Transcript SIDIS at EIC Harut Avakian (JLab) Gluons and the quark sea at high energies , INT Nov 9, 2010 •TMDs and spin-orbit correlations •PT-distributions •Higher.

SIDIS at EIC
Harut Avakian (JLab)
Gluons and the quark sea at high energies , INT Nov 9, 2010
•TMDs and spin-orbit correlations
•PT-distributions
•Higher twists in SIDIS
•Nuclear modifications
•Kaons vs pions
•MC-simulations
•Projections for observables
•Conclusions
H. Avakian, INT, Nov 9
1
Single hadron production in hard scattering
xF>0 (current fragmentation)
h
xF<0 (target fragmentation)
xF - momentum
in the CM frame
Target fragmentation
Current fragmentation
semi-inclusive
semi-exclusive
FF h
-1
Fracture Functions
h
h
DA
DA
h
M
exclusive
PDF
0
kT-dependent PDFs
PDF
GPD
1
xF
Generalized PDFs
Measurements in different kinematical regions provide complementary
information on the complex nucleon structure.
H. Avakian, INT, Nov 9
2
kT-dependent PDFs and FFs: “new testament”
Baccetta, Diehl, Goeke, Metz, Mulders, Schlegel EPJ-2007
No analog in twist-2
appear in sinf moment
of ALU and AUL
quark-gluon-quark correlations
responsible for azimuthal
moments in the cross section
H. Avakian, INT, Nov 9
3
Electroproduction kinematics: JLab12→EIC
Q2
collider experiments
EIC (4x60):
H1, ZEUS 10-4<xB<0.02
fixed target experiments
COMPASS 0.006<xB<0.3
HERMES 0.02<xB<0.3
gluons/valence and sea quarks
JLab
 0.1<xB<0.7
JLab@12GeV valence quarks
EIC
JLab12
EIC 10-4<xB<0.3
EIC
EIC provides access to:
gluons (and quarks)
•wide range of Q2 for a fixed x
•wide range of PT
•small x region
H. Avakian, INT, Nov 9
4
SIDIS: partonic cross sections
p┴
PT = p┴ +z kT
Ji,Ma,Yuan Phys.Rev.D71:034005,2005
How sensitive are SIDIS
observables to x-kT correlations?
H. Avakian, INT, Nov 9
5
Quark distributions at large kT: lattice
B.Musch et al arXiv:1011.1213
Higher probability to find a quark anti-aligned
with proton spin at large kT and bT
B.Pasquini et al
Higher probability to find a d-quark at large kT
H. Avakian, INT, Nov 9
6
Quark distributions at large kT: lattice
B.Musch et al arXiv:1011.1213
q
JMR model
Dq
Du/u
Sign change of Du/u consistent
between lattice and diquark model
H. Avakian, INT, Nov 9
7
H. Avakian, INT, Nov 9
8
Hadronic PT-distriutions
H. Mkrtchyan et al. Phys.Lett.B665:20-25,2008.
NJL model, H.Matevosyan
H. Avakian, INT, Nov 9
9
Flavor decomposition in SIDIS
DIS
Frascati, Oct 17
COMPASS
COMPASS data only
HERMES
SIDIS
COMPASS
HERMES
DIS vs SIDIS → additional hadron detection.
H. Avakian, INT, Nov 9
10
Acceptances and efficiencies
HERMES
EIC
How acceptance in f and PT affect
the A1 and Ds extractions in SIDIS?
H.Avakian,
H. Avakian,JLab,
INT, Oct
Nov29
9
11
cosf moment in ALL-PT-dependence
hep-ph/0608048
m02=0.25GeV2
mD2=0.2GeV2
CLAS PRELIMINARY
PT-dependence of cosf moment of double spin asymmetry is most sensitive to kTdistributions of quarks with spin orientations along and opposite to the proton spin.
H. Avakian, INT, Nov 9
12
A1
A1 PT-dependence
arXiv:1003.4549
Lattice
Anselmino
Collins
PT
PT
CLAS data suggests that width of g1 is less than the width of f1
New CLAS data would allow multidimensional binning to study kT-dependence for fixed x
H. Avakian, INT, Nov 9
13
A1 PT-dependence in SIDIS
M.Anselmino et al
hep-ph/0608048
m02=0.25GeV2
mD2=0.2GeV2
Perturbative limit calculations
available for
:
J.Zhou, F.Yuan, Z Liang: arXiv:0909.2238
•ALL (p) sensitive to difference in kT distributions for f1 and g1
•Wide range in PT allows studies of transition from TMD to perturbative approach
H. Avakian, INT, Nov 9
14
Beam SSA for exclusive pions
Sign flip at z ~ 0.5
W2>4 GeV2,Q2>1 GeV2
4.3 GeV
-t < 0.5GeV2
5.7 GeV
At z>0.5 struck quark
in pion
LUND-MC
H. Avakian, INT, Nov 9
15
Beam SSA: ALU from COMPASS & HERMES
CLAS @4.3 &5.7GeV
No x-dependence?
