An Update on Fermilab E906_ Drell-Yan Measurements of Nucleon by hcj

VIEWS: 3 PAGES: 30

									The Old Man and the Sea




                Donald Geesaman
Achievements and New Directions in Subatomic Physics
                 15 February 2010
In the quark model and QCD, it seems like the
valence quarks and glue get all the respect

 Valence quarks determine the charge and flavor of hadrons

 Seem to explain the magnetic moments.

 We thought, until 1990, that the valence quarks carried the spin

 New accelerators, like the JLAB 12 GeV upgrade get built to study high x
  quarks

 The glue dominates hadron structure at low x

 New accelerators, like the electron-ion collider are planned to study the
  glue.


                                                                              2
    Maybe the sea quarks will go away!
Motivated by desire to link to constituent quark or bag   It was then realized that
   models, the hope was that as some low scale, Q, of a
                                                          some valence-like sea
   few hundred MeV/c, valence-like quark distributions
   plus glue would describe the nucleon, and the sea      was needed.
   could be radiatively generated.                        GRV, ZPC53, 127(92)
 Gluck, Godbole, and Reya (Z. Phys. C, 66 (1989)




u             g
                              q


                             q
               u

      Then it was found that the sea was not flavor symmetric.

                                                                                      3
  Most of the information on the sea came from
  deep-inelastic lepton scattering, especially
  charged current neutrino experiments
   Q2 = (k-k’)2 = mass2 of the virtual boson
   x= Q2/(2m) is the fractional momentum nucleon carried by the
              parton
    = Ebeam- Escattered     y =  / Ebeam
   d
         l  qi  f i (x)
   dx   i


muon and electron scattering~               2 x(4 / 9[u  c  u  c ]  1 / 9[d  s  d  s ])
 charge current scattering ~               2 x[d  s  (1  y ) 2 (u  c )]
anti- c. c. scattering~                    2 x[u  c  (1  y ) 2 (d  s ))]
parity violating  scattering, F3~          2 x(d  s  u  c )
parity violating anti- scattering~         2 x(u  c  d  s )
 The high statistics  experiments are all done on nuclear targets
                                                                                                 4
    FNAL E866 Drell-Yan measurements on hydrogen and
    deuterium determined the x dependence of d-u
                                         Towell et al. Phys. Rev. D 64, 0522002 (2001)



    Small but very important                                         Q2=54 GeV2
    Since this is a flavor non-singlet
    quantity


          d ( x)  u ( x)dx  0
          1


          0


    is true at all scales.



 d ( x)  u ( x)dx  0.118  0.012
1


0



                                                                                    5
 The simplest explanation is the pion cloud


The proton spends part of its time as         LA-LP-98-
a neutron plus π+                                    56



|P> = α|uud> + β|udd> |uđ>


We know pion cloud effects are
important in quark models.




                                                          6
    Of course Tony Thomas and his collaborators knew
    all this. Indeed they invented much of it.
   1972 Sullivan
   1980 Cloudy Bag Model
    Pions have to be included to preserve
    chiral symmetry in bag or bag-like models

   1983 Tony used the calculated pionic
    content and measured DIS to conclude
    that the fraction of the momentum of the
    nucleon carried by pions was 5+/-1.5%
    and was consistent with a bag radius of
    0.87 +/-0.10 fm.

Even today this is not such a bad
   representation of

         x(d  u )
The problem is it also predicts the ratio

         d /u  5
as x goes to 1 from the charged and neutral
     pion Clebsch-Gordan coefficients
                                                       7
    Structure of the nucleon: What produces the
    nucleon sea?                                                     LA-LP-98-
   pQCD - Gluon splitting?                                                 56

   Meson Cloud? Chiral Solitons?
    Instantons?
   Models describe               well, but
    not           — pQCD becoming
    dominant????

                                                       Peng et al.




                                  No one has come up
                                  with a physical
                                  mechanism to make

                                        u d
                                                                                 8
A key seems to be the spin carried by the non-
singlet anti-quarks

             d ( x)  u ( x)dx  0.118  0.012
            1
E866
            0
Pion content – flavor non-singlet anti-quarks carry 0 net spin.
    Pions do affect the spin carried by the quarks through their
    interaction with the remnant baryon

Statistical Model - Bourelly and Soffer

        (d  u )  (d  u )
Instanton

        (d  u )  [5 / 3]( d  u )
Chiral quark-Soliton - Dresslar et al. EPJC18, 719 (2001) gives
    similar result.




