New Measurement of the EMC Effect in Light Nuclei and at Large x

1
New Measurement of the EMC Effect in Light Nuclei
and at Large x

• Dave Gaskell
• Jefferson Lab
• OU Institute of Nuclear and Particle Physics
  Seminar
• October 9, 2007

2
Outline

• Deep Inelastic Scattering and Nuclear Structure
  Functions
• The EMC Effect
• Review of Measurements
• Limitations in the Data
• JLab Experiment E03-103
• Motivation
• Preliminary results

3
Deep Inelastic Scattering
Cross section for inclusive lepton (electron)
scattering
In the limit of large Q2, structure functions
scale
4
F2 and Parton Distributions

• F2 interpreted in the quark-parton model as the
  charge-weighted sum over quark distributions
• At finite Q2, F2 not Q2 independent ? scaling
  violations can be predicted in pQCD
• At fixed x, scaling can be tested via logarithmic
  derivative of F2 w.r.t. to Q2
• In addition, corrections due to the finite mass
  of the nucleon lead to further scaling violations
  ? these can be partially accounted for by
  examining data in terms of Nachtmann variable, x

5
Quarks in the Nucleus

• Typical nuclear binding energies MeV while DIS
  scales ? GeV
• Naïve expectation
• More sophisticated approach includes effects from
  Fermi motion
• Quark distributions in nuclei not expected to be
  significantly different (below x0.6)

Bodek and Ritchie PRD 23, 1070 (1981)
6
First Measurement of the EMC Effect

• First published measurement of nuclear dependence
  of F2 by the European Muon Collaboration in 1983
• Observed 2 mysterious effects
• Significant enhancement at small x ? Nuclear
  Pions! (see my thesis)
• Depletion at large x ? the EMC Effect
• Enhancement at xlt0.1 later went away

Aubert et al, Phys. Lett. B123, 275 (1983)
7
Subsequent Measurements

• Initial observation of the nuclear dependence
  of F2 confirmed at SLAC and by BCDMS
  collaboration
• Later measurements saw smaller enhancement at
  x0.1-0.2 and a depletion at very small x ?
  shadowing

E139 (Fe) EMC (Cu) BCDMS (Fe)
8
EMC Effect Measurements at Large x
SLAC E139

• SLAC E139 most extensive and precise data set for
  xgt0.2
• Measured sA/sD for A4 to 197
• 4He, 9Be, C, 27Al, 40Ca, 56Fe, 108Ag, and 197Au
• Best determination of the A dependence
• Verified that the x dependence was roughly
  constant
• Data set could be improved with
• Higher precision data for 4He
• Addition of 3He data
• Precision data at large x

9
Explaining the EMC Effect

• Non-exotic models
• Fermi motion ? reproduces rise at large x
• Binding
• Fermi motion binding nuclear pions
• Exotic models
• Multiquark clusters
• Dynamical rescaling
• All of these models have varying degrees of
  success describing the EMC Effect in certain x
  regions none can describe the effect in its
  entirety

?
10
Binding and Nuclear Pions

• Start with a realistic description of nucleons
  in the nucleus
• ? Use a spectral functions rather than simple
  Fermi gas
• Start with convolution picture
• ? Allow virtual photon to scatter from quarks in
  pions in the nucleus
• Fair agreement is achieved at large x including
  nuclear pions improves agreement at lower x

Benhar, Pandharipande, and Sick Phys. Lett. B410,
79 (1997)
11
Multiquark Clusters

• Multiquark cluster model assumes that, in nuclei,
  quarks may combine into clusters that include
  more than 3 quarks
• Nuclear structure function is a convolution over
  contribution from nucleons (F2N) and contribution
  from 6 quark clusters (F26)
• Requires F2N ? F26 to get EMC effect

K.E. Lassila and U.P. Sakhatme Phys. Lett. B209,
343 (1988)
12
EMC Effect Model Issues

• Conventional nuclear physics based explanations
• Fermi motion alone clearly not sufficient
• Early attempts to include effects from binding
  were fairly simplistic
• Even more sophisticated approaches (spectral
  function) fail unless one includes nuclear
  pions
• Contributions from nuclear pions of the order
  typically assumed inconsistent with nuclear
  dependence of Drell-Yan
• Exotic effects
• Multiquark clusters, dynamical rescaling
  calculations often ignore contributions binding
• In both cases, calculations often only successful
  in particular x regions, or only explain part of
  the effect

13
Nuclear Physics and the EMC Effect
Existing data at large x and small A limited

• Fermi motion and binding clearly play a role in
  describing nuclear structure functions at large x
• Contributions from these effects extend to low x
  as well ? x0.2
• Binding calculations must be tested against
  nuclear structure ratios at large x where ratios
  are dominated by Fermi motion
• Small A measurements also important ? nuclear
  structure uncertainties small