Change the sign at
low z?
H. Avakian, INT, Nov 9
16
SSA at large xF
0 moves to lower xF with energy?
ANL
s=4.9 GeV
BNL
s=6.6 GeV
FNAL
s=19.4 GeV
H. Avakian, INT, Nov 9
RHIC
s=62.4 GeV
17
Beam SSA: ALU from CLAS @ JLab
Photon Sivers Effect Afanasev & Carlson,
Metz & Schlegel, Gamberg et al.
0.5<z<0.8
Beam SSA from initial
distribution (Boer-Mulders
TMD) F.Yuan using h1┴
from MIT bag model
Beam SSA from
hadronization (Collins
effect) by Schweitzer et al.
Collins contribution should be suppressed → g┴ wanted !!!
H. Avakian, INT, Nov 9
18
Chiral odd HT-distribution
How can we separate the HT contributions?
PT
fh
y
fS=p
x
HT function related to force on
the quark. M.Burkardt (2008)
Compare single hadron and dihadron SSAs
M.Radici
Only 2 terms with common
unknown HT G~ term!
H. Avakian, INT, Nov 9
19
Jet limit: Higher Twist azimuthal asymmetries
Twist-2
Twist-3
“interaction dependent”
No leading twist,
provide access to quarkgluon correlations
H.A.,A.Efremov,P.Schweitzer,F.Yuan
arXiv:1001.5467
First data available from lattice!
H. Avakian, INT, Nov 9
20
Modification of Cahn effect
Bag model
arXiv:1001.3146
Gao, Liang & Wang
total transverse
momentum
broadening squared
•Large
cosnf
moments
COMPASS
•Nuclear
modification
ofobserved
Cahn mayatprovide
info on kT broadening and proton TMDs
H. Avakian, INT, Nov 9
21
LEPTO/PEPSI: quark distributions
Event Generator with polarized
electron and nucleon: PEPSI,….
MC with Cahn
Acceptance check
Design parameters
Smearing/resolution routines
Use GEANT as input
z>0.1
Physics analysis using the
“reconstructed” event sample
z>0.3
•Implemented in PEPSI modification
to LEPTO done by A. Kotzinian
•Add different widths for q+ and q-
Need to model TMDs in LUND MC
H. Avakian, INT, Nov 9
22
EIC MC simulations
Different MC used in EIC simulations are consistent (Xin Qian)
H. Avakian, INT, Nov 9
23
PT-dependence of beam SSA
ssinfLU~FLU~ 1/Q (Twist-3)
In the
perturbative limit
1/PT behavior
expected
4x60 100 days,
L=1033cm-2s-1
Nonperturbative TMD
Perturbative region
Study for SSA transition from non-perturbative to perturbative regime.
EIC will significantly increase the PT range.
H. Avakian, INT, Nov 9
24
Q2-dependence of beam SSA
ssinfLU(UL) ~FLU(UL)~ 1/Q (Twist-3)
1/Q behavior
expected (fixed x bin)
Study for Q2 dependence of beam SSA allows to check the higher twist
nature and access quark-gluon correlations.
H. Avakian, INT, Nov 9
25
Sivers effect: pion electroproduction
GRV98, Kretzer FF (4par)
S. Arnold et al
arXiv:0805.2137
M. Anselmino et al
arXiv:0805.2677
GRV98, DSS FF (8par)
•EIC measurements at small x will pin down sea contributions to Sivers function
H. Avakian, INT, Nov 9
26
Sivers effect: Kaon electroproduction
EIC
CLAS12
•At small x of EIC Kaon relative rates higher, making it ideal place to study the
Sivers asymmetry in Kaon production (in particular K-).
•Combination with CLAS12 data will provide almost complete x-range.
H. Avakian, INT, Nov 9
27
Sivers effect: sea contributions
GRV98, DSS FF
M. Anselmino et al
arXiv:0805.2677
GRV98, Kretzer FF
S. Arnold et al
arXiv:0805.2137
•Negative Kaons most sensitive to sea contributions.
•Biggest uncertainty in experimental measurements (K- suppressed at large x).
