1. Introduction
Colliding two heavy nuclei at ultrarelativistic energies allows to create in the laboratory a bulk system with huge density, pressure and temperature and to study its properties
It is estimated that in Pb-Pb collisions at CERN-SPS we reach:
e ~ 3 GeV/fm3 ~ 20  nuclear density corresponding to:
T ~ 200 MeV ~ K
(for comparison, in the center of the Sun T ~ K)

2. At such densities, hadrons are so closely packed that they interpenetrate; novel physics phenomena are likely to appear
QCD predicts that under such conditions a phase transition from a system composed of colourless hadrons to a Quark-Gluon Plasma (QGP) should occur

3. A rich ultrarelativistic heavy-ion physics programme both at BNL-AGS (sNN ~ 5 GeV) and at CERN-SPS (sNN ~ 20 GeV)
since 1986: lighter ions such as 16O, 28Si, 32S
since 1992 (AGS) and 1994 (SPS): truly heavy ions (197Au, 207Pb)
in total about 1000 particle and nuclear physicists (~ 50-50)
The results obtained so far have led CERN to officially announce “evidence for a new state of matter”
A long-range programme of heavy-ion physics at higher energies is under way (BNL-RHIC, sNN ~ 200 GeV) and in preparation (CERN-LHC, sNN ~ 5.5 TeV)

4. QCD Theory of strong interactions: QCD (Quantum Chromo-Dynamics)
quarks carry a strong interaction charge (colour)
colour comes in three types, say red, green and blue
(antiquarks carry anticolour)
quarks interact among themselves via the exchange of the colour field quanta (gluons)
gluons themselves carry a colour charge, unlike the photon in QED (Quantum Electro-Dynamics), which carries no electric charge (the theory is “non-abelian”)
All known hadron states are colour singlets (“white”)
baryons: qqq states; mesons: qq states
in particular, no free quark has ever been detected: quarks seem to be permanently confined within the hadrons

5. Asymptotic freedom QCD is “asymptotically free”:
the short distance potential is of the type:
the coupling constant is “running” with (depends on) r in such a way that
Perturbation theory can be applied at short distance/high momentum transfer
[see e.g. Perkins, p. 291]

6. Confinement
At scales of the order of the hadron size (~ 1 fm) perturbative methods lose validity
Calculations rely on approximate methods (such as lattice theory or effective theories)
There are compelling arguments (but no rigorous proof) that the non-abelian nature of QCD is responsible for the confinement of colour
[see e.g. Gottfried-Weisskopf, p. 95]

7. Confining potential
In QCD, the field lines are compressed into a “flux tube” (or “string”) of constant cross-section, leading to a long-distance potential which grows linearly with r:
QED
QCD
with k ~ 1 GeV/fm

8. String breaking
If one tries to pull the string apart, when the energy stored in the string (k r) reaches the point where it is energetically favourable to create a qq pair, the string breaks…
...and one ends up with two colour-neutral strings (and eventually hadrons)
[illustration from Fritzsch]

9. Deconfinement
What if we compress/heat the system so much that the individual hadrons start to interpenetrate?
Lattice QCD predicts that if a system of hadrons is brought to sufficiently large density and/or temperature a deconfinement phase transition should occur
In the new phase, called Quark-Gluon Plasma (QGP), quarks and gluons are no longer confined within individual hadrons, but are free to move around over a larger volume

10. Restoration of bare masses
Confined quarks acquire an additional mass (~ 350 MeV) dynamically, through the confining effect of strong interactions
Deconfinement is expected to be accompanied by a restoration of the masses to the “bare” values they have in the Lagrangean
As quarks become deconfined, the masses go back to the bare values; e.g.:
m(u,d): ~ 350 MeV  a few MeV
m(s): ~ 500 MeV  ~ 150 MeV
(This effect is usually referred to as “Partial Restoration of Chiral Symmetry”. Chiral Symmetry: fermions and antifermions have opposite helicity. The symmetry is exact only for massless particles, therefore its restoration here is only partial)