                                                                   9
What are the correlations between the q and q pairs in the sea?


                                  g
                                             q
   Gluon 1-, 3S1
          Flavor neutral

                                             q

   Meson 0-, 1+




   Vacuum 0+, 3P0
         Flavor neutral




                                                                  10
What do the data tell us ?


   E866 - PR D64, 052002 (2001) Q2=54 GeV2

          d ( x)  u ( x)dx  0.118  0.012
         1


         0

   HERMES - PR D71, 012003 (2005)
         0.3

          (d  u )dx  0.048  0.057  .028
        0.023
                                                           To be compared with
   COMPASS- arXiv:0909.3729v1 Q2=3 GeV2                   0, -1, -5/3 * flavor asymmetry
        0.3

          (d  u )dx  0.052  0.035  .013
       0.004

   de Florian et al - PRL 101, 072001 (2008) Q2=10 GeV2
        1
                                                               3 σ from zero
         (d  u )dx  0.117  0.036
        0
                                                               2 σ from .197=Chiral soliton



                                                                                            11
COMPASS and HERMES Data




     1

      (d  u )dx  0.117  0.036
     0




                                       1

           DNS 2005                     (d  u )dx  -.03 to -.19
                                       0
                                       1

           DSSV 2008                    (d  u )dx  0.117  0.036
                                       0



                                                                         12
What does Tony say now?


Myhrer-Thomas picture of proton spin

   Relativistic valence quarks - orbital motion accounts for 35%

   quark-quark hyperfine interaction

   Pion cloud



Only the hyper-fine interaction could contribute to   d  u
so I believe the prediction is small.




                                                                    13
HERMES has a new slant on the strange quark distributions.
A. Airapetian et al Phys. Lett. B 666, 446 (2008)

Usually s(x)+sbar(x) ~ κ (ubar+ dbar) with κ~ 0.5

Best handle has been considered to be multi-muon events in neutrino scattering.

HERMES looks at polarized DIS on deuterium and compares inclusive with semi-
inclusive kaon multiplicities

        d 2 N DIS ( x)
                        U ( x, Q 2 )5Q( x)  2 S ( x)
          dxdQ2


        d 2 N K ( x)
         dxdQ 2
                                   
                      U ( x, Q 2 ) Q( x)  DQ ( z )dz  S ( x)  DSK ( z )dz
                                              K
                                                                                 

        Q( x)  u ( x)  u ( x)  d ( x)  d ( x)
        S ( x)  s ( x)  s ( x)
                                                                                     14
HERMES sees little strange quark content for x>0.1
and s(x)+sbar(x) ~ ubar(x)+dbar(x) at x< 0.03!



      A. Airapetian et al Phys. Lett. B 666, 446 (2008)   Q2=2.5 GeV2




                                                                        15
How is this consistent with years of neutrino multi-muon
data? ν + s → μ+ + c →μ-
    NUTEV, PRD 64 112006(2001)      CTEQ, JHEP 42, 89 (2007)




                                                               16
NuTeV Data Suggest Small Strange vs Anti-strange
Asymmetry
PRL 99, 192001 (07)




                                                   17
  Comparison of ubar+dbar-s-sbar with dbar-ubar
                                      x[d ( x)  u ( x)] vs 0.25 *HERMES x[u ( x)  d ( x)  s( x)  s ( x)]

  Based on the
  HERMES result and
  assuming the strange
  quark distribution
  represents the
  gluon-splitting
  induced distribution,
  the shape of the
  non-perturbative
x[u ( x)  d ( x)  s( x)  s ( x)]

  is similar to
    x[d ( x)  u ( x)]

                                                                                                               18
Nuclear corrections in charged lepton and neutrino
scattering are different

      Charged lepton Fe/D                Neutrino Fe/D




                                                         19
 Parton Distributions in Nuclei

 1984 – Parton distributions are
  different
   EMC effect – nucleon carries smaller
      fraction of momentum or changes




                                                                          Alde et al (Fermilab E772) Phys. Rev. Lett. 64 2479 (1990)
      structure
   Shadowing
 Expected large pion-cloud effects
 1990 – little change in sea quarks for
  x>0,1




                                      My one publication with Tony
                                      6th on his citation list
                                                                     20
                                                  average spacing at ρnm ~ 1.8 fm
Our visual images of a nucleus                    Radius of a nucleon      ~ 0.8 fm
                                                  average spacing at 3ρnm ~ 1.3 fm




                                           OR

“nucleons” held apart by short range repulsion
but even in 208Pb, half the nucleons are in the surface
                                                          Remember 1983 Thomas result
                                                          favored a bag radius of 0.9 fm 21
We want to describe a nucleus

   Hadronic Description                          Pure QCD Description
     – exemplified by ab initio calculations        – what are the clusters of quarks in a
       with potentials                                nucleus?
          •   NN                                    – know the parton distributions change
          •   NNN + NNNN +                               • EMC effect
          •   Bare form factors                          • shadowing
          •   Meson exchange currents                    • x>1


   Past two decades have shown this is           One problem is always whether our
    remarkably successful                          description of a bare proton is good
                                                   enough. The second is how to
                                                   actually calculate many body effects
                                                   beyond mean field?