H. Avakian, INT, Nov 9
28
Kaon production in SIDIS
FAST-MC
L
S
CLAS12
S*
EIC4x60
detected
s(p) = 0.05 + 0.06*p [GeV] %
Identification using the missing mass may be possible
H. Avakian, INT, Nov 9
29
EIC 4x60 (Lumi 1033,cm-2sec-1 , ~1 hour)
<x>=0.1, <Q2>=4
s(p) = 0.05 + 0.06*p [GeV] %
K*s can be studied with EIC
H. Avakian, INT, Nov 9
30
Kaon <cos2f> @ HERMES
H.Avakian,
H. Avakian,JLab,
INT, Oct
Nov29
9
31
Collins asymmetry - proton
“Kaon puzzle” in
spin-orbit correlations
Is there a link between HERMES and BRAHMS Kaon vs pion
moments (K- has the same sign as K+ and pi+, comparable with K+)?
H. Avakian, INT, Nov 9
32
Collins effect
Simple string fragmentation (Artru model)
p+
z
Leading pion out of
page (
- direction )
L
L
p-
kicked in the opposite
to the leading pion(into
the page)
Sub-leading pion opposite
to leading (double kick
into the page)
If unfavored Collins fragmentation dominates
measured p- vs p+, why K- vs K+ is different?
H. Avakian, INT, Nov 9
33
Boer-Mulders Asymmetry with CLAS12 & EIC
5-GeV
-
e
p
50 GeV
Transversely polarized quarks in
the unpolarized nucleon
sin(fC) =cos(2fh)
CLAS12
EIC
Perturbative limit calculations
available for
Nonperturbative TMD
Perturbative region
:
J.Zhou, F.Yuan, Z Liang:
arXiv:0909.2238
CLAS12 and EIC studies of transition from non-perturbative to perturbative regime will
provide complementary info on spin-orbit correlations and test unified theory (Ji et al)
H. Avakian, INT, Nov 9
34
From CLAS12 to EIC: Kotzinian-Mulders Effect
KM
sUL
~
Transversely polarized
quarks in the longitudinally
polarized nucleon
Worm
gear
Study Collins fragmentation using transversely
polarized quarks in a longitudinally polarized nucleon.
H. Avakian, INT, Nov 9
35
Pretzelosity @ EIC
5x50 epX
positivity bound
p+
p-
helicity-transversity=pretzelosity
In models (bag, diquark)
pretzelosity defines the OAM
•EIC measurement combined with CLAS12 will provide a complete
kinematic range for pretzelosity measurements
H. Avakian, INT, Nov 9
36
Collins Effect: from asymmetries to distributions
need
Combined analysis of Collins fragmentation asymmetries from proton
and deuteron may provide independent to e+e- (BELLE/BABAR)
Information on the underlying Collins function.
H. Avakian, INT, Nov 9
37
L production in the target fragmentation
J.Ellis, D.Kharzeev, A. Kotzinian ‘96
W.Melnitchouk and A.W.Thomas ‘96
L polarization in TFR provides information on
contribution of strange sea to proton spin
(ud)-diquark is a spin and
isospin singlet s-quark carries
whole spin of L L  uds
xF - momentum
xF(L)
in the CM frame
Study polarized diquark fracture
functions sensitive to the correlations
between struck quark transverse
momentum and the diquark spin.
EIC
CLAS12
Wide kinematical coverage of EIC would allow studies of
hadronization in the target fragmentation region (fracture functions)
H. Avakian, INT, Nov 9
38
Sivers effect in the target fragmentation
xF>0 (current
fragmentation)
A.Kotzinian
xF<0 (target
fragmentation)
h
M
Fracture Functions
Separation of current and target fragmentation at EIC will
allow studies of kinematic dependences of the Sivers effect in
target fragmentation region
H. Avakian, INT, Nov 9
39
Summary
Studies of spin and azimuthal asymmetries in semi-inclusive
processes at EIC :
•Provide detailed info on partonic spin-orbit correlations
•Measure transverse momentum distributions of partons at
small x, in a wide range of Q.
•Study quark-gluon correlations (HT) in nucleon and nucleus
•Need realistic MC simulations (LUND,Geant) to check sensitivity to
various effects related to the transverse structure of the nucleon
•Need more theory (+lattice) support for HT
EIC: Measurements related to the spin, spin orbit and quark-gluon
correlations combined with JLab12 HERMES,COMPASS,
RHIC,BELLE,BABAR,Fermilab,J-PARC,GSI data will help construct a
more complete picture about the spin structure of the nucleon beyond
the collinear approximation.
H. Avakian, INT, Nov 9
40
Support slides….
