1. Drell-Yan with Fixed Target at RHIC
“RHIC Spin: The Next Decade”
at Iowa State University
May 15th, 2010
Yuji Goto (RIKEN/RBRC)

2. Drell-Yan experiment The simplest process in hadron-hadron reactions
No QCD final state effect
Fermilab E866
Flavor asymmetry of sea-quark distribution
DIS
Drell-Yan
May 15, 2010

3. Drell-Yan experiment Flavor asymmetry of sea-quark distribution
Possible origins
meson-cloud model
virtual meson-baryon state
chiral quark model
instanton model
chiral quark soliton model
+ in the proton as an origin of anti-d quark
pseudo-scaler meson should have orbital angular momentum in the proton…
Polarized Drell-Yan experiment
Not yet done!
Many new inputs for remaining proton-spin puzzle
flavor asymmetry of the sea-quark polarization
transversity distribution
transverse-momentum dependent (TMD) distributions
Sivers function, Boer-Mulders function, etc.
May 15, 2010

4. Single transverse-spin asymmetry
Sivers effect
Correlation between nucleon transverse spin and parton transverse momentum (Sivers distribution function)
Related to the orbital angular momentum in the proton (and the shape of the proton)
Multi-dimensional structure of the proton
Initial-state or final-state interaction with remnant partons
J-PARC
50-GeV
polarized beam
s ~ 10 GeV
RHIC
collider
s = 200 GeV
Anselmino, et al., PRD79, (2009)
May 15, 2010

5. Sivers function measurement
Sign of Sivers function determined by single transverse-spin (SSA) measurement of DIS and Drell-Yan processes
Should be opposite each other
Initial-state interaction or final-state interaction with remnant partons
< 1% level multi-points measurements have already been　done for SSA of DIS process
x = – 0.3 (more sensitive in lower-x region)
comparable level measurement needs to be done for SSA of Drell-Yan process for comparison
May 15, 2010

6. J-PARC and/or RHIC J-PARC RHIC Advantage Disadvantage Uncertainties
High intensity = high luminosity
Disadvantage
Smaller cross section (at the same invariant mass)
Uncertainties
Availability of 50 GeV beam
Polarized proton beam?
RHIC
Polarized proton beam available
Larger cross section (at the same invariant mass)
Luminosity?
Collider or fixed-target (internal-target) experiment
Before discussing the RHIC case, I compare the J-PARC experiment and RHIC experiment.
May 15, 2010

7. FNAL-E906 dimuon spectrometer
Solid Iron Focusing Magnet, Hadron absorber and beam dump
4.9m
Mom. Meas.
(KTeV Magnet)
Hadron Absorber
Station 1:
Hodoscope array
MWPC tracking
Station 2 and 3:
Drift Chamber tracking
Station 4:
Prop tube tracking
Liquid H2, d2, and solid targets
May 15, 2010

8. PYTHIA simulation with IP2 configuration
IP2 configuration shown in the next slide
detector components from FNAL-E906 apparatus
E906: total z-length ~25 m
IP2: available z-lengh ~14 m
z-length of FMAG (1st magnet) is shortened
because of limited z-length at IP2
PYTHIA simulation
s = 22 GeV (Elab = 250 GeV)
luminosity assumption 10,000 pb-1
 10 times larger luminosity necessary than collider experiments
because of ~10 times smaller cross section  acceptance (in the same mass region)
M = 4.5  8 GeV
acceptance for Drell-Yan dimuon signal is studied
momentum resolution (geometric) 6  10-4  p (GeV/c)
background rate needs to be studied
May 15, 2010

9. IP2 (overplotted on BRAHMS)
Internal target position
St.4
MuID
St.1
St.2
St.3
FMAG 3.9 m
Mom.kick
2.1 GeV/c
KMAG 2.4 m
Mom.kick
0.55 GeV/c
14 m
18 m
May 15, 2010