                                                                                          22
  Drell-Yan scattering:
  A laboratory for sea quarks




                                                   xtarget   xbeam

Use a proton beam: primarily u quarks at high x. Detector
   acceptance chooses xtarget and xbeam.
 Fixed target  high xF = xbeam – xtarget
 Valence Beam quarks at high-x.
   (e2 u)/(e2d) > 8 Dominated by u quarks
 Sea Target quarks at low/intermediate-x.

                                                                     23
         Advantages of 120 GeV Main Injector
               The (very successful) past:                             The future:
                 Fermilab E866/NuSea                                Fermilab E906
        Data in 1996-1997                               Data taking 2010-2012
        1H, 2H, and nuclear targets                     1H, 2H, and nuclear targets

        800 GeV proton beam                             120 GeV proton Beam




       Cross section scales as 1/s
         – 7 x that of 800 GeV beam
       Backgrounds, primarily from J/ decays
        scale as s                                                                      Tevatron
                                                                                        800 GeV
         – 7 x Luminosity for same detector rate as                Main
           800 GeV beam                                           Injector
                                                                 120 GeV
                  50 x statistics!!


                                                                                                   24
 Fermilab E906/SeaQuest Collaboration
             Abilene Christian University                          Ling-Tung University
    Donald Isenhower, Mike Sadler, Rusty Towell,                     Ting-Hua Chang
                    Shon Watson
                                                             Los Alamos National Laboratory
               Academia Sinica                          Gerry Garvey, Xiaodong Jaing, Mike Leitch,
   Wen-Chen Chang, Yen-Chu Chen, Da-Shung Su              Ming Liu, Pat McGaughey, Joel Moss

            Argonne National Laboratory                           University of Maryland
   John Arrington, Don Geesaman*, Kawtar Hafidi,     Prabin Adhikari, Betsy Beise, Kazutaka Nakahara
     Roy Holt, Harold Jackson, David Potterveld,
   Paul E. Reimer*, Josh Rubin, Patricia Solvignon               University of Michigan
                                                          Wolfgang Lorenzon, Richard Raymond
               University of Colorado
                    Ed Kinney                                              RIKEN
                                                       Yuji Goto, Atsushi Taketani, Yoshinori Fukao,
       Fermi National Accelerator Laboratory                         Manabu Togawa
          Chuck Brown, Dave Christian
                                                                    Rutgers University
               University of Illinois                Lamiaa El Fassi, Ron Gilman, Elena Kuchina, Ron
        Naomi C.R Makins, Jen-Chieh Peng                        Ransome, Elaine Schulte

                          KEK                                   Texas A & M University
                    Shin'ya Sawada                            Carl Gagliardi, Robert Tribble

                 Kyoto University                     Thomas Jefferson National Accelerator Facility
             KenIchi Imai, Tomo Nagae                                Dave Gaskell
*Co-Spokespersons                                             Tokyo Institute of Technology
                                                           Toshi-Aki Shibata, Yoshiyuki Miyachi


                                                                                                       25
Projected errors on ratios of D to H




Errors on ratio ~ 1% until
statistics become a factor.

The absolute cross section
on deuterium measures

      d u
Errors limited by beam
normalization and acceptance
~ 5%
                                       26
Does deuterium structure affect the results at
higher x




                                                 27
Structure of nucleonic matter:

 Nucleon motion in the nucleus
  tends to reduce parton
  distributions – f(y) peaked below
  y=1.
 Rescaling effects also reduce
  parton distribution for x>0.15
 Antiquark enhancement expected
  from Nuclear Pions.

 This data also constrains the
  maximum effects for deuterium.




                                      28
Summary
 The origin and structure of the sea remain critical themes in the physics of
  the nucleon and nucleus

 We need to push to higher x values and E906/SeaQuest is especially well
  suited for this. We start this summer and run for two years.

 The other really key measurement is improved precision in the spin carried
  by the sea quarks and the spin-correlations in the sea.
   – COMPASS, RHIC, J-PARC, JLAB 12 GeV

 This is difficult and may require the next generation of polarized Drell-Yan
  experiments

 Whatever we measure, Tony Thomas will have thought of it first and helped
  stimulate the experiments

 And there is a chance, he may even have got it right.
                                                                                 29
The models all have close relations between
antiquark flavor asymmetry and spin




• Statistical Parton Distributions   u ( x)  d ( x)  d ( x)  u ( x)
                                                                           30

								
To top