The Strange Vector Current of the Nucleon

Forward Angle Experiment

Jianglai LiuUniversity of Maryland

Strangeness, very briefly Parity violation as an experimental probe The G0 experiment Emerging physics picture

Fermi Lab seminar, 11-15-2005

Strangeness in the Nucleon?

Quark models:Only u and d quarks in nucleons. No strangeness!

Quark-antiquark pairs and gluons make up the QCD vacuum (“sea”).

ss are “virtual pairs”, so the net strangeness is zero.

s and s might not have identical distributions. So strangeness might manifest locally. Analogous to the charge distribution in neutron!

“Full QCD description”

What’s the Big Deal of Strangeness?

s quark belongs to the 2nd generation. So “ Lamb shift in QCD”!!! Would have been hugely suppressed in a perturbative theory.

Tough to calculate!!!

Different from QED, however, QCD is non-perturbative! So the vacuum fluctuation could be sizable, so does strangeness!

Nucleons are the “hydrogen atom” of QCD. ss gives direct access to the “loops” of QCD sea. Recall the QED loops and the famous Lamb shift, g-2 …

What Do We “Know” Already?

Contribution of s quark to the longitudinal momentum

Contribution to the nucleon mass

Contribution to the nucleon spin

%2~))()((1

0dxxsxsx DIS N charm production:

MeV130~|| NssNms

N scattering + hyperon mass splitting:

likely 100% uncertainty

1.01.0~

|| 5

s

sNssN polarized inclusive DIS elastic N scattering polarized semi-inclusive DIS

All these indicate s quarks contribute sizably in nucleon structure, however with large uncertainties.

Strange Vector Current

qqeJ qEM

EM quark current of the nucleon

MEEM GNJN ,

Define vector (EM) form factors:

sME

uME

dME

nME

sME

dME

uME

pME

GGGG

GGGG

,,,,,

,,,,,

3

1

3

1

3

23

1

3

1

3

2

nspsnupdndpu GGGGGG ,,,,,, ;; Neglected heavier quarks Charge symmetry

Strange quark contributes to nucleon charge and magnetism?

Distribution of nucleon’s charge and magnetization.

Need one more constraint …

sdu GGGG3

1

3

1

3

2

Neutral-weak Current Additional Constraint

NC contains a vector and axial piece

Q cV cA

e -1 -1+4sin2W 1

u 2/3 1-8/3sin2W -1

d,s -1/3 -1+4/3sin2W 1

qccqJ AVNC )( 5

sMEW

dMEW

uMEW

pZME GGGG ,

2,

2,

2,, sin

3

41sin

3

41sin

3

81

So define NC vector form factor:

Charges in the unified electroweak theory in SM

ZME

VNC GNJN ,

,

sin2W =.2312 ± 0.00015

Kaplan and Manohar, 1988 pZnpW

psMEMEMEME GGGG ,,,2,

,,,, sin41

s-quark form factors calculations at Q2=0

NN

K

NN

0

2

22

0 2

)(4

)(

)0(

Q

sE

N

sE

s

sMs

dQ

QdGM

d

dG

G

N

e

N

e

Elastic e-N scattering

Measuring the NC Form Factor: Parity Violation

2

e e pp

NC amplitude suppressedby ~10-4

Difficult to see in cross-section measurement

However, if one measures the parity violation in the elastic scattering, one accesses the interference between EM and NC interactions “amplify” the relative experimental sensitivity to NC interaction.

PC PV

Mckeown and Beck, 1989

60Co 60Ni* + e- + e

C. S. Wu

Measurement of Parity Violation

60Co

B

detector

)1(~pσAPV O

24

LR

LRPV 10~

σσ

σσpσA Q

First observation of parity violation in weak interaction; Madam Wu’s famous 1957 60Co beta decay experiment.

In parity violating e-p scattering, the spin (helicity) of the electron is flipped back and forth.

e p

spinRL

p’

e’

detector

Parity Violating Asymmetry

22

2

)()(24 pM

pE

AMEF

LR

LR

GG

AAAQGA

)()()sin41(

)()()(

)()()(

222

222

22

QGQGA

QGQGQA

QGQGA

MeAWA

MZMM

EZEE

eA

sM

sE

GGG

Q2

4M 2

1 2(1 )tan2 2 1

(1 2) (1 )

kinematic factors

forward ep

backward ep

backward ed

Assuming EM and axial form factors are known (with errors), each measurement yield GE

s+GMs where )/( p

EpM GG

publishing, running

x2,

publishing, running

publishing, running

published x2, running

Summary of PV Electron Scattering Experiments

From D.H. Beck

Strange FF: Results at Q2=0.1

2=1

95% c.l.

