New results from PHENIX
on energy loss and direct photons
Zimanyi Memorial Workshop
July 2-4, 2007 – Budapest
G. David, BNL
1
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
Reminder: PHENIX data sets
p+p
d+Au
(19.6)
62.4 GeV
130 GeV
energy
dependence
22.4 GeV
Au+Au Cu+Cu
++
200 GeV
reference
points
system size
dependence
Upgrades, new possibilities
2
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
High pT π0-s in Au+Au – birds-eye view
Run-1, 130GeV Au+Au:
- First observation of large π0 suppression above 2GeV (PRL 88, (2002) )
(“world data” interpolation as reference)
Run-2, 200GeV Au+Au:
- spectra up to 14GeV, suppression is constant (pQCD reference)
Run-4, 200GeV Au+Au + Run-3 p+p
- spectra up to 18GeV, suppression is constant (measured reference)
Run-4, 62GeV Au+Au
- “fast track analysis”, smaller suppression than at 200GeV
(“world data fit” as reference)
Run-5, 200GeV Cu+Cu and p+p:
- suppression commensurate to 200GeV Au+Au at similar Npart
Run-4, 62GeV Au+Au RAA revisited (Au+Au data points are the same!)
- suppression very similar to 200GeV if measured p+p is used as ref.
3
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
π0 in 200GeV Au+Au and Cu+Cu
Suppression pattern similar for comparable Npart
(despite very different geometry – limitations of φ-integrated measurements)
π0 RAA
Compare this to…
in Au+Au and Cu+Cu
4
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
π0 RAA in 62GeV Au+Au – world data reference used
“Fast track analysis”, 2 months after data taking
π0 Au+Au, 62.4 GeV
I. Vitev nucl-th/0404052
dNg/dy = 650-800
Suppression even in the most central collisions appears to be
5
significantly smaller than at 200 Gev! (Note: but Cronin not observed,
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
not even in d+Au)
π0 at 62GeV p+p – world data reference
6
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
π0 RAA in 62GeV Au+Au – measured reference used
NEW
Nuclear modification now quite comparable to 200GeV!
(the medium is almost the same(?))
7
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
RAA
π0 RAA in 62GeV Au+Au – world data vs PHENIX data
1.4
PHENIX preliminary
1.2
1
vertical normalization error on p+p ref. not shown:
25% (fit world data), 19% (PHENIX data)
62.4 GeV (world p+p fit ref.)
0.8
62.4 GeV (p+p data ref.)
0.6
0.4
0.2
0
0
π0 Preliminary, 0-10 % central
1
2
3
4
5
6
7
8
p T (GeV/c)
The physics message changes quite dramatically
(paranoid question: who is right ???)
Same experiment, same acceptance, many systematic errors the same! 8
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
What do we want to learn from this –
and what precision is needed? / 1
Using π0 RAA in 200GeV Au+Au to constrain parameters in theories
Theory: Loizides,
hep-ph/0608133v2
PQM, C. Loizides
hep-ph/0608133v2
6 ≤ qˆ ≤ 24 GeV 2 /fm
(Probability > 10%)
Nice first try, but not very restrictive…
9
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
What do we want to learn from this –
and what precision is needed? / 2
GLV, I. Vitev
Phys.Lett.B639:38-45,2006
1000 ≤
dN g
≤ 2000 GeV 2 /c 2 fm
dy
(Probability > 10%)
10
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
What do we want to learn from this –
and what precision is needed? / 3
WHDG, William Horowitz
600 ≤
dN g
≤ 1600
dy
(Probability > 10%)
11
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
Precision, physics
theorists and experimentalists
Our field is very rich, and doesn’t cease to surprise us:
- some much anticipated signals of the QGP (like thermal photons,
medium modifications of vector mesons, γ-jet correlations, …)
are somewhat lagging behind for technical reasons
- other – less anticipated – signals became the rage overnight, sparking
tremendous theoretical activity (like hadron suppression, parton scaling
of flow, “baryon anomaly”, charm suppression and flow, J/Psi
suppression commensurate to SPS, …)
- still we are far from a coherent picture: none of the results is powerful
enough to reject all but one explanation
“Powerful enough” usually means “precise”. But “precision” is not a goal
for itself. The goal is to understand the physics, and sometimes it is
best served by pushing, say, an inclusive spectrum to the limits and beyond,
other times may be better served with a less precise, but more complex
and more constraining measurement (double differential, correlation, …)
No single recipe here: this is decided by the intense interaction between
theory and experiment, which was excellent in the past years.
Let’s keep it that way!
12
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
One way out: differential quantities
13
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
High pT direct photons in Au+Au – birds-eye view
Run-2, 200GeV Au+Au:
- spectra up to 14GeV, excess (double ratio) consistent with NLO pQCD
- integrated RAA (using NLO pQCD as reference) consistent with 1
(QM’04, PRL 94, 232301 (2005))
Run-4, 200GeV Au+Au:
- spectra up to 18GeV
- double ratio at mid-pT above TAA scaled NLO pQCD,
but below it at high pT (?)
