1 Simonetta Liuti University of Virginia Structure of Nucleons and Nuclei Workshop Como, June 10 th- 14 th , 2013 N&N-Structure Simonetta Liuti Generalized TMDs sented by Osvaldo Gonzalez Hernandez - Torin
Jan 18, 2016
1
Simonetta LiutiUniversity of Virginia
Structure of Nucleons and Nuclei Workshop
Como, June 10th-14th, 2013
N&N-StructureSimonetta Liuti
Generalized TMDs
presented by Osvaldo Gonzalez Hernandez - Torino
N&N-StructureSimonetta Liuti 2
In collaboration with:
Aurore Courtoy (Liege U.)Gary Goldstein (Tufts U.)Osvaldo Gonzalez Hernandez (INFN Torino since Fall 2012)
Graduate StudentsKunal Kathuria (U.Va.)Evan Askanazi (U.Va.)Abha Rajan (U.Va.)
Question of what components do Lq and measure(Ji, Xiong, Yuan, 2011, 2012, Burkardt 2012)
N&N-StructureSimonetta Liuti 3
A Sum Rule was constructed which identified components of the Energy Momentum Tensor (EMT) with the Angular Momentum carried by quarks and gluons. (Jaffe&Manohar (JM))
A Saga in Several Episodes: Developing a Sum Rule for Angular Momentum
1990
N&N-StructureSimonetta Liuti 4
N&N-StructureSimonetta Liuti 5
Partonic picture:work directly in A+=0 gauge quark and gluon spin components are identified with the n=1 moments of spin dependent structure functions from DIS, ΔΣ and ΔG.
ΔGΔΣ
New Relation (X.Ji)1997
q
p’+= p –Δ
q'=q+Δ
p+q
P’
e
p+
New processes (DVCS …) were thought of, whose structure functions – the GPDs - admit n=2 moments that were identified with the (spin+OAM) quark and gluon components of the SR P
H,E
N&N-StructureSimonetta Liuti 6
N&N-StructureSimonetta Liuti 7
Lots of work followed in a series of papers by Wakamatsu, Leader and collaborators, Chen et al., Hatta, and Burkardt to understand the origin of this discrepancy.
N&N-StructureSimonetta Liuti 8
Conceptual issues: “What type of information can one obtain?”The goal is to give a partonic interpretation of the observables
Once a scheme, whichever, is defined to connect with experiment, we stick to it. Corollary: If Lcan is not observable, does it exist?
Practical issues: “how can one extract information from experiment”?Example: GPDs are hard to extract because they need to be “deconvoluted” from Compton Form Factors…
Both points pose important theoretical problems.
We would like to do some phenomenology
N&N-StructureSimonetta Liuti 9
p’p
x,kT’
bin
GPD
bout
x,kT
N&N-StructureSimonetta Liuti 10
zT
p’p
bout
bin
TMD
x,kT
x,kT
p’=p, kT’=kT
N&N-StructureSimonetta Liuti 11
(using constraints from Jlab flavor separated form factor data, G. Cates et al, 2011)
Transverse coordinate density distributions
PauliDirac
dd
u
u
GPD
What do we obtain (or wish to obtain) from experiment
N&N-StructureSimonetta Liuti 12
TMDs
Calculations done using reggeized diquark model,O. Gonzalez et al.,arXiv:1206.1876
What do we obtain (or wish to obtain) from experiment
Bacchetta, Conti Radici
Gonzalez et al.
Qo2≈ 0.3 GeV2
d
u
This needs to be evolved to the scale of the data(Collins, Rogers, Boglione, Prokudin, …)
N&N-StructureSimonetta Liuti 13
Unintegrated GPDs GTMDs
What partonic configurations do they correspond to?
Fourier Transform wrt ΔT Wigner Distributions
Notice! Two transverse momenta, simultaneously present
N&N-StructureSimonetta Liuti 14
zT
p’p
GTMD
b
GTMDs correlate partonic configurations with both:• a shift in transverse position from the initial to final state zT
• an average transverse position b.
averageshift
When can these configurations exist in the “impulse approximation”, and when do we need to introduce FSI?