14
JLab Experiment E03-103

• Measurement of the EMC Effect in light nuclei
  (3He and 4He) and at large x ?spokespersons DG
  and J. Arrington, graduate students J. Seely and
  A. Daniel
• 3He, 4He amenable to calculations using exact
  nuclear wave functions
• Large x dominated by binding, conventional
  nuclear effects
• Ran in Hall C at JLab
  summer and
  fall 2004
  (w/E02-109, xgt1)
  
  
• A(e,e) at 5.77 GeV
• Targets H, 2H, 3He,
  4He,
  Be, C, Al, Cu,
  and Au
• Six angles to measure
  Q2 dependence

15
E03-103 Details

• High Momentum Spectrometer (HMS) used for e-
  detection
• Superconducting QQQD configuration
• Typical detector stack for event ID (drift
  chambers, threshold gas Cerenkov, calorimeter)

Short Orbit Spectrometer (SOS) used for
background measurements
Hall C
HMS
SOS

• Solid target data, H,D data taken during same
  running period
• 3He, 4He data not taken at same time as D ?
  Carbon data at all settings to check time
  dependence of yields

Target chamber
16
E03-103 Analysis Status
Analysis 95 complete ?Extraction of
pre-corrected cross sections finalized ?Iterating
input model for radiative corrections ?Evaluating
systematic errors All results I will show are
preliminary, but I do not expect them to change
17
Challenges at Low Energy

• Most previous measurements of the EMC Effect were
  made at high lepton energies ? 10s to 100s of
  GeV
• JLab tops out at 6 GeV this makes life more
  complicated in several ways
• Kinematics
• At low energy, large Q2 needed to access scaling
  requires larger electron scattering angle
• Even at large angle, significant fraction of our
  data in canonical resonance region
• Large angle results in experimental
  difficulties
• Big contribution from charge symmetric processes
• Cannot ignore acceleration of electrons in
  Coulomb field of the nucleus

18
Charge Symmetric Processes
Pion electroproduction yields charge symmetric
background
q40 degrees
q50 degrees
Positrons measured in HMS at identical kinematics
yield directly subtracted
19
Coulomb Corrections

• Initial (scattered) electrons are accelerated
  (decelerated) in Coulomb field of nucleus with Z
  protons
• Not accounted for in typical radiative
  corrections
• Usually, not a large effect at high energy
  machines not true at JLab (6 GeV!)
• E03-103 uses modified
  Effective Momentum
  
  Approximation (EMA),
  
  Aste and Trautmann,
  
  Eur, Phys. J. A26,
  167-178(2005)
• E ? ED, E?ED
• D -¾ V0, V0 3a(Z-1)/(2rc)
• EMA tested against DWBA
• calculation for QE scattering
• ? application to inelastic
• scattering appropriate?

20
Deep Inelastic Scattering at low W
Canonical DIS regime Q2gt1 GeV2 AND
W2 gt 4 GeV2 ? This requirement ensures that we
are scattering from quarks in the nucleon or
nucleus

• At JLab, we have access to large Q2, and W2gt4 up
  to x0.6
• At xgt0.6, we are in the resonance region ?
  excited, bound states of the nucleon
• Are we really sensitive to quarks in this regime?

21
Quark-Hadron Duality

• High energy processes?quarks and gluons
• Low energy processes ? hadronic degrees of
  freedom
• Low energy effective picture is dual to high
  energy picture if enough degrees of freedom are
  included
• Initially observed in electron scattering (Bloom
  and Gilman in 70s) new JLab confirms with high
  precision
• Structure function scaling seen at high Q2 is a
  good average over resonance region at low Q2

I. Niculescu et al., PRL851182 (2000)
22
Quark-Hadron Duality in Nuclei
J. Arrington, et al., PRC73035205 (2006)

• Free nucleon
• average over resonance region DIS scaling limit
• Bound nucleon
• Fermi motion does the averaging for us
• Resonances much less prominent in nuclear
  structure functions
• Nuclear structure functions appear to scale to
  lower Q2 than their free nucleon counterparts
  with no explicit resonance averaging