(Hard Probes 2006, Nucl. Phys. A 783 (2007) p359 and p569)
Run-5, 200GeV p+p:
- spectra 5-23GeV, shape different from NLO pQCD, new fit to the data
Run-4, 200GeV Au+Au RAA revisited (200GeV photon data still the same):
- suppression in photon RAA at high pT, even beyond isospin-effects?
Run-4, 62GeV
- photon RAA with NLO pQCD as it did in step 2… analyzing 62GeV p+p
14
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
First: direct photon excess, compared to NLO pQCD
Run-2, 200GeV Au+Au
0
Curve, bands:
NLO pQCD, scaled
with TAB
Reasonable agreement
0
(γ/π0)meas/(γ/π0)had
(γ / π )Measured / (γ / π )Background
Inclusive/decay photons
Technically:
“double ratio”
Large signal due to
π0 suppression
No statement on
intermediate pT
6
5 0-10%
4
3
2
1
0
6
5 20-30%
4
3
2
1
0
6
5 60-92%
4
3
2
1
0
0
2
4
10-20%
30-40%
MinBias
6
8 10 12 14 0
pT (GeV/c)
2
4
6
8 10 12 14
pT (GeV/c)
15
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
Integrated photon RAA (using NLO pQCD)
RAA (pT > 6.0 GeV/c)
Consistent with unity at all centralities
Measurement up to 14GeV/c
Dominated by hard scattering (>6GeV/c) and unaffected by the medium
TAB scaling is a valid concept
Integrated above 6GeV
Left the door open for
- extra sources at
medium pT
- shape differences
-…
200 GeV Au+Au Direct Photon
200 GeV Au+Au π0
2
1.5
1
0.5
0
0
50
100
150
200
250
300
350
Npart
16
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
Direct γ in p+p, PHENIX – data vs theory, old vs new
p-p
p-p
PHENIX PRL 98 (2007) 012002
Published results 3<pT<15 GeV/c
Preliminary results for 5<pT<24 GeV/c
17
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
Direct γ in p+p, PHENIX – comparison w/ other data and pQCD
Aurenche et al Eur. Phys. JC9 (1999)107
Plot by Monique Werlen at
RHIC&AGS users meeting 2005
Published PHENIX data
(“old”, Run-3)
Longstanding issue with theory
seems to be settled now
But note: the new, Run-5 prelim.
data are ~20% higher; also
seem to have a different shape
than pQCD
18
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
What Is The p+p Reference? Calculation vs Data
Photons in 200GeV p+p (Run-5)
0.5pT favored, but even this misses the shape
Black circles: Run-5 data divided by
an empirical fit.
Blue lines: NLO pQCD
(different both in magnitude and shape)
19
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
Photon RAA central, 200 GeV with fit to p+p data
One of the big sensations at Quark Matter 2006: our direct photon RAA in Au+Au
200 GeV Au+Au Direct Photon
200 GeV Au+Au π0
2
1.5
Most central collisions
Direct Photon Au+Au sNN = 200GeV, 0-10%
2
RAA
RAA (pT > 6.0 GeV/c)
Remember: integrated
RAA with NLO pQCD
(as published)
1
1.8
0.5
1.6
0
0
50
100
150
200
250
300
350
Npart
Is the high pT
suppression real?
PHENIX
preliminary
PbSc + PbGl
1.4
1.2
1
0.8
Is it suppression at all?
0.6
0.4
Are p+p data the right
thing to normalize
photon RAA?
0.2
0
0
2
4
6
8
10
12
14
16
18
20
pT (GeV/ c )
20
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
Comparisons to predictions
•
Turbide et al.
– Jet photons + pQCD + thermal
– AMY formalism for jet-quenching
of fragmentation photons
– Data systematically below theory
– Phys. Rev. C72 (2005) 014906 +
private communication
•
F. Arleo
– pQCD photons only
– High-pT suppression due to
isospin effect, shadowing, and
energy loss
– BDMPS for jet-quenching
– JHEP 0609 (2006) 015
21
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
Why isn’t the γ RAA unity at high pT?
Evolution of the gluon structure function?
“Isospin” effect (different charge squared
content of p and n)?
Note
- pions care about color, not charge
- pion xT is 2pT/ sqrt(s),
photon xT is 2pT/z*sqrt(s)
Your favorite model here ☺
Yet another dreaded explanation:
it could be an experimental error
so we started looking for a cross-check
Eskola,Kolhinen,Ruuskanen
Nucl. Phys. B535(1998)351
22
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
PHENIX – “Isospin Effect” (?)