N&N-StructureSimonetta Liuti 15
How would we observe GTMDs?
q, Λγ
k’= k –Δ. λ’
q’=q+Δ, Λγ’
k, λ
p’= k-Δ, Λ’
p, Λ
p, Λ
u-channel two body scattering
k’= k –Δ. λ’ k, λ
ϑ
N&N-StructureSimonetta Liuti 16
k’CM= kCM –Δ. λ’
kCM, λ
p’CM= kCM-Δ, Λ’
pCM, Λ
ϑ This angle determined by ΔT
z
y
x
CoM motion determined by
Transverse Plane
N&N-StructureSimonetta Liuti 17
In CoM the scattering happens in one plane therefore there is only one transverse direction defined by ΔT
In CoM frame ΔT and individual partons kT are parallel to one another
Average kT describes the motion of the CoM
N&N-StructureSimonetta Liuti 18
λ'λ
same as GPDs
“New” terms can exist
GTMDs in Cartesian BasisMeissner, Metz, Schlegel, JHEP 2009
N&N-StructureSimonetta Liuti 19
Spin(Helicity/Transversity) Basis
Quark-proton helicity amplitude
N&N-StructureSimonetta Liuti 20
Chiral Even Sector
G14
F11
2F13-F11
2G13-G14
Us Helicity amps contentMetz
N&N-StructureSimonetta Liuti 21
Can the new GTMDs F14 and G11 be defined at leading twist?
In two-body scattering these terms are Parity violating(valid in CM frame, demonstration in general frame hinges on appropriate Lorentz Transformation)
In a quark-target model only H, H are present~
N&N-StructureSimonetta Liuti 22
Why are they important: the idea has been emerging that canonical OAM can be related to F14
Ji, Xiong and Yuan, 2012
Lorce and Pasquini, 2011
N&N-StructureSimonetta Liuti 23
However F14 drops out of the observables
Not because it contains “even more information on the structure of the proton than what can be measured”, but for a good reason…
N&N-StructureSimonetta Liuti 24
The demonstration is rather technical
Out of all the functions that parameterize the generalized correlator (MMS’09)…
… the matrix elements for the “type 3” ones transform the opposite way underParity. These define F14
N&N-StructureSimonetta Liuti 25
Summary so far…
We confirm that canonical AM cannot be measured if defined as
because it corresponds – in two body scattering – to a combination of helicity amplitudes that violates Parity
N&N-StructureSimonetta Liuti 26
Consider now the angular momentum sum rule including twist 3 contributions Belitsky and Mueller (2000), Polyakov et al. (2000), Hatta (2011)
-Jq Sq-Lq
Twist 3 decomposition of hadronic tensor
OAM is a twist three contribution
N&N-StructureSimonetta Liuti 27
In order to describe GTMDs including partonic interactions, i.e. moving out of a 2 body scattering picture, we need a situation whereeven in the CoM, one can define two independent transverse vectors as if there were a third particle, so that the additional combinations giving rise to “LU”-type contributions can exist
N&N-StructureSimonetta Liuti 28
Twist 3 GTMDs
spin flip
spin non flipMMS’09
N&N-StructureSimonetta Liuti 29
N&N-StructureSimonetta Liuti 30
Finally…a prerequisite for all these calculations is that we use a model whose parameters are constrained by a quantitative fit to all data which are relevant for GPDs (Form Factors, DIS and DVCS)
N&N-StructureSimonetta Liuti 31
Sections of Wigner distributions: Completely Unpolarized Case
kT
u quark
N&N-StructureSimonetta Liuti 32
All together: integrated over x and kT (this is a fit to real data: DVCS+G.Cates et al)
d quarku quark
N&N-StructureSimonetta Liuti 33 APS Denver 4/12/13Simonetta Liuti 33
Sections of Wigner distributions: Transversely Polarized Case
kT
u quark
N&N-StructureSimonetta Liuti 34
All together: integrated over x and kT
Δ1Eu - Δ1Ed
N&N-StructureSimonetta Liuti 35
Twist 3 component G2
N&N-StructureSimonetta Liuti 36
Summary of this part and Open QuestionsG2 can be measured directly, and it does have a partonic interpretation, if one views it as a twist 3 quantity.