Combining world data

(backward and forward) at

Q2=0.1 GeV2 allows one to

separate GE

sand GMS

GMs= 0.550.28

GEs= -0.010.03

The Jefferson Laboratory

A B C

G0

polarized source

linac

The G0 Collaboration D.S.Armstrong1, J.Arvieux2, R.Asaturyan3, T.Averett1, S.L.Bailey1, G.Batigne4, D.H.Beck5, E.J.Beise6, J.Benesch7, L.Bimbot2, J.Birchall8, A.Biselli9, P.Bosted7, E.Boukobza2,7, H.Breuer6, R.Carlini7, R.Carr10,

N.Chant6, Y.-C.Chao7, S.Chattopadhyay7, R.Clark9, S.Covrig10, A.Cowley6, D.Dale11, C.Davis12, W.Falk8, J.M.Finn1, T.Forest13, G.Franklin9,

C.Furget4,D.Gaskell7, J.Grames7, K.A.Griffioen1, K.Grimm1,4,B.Guillon4, H.Guler2, L.Hannelius10, R.Hasty5, A. Hawthorne Allen14, T.Horn6,

K.Johnston13, M.Jones7, P.Kammel5, R.Kazimi7, P.M.King6,5, A.Kolarkar11, E.Korkmaz15, W.Korsch11, S.Kox4, J.Kuhn9, J.Lachniet9, L.Lee8, J.Lenoble2, E.Liatard4, J.Liu6, B.Loupias2,7, A.Lung7, G.A.MacLachlan16, D.Marchand2,

J.W.Martin10,17, K.W.McFarlane18, D.W.McKee16, R.D.McKeown10, F.Merchez4, H.Mkrtchyan3, B.Moffit1, M.Morlet2, I.Nakagawa11, K.Nakahara5, M.Nakos16,

R.Neveling5, S.Niccolai2, S.Ong2, S.Page8, V.Papavassiliou16, S.F.Pate16, S.K.Phillips1, M.L.Pitt14, M.Poelker7, T.A.Porcelli15,8, G.Quéméner4, B.Quinn9,

W.D.Ramsay8, A.W.Rauf8, J.-S.Real4, J.Roche7,1, P.Roos6, G.A.Rutledge8, J.Secrest1, N.Simicevic13, G.R.Smith7, D.T.Spayde5,19, S.Stepanyan3,

M.Stutzman7, V.Sulkosky1, V.Tadevosyan3, R.Tieulent4, J.van de Wiele2, W.van Oers8, E.Voutier4, W.Vulcan7, G.Warren7, S.P.Wells13,

S.E.Williamson5, S.A.Wood7, C.Yan7, J.Yun14

1College of William and Mary, 2Institut de Physique Nucléaire d'Orsay,3Yerevan Physics Institute, 4Laboratoire de Physique Subatomique et de

Cosmologie-Grenoble, 5University of Illinois, 6University of Maryland, 7Thomas Jefferson National Accelerator Facility, 8University of Manitoba,

9Carnegie Mellon University, 10California Institute of Technology, 11University of Kentucky, 12TRIUMF, 13Louisiana Tech University, 14Virginia

Tech, 15University of Northern British Columbia, 16New Mexico State University, 17University of Winnipeg, 18Hampton University, 19Grinnell

College

Measure forward & backward asymmetries 16 scintillator rings recoil protons for forward

measurement electrons for backward

measurements elastic/inelastic for 1H, quasi-elastic for 2H Q2 = 0.12~1.0 for forward Q2 = 0.3, 0.5, 0.8 for

backward

Forward measurements complete (101 C electrons). 1013 protons per detector ring!

