Raluca Muresan

1
Bose-Einstein Correlation Studies at HERA-B

• Raluca Muresan
• NBI
• Copenhagen

2
Outline

• HERA-B experiment
• BEC introduction
• Why BEC at HERA-B
• Main ingredients for BEC analysis
• Preliminary results pp in pA with AC, Ti, W
• On-going studies
• Summary

3
HERA-B experiment
Located at the proton-electron collider HERA at
DESY. Fixed target experiment 920 GeV proton
collisions on various targets. Large acceptance
at mid-rapidity x?(15, 220) mrad, y ?(15-160) mrad
4
Target wires

• C, Ti, W, Pd, Al wires inserted in to the halo
  of HERA proton beam
• Mounted in 2 stations separated by 4 cm
• Each wire can be inserted independently

5
VDS

• 64 double-sided silicon microstrip detectors
  aranged in 8
• superlayers, divided in 4 quadrants
• quadrants located in Roman pots allowing the
  detectors
• to be retracted during injection (except SI08)

6
OTR
Superlayers (MCs, PCs, TCs) are made of few
modules containing several layers with angles of
0 and 5 degrees. The layers are made of honeycomb
drift cells.
7
RICH
Large vessel filled with C4F10 radiator gas. The
photons are reflected, by a serie of mirrors, to
the photon detectors placed outside the detector
acceptance.
8
BEC

• Symmetric wave functions of bosons, no exclusion
  principle
• Enhanced probability for the identical bosons to
  be emitted with small relative momenta
• Quantum statistical correlations between pairs of
  identical particles.
• Presuming that only particles emitted from
  the same or very close sources exhibit this
  behaviour ? from studies of BEC one can obtain
  information about the size, shape and spacetime
  development of the particle emitting source.

9
BEC

• For two identical bosons the Bose-Einstein
  correlation is defined as

probability density of two particles to be
produced with 4-momentum p1 and p2
probability densities for a single particle to
be produced with 4-momentum p1 or p2 ,
difficult to build in practice ? reference
sample
10
BEC function parametrisation
invariant four-momentum difference,
related to the fraction of identical bosons
which do interfere,
interpreted as the geometrical radius of the
presumably spherical boson emitting source (just
an approximation),
overall normalization,
linear background.
11
Why BEC at HERA-B ?

• Opportunity to measure BEC parameters at
  sqrt(s)41,6 GeV, intermediary value between SPS
  and RHIC
• Possibility of studying the BEC parameters
  dependence on the target material (C, Ti, W).
• BEC can be studied for both pion and kaon pairs
  (RICH selection) .
• Large minimum bias sample(single wire runs)
• 103 million pC events,
• 74 million pW events,
• 28 million pTi events.
• large enough for differential studies
  (directional dependence, multiplicity,
  transverse mass dependence ).

This talk is presenting only a feasibility
study, made on a small sample (few million
events).
12
Reference sample

• Mixed sample tracks from different events, some
  other kind of correlations are also disappearing
  (long range correlations , energy-momentum ...)

To correct for the effects introduced by mixing
double ratio.
13
Event, track, pair selection
Pion and kaon tracks - RICH likelihood.
Remove from the analysed sample the pairs of
tracks too close in space
abs(tx1-tx2)lt0.0008, txpx/pz
abs(ty1-ty2)lt0.0008, typy/pz and
momentum abs(p1-p2)lt0.5 GeV
pabs( ) to be resolved.
The C2(Q) distributions for pions corrected for
Coulomb interaction (Gamow), only small
variation in the parameter values occured.
14
MC studies -pC

• BEC JETSET
• MSTJ(51)2 the shape of correlation function
  Gaussian
• MSTJ(52)9 BEC for p , K , h.
• PARJ(92)1 meaning that particles that can be
  subject to BEC
• are subject to BEC.
• PARJ(93) 2 fm R.
• PARJ(91) 0.020 GeV minimum particle width
  above which the
• particle decays are assumed to take place before
  the stage when BE
• effects are introduced. Particles with broader
  width than 0.020 are
• assumed to have time to decay before BE effects
  are to be considered