12. Screening by the QGP

13. QCD phase diagram ec ~ 1 GeV/fm3 ~ 10 ms after Big Bang LHC
Early Universe
Baryon density
Temperature
RHIC
Quark-Gluon
Plasma
Hadron gas
Nuclear
matter
Tc ~ 170 MeV
SPS
AGS
Neutron Star
~ nuclear r
ec ~ 1 GeV/fm3

14. Heat is also a window back in time

15. Deconfinement and Astrophysics
It has been speculated that the cores of some dense neutron stars might actually be in a deconfined phase
For fast rotating neutron stars (pulsars) the phase transition might actually be observable from Earth:
pulsars become more and more compressed as they emit energy and spin down
if in the process the central density is raised above the critical density, the star may start to collapse into a quark star
the resulting shrinkage would then cause it to increase its rotational (and therefore its pulsating) frequency (“spinning skater” effect)
the observation of a “spinning up” pulsar may therefore be interpreted as the result of quark deconfinement
[see e.g.: F.Weber, J. Phys. G 27 (2001) 465]

16. Nucleus-nucleus collisions
How do we test this theory in the lab?
How can we compress/heat matter to such astronomical energy densities?
By colliding two heavy nuclei at ultrarelativistic energies we hope to be able to recreate, for a short time span (about 10-23s, or a few fm/c) the appropriate conditions

17. Even if a QGP is formed, as the system expands and cools down it will hadronize again, as it did at the beginning of the life of the Universe: we end up with confined matter again
This is the experimental challenge: to observe in the final state the signatures of the phase transition: physical effects which:
are consequences of the phase transition
cannot be explained otherwise

18. Hot and Dense Laboratory Matter: Heavy Ion Collisions
But: Will these fast violent collisions
teach us anything?
2x1012 K
10-23 seconds, liters

19. Fasi dell’urto
Hard scattering;
Formazione di QGP a T ~ 170 MeV;
Chemical freeze out ( I tipo di particelle sono stabiliti : rapporti p/p/K, )
avviene intorno a T = ( 160 ± 10 ) MeV;
Kinetic freeze out ( il momento delle singole particelle e’ fissato)
Avviene intorno a T ~ 120 – 140 MeV

20. Collision centrality
How far the centers of the two colliding nuclei passed each other
Usually expressed in terms of:
b (impact parameter)
cross section s(b)
number of participants Npart(b)
[sometimes one speaks of “number of wounded nucleons”: NW(b) ]
rapidity
y=0 b
participants
spectators

21. Experimentally, the centrality is evaluated by measuring one or more of these variables:
Nch: number of charged particles produced in a given rapidity interval (near mid-rapidity)
increases (~ linearly) with Npart
ET: transverse energy = S Ei sin qi
EZDC: energy collected in a “zero degree” calorimeter
increases (~ linearly) with Nspectators

22. Bjorken’s formula dV = S dz = S c t dy dE = e dV
To have an estimate of the energy density reached in the initial stages of the collisions, we can project back in time the energy carried by the collision products (“Bjorken’s estimate”)
Consider a thin cylindrical slab of transverse dimension S of expanding matter contained within a thickness dz at time t.
v=0
-dv dv dz dE
dv = c db = (c/2) dy (non rel.: y = b)
dz = 2t dv = c t dy
dV = S dz = S c t dy
dE = e dV
Bjorken’s formula
In the center-of-mass frame v=0 at the center of the slab

23. Initial energy density
Estimate for central (head-on) Pb-Pb collisions at the SPS
Initial time t0 : usually taken to be ~ 1 fm/c
i.e.: equal to the “formation time”: the time it takes for the energy initially stored in the field to materialize into hadrons
Bjorken’s formula
Transverse dimension S :
e ~ (400/160) GeV/fm3 ~ 2.5 GeV/fm3
Enough for deconfinement!
Published estimate from NA49:
e = 3.2  0.3 GeV/fm3
[Phys. Rev. Lett. 75 (1995), 3814]

24. Transverse mass distributions
Usually fitted to thermal distributions:
T = “inverse slope” or “apparent temperature” or “mT slope”
R.Stock