H. Avakian, INT, Nov 9
41
H. Avakian, INT, Nov 9
42
MC simulations using NJL
H. Avakian, INT, Nov 9
43
M.Osipenko
H. Avakian, INT, Nov 9
44
Tang,Wang & Zhou
Phys.Rev.D77:125010,2008
kT and FSI
l’
l
BHS 2002
Collins 2002
Ji,Yuan 2002
x,kT
proton
l’
l
total transverse momentum
broadening squared
x,k’T
lT
spectator
system
nucleus
l’T
spectator
system
soft gluon exchanges included in
the distribution function (gauge link)
•The difference is coming from final state interactions (different remnant)
H. Avakian, INT, Nov 9
45
Nuclear broadening Hadronic PT-distriutions
Large PT may have significant nuclear contribution
H. Avakian, INT, Nov 9
46
Azimuthal moments with unpolarized target
quark polarization
H. Avakian,
JLab, Nov
INT,25Nov 9
47
Azimuthal moments with unpolarized target
quark polarization
H. Avakian,
JLab, Nov
INT,25Nov 9
48
SSA with unpolarized target
quark polarization
H. Avakian,
JLab, Nov
INT,25Nov 9
49
SSA with unpolarized target
quark polarization
H. Avakian,
JLab, Nov
INT,25Nov 9
50
SSA with long. polarized target
quark polarization
H. Avakian, INT, Nov 9
51
SSA with long. polarized target
quark polarization
H. Avakian, INT, Nov 9
52
SSA with unpolarized target
quark polarization
H. Avakian, INT, Nov 9
53
SSA with unpolarized target
quark polarization
H. Avakian, INT, Nov 9
54
Twist-3 PDFs : “new testament”
H. Avakian, INT, Nov 9
55
Struck quark kinematics (EIC 4x60)
qq
High energy quarks at small angles
H. Avakian, INT, Nov 9
56
Quark distributions at large kT
bigger effect at large z
PT = p┴ +z kT
Higher probability to find a
hadron at large PT in nuclei
H. Avakian, INT, Nov 9
kT-distributions may be
wider in nuclei?
57
Hadronic PT-distriutions
H. Mkrtchyan et al. Phys.Lett.B665:20-25,2008.
H. Avakian, INT, Nov 9
58
SIDIS (g*p->pX) x-section at leading twist
TMD PDFs
•Measure Boer-Mulders distribution functions and probe the polarized fragmentation function
•Measurements from different experiments consistent
H. Avakian, INT, Nov 9
59
Transverse force on the polarized quarks
Quark polarized in the x-direction with
kT in the y-direction
Force on the active quark right after scattering (t=0)
Interpreting HT (quark-gluon-quark correlations) as force on
the quarks (Burkardt hep-ph:0810.3589)
H. Avakian, INT, Nov 9
60
EIC: Kinematics Coverage
5 GeV
e
p
50 GeV
e’p+X
xF>0 (CFR)
EIC-MC
xF<0 ( TFR)
all
xF>0
z>0.3
EIC-MC
Major part of current particles at large angles in
Lab frame (PID at large angles crucial).
H. Avakian, INT, Nov 9
61
EIC medium energy
EIC@JLab
EIC@RHIC
•
Electron energy: 4-20 GeV
•
Proton energy: 50-250 GeV
–
•
•
Electron energy: 3-11 GeV
•
Proton energy: 20-60 GeV
–
More symmetric kinematics provides better
resolution and particle id
Luminosity: ~ 1033 cm-2 s-1
–
Main Features
•
More symmetric kinematics provides better
resolution and particle id
Luminosity: ~ 1034 cm-2 s-1
–
in range around s ~ 1000-10000 GeV2
in range around s ~ 1000 GeV2
•
Polarized electrons and light ions
•
Polarized electrons and light ions
•
90% of hardware can be reused
•
Potential upgrade with high-energy ring
Slides
are for
a “generic” US version of an EIC (5x50
or 4x60):
– longitudinal
and transverse
– longitudinal
and transverse
• Limited
R&D needs
polarized
> 70%)
• • Limited
R&D needsbeams (longitudinal and transverse,
• 3 interaction regions (detectors)
luminosities
of at least 1033
• • ? interaction
regions (detectors)
H. Avakian, INT, Nov 9
62
JETSET:Single particle production in hard scattering
LUND Fragmentation Functions
- Before
- After Target remnant
quark
Lund-MC should be modified to allow checks of sensitivity of
measurements to different effects related to the transverse structure
H. Avakian, INT, Nov 9
63
SIDIS kinematical plane and observables
Cross section is a function of
scale variables x,y,z
z