Not all the pions are correlated and PARJ(92) and
the measured lambda (obtained by fitting) are
different
15
MC study -Influence of the non-belonging pairs-
pC
C2 (Q)
P1N, P2l, P3R, P4d
Non-belonging pairs pairs in which, at least
one particle is not a pion (based on MC
truth). Is a junk, a kaon, a proton etc...
Q(GeV)
The junks and the possible pion contamination do
not affect our result.
16
MC studies pC
To test the procedure
C2 (Q)
Studies of systematic errors are necessary but so
far it seems that the procedure gives the correct
results. The radius used as MC example is bigger
than the radius we measure.
Q(GeV)
P1N, P2l, P3R, P4d
17
C2(Q) - Data
P1N, P2l, P3R, P4d
C2 (Q)
C2 (Q)
W
C
Q(GeV)
Q(GeV)
l 0.276 0.016 R 1.0570.056 (fm)
l 0.293 0.021 R 1.3070.076 (fm)
18
C2(Q) - data
P1N, P2l, P3R, P4d
C2 (Q)
l 0.253 0.020 R 1.2980.010 (fm)
Ti
Q(GeV)
19
Fit results- Dependence on the fit range
About 1.000.000 events for each sample Weak
dependence, if any, on the target material ( same
as Na44- S on S, Ag, Pb). Radius value smaller
than the Na44 one, pPb R2.890.30, but pt larger
at HERA-B, source size decreases.
20
LCMS
The spherical source shape just an
approximation, difficult to interpret. LCMS -
the spatial dimension of the source
couples to all components
- the temporal component couples
only to Qt,out.
beam
21
Multidimensional correlation function
2-dim
3-dim
22
Even more preliminary results 2-dim pp
About 5.5 million pC events
23
On-going studies - Kaons
First results (about 18 million pC events)
indicate, as it was expected that the BEC radius
for kaons is smaller than for pions (about one
half).
24
Plans - Multiplicity dependence
Code ready, results on the way
25
Plans - mt dependence of the 3-dim BEC parameters
Code ready, results on the way
26
Summary

• BEC correlations were observed both for pp and KK
  pairs.
• Preliminary results, pp, were presented for both
  1-dim BEC
• parameters (pA - AC, Ti, W) and two-dim BEC
  (pC).
• Studies on-going pp - 3-dim, multiplicity and
  mt dependence KK.

It still a lot to be done in terms of running
over all the data set, careful study of
systematic errors, undestanding the results and
comparing them with other experimental results.

• HERA-B is the place to do interesting studies of
  BEC correlations
• More results are expected very soon

27

• Back-up sildes

28
HERA-B detector
29
Gamow correction

• To measure the undisturbed BEC we have
  to substract all other correlation effects from
  our distributions. For the correlated
  distributions of charged particles one has to
  correct for Coulomb interaction by weighting the
  Q distribution with the inverse Gamow factor.

30
Gamow correction

• Before
• ?0.247 0.012
• R0.958 0.043 fm
• After
• ?0.288 0.013
• R1.003 0.042 fm

P1N, P2l, P3R, P4d
31
Cut abs(p1-p2)gt0.5 GeV.
abs(tx1-tx2)lt0.0008 abs(ty1-ty2)lt0.0008
32
Pion selection
medium
p(GeV)
p(GeV)
soft
The identification probability is not constant
over the momenta range in the medium selection.
In the soft selection is constant for 5ltplt40GeV
33
Only the pairs with 0ltQlt1.2 GeV We must not
worry about momenta gt 40 GeV and lt2.6 GeV
34
Purity of the sample MC
0ltQlt1.2 and particles that pass the soft pion
selection Here the truthid of the track is the
LUND code of the truth. Tracks for which no
MCtruth was found were considered junk and their
truthid was set to 0 We need to look around
2000 and in the region where The code is smaller
than 500
35
Purity of the sample
Junk tracks tracks without MCGEN correspondent
Two pions
1 junk track and 1 pion
1 pion and 1 junk track
2 junk tracks
36
Purity of the sample
Two protons
37
Data