10. Drell-Yan at IP2 rapidity E1 vs E2 (accepted events)
all generated dumuon
from Drell-Yan
passed through FMAG
passed through St. 1
passed through St. 2
passed through St. 3
accepted by all detectors
E1, E2: energy of dimuons
energy cut shown ~10 GeV
May 15, 2010

11. Drell-Yan at IP2 rapidity x1 xF pT accepted 4.5 - 8 GeV/c2
May 15, 2010

12. Drell-Yan at IP2 About 50,000 events for 10,000 pb-1
Mass
(GeV/c2)
Total
Rapidity
45 – 50
50 – 60
60 – 80
45 – 80
-0.4 – 0
3.1 K
1.4 K
7.6 K
0 – 0.4
6.2 K
6.1 K
3.0 K
15.3 K
0.4 – 0.8
6.4 K
2.3 K
16.3 K
0.8 – 1.2
4.4 K
2.5 K
0.4 K
7.3 K
About 50,000 events for 10,000 pb-1
x1 vs x2 (accepted events)
x1: x of beam proton (polarized)
x2: x of target proton
May 15, 2010

13. Collider vs fixed-target
x1 & x2 coverage
Collider experiment with PHENIX muon arm (simple PYTHIA simulation)
s = 500 GeV
Angle & E cut only
1.2 < || < 2.2 (0.22 < || < 0.59), E > 2, 5, 10 GeV
(no magnetic field, no detector acceptance)
luminosity assumption 1,000 pb-1
M = 4.5  8 GeV
Single arm: x1 = 0.05 – 0.1 (x2 = – 0.002)
Very sensitive x-region of SIDIS data
Fixed-target experiment
x1 = 0.25 – 0.4 (x2 = 0.1 – 0.2)
Can explore higher-x region with better sensitivity
PHENIX muon arm (angle & E cut only)
Fixed-target experiment
This is a comparison of x of polarized beam, x_1 and x of target, x_2 with my simulation of collider experiment with PHENIX muon arm.
The collider experiment covers x_1 from 0.05 – 0.1 which is also a very sensitive region of the SIDIS data.
The Fixed target experiment covers a larger x_1 region from 0.25 to 0.4, and the measurement can explore higher-x region with better sensitivity. x2 x2 x1 x1
May 15, 2010

14. Fixed-target experiment
Phase-1: parasitic experiment (with other collider experiments)
Beam intensity 2×1011×10MHz = 21018/sec
Cluster or pellet target 1015/cm2
50 times thinner than RHIC CNI carbon target
Luminosity 2×1033/cm2/sec
10,000pb-1 with 5×106sec
8 weeks, or 3 years (10 weeks×3) with efficiency and live time
Reaction rate 2×1033×50mb = 108/sec = 100MHz
Beam lifetime 2×1011×100bunch / 108 = 2×105sec (at longest)
More factor to make the lifetime shorter, but expected to be still long enough
May 15, 2010 14

15. Fixed-target experiment
Phase-2: dedicated experiment
Beam intensity 2×1011×30MHz = 6×1018/sec
Cluster, pellet or solid target 1016/cm2
5 times thinner than RHIC CNI carbon target
Luminosity 6×1034/cm2/sec
10,000pb-1 with 2×105sec
3 days, or 2 weeks with efficiency and live time
Or 50,000pb-1 with 106sec
2 weeks, or 8 weeks with efficiency and live time
Reaction rate 6×1034×50mb = 3×109/sec = 3GHz
Beam lifetime 2×1011×300bunch / 3×109 = 2×104sec ~ 6hours
More factor to make the lifetime shorter
Special short-interval operation necessary
May 15, 2010

16. Physics Experimental sensitivity Phase-1 (parasitic operation)
L = 2×1033 cm-2s-1
10,000 pb-1 with 5106 s ~ 8 weeks (1 year), or 3 years of beam time by considering efficiency and live time
Phase-2 (dedicated operation)
L = 6×1034 cm-2s-1
50,000 pb-1 with 106 s ~ 2 weeks, or 8 weeks (1 year) of beam time by considering efficiency and live time
Theory calculation:
U. D’Alesio and S. Melis, private communication;
M. Anselmino, et al., Phys. Rev. D79, (2009)
10,000 pb-1 (phase-1)
50,000 pb-1 (phase-2)
May 15, 2010