Ebeam = 3.03 GeV, 0.33 - 0.93 GeV

Ibeam = 40 uA, 80 uA

Pbeam = 75%, 80%

= 52 – 760 , 104 – 1160

= 0.9 sr, 0.5 sr

ltarget = 20 cm

L = 2.1, 4.2 x 1038 cm-2 s-1

A ~ -1 to -50 ppm, -12 to -70 ppm

Overview of the G0 experiment

G0 in Hall C

beammonitoring girder

superconducting magnet (SMS)

scintillation detectors

cryogenic supply

cryogenic target service module

electron beamline

Lumi monitors

Beam Properties

New Tiger laser system for G0i

ifalse p

p

Y

YA

2

1

40uA, 32 ns time structure, much higher bunch charge. Pbeam = 73.71.0%. Need to minimize the helicity correlated beam properties to avoid false asymmetry.

Accelerator

“IA” Pockels Cell (charge) and PZT mirrors (position) feedback system to minimize helicity correlated beam properties

Spectrometer

Toroidal magnet, elastic protons dispersed in Q2 along focal surface

Acceptance 0.12<Q2<1.0 GeV for 3 GeV incident beam

16 scintillator rings at the focal plane. 8 octants.

Azimuthally symmetric! Detector 15 acceptance:

0.44-0.88 GeV2

Detector 14: Q2 = 0.41, 1.0 GeV2

Detector 16: “super-elastic”, crucial to measure the background

lead collimators

elastic protons

detectors

targetbeam

Timing in the Experiment

Accelerator pulse structuret

32 ns

Ibeam

BeamHelicity

+1

-1

ON

DAQ

OFFt

1/30 s ~500 s “Macropulse”

Measurement timing

Typical t.o.f. spectrum

Electronics

High rate counting experiment, coinc. rate ~1MHz per scintillator pair. Electronics deadtime well-understood.

Fast time encoding (ToF histogramming electronics. beam pick-off signal T=0)

DAQ rate, every helicity flip (30Hz)

PMT Left

PMT Right

PMT Left

PMT Right

MeanTimer

MeanTimer

CoincTDC /LTD

Front

Back

Scalers:Histogramming

Aphys

Blinding Factor

Analysis Overview

Raw Asymmetries, Ameas

Beam and instrumental corrections:Deadtime

Helicity-correlated beam propertiesLeakage beam

Beam polarization

Background correction

Q2

nucleon form factors

Unblinding

GEs+GM

s

Electronics Deadtime Corrections

)(

%10~,))(1(

physQfalse

tm

AAfA

RfYRfY

To the first order

This shows up as 1) a decrease of normalized yield

with Ibeam (almost like a target density reduction)

2) a correlation between Am and AQ .

The DT effect is largely corrected based on the model of the electronics

The residual AQ correlation is removed by the linear regression

The final residual Aphys dependence is corrected based on Aphys and fresidual (~0.050.05 ppm).

499MHz Leakage Beam

Use “cut0” region in actual data to measure leakage current and asymmetry throughout run. Worked out nicely.

~ 50 nA 499 MHz beam leaks into G0 beam (~ 40 uA)

Leakage current has large, varying asymmetry (A ~ 600 ppm).

Aleak = -0.710.14 ppm (global uncertainty!)

“G0” AQ = 1000 ppm

“Leakage” AQ = -1000 ppm

0 16 32 ns

Leakage demo, extreme case

BCM integrates charge, no sensitivity to the micro-structure!

Positive background asymmetry?

Where do they come from ????

Physics Origin of the Positive Background Asymmetries

GEANT simulationWeak decay-particles rescatter inside the spectrometer that make into the detector. Low rate, large asymmetry!

Hyperons being produced in the scattering is highly polarized. Y → N+π is a PV decay with very large asymmetry (~1) However the decay-nucleon is highly supressed by the acceptance.

Results of the Hyperon Simulation

Source explained; used measured data in the correction.

Background Correction Yield & Asymmetry “2-step” Fit

Extract bin-by-bin dilution factor fb(t) by fitting time-of-flight spectra (gaussian signal + 4nd order polynomial background)

Use results to perform asymmetry fit: Ae = const, Ab = quadratic

2/=31.1/40 2/=37.5/44

)()())(1()( bbebm tAtfAtftA )()(

)()(

be

bb tYtY

tYtf

constant quadratic

detector 8

Det. 15 Background Corrections

Elastic peak broadened (~6 ns) because of increased Q2 acceptance.

Smooth variation of the background yield and asymmetry over detector range 12-14, 16. So make linear interpretation over detector number to determine fb(t) and Ab(t).