23
EMC Effect in Resonance Region
JLab E89-008 Q2 4 GeV2 1.3ltW2lt2.8 GeV2 data in
the resonance region ? In region of overlap
agrees well with DIS data
E03-103 data are at higher Q2, will test scaling
with precise measurement of Q2-dependence
J. Arrington, et al., PRC73035205 (2006)
24
Carbon EMC Ratio and Q2 Dependence
E03-103 Results
Q22.3 GeV2 W21.9 GeV2
Preliminary
Q23.6 GeV2 W22.4 GeV2
W2 gt 1.5 GeV2
Small angle, low Q2 ? no scaling above x0.55
25
Carbon EMC Ratio and Q2 Dependence
E03-103 Results
Q22.3 GeV2 W21.9 GeV2
Preliminary
Q25.7 GeV2 W23.3 GeV2
W2 gt 1.5 GeV2
At larger angles (Q2) ? ratio appears to scale
26
Carbon EMC Ratio and Q2 Dependence
Preliminary
W2 gt 1.5 GeV2
Good agreement with SLAC e139 data
27
Detailed Scaling Tests Fixed x
Scaling test
SLAC
E03-103
Curves fit to large Q2 SLAC data ? compared to
E03-103 at lower Q2
At x0.8 (x0.88 at 50 degrees), F2 for deuterium
scales down to Q24 GeV2, even though well below
W24 GeV2
28
EMC Effect in 4He
JLab results consistent with SLAC E139 ? Improved
statistics and precision
29
EMC Effect in 4He

• JLab results consistent with SLAC E139
• Improved statistics and precision
• Challenging existing models none shown agree
  with data over full range

Cloet private communication Smirnov Burov,
Molochkov and Smirnov Phys. Lett. B 466, 1
(1999) Benhar private communication
30
Carbon to 4He Comparison
Globally magnitude of the EMC Effect very
similar ?In this context, 4He behaves like a
real nucleus
Preliminary
Some hint of difference in shape, but hard to
tell with existing errors
31
Isoscalar Corrections

• When extracting structure function ratios, want
  to compare a nucleus with ZN protons and
  neutrons to deuterium (Z1, N1)
• In some cases, nature is kind enough to provide
  this for us
• As A gets large, typically have more neutrons
  than protons
• sA/sD must be corrected for non-isoscalarity of
  nucleus

32
E139 Isoscalar Correction
A. Bodek, et al., PRD201471 (1979)
1-0.8x
sn/sp extracted from sD/sp
33
E03-103 Isoscalar Correction?

• For initial comparisons, desirable to use same
  sn/sp as SLAC e139

• Ideally, we could show
• (sD/sp)E03103(sD/sp)SLAC
• implying
• (sn/sp)E03103(sn/sp)SLAC
• Unfortunately, E03-103
• proton data at large x in
• resonance region
• ? D/p ratio for xgt0.7
• exhibits a lot of structure

E01-103 data at 40 degrees
Preliminary
Fit to NMCSLAC data
34
Smeared sn/sp

• Cannot directly compare E01-103 D/p to SLAC to
  demonstrate n/p the same
• Even worse, free n/p at our kinematics also
  likely shows a lot of structure

Smeared E03-103 agrees with free SLAC

• However, we are
• correcting nuclei dont
• want free n/p anyway
• ? ideally wed like
• bound n/p for relevant
• nucleus
• This is difficult start with
• bound n/p in deuterium
• in first approximation

35
Effect of Isoscalar Corrections

• SLAC param. (1-0.8x)
• CTEQ
• NMC fit

Isoscalar correction applied to data
Au
3He
36
3He EMC Ratio
Large proton excess correction
Good agreement with HERMES in overlap region
Preliminary
37
3He EMC Ratio HERMES Comparison
Fair
Good agreement with HERMES in overlap region
Preliminary
HERMES uses different param. for isoscalar
correction!
38
3He EMC Ratio
All calculations shown use convolution formalism
at some level
Preliminary
Melnitchouk Afnan et.al. PRC68 035201
(2003) Smirnov Molochkov and Smirnov
arXivnucl-th/9904050 (1999) Benhar private
communication
39
EMC Measurements for Heavy Nuclei
E03-103 data corrected for coulomb distortion
Shape independent of A ? especially at large x
(x)
Preliminary
Nachtmann variable
40
EMC Measurements for Heavy Nuclei
E03-103 data corrected for coulomb distortion
E03-103 Copper data agrees with Coulomb Corrected
Fe data from SLAC
Preliminary
41
EMC Measurements for Heavy Nuclei
E03-103 data corrected for coulomb distortion
Gold data also show nice agreement
Preliminary
42
Nuclear Dependence of the EMC Effect

• After correction for Coulomb effects, e139 and
  E03-103 data show nice agreement
• Original e139 paper parameterized in terms of A
  or rnuclear density assuming uniform sphere of
  radius Re (r3A/4pRe3)

x0.6
E03-103 data with CC SLAC data with CC
Preliminary
43
Nuclear Dependence of the EMC Effect

• At SLAC kinematics, Coulomb corrections are
  smaller but still non-zero (1-2 for gold)
• Has potential impact for extrapolation to nuclear
  matter

x0.6
E03-103 data with CC SLAC data no CC
44
Extrapolation to Nuclear Matter

• Re-analysis of extrapolation to nuclear matter,
  technique described in PLB274 (1992)
• Local Density Approximation effect of nuclear
  medium depends on density at interaction point