17. pQCD studies of Drell-Yan cross section
pQCD correction can be controlled
NNLO calculation
Hamberg, van Neerven, Matsuura,
Harlander, Kilgore
NLL, NNLL calculation
Yokoya, et al...
by Hiroshi Yokoya
For the J-PARC experiment, it is important to discuss perturbative-QCD studies of unpolarized Drell-Yan cross section because the J-PARC energy is low. The perturbative QCD correction has to be controlled theoretically at the J-PARC energy, and the measured cross section has to be explained by the perturbative QCD before discussing the polarized results theoretically.
Correction calculations have been done by these theorists in NNLO calculation, and Yokoya has given NLL and NNLL calculation as shown by these figures.
According to Yokoya’s calculation, …
May 15, 2010 17

18. Letter of Intent Author list ANL UIUC KEK LANL RIKEN RBRC
D. F. Geesaman, R. E. Reimer, J. Rubin
UIUC
M. Grosse Perdekamp, J.-C. Peng
KEK
N. Saito, S. Sawada
LANL
M. Brooks, X. Jiang, G. Kunde, M. J. Leitch, M. X. Liu, P. L. McGaughey
RIKEN
Y. Fukao, Y. Goto, I. Nakagawa, R. Seidl, A. Taketani
RBRC
K. Okada
Stony Brook Univ.
A. Deshpande
Tokyo Tech
K. Nakano, T.-A. Shibata
Yamagata Univ.
N. Doshita, T. Iwata, K. Kondo, Y. Miyachi
May 15, 2010

19. Fixed-target experiment
Internal target
Storage cell target (polarized)
H2, D2, 3He
1014 / cm2
HERMES
Cluster jet target
H2, D2, N2, CH4, Ne, Ar, Kr, Xe, …
1014 – 1015 / cm2
CELSIUS (Uppsala), FNAL-E835, Muenster, COSY, GSI, …
Pellet target
H2, D2, N2, Ne, Ar, Kr, Xe, …
1015 – 1016 / cm2
CELSIUS (Uppsala), COSY, GSI, …
Double-spin experiment possible
Low density = low luminosity
Issue to be used in the RHIC ring
High luminosity
Only single-spin experiment possible
May 15, 2010 19

20. Cost and schedule To be considered… Cost Schedule
Accelerator & experimental hall (IP2)
the most expensive part?
Internal target
e.g. PANDA pellet target: $1.5M R&D, infrastructure + $1.5M construction
Experimental apparatus
FNAL-E906 magnet (modification) or other existing magnet at BNL
FNAL-E906 apparatus (modification) and/or new/existing apparatus
Schedule
FNAL-E906 beam time
accelerator R&D, internal target R&D + construction
2014 experimental setup & commissioning at RHIC
phase-1: parasitic experiment
2018 phase-2: dedicated experiment
May 15, 2010

21. Issues (as a summary) Requirement for target thickness
In phase-1 (parasitic operation), 1015 /cm2 thickness is necessary to achieve L = 21033 cm-2s-1
Pellet target (~1015 /cm2) or cluster-jet target (up to 81014/cm2 ?) is necessary
If it is not achieved, the internal-target experiment would not be competitive against collider experiments
In collider experiments, L = 2(or 3)1032 is expected and cross section ( acceptance) is >10 times larger
In phase-2 (dedicated operation), 10-times larger thickness, 1016 /cm2 is expected
Requirement for accelerator
Compensation of dipole magnets in the experimental apparatus
both beams need to be restored on axis in two colliding-beam operation
Size of the experimental site and possible target-position (with proper beam operation)
Reaction rate (peak rate) and beam lifetime
Radiation issues
Beam loss/dump requirement
May 15, 2010

22. Backup slides

23. Polarized Drell-Yan experiment
Single transverse-spin asymmetry
Sivers function measurement
Transversity  Boer-Mulders function
Double transverse-spin asymmetry
Transversity (quark  antiquark for p+p collisions)
Double helicity asymmtry
Flavor asymmetry of quark polarization
Other physics
Parity violation asymmetry?
May 15, 2010