The isospin effect (charge square difference between uud and udd) SHOULD be there,
but is this (and only this “trivial effect”) what we see?
Or do we see in addition some genuine photon suppression?
No contradiction here: only “primordial” photons should be unaltered,
“medium-induced” photons can be enhanced or suppressed
σ AA /N coll = (1/A 2 ) × (Z2σ pp + 2Z(A - Z)σ pn + (A - Z)2 σ nn )
F. Arleo, JHEP09 (2006) O15
W. Vogelsang, NLO pQCD + isospin
RAA with pQCD
RAA with p+p data
23
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
Isospin Effect – Statistics?
If this is real, we should convincingly see it in ALL centralities
(need much higher statistics – and smaller systematics… - to reduce errors)
24
Isospin Effect – xT scaling
Unfortunately the suppression is seen in a region where we are very sensitive
to detector bias (cluster merging).
Also, so far it was seen only in one of the detectors (the one more prone to merging)
xT scaling to the rescue?
The reason:
certain known detector
imperfections
(like shower merging,
nonlinearity…) are smaller at lower pT!
Yes, we do our best to correct
for them but nothing beats
not having the problem
in the first place…
The catch:
sources at intermediate pT
(like jet conversion)
that are so far of unknown
magnitude,
come into play, too!
25
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
Brand new: direct photons in 62GeV Au+Au
Lines: TAB-scaled pQCD
d2N (GeV/c)-2
1
2π pTNevt dpTdy
So we analyzed our
62GeV AuAu data,
and that’s what we got
1
-1
10
PbSc direct photon pT spectra
Au+Au sNN = 62.4GeV
PHENIX Preliminary
10-2
min. bias × 10
0-10% × 10
3
1
10-30% × 10
0
30-60% × 10
-1
NLO pQCD × T AB
m = 0.5pT, 2pT
10-3
10-4
Note the same tendency
as for π0
(and 200GeV p+p):
favors 0.5pT scale
10-5
10-6
-7
10
10-8
10-9
3
4
5
6
7
8
pT (GeV/c)
26
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
Photon excess (double ratios) in 62GeV Au+Au
(γ/π0)meas/(γ/π0)had
3.5
Au+Au
sNN = 62.4GeV
0-10%
π0
4
5
(Nall/N )meas/(Nbg/N )MC
π0
4.5
γ
PHENIX Preliminary
γ
(Nall/N )meas/(Nbg/N )MC
5
4
3.5
π0
π0
2.5
2
γ
3
2.5
2
1.5
1.5
1
1
3
4
5
6
7
8
pT (GeV/ c )
Au+Au
sNN = 62.4GeV
Min. bias
PHENIX Preliminary
γ
3
4.5
3
4
5
6
7
8
pT (GeV/ c )
Remarkably high – keep in mind pion suppression!
27
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
3
Data/(NLO pQCD (m=pT) × TAB)
Data/(NLO pQCD (m=pT) × TAB)
Photon RAA (with pQCD) in 62GeV Au+Au
Au+Au sNN = 62.4GeV
0-10%
2.5
2
1.5
1
0.5
3
Au+Au sNN = 62.4GeV
10-30%
2.5
2
1.5
1
0.5
(m=0.5p T ) / (m=pT )
PHENIX Preliminary
4
5
6
7
8
pT(GeV/ c )
3
Au+Au sNN = 62.4GeV
30-60%
2.5
0
3
Data/(NLO pQCD (m=pT) × TAB)
Data/(NLO pQCD (m=pT) × TAB)
0
3
(m=0.5p T ) / (m=pT )
PHENIX Preliminary
(m=2p T ) / (m=pT )
2
1.5
1
5
6
7
8
pT(GeV/ c )
3
Au+Au sNN = 62.4GeV
Min. bias
2.5
2
1.5
1
0.5
0.5
(m=0.5p T ) / (m=pT )
(m=0.5p T ) / (m=pT )
PHENIX Preliminary
0
3
4
(m=2p T ) / (m=pT )
4
PHENIX Preliminary
(m=2p T ) / (m=pT )
5
6
7
8
pT(GeV/ c )
0
3
4
(m=2p T ) / (m=pT )
5
6
7
28 8
pT(GeV/ c )
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
Photon RAA : don’t take this literally!
29
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL
Summary
Waning (albeit not the end) of the era of single inclusive measurements
Tremendous benefits from measuring the reference in the same experiment
Physics message of 62 GeV Au+Au π0 RAA revised – medium almost the same
as the one at 200GeV Au+Au
Photons suppression at high pT in 200GeV Au+Au will likely be confirmed
by the final 62GeV results
Nothing beats a detailed energy/species scan in the same experiment:
RHIC-II is coming!
30
Zimanyi 75 Memorial Workshop, Budapest – July 3, 2007 – G. David, BNL