1.  production in p-A and In-In collisions Motivation Apparatus Collected data Results for     Ongoing work for    KK Alessandro De Falco – University and INFN Cagliari, Italy For the NA60 collaboration Quark Matter 2005 – August 4-9 – Budapest

2. Motivation The study of  meson production in heavy ion collisions carries information about strangeness enhancement Previous measurements of   production at the SPS: –NA49 studied the   KK channel Good mass resolution Broad p T coverage, dominated by low p T –NA50 studied the    channel Muons insensitive to medium Acceptance limited to high p T NA49 and NA50 observed different T slope values, as a function of N part   puzzle NA60 is well placed to help solving the  puzzle: –it can cleanly measure the  channel down to zero p T –and has access to   KK

3. ~ 1m Muon Spectrometer MWPC’s Trigger Hodoscopes Toroidal Magnet Iron wall Hadron absorber Target area ZDC beam   ZDC dimuon studies vs. collision centrality  Matching in coordinate and momentum space Improved dimuon mass resolution  MUON FILTER BEAM TRACKER TARGET BOX VERTEX TELESCOPE Dipole field 2.5 T BEAM IC not to scale Layout of the NA60 apparatus

4. Detector performance Vertexing: clear separation of the individual targets –Z vertex resolution: ~ 600–900  m in p-A; < 200  m in Indium-Indium target box windows 7 In targets z-vertex (cm) Indium beam 158 A GeV Transverse vertexing with 20 µm accuracy Beam tracker station p beam 400 GeV InPb 3 x Be Be p T coverage: the dipole field in the target region leads to much better acceptances at low mass and low p T than previous dimuon measurements A(%)

5. Data collected in 4 days during 2002 run statistics in 2004 is around 100 times higher Muons in the spectrometer are matched to the tracks in the vertex telescope The combinatorial background due to uncorrelated  and K decays is estimated through an event mixing technique (matching improves S / B by up to a factor ~ 4) ~ 15000 OS dimuons in final sample matching   p-A data: event selection

6. Mass spectra described as a superposition of light meson decays + charm + Drell-Yan A fit to the mass spectra allows us to extract the  ratio    BG+charm Mass spectra in p-Be, p-In, p-Pb  p-Be 0.062  0.04 p-In 0.083  0.07 p-Pb 0.081  0.06

7. In-In data selection Data sample: 570 000 events after background subtraction (around 50% of the total statistics) Overall Signal to Background ratio = 1 / 4 Fake matches (tracks in the muon spectrometer incorrectly matched to the tracks in the vertex telescope) estimated superimposing simulated dimuons on real data tracks   mass resolution: 23 MeV (not depending on centrality) signal + fakes fakes signal opposite-sign pairs combinatorial background signal + fakes opposite-sign pairs combinatorial background signal pairs (inc. fakes)

8. Extraction of  cross section ratio vs. N part Mass spectra are obtained in four centrality bins, using the charged track multiplicity Corresponding  N part  values are evaluated from the E ZDC spectrum Detector acceptances and efficiencies are accounted through MC simulations –GENESIS code describes the light meson decays –Empirical continuum source added Mass spectra are fit with the expected sources leaving as free parameters  and the continuum normalization Centrality bin dN ch /d  range 〈 dN ch /d  〉〈 N part 〉 Peripheral4–28 16~ 20 Semi-peripheral28–92 70 91 Semi-central92–160145161 Central> 160200197 peripheral all p T Signal Cocktail     D  D 

9. 0.5 < p T < 1.0 GeV/cp T > 1.0 GeV/c The normalizations of the hadron decay cocktail and of the continuum are independently fit in each p T bin   and   ratios: nearly p T independent In general, the peripheral bin is very well described in terms of expected sources (but there are “too many” low p T  mesons) p T < 0.5 GeV/c Signal Cocktail Peripheral collisions in 3 p T bins

10.  mass spectra in peripheral collisions Vacuum  contribution (  annihilation) important at low p T even in Indium-Indium peripheral collisions Vacuum + cocktail  Cocktail  Vacuum + cocktail  Cocktail  Vacuum + cocktail  Cocktail 

11. Mass spectrum in semi-central In-In collisions Complicated continuum under the  in more central collisions… However, the excellent mass resolution of NA60 allows us to extract a robust  yield, in particular at high p T

12.  cross section ratio vs. N part  obtained for p T > 1 GeV/c Increase by a factor ~ 2 from peripheral to central collisions 

13.  comparison between NA60 and NA50 A direct comparison is impossible, due to the contribution from pion annihilation, which NA50 cannot isolate. This contribution must be even higher in Pb-Pb collisions. NA50 points converted to the window p T >1.1 GeV/c assuming T = 228 MeV      used (lower limit for NA50  ) 

14.  NA60 ratio compared with NA49  Same trend as a function of N part  constant If we set the ratio  to 0.07–0.08, as suggested by statistical models, then the NA60  yield is a factor 1.5–2 higher than the NA49 value

15. Extraction of   p T spectra Continuum under  subtracted using 2 side windows defined symmetrically around the   peak The net p T spectrum is corrected for the acceptance using a 2-D acceptance correction (y vs. p T )

16. Characterized by the extracted inverse p T slope parameter, T  p T spectra vs. rapidity and multiplicity p T distribution vs. rapidity p T distribution vs. multiplicity Only statistical errors No significant variation with rapidityClear increase with multiplicity

17.  T parameter: NA60 versus NA50 and NA49 Average T(  ) for In-In collisions: 1) all p T : 253  2 MeV 2) p T < 1.50 GeV (NA49 range) : 260  5 MeV 3) m T > 1.65 GeV (NA50 range) : 244  5 MeV Pb-Pb In-In Si-Si C-C pp The In-In measurement of NA60 follows the NA49 systematics, contrary to the NA50 Pb-Pb point It seems that the difference between NA50 and NA49 is not due to the different decay channels probed NA49 Pb-Pb NA60 In-In NA50 Pb-Pb NA49 pp

18. Preliminary study of the   KK channel Consider all charged tracks associated to a vertex Assume they are kaons (no particle id.) Extract the invariant mass spectrum Subtract the (huge) combinatorial background with event mixing technique Carefully select candidate charged tracks, using pseudo-rapidity, p T, p and opening angle –to stay away from acceptance borders –to improve signal over background ratio p T (GeV/c) K from  All tracks MC

19. p T (GeV/c) Extraction of    KK signal from Monte Carlo Full simulation of In-In collisions (VENUS) Tracks in the same event are combined to build the mass spectrum Tracks from different events are mixed if events belong to the same centrality class and their vertices are closer than 2  x, 2  y, 2  z Gaussian fit gives the mass and peak width  Extracted values agree very well with the inputs of the simulation MC (VENUS) semi-peripheral opposite sign background

20.    KK mass spectrum in In-In collisions The real data follows the same analysis procedure Statistics sufficient to extract the  p T distribution in 4 centrality bins … but the background is not yet under sufficient control semi-peripheral bin

21. Summary and Outlook   ratio: –Rise with N part consistent with NA49 and NA50 –Absolute values between NA49 and NA50 Inverse slope T of the  p T distribution: –Our    values agree with NA49:  the difference between NA49 and NA50 is not due to the different decay channels probed   KK –Full MC simulation shows feasibility of the study –Final tuning still needed for background subtraction in real data  flow also under study