Transcript The relevance of proton-proton physics for the understanding

The relevance of pp results to the
understanding of soft physics in AA
collisions at RHIC and LHC
R. Bellwied (Wayne State University)
Is hadron production in medium different than production
in vacuum ?
1st Workshop of soft physics in ultrarelativistic heavy ion collisions,
Catania, Italy, Sept.27-29,2006
The Physics Questions
What do we know about fragmentation ?
Hadronization process studies
Baryon vs. meson production in pp
Flavor production in pp
Alternatives to string fragmentation ?
Is there collectivity in pp ?
Hadronization in QCD
(the factorization theorem)
Jet: A localized collection of hadrons
which come from a fragmenting parton
hadrons
c
a
Parton Distribution Functions
Hard-scattering cross-section
b
d
hadrons
Fragmentation Function
leading
particle
High pT (>~ 2.0 GeV/c) hadron production in pp collisions:
h
d pp
0
D
d

2
2
h/c

K
dx
dx
f
(
x
,
Q
)
f
(
x
,
Q
)
(
ab

cd
)

a
b a
a
b
b
2

dyd pT
dtˆ
zc
abcd
“Collinear factorization”
Do we understand hadron production
in elementary collisions ? (Ingredient I: PDF)
RHIC
Ingredient II: Fragmentation functions
KKP (universality), Bourrely & Soffer (hep-ph/0305070)
Non-valence quark
contribution to parton
fragmentation into
octet baryons at low
fractional momentum
in pp !!
Quark separation in
fragmentation models
is important. FFs are
not universal.
z
z
Depend on Q, Einc,
and flavor
The Lessons from RHIC
(I) unidentified particles
Is there anything interesting in the
non-identified charged particle spectra ?
Deviations from
a power-law
as a function
of multiplicity
Conclusions:
a.) hard component yield increases with nch
b.) not more energetic partons but high frequency
of events with single hard scattering (mean and
Deviation from a
two component fit: width stays the same)
c.) Levy function (soft component) =
Levy function
(soft) + Gaussian
thermal radiation from moving sources
(hard)
d.) low Q2 parton scattering dominated by mini-jets
nucl-ex/0606028
Transverse parton fragmentation = hard
Longitudinal string fragmentation = soft (LUND ?)
Is there anything interesting in the
non-identified two particle correlations ?
see T.Trainor’s talk on Friday
The Probe
Identified particle spectra:
- Meson / baryon spectra
- Strangeness / heavy flavor spectra
- Resonance spectra
- Correlations (HBT etc.)
The Lessons from RHIC
(II) identified particles
How to measure PID ?
• Initial PID: charged hadrons vs. neutral pions
• Detailed PID:
–
–
–
–
V0 topology
dE/dx
rdE/dx
TOF / RICH / TRD
Why measure these effects with
K and L instead of  and p ?
Particle identification benefits from fact that the
topological reconstruction method has no
intrinsic momentum cut-off compared to dE/dx.
…but the use of rdE/dx might change
that at least for inclusive measurements
0 in pp: well described by NLO (& LO)
p+p->0 + X
Thermallyshaped Soft
Production
“Well Calibrated”
Hard
Scattering
• Ingredients (via KKP or Kretzer)
– pQCD
– Parton distribution functions
– Fragmentation functions
• ..or simply PYTHIA…
hep-ex/0305013 S.S. Adler et al.
pp at RHIC: Strangeness formation in QCD
nucl-ex/0607033
Strangeness production not described by leading order calculation
(contrary to pion production).
It needs multiple parton scattering (e.g. EPOS) or NLO corrections to
describe strangeness production.
Part of it is a mass effect (plus a baryon-meson effect) but in addition
there is a strangeness ‘penalty’ factor (e.g. the proton fragmentation
function does not describe L production). s is not just another light quark
How strong are the NLO corrections ?
STAR
• K.Eskola et al.
(NPA 713 (2003)):
Large NLO
corrections not
unreasonable at
RHIC energies.
Should be negligible
at LHC.
New NLO calculation based on STAR data
(AKK, hep-ph/0502188, Nucl.Phys.B734 (2006))
K0s
apparent Einc dependence of separated
quark contributions.
Non-strange baryon spectra in p+p
Pions agree with LO (PYTHIA)
Protons require NLO with
AKK-FF parametrization
(quark separated FF contributions)
PLB 637 (2006) 161
The Lessons from RHIC
(III) baryon / meson physics
Non-strange particle ratios – p+p collisions
PLB 637 (2006) 161
Collision energy dependence of baryon
vs. meson production
630 GeV
Peak amplitude doubles in pp from 200 to 630 GeV
Bump is intrinsic in pp, enhancement is unique to AA
Baryon/Meson ratio @ 630 and 1800 GeV
(Boris Hippolyte, Hot Quarks 2006)
Extracting mixed ratio from UA1 spectra (1996) and from CDF spectra (2005)
Ratio vs pT seems very energy dependent (RHIC < SPS > FNAL ?)
Mt scaling in pp
Breakdown of mT scaling in pp ?
mT slopes from PYTHIA 6.3
Gluon dominance at RHIC
PYTHIA: Di-quark structures in baryon production cause mt-shift
Recombination: 2 vs 3 quark structure causes mt shift
Baryon production mechanism
through strange particle correlations
 0
e e  Z  qq  jets…
Test phenomenological fragmentation
models
OPAL ALEPH and DELPHI measurements:
Yields and cosQ distribution between
correlated pairs distinguishes between
isotropic cluster (HERWIG) and
non-isotropic string decay (JETSET)
for production mechanism.
Clustering favors baryon production
JETSET is clearly favored by the data.
Correlated L-Lbar pairs are produced
predominantly in the same jet, i.e. short
range compensation of quantum numbers.
The Lessons from RHIC
(III) flavor physics
Strange enhancement vs. charm suppression ?
A remarkable difference
between RAA and RCP
(Helen Caines talk)
‘Canonical suppression’ in pp
is unique to strange hadrons.
But is it a flavor effect ?
Kaon behaves like D-meson,
we need to measure Lc
Charm cross-section measurements in
pp collisions in STAR
– Charm quarks are believed to be produced at early stage
by initial gluon fusions
– Charm cross-section should follow number of binary
collisions (Nbin) scaling
Measurements
direct D0
(event mixing)
c→+X
(dE/dx, ToF)
c→e+X
(ToF)
c→e+X
(EMC)
pT (GeV/c)
0.1-3.0
0.17-0.25
0.9-4.0
 1.5
constraint
, d/dpT