We can distinguish in models a WW part defining OAM, and a genuine twist 3 part.
Canonical angular momentum cannot be measured directly, but it can be linked to G2 through M. Burkardt’s “force-type” correction
We need to devise experiments that are sensitive to this contributions (they might be already in the available data!)
Question of L vs. T interpretation
We focused mainly on quarks, question of gluon component
One way to shed light on these questions is to look at deuteron
N&N-StructureSimonetta Liuti 37
Spin 1 systems, due to
1)The presence of additional L components (D-waves) 2)Isoscalarity
provide a crucial test the working of the angular momentum sum rules
Nucleon DVCS
Deuteron
HISO=Hu+Hd EISO=Eu+Ed
λN
λq
λ
z=pN+/P+
D
What are the quark and gluon angular momenta in the deuteron?
N&N-StructureSimonetta Liuti 38
Sum Rules in Deuteron (K.Kathuria)
OAM
Momentum
Spin 1/2
Spin 1/2
Form factors from the respective Energy Momentum tensors, Tμν
N&N-StructureSimonetta Liuti 39
If f++(z) = f+0(z)=δ(1-z) then H2=H+E
F1+F2= GM GM
Differently from the nucleon, in the transverse case, we are not finding the same relation (other GPDs describing charge and tensor component enter…). More details later…
Longitudinal
Transverse Deuteron
DeuteronNucleon
N&N-StructureSimonetta Liuti 40
How does Ji sum rule differ from JM in the deuteron?
Effect of evolution
Models are important Model calculations of L with w.f.s’ seem to lead to similar conclusions as M.Burkardt, more to explore here… avenue to compare different schemes?
Using GPDs from Goldstein, Gonzalez, SL, PRD84
N&N-StructureSimonetta Liuti 41
Nuclear effect much larger than in unpolarized scattering
Needs to be treated systematically…N&N-StructureSimonetta Liuti 42
Observables: DVCS from deuteron
subleading
Can the deuteron help us understand the role of gluon OAM?(Brodsky, Gardner, 2006)
By connecting Lg to SSA in
≈ 0
Both L q and Lg contribute! Since Lq disappears because of isospinsymmetry, if AUT
π is 0 then Lg is 0
N&N-StructureSimonetta Liuti 43
Two ways to go beyond the two body scattering scenario:
312
1
2 3
Our model for fracture functions (work in progress…)
scattered quark
outgoing qqqq-bar
Initial proton
h
q-bar frag.
BB
AA
N&N-StructureSimonetta Liuti 46
GG, S. Liuti and O. Gonzalez, GPDS, THEIR RELATIONSHIPS WITH TMDS & RELATED TOPICSProceedings of QCD-N12, Bilbao, Spain - 2012
N&N-StructureSimonetta Liuti 47
ConclusionsWe conclude that the distribution of an unpolarized quark in a longitudinally polarized nucleon does indeed measure OAM
We provide a theoretical basis for the “simple moment” of Wigner distributions, connecting to twist 3 GPDs Belitsky, Mueller, Polyakov, Hatta,…
Now one can think seriously of observables (DV processes including twist 3 GPDs)
Our argument stresses the fact that in order to measure OAM one has to go beyond a two body scattering picture
Transverse vs. Longitudinal sum rule? Gluon components? we should look into Spin 1 systems also
N&N-StructureSimonetta Liuti 48
G. Goldstein, O. Gonzalez Hernandez, S.Liuti, J.Phys. G, 39 (2012) 11500G. Goldstein, O. Gonzalez Hernandez, S.Liuti, Phys. Rev. D84, 034007 (2011)S.K. Taneja, K. Kathuria, S.Liuti, G. Goldstein, Phys. Rev. D86 (2012)G. Goldstein, O. Gonzalez Hernandez, S.Liuti, K. Kathuria, arXiv:1206.1876, subm. PRCG. Goldstein, O. Gonzalez Hernandez, S.Liuti, K. Kathuria, “Proceedings of 4th QCD Evolution Workshop”, Jefferson Lab, May (2012)
N&N-StructureSimonetta Liuti 49
Backup
N&N-StructureSimonetta Liuti 50
Sections of Wigner distributions: Mixed Transverse/Longitudinal Basis