25. Partial summary We saw that in Pb-Pb collisions at the SPS:
we initially create a “fireball” with e ~ 3 GeV/fm3, in principle enough for deconfinement
when the system finally “freezes out”:
it has a temperature of about 120 MeV
it is “exploding” at about 1/2 speed of light, indicating that a large pressure was generated inside the fireball
In order to explore what happens in between, we shall now turn to more specific observables:
charmonium production
strangeness production

26. Strangeness enhancement
restoration of c symmetry -> increased production of s
mass of strange quark in QGP expected to go back to current value
mS ~ 150 MeV ~ Tc
copious production of ss pairs, mostly by gg fusion
[Rafelski: Phys. Rep. 88 (1982) 331]
[Rafelski-Müller: P. R. Lett. 48 (1982) 1066]
deconfinement  stronger effect
for multi-strange
can be built recombining uncorrelated
s quarks produced in independent
microscopic reactions
strangeness enhancement increasing
with strangeness content
[Koch, Müller & Rafelski: Phys. Rep. 142 (1986) 167] s u d K+ W+ p+ p- p L X-

27. Charmonium as a Probe of QGP
Matsui and Satz predicted J/y production suppression in Quark Gluon Plasma because of color screening

28. Charmonium suppression
QGP signature proposed by Matsui and Satz, 1986
In the plasma phase the interaction potential is expected to be screened beyond the Debye length lD (analogous to e.m. Debye screening):
Charmonium (cc) and bottonium (bb) states with r > lD will not bind; their production will be suppressed
lD , and therefore which onium states will be suppressed, depends on the temperature

29. Debye screening
In an electromagnetic plasma, the potential of a charge is screened by the field of the electrons that surround it [see e.g.: Jackson p. 494] :
with
n0 = density of electrons
in the plasma
In a QGP, the colour field is likewise going to be screened.
In order to have a back-of-envelope estimate the screening length, one can take the above formula, and substitute:
e2 (Gauss system)  aQCD ~ 1
getting:
n = MeV3
n0  n = 3.6 T3
(Stefan-Boltzmann law for QGP)
kT ~ 200 MeV
and, using: 1 MeV-1 = fm: lD  0.15 fm

30. NO QGP
QGP c c c c c
V= kr - a / R
V = -a e – r/ld R

31. Come si identifica una risonanza (esempio J/Y ) ?
Identificazione dei leptoni ( canale a 3p molto difficile);
misura del loro momento;
Calcolo della massa invariante e selezione.
minv = sqrt( (E1+E2)2 – (p1+p2)2)

32. NA38

33. Dimuon Spectrum
The measured dimuon spectrum is fitted to a source cocktail in order to extract the J/y, y’ and Drell-Yan contributions
NA50 dimuon spectrum (Pb-Pb, 158 A GeV/c)
Quarkonium production is usually normalised to Drell-Yan production (which is not influenced by strong interactions) m+ m-

34. Why do we keep using Drell-Yan ?
Drell-Yan (muon pairs) is a well known computable process, proportional to the # of elementary nucleon-nucleon collisions, with the following priceless advantages:
identical experimental biases
identical inefficiencies
identical selection criteria
identical cuts
as J/
Therefore the corrections cancel out in the ratio
 (J/)
 (DY)
which is insensitive to normalization factors/uncertainties
PUNTI NEGATIVI : 1) statistica DY << statisticaJ/y 2) normalizzazione isospin

35. Nuclear absorption
Branching to muons
There is a “normal” suppression of the production of J/y, observed already in pA and lighter ion collisions and attributed to nuclear absorpion
The Pb-Pb point falls below the nuclear absorption curve (“anomalous” suppression)

36. From a fit of experimental p-A data (NA38,NA50):
sabs= mb (hep-ex/ )
In S-U collisions the same suppression is observed
The normal suppression is interpreted as the absorption, in the nuclear environment, of the
c-cbar pair before the J/y (or y’ or c) formation : preresonance absorption.

