/separate by ECAL πo, η, ω and other mesons, produced in QCD and mini- bias background events (like it was done at ISR, SppS,WA98, PHENIX, and other collaborations …). .

22 )( TpQ 24GeV

For this region PYTHIA predicts about 5 x109 !!! signal events per year.

Double parton (DP) interaction measurements at PANDA

36

1.The inelastic scattering of nucleons need not to occur only through a single parton-parton interaction (A) and the contribution from double parton collisions (A and B) can be significant.

2.The probability of the second parton-parton collision (B) will be larger or smaller depending on the parton spatial distribution.

If partons were concentrated within small region inside the proton then the second collision (B) would be more likely to occur for a given A collision.

By contrast, in a case of uniform parton density throughout the proton, the presence of A would not increase the probability of B.

37

History of Measurements of Double Parton

scatterings

A new measurement was done by our group at D0:

Phys. Rev. D 81:052012, 2010; arXiv:0912 5104 [hep-ex]

39

Outback

42

DP: Signal Variables

Calculated for the pair that gives the minimum value of S:

43

Conclusion

The proposed cuts:1. Events with only 2 leptons of the opposite sign

and El > 0.2 GeV, PTl > 0.2 GeV 2. The vertex of origin lies within the

distance from the interaction point < 15 mm 3. Minv (l +,l -) > 0.9 GeV

4. Isolation criteria E (R isolation = 0.2) = 0.5 GeV

Allow to suppress QCD & Mini-bias bkgd to:completely for muons; S/B = 9 for electrons


45

Double parton (DP) events at PANDA

Signal process Backgroud process

PT- inbalanced lepton pairs PT-balanced muon pairs

Preliminary PYTHIA estimations (model dependent) has shown a moderate reduction

of the total cross section as comparing to lepton-pair production (by ~10-2 ).

To perform the analogous measurement we need only to substitute jets and photon by leptons: e/muons;

The electrons and muons with E_l > 0.8 can be used for such kind of measurement;

The measurement with leptons (which are much more better measured: Energy, Vertex position, no need of jet energy calibration) can provide a much higher preci-sion.

Double parton (DP) events at PANDA

Conclusions:

1. PANDA energy range is quite promising for studying of a large

number of very interesting Deep Inelastic QCD processes.

2. The study of single lepton pair production can provide the informa-

tion about valence quark distribution in a new kinematical region

of the intermediate “time-like” momentum transferred q2 > 0.

3. Direct photon production shall allow to measure the gluon

distribution component of proton structure function at q2 > 0 .

4. Double parton interactions can bring the information,

may be the most clean one (jets are substituted by e/muons)

about the spatial distribution of partons inside the proton.

Outback

48

49

Θl ”slow” / Θl ”fast” correlation

Muon system geometry cuts :

Θl_”slow” & Θl_”fast” ≤ 20º 14% events saved



Θl_”slow” & Θl_”fast” ≤ 90º 94% events saved ( 6% of events loss )


50

El ”slow” / El ”fast” correlation

Applying the Energy cuts:

El_”fast” & El_”slow” > 0.5 GeV

loss of ~25% of signal events

El_”fast” & E l_”slow” > 1 GeV

loss of ~45% of signal events


51

1. for El > 0.5 GeV “slow “ “fast”

Nevent ≈ 70% Nevent ≈ 98%

2. for El > 4 GeV

Nevent ≈ 2% Nevent ≈ 40%

3. for Θl < 60°

Nevent ≈ 78% Nevent ≈ 99.96%

Θl/El correlation for signal l±

Θl ”slow” / E l ”slow” Θl ”fast” / E l ”fast”


52

Some first estimations were

Published (2006) as PANDA

Note PHY-003:

and hep-ph/0506139

A.N.Skachkova, N.B.Skachkov

“Monte-Carlo simulation of lepton pair production in “ppbar l+l - + X” events at Ebeam = 14 GeV”


53

Cuts efficiency for Minimum-Bias background events

e+e- production

N of cuts S/B ratioS/B ratio EfficiencyEfficiency

1 (exactly 2 leptons (exactly 2 leptons

withwith EEl l > 0.2 GeV, PT> 0.2 GeV, PTll > 0.2 GeV > 0.2 GeV ) )

5.3 * 10 - 4- 4 1.78 * 10 - 4

2 (2 leptons are of the opposite sign)(2 leptons are of the opposite sign) 5.4 * 10 - 4- 4 0.98

3 ((The vertex of origin lies withinThe vertex of origin lies within

thethe R < 15 mm))

5.5 * 10 - 4- 4 0.98

4 (M(Minvinv(l(l11,l,l22) > 0.9)) > 0.9) 0.09 0.006


54

Efficiency of M inv (l+,l-) cut

Minv(l+,l-)> S / B Efficiency The rest of

signal events

0.9 0.09 0.0057 84 %

1.0 0.15 0.0034 81 %

1.03 0.38 0.0013 73 %

1.05 0.40 0.0012 69 %

1.1 0.83 0.0005 58 %

1.2 2.00 0.0002 42 %

e+e- case


55

Cuts influence on signal events Applying these cuts we have loss of the signal

events:

N of cuts e+e- production μ+μ- production

1 14.330 % 16.525 %

1 & 2 14.340 % 16.805 %

1 & 2 & 3 14.341 % 17.108 %

The rate of events with left fake leptons is negligible !

0.008 % 0.001 %


a.n.skachkova, n.b.skachkov (jinr, dubna)

Documents