• Contributions to cross
• section scale like A in the
• nuclear volume, scale
• like A2/3 at the surface
• s a1Aa2A2/3 ?
• s/Aa1a2A-1/3
• Extrapolate to A8 for
• nuclear matter

45
EMC Effect in Nuclear Matter
Solutions to realistic nucleon potentials
straightforward for small A or A? ? NM
calculations potentially simpler
Extrapolation only includes A 12
Sick and Day Phys. Lett. B274, 16 (1992)
46
EMC Effect in Nuclear Matter
New analysis with updated data set and Coulomb
corrections (P. Solvignon, ANL)
Sick and Day Phys. Lett. B274, 16 (1992)
47
EMC Effect in Nuclear Matter
New analysis with updated data set and Coulomb
corrections (P. Solvignon, ANL)
Global fit including E03-103 data
Sick and Day Phys. Lett. B274, 16 (1992)
48
E0-103 Impact

• Measurements from light nuclei
• First measurement of EMC effect in 3He above
  x0.4
• Improved 4He measurement
• These results will serve as excellent testing
  ground for convolution calculations ? virtually
  no uncertainty in nuclear wave function
• Measurements at large x
• Assuming one believes in scaling for W2lt4 GeV2,
  our heavy target data improves the precision for
  xgt0.75 where Fermi motion dominates
• Both of these combined should help settle to what
  degree conventional nuclear physics plays a role
  in the EMC effect
• Once this is understood, we are in a better
  position to quantify to what extent we must
  introduce additional mechanisms

49
Future of the EMC Effect

• Will E03-103 data settle all the questions
  relating to modification of quark structure
  functions in nuclei?
• No
• What else is there to learn?
• Flavor dependence ? u(x) changed in the same way
  as d(x)?
• Anti-quarks ? how the sea quarks are affected
• Spin dependence ? how will the polarized quark
  distributions change in the nucleus?

50
Flavor Dependence of the EMC Effect
Semi-inclusive DIS can be used to probe the
flavor dependence of reactions in DIS ? e.g.
polarized quark distributions

• Can semi-inclusive DIS be used to probe flavor
  dependence of EMC effect?
• Hadron attenuation effects make this difficult
• Need large n or precise knowledge of medium
  modification of fragmentation process
• Perhaps at electron ion collider?

51
A Dependence of Anti-quark Distributions

• Drell-Yan proccess sensitive
• to anti-quark distributions in
• the target
• E772 measured no A
• dependence over limited x
• range, with limited precision
• E906 will measure up to x0.4

D.M. Alde et al., PRL64 2479 (1990)
52
Polarized EMC Effect
New calculation describes EMC Effect in terms
convolution medium modification to quark
distributions in nucleon
Cloet, Bentz, and Thomas Phys. Lett. B 642 210
(2006)

• Same model predicts significant modification to
  polarized quark distributions in nucleus
• Experimentally difficult to define cannot
  polarize ALL the nucleons in a nucleus
• In addition, asymmetries will be rather small

53
Summary

• The EMC effect has been with us for more than
  years yet there is no clear consensus regarding
  the source of the effect
• A minimal ingredient in any model that claims to
  describe the EMC effect is conventional nuclear
  physics, i.e., binding and Fermi motion
• Once these are under control, we can explore the
  contributions from exotic phenomena
• E03-103 provides new data that will help
  constrain to what degree these conventional
  effects play a role
• Theres still more to learn regarding the flavor
  dependence and spin dependence, and to what
  degree sea quarks are affected

54
Backup
55
JLab at 12 GeV

• Measurement at 11 GeV would allow W2 GeV in
  crossover region
• A reasonable large x measurement with a variety
  of targets could be done relatively quickly (on
  the order of weeks of beam)
• No new equipment needed HMS only!

56
Simple Description of EMC Effect at Large x

• Nuclear structure function modeled as a simple
  convolution of nucleon distribution and F2
• Expanding F2(x/y) in a Taylor series
• At large x, y is biased to values larger than 1,
  hence

57
Calculations of the EMC Effect at Large x

• Large x dominated by nucleon Fermi motion
• For Agt12, Fermi momentum changes very little with
  A shape at large x should be relatively
  independent of nucleus
• Calculation of Oset et al. (interacting Fermi sea
  p r) confirms this intuitive expectation

58
Large x Crossover from E139

• Simple by hand estimate of crossover from SLAC
  data
• Gross and Liuti (PRC45 1374 1992) predict
  significant change to large x crossover with A
• Existing data cannot differentiate between
  Marco/Gross