Detector 15 Asymmetry

Compare interpolated background asymmetry and data

)()())(1()( bbebm tAtfAtftA

Assume elastic asymmetry in each bin is a constant. Take interpolated fb and Ab, fit three Ae.

Elastic Asymmetries

“No vector strange” asymmetry, ANVS, is A(GEs,GM

s = 0)

Nucleon EM form factors: Kelly PRC 70 (2004) 068202

Inner error bars: stat; outer: stat & pt-pt sys

Primary contributors to the global error band: Leakage correction (low Q2) background correction (high Q2)

GEs+GM

s, Q2 = 0.12-1.0 GeV2

A 2 test based on the random and correlated errors: the non-vector-strangeness hypothesis is disfavored at 89%!

“Model” uncertainty is the EW rad. corr. unc. Dominated by the uncertainty of GA

e.

3 nucleon form factor fits; spread indicate uncertainties. |Kelly-FW| heavily driven by the difference in GE

n.

Are the data consistent with zero?

15.028.054.0

008.0027.0011.0

sM

sE

G

G

World Data with G0

Q2=0.48 GeV2

82.079.0

16.014.0

sM

sE

G

G

Strange quark contributes to p at -10% level.

“Complete” picture with G0 data

Emerging picture: At low Q2 end, despite the small , the data are positive,

consistent with a large and positive GMs.

Data go toward zero around Q2~0.2, which suggests GEs might

have a negative bump there. Data curve back up again for Q2>0.3 (note the growing ),

indicating that GMs stays positive.

For G0, ~ 0.94Q2

Global Fits to World Data

à la Kelly

222

2

2

2

33

221

12

1

4,

1

s

ssM

p

sE

QQG

M

Q

bbb

aQG

Toy model, minimal physics input

dipole form

Kelly form ensures GEs(0)=0, GE

s1/Q4 when Q2 large Fixed b3 = 1, all other variables float Fit all 24 data points: 18 G0, 3 HAPPEX, 2 A4, 1 SAMPLE

Results of the fit

Excellent fit2 = 14.919 d.o.f.

194.010.035.027.0

94.0141.027.012.0

10.041.0117.043.0

35.027.017.0194.0

27.012.043.094.01

2

1

1

2

2112

b

b

a

bba

s

s

ss

)0.44(3.336

)7.2(9.35

)06.0(05.0

)09.0(16.0

)10.1(51.1

2

1

1

2

b

b

as

s

Correlation Coeff.

Compare GEs with GE

n, and GMs with GM

p

Recall the factor of -1/3

GEs and GM

s Separately

-1/3s/p = -18%-1/3GEs(0.2)/GE

n(0.2)~40%

Very Naïve Interpretation (Conclusion?)

Hannelius, Riska + Glozman, Nucl. Phys. A 665 (2000) 353

Interpret GEs and GM

s results from the momentum space into spatial distribution.

So the GEs and GM

s results are both favor the picture that nucleon has an s quark “core” and s quark “skin”

Naïve kaon cloud (s quark spatially outside) leads to a negative s. Disfavored by the data “s quark skin”.

R. Jaffe, PLB 229 (1989) 275

Recall the well-known charge distribution in the neutron: positive bump in GE

n neutron has a positive core and negative “skin”.

-1/3GEs has a positive bump s quark spatially outside on

average!

Strange FF in the Near Future

G0 backward: detect electrons at = 108°

Q2 = 0.3, 0.5, 0.8 GeV2

both LH2 and LD2 targetsPV-A4 backward: = 145°

Q2 = 0.23, 0.47 GeV2 (underway)

HAPPEX (H and He4) running nowhigh precision at Q2 = 0.1 GeV2

high precision at Q2 = 0.6 GeV2 (proposed July 2005)

Prospective G0 Data @ Q2 = 0.8, 0.23 GeV2

• Run in Spring 06 at Q2 = 0.79 GeV2 (H and D targets)

• Possible run at Q2 = 0.23 GeV2 next (H alone?)