Target polarization
Beam polarization
U unpolarized
L long.polarized
T trans.polarized
sin2f moment of the cross section for
unpolarized beam and long. polarized target
H. Avakian, INT, Nov 9
64
Kaon production in SIDIS
FAST-MC
L
S
CLAS12
S*
EIC4x60
detected
s(p) = 0.05 + 0.06*p [GeV] %
Identification using the missing mass may be possible
H. Avakian, INT, Nov 9
65
Azimuthal moments with unpolarized target
quark polarization
H. Avakian,
JLab, Nov
INT,25Nov 9
66
Azimuthal moments with unpolarized target
quark polarization
H. Avakian,
JLab, Nov
INT,25Nov 9
67
SSA with unpolarized target
quark polarization
H. Avakian,
JLab, Nov
INT,25Nov 9
68
SSA with unpolarized target
quark polarization
H. Avakian,
JLab, Nov
INT,25Nov 9
69
SSA with long. polarized target
quark polarization
H. Avakian, INT, Nov 9
70
SSA with long. polarized target
quark polarization
H. Avakian, INT, Nov 9
71
SSA with unpolarized target
quark polarization
H. Avakian, INT, Nov 9
72
SSA with unpolarized target
quark polarization
H. Avakian, INT, Nov 9
73
Single hadron production in hard scattering
xF>0 (current fragmentation)
h
xF<0 (target fragmentation)
xF - momentum
in the CM frame
Target fragmentation
Current fragmentation
semi-inclusive
semi-exclusive
FF h
-1
Fracture Functions
h
h
DA
DA
h
M
exclusive
PDF
0
kT-dependent PDFs
PDF
GPD
1
xF
Generalized PDFs
Measurements in different kinematical regions provide complementary
information on the complex nucleon structure.
H. Avakian, INT, Nov 9
74
L production in the target fragmentation
L polarization in TFR provides information on
contribution of strange sea to proton spin
(ud)-diquark is a spin and
isospin singlet s-quark carries
whole spin of L L  uds
xF - momentum
xF(L)
in the CM frame
Study polarized diquark fracture
functions sensitive to the correlations
between struck quark transverse
momentum and the diquark spin.
EIC
CLAS12
Wide kinematical coverage of EIC would allow studies of
hadronization in the target fragmentation region (fracture functions)
H. Avakian, INT, Nov 9
75
Collins effect
p+
Simple string fragmentation
for pions (Artru model)
z
leading pion out of
page
L
r production may
produce an opposite
sign AUT
Fraction of r in
epX
% left from epX
asm
20%
40%
~75%
~50%
L
r
z
Leading r opposite to
leading p(into page)
hep-ph/9606390
Fraction of direct kaons may
be significantly higher than
the fraction of direct pions.
LUND-MC
H. Avakian, INT, Nov 9
76
K/K* and L/S separations
Detection of K+ crucial for separation of different final states (L,S,K*)
H. Avakian, INT, Nov 9
77
Sivers effect in the target fragmentation
A.Kotzinian
High statistics of CLAS12 will allow studies of kinematic
dependences of the Sivers effect in target fragmentation region
H. Avakian, INT, Nov 9
78
Q2
Hard Scattering Processes: Kinematics Coverage
Q2
HERA
collider experiments
H1, ZEUS (EIC)
10-4<xB<0.02 (0.3): gluons (and quarks)
in the proton
EIC
ENC
ENC
fixed target experiments
COMPASS, HERMES
 0.006/0.02<xB<0.3 : gluons/valence
and sea quarks
JLab/JLab@12GeV
 0.1<xB<0.7 : valence quarks
Study of high x domain requires high
luminosity, low x higher energies
H. Avakian, INT, Nov 9
79
Q2
Hard Scattering Processes: Kinematics Coverage
HERA
Q2
collider experiments
H1, ZEUS (EIC)
10-4<xB<0.02 (0.3): gluons (and quarks)
in the proton
EIC
ENC
ENC
EIC
ENC
JLab12
fixed target experiments
COMPASS, HERMES
 0.006/0.02<xB<0.3 : gluons/valence
and sea quarks
JLab/JLab@12GeV
 0.1<xB<0.7 : valence quarks
Study of high x domain requires high
luminosity, low x higher energies
H. Avakian, INT, Nov 9
80
hep:arXiv-09092238
H. Avakian, INT, Nov 9
81
TMDs: QCD based predictions
Large-x limit
Burkardt (2007)
Brodsky & Yuan (2006)
Large-Nc limit (Pobilitsa)
Do not change sign (isoscalar)
All others change sign
u→d (isovector)
H. Avakian, INT, Nov 9
82