24. Sivers function measurement
Sign of Sivers function determined by single transverse-spin (SSA) measurement of DIS and Drell-Yan processes
Should be opposite each other
Initial-state interaction or final-state interaction with remnant partons
Test of TMD factorization
Explanation by Vogelsang and Yuan…
From “Transverse-Spin Drell-Yan Physics at RHIC,” Les Bland, et al., May 1, 2007
DIS
Drell-Yan
“toy” QED
QCD
attractive
repulsive
May 15, 2010

25. Single transverse-spin asymmetry
Sivers function measurement
Transversity  Boer-Mulders function measurement
by Stefano Melis
May 15, 2010

26. Polarized and unpolarized Drell-Yan experiment
ATT measurement
h1(x): transversity
quark  antiquark in p+p collisions
Unpolarized measurement
angular distribution of unpolarized Drell-Yan
Boer-Mulders function
violation of the Lam-Tung relation
With Boer-Mulders function h1┴:
ν(π-Wµ+µ-X)~valence h1┴(π)*valence h1┴(p)
ν(pdµ+µ-X)~valence h1┴(p)*sea h1┴(p)
Another important topics is measurements of transversity and Boer-Mulders function.
Transversity is a remaining leading-order distribution function of the nucleon.
Transversity is measured by the double transverse-spin asymmetry A_TT measurement.
From SSA measurement, we can also measure a product of transversity and Boer-Mulders function.
L.Y. Zhu,J.C. Peng, P. Reimer et al.
hep-ex/
May 15, 2010

27. Polarized Drell-Yan experiment
Longitudinally-polarized measurement
ALL measurement
flavor asymmetry of sea-quark polarization
SIDIS data from HERMES, and new COMPASS data available
W data from RHIC will be available in the near future
Polarized Drell-Yan data will be able to cover higher-x region
Bhalerao, statistical model
Cao & Signa, bag model
Cao & Signa, meson cloud
Wakamatsu, CQSM SU(2)
Wakamatsu, CQSM SU(3)
HERMES, PRD71: , 2005
By using polarized beam and target, the experiment can be extended to measure Delta-dbar / Delta-ubar ratio.
In chiral quark soliton model, Delta-dbar – Delta-ubar is predicted to have a larger magnitude than dbar – ubar.
This is an expected statistics for A_LL measurement of 120-day run at J-PARC.
May 15, 2010

28. FNAL-E906 dimuon spectrometer
Station 2 and 3:
Hodoscope array
Drift Chamber tracking
Station 1:
Hodoscope array
MWPC tracking
Hadron Absorber
Solid Iron Focusing Magnet, Hadron absorber and beam dump
Mom. Meas.
(KTeV Magnet)
Station 4:
Hodoscope array
Prop tube tracking
4.9m
25m
Solid iron magnet
Reuse SM3 magnet coils
Sufficient Field with reasonable coils (amp-turns)
Beam dumped within magnet
Liquid H2, d2, and solid targets
May 15, 2010

29. pQCD studies of Drell-Yan cross section
pQCD correction can be controlled at J-PARC energy
NNLO calculation
Hamberg, van Neerven, Matsuura,
Harlander, Kilgore
NLL, NNLL calculation
Yokoya, et al...
by Hiroshi Yokoya
For the J-PARC experiment, it is important to discuss perturbative-QCD studies of unpolarized Drell-Yan cross section because the J-PARC energy is low. The perturbative QCD correction has to be controlled theoretically at the J-PARC energy, and the measured cross section has to be explained by the perturbative QCD before discussing the polarized results theoretically.
Correction calculations have been done by these theorists in NNLO calculation, and Yokoya has given NLL and NNLL calculation as shown by these figures.
According to Yokoya’s calculation, …
May 15, 2010