, d/dpT
d/dpT
LO / NLO / FONLL?
•A LO calculation gives you a rough estimate of the cross section
•A NLO calculation gives you a better estimate of the cross section and a rough
estimate of the uncertainty
•Fixed-Order plus Next-to-Leading-Log (FONLL)
– Designed to cure large logs in NLO for pT >> mc where mass is not relevant
– Calculations depend on quark mass mc, factorization scale F (typically F = mc
or 2 mc), renormalization scale R (typically R = F), parton density functions
(PDF)
– Hard to obtain large  with R = F (which is used in PDF fits)
LO:
FONLL RHIC:
400
381
 cFONLL
256-146
b;  cNLO
c
c 244 -134 b
99
 bbFONLL 1.87-00..67
b
NLO:
from hep-ph/0502203
Charm - Experiment vs. Theory
• The non-perturbative charm
fragmentation needed to be tweaked in
FONLL to describe charm. FFFONLL is
much harder than used before in ‘plain’
NLO  FFFONLL ≠ FFNLO
RHIC: FONLL versus Data
 cc (STAR from D 0  eTOF   )
 cc ( FONLL)
• Matteo Cacciari
(FONLL):
• factor 2 is not a
problem
hep-ex/0609010
• factor 5 is !!!
nucl-ex/0607012
–This
Spectra
in pp between
seem to require
a bottom
contribution
difference
STAR and
PHENIX
in the pp data
–(f=2.5),
Does the
5 excess
in the difference
charm cross-section
between
willfactor
lead to
a significant
in the R(AA)
spectra
FONLL
and STAR
also apply
cross-section?
between
STAR
and PHENIX
forto
thebottom
non-photonic
electrons
Conclusions
• We need to establish the energy dependence of the
hadronization process in vacuum and the
factorization theorem as a function of flavor.
• This is an interesting overlap topic with high energy
physicists. Not everybody is involved in the Higgs
search.
• Fragmentation studies are a link between pp and
AA, between nuclear physics and high energy
physics. Is there recombination in pp ?
• Novel ideas of nuclear physics need to be applied to
pp (HBT, blastwave, v2). How collective is pp ?
Is there a radial flow component ?
(blastwave fits to STAR data)
There is an elliptic flow component
There is an interesting HBT component, see Mike Lisa’s talk
First publications
• It only takes a handful of events to measure a few important global
event properties (dN/dh, d/dpT, etc.) – after LHC start-up, with
few tens of thousand events we will do:
Claus Jorgensen
Pseudorapidity
density dN/dη
CDF:
Phys. Rev.
D41, 2330 (1990)
30000 events at √s=1.8TeV
9400 events at √s=640TeV
Multiplicity
distribution
UA5:
Z. Phys
43, 357 (1989)
6839 events at √s=900GeV
4256 events at √s=200GeV
pT spectrum
of charged
particles
CDF:
Phys. Rev. Lett.
61, 1819 (1988)
55700 events at √s=1.8TeV
Mean pT vs
multiplicity
CDF:
Phys. Rev.
D65,72005(2002)
3.3M events at 1.8TeV
2.6M events at 630GeV
Outlook
• There are significant questions regarding the
fragmentation process at LHC energies
• Topological V0 and rdE/dx analysis will allow us to
measure many properties particle identified.
• There is no ‘statistics’ problem out to 20 GeV/c.
• There is a viable physics program besides being a
reference for AA:
– Hadronization (baryons vs. mesons ?)
– Fragmentation (universality ?, applicability ?)
• The collision energy dependence is crucial.
The Black Hole search…..
(Humanic, Koch, Stoecker, hep-ph/0607074)
NOT Year-1 physics. For later…