38. The CERN Pb ion programme
Started in 1994
Pb nuclei accelerated to 158 A GeV/c (40 A GeV/c in 1999) collide on fixed targets (typically 4-6 weeks/year)
7 experiments:
NA44 (single arm spectrometer: particle spectra, interferometry, particle correlations)
NA45 (e+e- spectrometer: low mass lepton pairs)
NA49 (large acceptance TPC: particle spectra, strangeness production, interferometry, …)
NA50 (dimuon spectrometer: high mass lepton pairs, J/y production)
NA52(focussing spectrometer: strangelet search, particle production)
WA97/NA57 (silicon pixel telescope spectrometer: production of strange and multiply strange particles)
WA98 (photon and hadron spectrometer: photon and hadron production)

39. A Pb-Pb collision at the SPS
“Busy” events! (thousands of produced particles)
High granularity detectors are employed (TPC, Si Pixels,...)

40. NA38 first results O+U at 200 GeV/c: Factor 2 suppression…
but… including:
normal nuclear absorption !!!
IMR charm-like excess !!! (fit starts from 1.7 GeV/c2 !!)

41. Experiment NA50 Aim: study the production of J/y in Pb-Pb collisions
Experimental technique:
absorb all charged particles produced in the collision except muons
detect J/y by reconstructing the decays J/y  m+m- (B.R.  5.9 %)

42. Anomalous J/y suppression
J/y normalized to Drell-Yan as a function of the transverse energy (i.e. centrality)
The data points deviate from the solid curve, which indicates the prediction for nuclear absorption
The deviation increases with increasing collision centrality

43. Attempts at describing the NA50 data within purely hadronic models without deconfinement
dissociation of the J/y in final state hadronic interactions with comovers
try harder...

44. J/y L
Projectile
Target

45. Alcune delle motivazioni dell’esperimento:
NA60: stesso rivelatore di NA50 con aggiunta di rivelatori al Silicio (tracking).
Alcune delle motivazioni dell’esperimento:
If the J/psi suppression pattern in Pb-Pb collisions indicates that central Pb-Pb collisions produce a state
of matter where colour is no longer confined, we should move on to the detailed understanding of how
deconfinement sets in, and what physics variable governs the threshold behaviour of charmonia (cc)
suppression: (local) energy density, density of wounded nucleons, density of percolation clusters, etc.
This requires collecting data with smaller nuclear systems like In-In.
The J/psi data collected in central Pb-Pb collisions indicate that we are already beyond the point where the phase transition takes place, but do not provide any information on the value of the critical temperature. Finite temperature lattice QCD tells us that the strongly bound J/psi ccbar
state should be screened when the medium reaches temperatures % higher than T_c, while the large and more loosely bound psi' state should melt near T_c. The NA38 experiment has shown that the psi' is significantly suppressed when going from p-U to peripheral S-U collisions. We need to
see if this suppression
follows a smooth pattern or a sudden transition, within a single collision system rather than comparing p-U to S-U data.
If the Y' suppression is due to Debye screening, its suppression pattern could provide a clear measurement of T_c. However, the hadronic "comovers" produced in S-U collisions may "absorb" the psi' mesons, since its binding energy is only around 40 MeV. What mechanism is responsible for the Y' suppression? The presently existing results are not clear in what concerns the onset and pattern
of the psi' suppression. A new measurement is needed, with improved mass resolution to have a cleaner separation of the psi‘ peak with respect to the J/Y shoulder, and which scans the energy density region from the p-U to the S-U data. In-In collisions are also well placed for this study.

46. Matching in coordinate and in momentum space
NA60’s detector concept
Idea: place a high granularity and radiation-hard silicon tracking telescope in the vertex region to measure the muons before they suffer multiple scattering and energy loss in the absorber
beam
~ 1m
Muon Spectrometer
MWPC’s
Trigger Hodoscopes
Toroidal Magnet
Iron
wall
Hadron absorber
ZDC
Target area m
MUON FILTER
BEAM TRACKER
TARGET
BOX
VERTEX
TELESCOPE
Dipole field 2.5 T
BEAM IC
not to scale
Origin of muons can be accurately determined
Improved dimuon mass resolution

Matching in coordinate and in momentum space
ZDC 
allows studies vs.
collision centrality