30. Collider experiment Very simple PYTHIA simulation for PHENIX muon arm
s = 500 GeV
Angle & E cut only
1.2 < || < 2.2 (0.22 < || < 0.59)
E > 2, 5, 10 GeV
(no magnetic field, no detector acceptance)
RHIC-II luminosity assumption 1,000 pb-1
M = 4.5  8 GeV
110K events total with 2-GeV cut
May 15, 2010

31. Collider experiment “Transverse-Spin Drell-Yan Physics at RHIC”
Les Bland, et al., May 1, 2007
s = 200 GeV
PHENIX muon arm
STAR FMS (Forward Muon Spectrometer)
Large background from b-quark
s = 500 GeV may be better…
Higher luminosity and larger cross section
May 15, 2010

32. Collider experiment
Very simple PYTHIA simulation for a dedicated collider experiment
s = 500 GeV
Angle & E cut only
|| < 2.2
E > 2, 5, 10 GeV
(no magnetic field, no detector acceptance)
RHIC-II luminosity assumption 1,000 pb-1
M = 4.5  8 GeV
550K events total
with 2-GeV cut
May 15, 2010

33. Fixed-target experiment
Simple PYTHIA simulation for a fixed target experiement
s = 22 GeV (Elab = 250 GeV)
Angle & E cut only
0.03 <  < 0.1
E > 2, 5, 10 GeV
(no magnetic field, no detector acceptance)
luminosity assumption 10,000 pb-1
 10 times larger luminosity necessary than collider experiments
because of ~10 times smaller cross section  acceptance
M = 4.5  8 GeV
Red: E > 2 GeV cut
Green: E > 5 GeV cut
Blue: E > 10 GeV cut
35K events total with 10-GeV cut
For the case study of RHIC fixed-target experiment, I performed a simple PYTHIA simulation.
250-GeV beam energy was assumed which corresponds to root(s) = 22-GeV.
Only angle cut and muon energy cut were applied and no magnetic field and no detector acceptance were not applied.
10-times larger luminosity is necessary for Drell-Yan events comparable to the collider experiments, because of 10 times smaller cross section times acceptance. So 10,000 pb^-1 luminosity was assumed, and the number of Drell-Yan events was counted for the same dimuon mass range from 4.5 to 8 GeV.
As a result, 35 K events are expected for 10-GeV energy cut.
May 15, 2010

34. FNAL-E906 First Magnet (FMag)
May 15, 2010

35. FNAL-E906 First Magnet (FMag)
May 15, 2010

36. FNAL-E906 Second Magnet (KMag)
May 15, 2010

37. experiment particles energy x1 or x2 luminosity COMPASS  + p
160 GeV
s = 17.4 GeV
x2 = 0.2 – 0.3
2 × 1033 cm-2s-1
COMPASS (low mass)
x2 ~ 0.05
PAX
p + pbar
collider
s = 14 GeV
x1 = 0.1 – 0.9
2 × 1030 cm-2s-1
PANDA
(low mass)
pbar + p
15 GeV
s = 5.5 GeV
x2 = 0.2 – 0.4
2 × 1032 cm-2s-1
J-PARC
p + p
50 GeV
s = 10 GeV
x1 = 0.5 – 0.9
1035 cm-2s-1
NICA
s = 20 GeV
x1 = 0.1 – 0.8
1030 cm-2s-1
SPASCHARM
p + p
60 GeV
s = 11 GeV
x2 = 0.05 – 0.2
34 GeV
s = 8 GeV
x2 = 0.1 – 0.3
RHIC PHENIX
Muon
s = 500 GeV
x1 = 0.05 – 0.1
RHIC Internal
Target phase-1
250 GeV
s = 22 GeV
x1 = 0.25 – 0.4
Target phase-2
6 × 1034 cm-2s-1
May 15, 2010

38. Accelerator issues Experimental dipole magnets
Beam axis restoration (one beam on target, the other beam displaced)
Size of the experimental site and possible target-position (with proper beam operation)
IP2 (BRAHMS) area?
Reaction rate (peak rate) and beam lifetime
Radiation issues
Beam loss/dump requirement
May 15, 2010