Summary

The successful G0 forward angle experiment yield the first measurement of parity-violating asymmetries over broad Q2 range. PRL 95, 092001(2005)Emerging picture: GM

s and GEs are both

likely nonzero

o GMs positive at low Q2 and stays

positive up to Q2=1.0 (GeV/c)2

o GEs might have a negative bump at

Q2~0.2 (GeV/c)2

o s quark skin of the nucleon???Stay tuned for G0 backward results

Target

20 cm LH2, aluminum target cell

longitudinal flow, v ~ 8 m/s, P > 1000 W!

negligible density change < 0.5%

measured small boiling contribution:

260 ppm : 1200 ppm (stat. width)

Formalism Including EW Rad. Corr.

eAp

MW

VsM

sE

pE

nV

nM

pM

nE

pE

pV

pM

pEWp

MpE

F

GG

RGGGRGGGG

RGGGG

QGA

2

)0(

22222

2

sin41

11

1sin411

24

222

)0(01

)/1(

1

132

11

A

DA

DAA

TA

TA

V

AeA

QG

GRsRDFRg

gG

Where pE

pM

G

G

and

Each asymmetry measurement can be cast into a linear combination of GE

s and GMs, assuming everything else is

known. In forward angle, use theoretical value and uncertainty of GA

e. Uncertainty dominated by the “anapole” term.

At tree level, R’s are zeros.

M.J. Mosolf et al, Phys Rep. 239, No. 1(1994)S.L Zhu et al, PRD 62,033008(2000)

Feedback Performance

All parameters coverage to zero!AQ (ppm) -0.14

x (nm) 3 ± 4

y (nm) 4 ± 4

Acceptance

Large and continuous acceptance for protons.

Helicity Correlated Beam Properties and Their Corrections

ii

false 2

1P

P

Y

YA

i

So require

Small ΔPi

Small sensitivity to Pi

Azimuthal symmetry large reduction of detector sensitivity to beam positions

Response of spectrometer to beam changes well understood

False asymmetries (and the uncertainty) due to helicity-correlated beam parameters

very small (~-0.02 ppm)

Measured Asymmetry upon Beam Spin Reversal

Detector 1-14 Background Uncertainty

Allowed background yield varied within

“lozenge”. Similar approach for asymmetry.

Separated point-to-point (pt-pt) uncertainties in background correction from global uncertainties. E.g. linear quadratic model of Ab move Ae downward for detectors 1-14

syseglobe

syseptpte

globeptptesyse

AA

AA

AAA

,2

,2

,2

,2

,2

,2

,2

4

14

3

Det. 15 Asymmetry

Compare interpolated background asymmetry and data

In detector 15, the global uncertainties larger because bins are continuous. syseglobe

syseptpte

globeptptesyse

AA

AA

AAA

,2

,2

,2

,2

,2

,2

,2

2

12

1

)()())(1()( bbebm tAtfAtftA

Assume elastic asymmetry in each bin is a constant. Take interpolated fb and Ab, fit three Ae.

Different Nucleon EM FF Parametrizations

Interpolate G0 Data

Three overlapping Q2 with other experiments:

Q2 = 0.1(HAPPEX, SAMPLE, A4), Q2 = 0.23 (A4), Q2=0.48 (HAPPEX)

Q2 = 0.1

extrapolate G0 using Ai/Q2i for first 3 Q2 points

Q2 = {0.122, 0.128, 0.136} Q2 = 0.23 (PVA4-I), 0.477 (HAPPEX-I) GeV2

Interpolate Ai/Q2i for Q2 = {0.210, 0.232, 0.262},

{0.410, 0.511, 0.631} Average the results of flat and linear interpolation.

Use the ½ difference as an additional “model” uncertainty.

Combine World Data

I. Start from the experimental asymmetries and uncertainties

II. Use a common set of form factor and EW parameters

III. Calculate GEs+GM

s

IV. Separate GEs and GM

s

V. The sensitivity to nucleon FF and EW parameters are evaluated separately by changing the model input globally and repeat I-IV

General procedure:

G0 Backward Angle Measurements

Match forward angle range with measurements at 3

momentum transfers

New detectors and electronics: Cryostat Exit Detectors (CEDs): separate elastic and inelastic electrons by trajectory Cerenkov Detectors: pion detectors Counting electronics only (no ToF separation)

Trigger change to run with standard beam (499 MHz)

Q2 Beam Energy Target Rate Asymmetry

(GeV2) (GeV) (MHz) (ppm)

0.3 0.424 H2 2.03 -18

D2 2.80 -25

0.5 0.576 H2 0.718 -32

D2 1.10 -43

0.8 0.799 H2 0.190 -54

D2 0.274 -72

Scheduled:Mar 06 – May 06

resume at Sep 06