47. Specific questions that remain open
Is the anomalous suppression also present in lighter nuclear systems?
Study collisions between other systems, such as Indium-Indium
S-U
In-In
Pb-Pb
Npart
L (fm)
pure Glauber calculation
Which is the variable driving the suppression?
Study the J/ suppression pattern as a function of different centrality variables, including data from different collision systems
What is the normal nuclear absorption cross-section at the energy of the heavy ion data?
Study J/ production in p-A collisions at 158 GeV
What is the impact of the cc feed-down on the observed J/y suppression pattern?
Study the nuclear dependence of cc production in p-A collisions
New and accurate measurements are needed to answer these questions

48. Comparison with previous results
An “anomalous suppression” is present already in Indium-Indium
The normal absorption curve is based on the NA50 results. Its uncertainty (~ 8%) at 158 GeV is dominated by the (model dependent) extrapolation from the 400 and 450 GeV data
 need p-A measurements at 158 GeV: data collected in 2004 (analysis under way)

49. Spiegazioni alternativa al QGP
1) Comovers
Comovers model. La J/Y puo’ interagire con gli adroni “comovers”. La sezione
d’urto e’ molto difficile da stimare.
Na50 J/y suppression can be reproduced by DPM
with absorption by comovers. The number of
comovers in Capella model is proportional to
number of participants and also to number of
collisions.
A. Capella, D. Sousa, nucl-th/

50. Suppression by produced hadrons (“comovers”)
The model takes into account nuclear absorption and comovers interaction with sco = 0.65 mb (Capella-Ferreiro)
158 GeV
J/y / NColl
nuclear absorption
comover + nuclear absorption
(E. Ferreiro, private communication)
NA60 In-In 158 GeV preliminary
The smeared form (dashed line) is obtained taking into account the resolution on NPart due to our experimental resolution
158 GeV

51. contiene partoni abbastanza “hard”, puo’ dissociare la J/Y
2) Percolation
[First works: Baym , Physica (Amsterdam) 96A, 131 (1979)
Celik et al., Phys. Lett. 97B (1980) 128]
Forma di deconfinamento geometrica, di pre-equilibrio. Pre-requisito al deconfinamento
Vero e proprio, applicabile ai sistemi finiti. Se il condensato di partoni
contiene partoni abbastanza “hard”, puo’ dissociare la J/Y
N dischi di raggio r<< R
Densita’ n= N/pR2
Superficie di area pR2
Aumentanto la densita’ si trovano cluster
di area sempre maggiore
Quando N,R  ∞ e n finito , la cluster size
diverge a n= 1.13/ pR2. Per N,R finiti si ha
Percolazione quando il cluster piu’ largo
Riempie tutta la superficie. R 1
Percolation
probability
0.5
r/R=1/100
n=n(r/R)2 1
1.5

52. A causa dell’overlap, alla soglia di percolation, solo 2/3 dell’area e’ riempita.
Local Percolation: Hard probe, come gli stati c-cbar risentono del mezzo localmente.
Si ottiene a 1.72/pr2

53. Osservazione sperimentale (Gazdzicki and Gorenstein)
Regeneration models[2,3,4] predict an enhanced production of hidden charm states for sufficiently high charm densities
This would imply thermalization of charm quarks, and by extension, the light quarks that comprise the QGP
3) Regeneration
Osservazione sperimentale (Gazdzicki and Gorenstein)
Il rapporto di J/Y / p- = cost
“a dominant fraction of the Jc mesons produced in hadronic and nuclear collisions at CERN SPS energies is created at hadronization according to the available hadronic phase space”
L. Grandchamp and R. Rapp, Phys. Lett. B (2001)
regeneration
suppression

55. “Two component” model: suppression in hadronic and QGP phase + statistical
production at hadronization

56. J/y at SPS J/y in NA60 poorly reproduced by models which fit NA50 data
Satz, Digal, Fortunato
Rapp, Grandchamp, Brown
Capella, Ferreiro

63. color screening, comover
- model discribe the NA50 results well
- But over predict the suppression at RHIC

64. Understanding J/ suppression in Au+Au Collisions
Regeneration
statistical hadronization.
In-medium formation.
Agree with the data except in the most central collisions
Kostyuk (hep-ph/ )
Andronic (nucl-th/ )
Grandchamp, Rapp, Brown
hep-ph/
Thews, hep-ph/

65. Johanna Stachel

66. Model prediction