A Measurement of Two-Photon Exchange in Unpolarized Elastic

Electron-Proton Scattering

John Arrington and James JohnsonNorthwestern University & Argonne National Lab

For the Rosen07/E05-017 Collaboration

Outline

• The electromagnetic interactions of the proton are described by two form factors, GE (Q2) and GM(Q2)

• Two methods of extraction, but their results don’t agree

• Leading candidate is two-photon exchange

Prior Experiments

Rosenbluth Scattering• Measure electron-

proton scattering• Factor out Mott cross

section, and get a function linear in the squares of the form factors

τGM2 + εGE2

Polarization Transfer• Scatter longitudinally

polarized electrons from unpolarized protons

• The ratio GE/GM is proportional to pT/pL

• Does not give form factors directly

Disagreement

• Rosenbluth gives a ratio that stays flat– The errors on GE increase

with Q2

• Polarization transfer shows a decreasing ratio– Smaller errors at high Q2

– Implies a difference between charge and magnetic distributions

J. Arrington, Phys. Rev. C69:022201, 2004

M. Jones et al, Phys. Rev. Lett. 84:1398-1402, 2000

O. Gayou et al, Phys. Rev. Lett. 88:092301, 2002

“Super-Rosenbluth”JLab E01-001

• E01-001Detect scattered protons instead of electrons

• Same reaction, smaller angular-dependant corrections

• Precision comparable to polarization transfer

• Agrees with electron Rosenbluth– The disagreement is real– High-precision measurement of

the discrepancy– Tests radiative corrections I. A. Qattan et. al, Phys. Rev. Lett.

94:142301, 2005

Two-Photon Exchange

• Both methods account for radiative corrections, but neither considers two-photon exchange

• Difficult to Calculate– Rough qualitative

agreement– Different ε dependence– Scale not predicted

Magnitude of the Discrepancy

Solid line – fit to E01-001 ‘Super-Rosenbluth’

Dashed line – taken from polarization transfer ratio

Quantify difference, look for nonlinearity

Rosenbluth 2007JLab E05-017

• HMS in Hall C at Jefferson Lab

• 4cm liquid hydrogen target for elastics

• 4cm aluminum dummy for endcap subtraction

• May 8 – July 13, 2007

Rosenbluth 2007

102 Kinematics points

Q2 0.40-5.76 GeV2

13 points at Q2=0.983

10 points at Q2=2.284

Aerogel Calibration

• Aerogel distinguishes π+ from heavier particles• Fit the position of the 1-photoelectron peak

– Not possible on runs with low pion count due to interference from the pedestal

• Noisy ADC signals– Not needed for pion rejection at most (all?) settings– Mainly check TOF efficiency, pion contamination

Time of Flight Calibration

• Acceptance cuts– Solid – full delta-β

spectrum– Small dashes - Aerogel

cut to exclude pions– Large dashes - Beta cut

to exclude deuterons

Time of Flight Calibration• Six total calibrations

– Three momentum ranges– Before/After discriminator

replacement• Solid line – uncalibrated• Dashed line – calibrated

No kinematic, aerogel cuts-cut on elastic peak supresses pions and deuterons

No dummy subtraction-removes deuterons and tritons

Analysis Steps

• Sum data & dummy runs at selected kinematic• Simulate elastics, pion photoproduction, compton

scattering• Scale all to corrected charges• Fit dummy + simulations to the data

– Extract ratio of simulation cross-section to actual cross-section

Charge Correction• Included so far

– Computer & electronics deadtime– HSCIN (¾ scintillator) efficiency– default tracking efficiency (“HMS w/DC cuts”)– prescale factor

• Not yet included– Final BCM Calibration*– Target boiling*– Particle Identification efficiency*– Proton Absorption*– Beam offset

* Should be -independent

Unpeeling

• Hydrogen elastics– Compare to simulated

elastics• Background

– ‘Dummy’ runs for endcap subtraction

– Simulated 0 photoproduction

Unpeeling

• Hydrogen elastics– Compare to simulated

elastics• Background

– ‘Dummy’ runs for endcap subtraction

– Simulated pi-0 photoproduction

Low setting (p = 12.5o)

“High” setting

Nonlinearity Tests• Born approximation

linear ε dependence, TPE could cause a deviation

• E01-001 and NE11 show quadratic terms consistent with zero

• Project P2 within ±0.020 for E05-017

• Much better limits over wide Q2 range

NE11: L. Andivahis et al, Phys. Rev. D50:5491, 1994

Conclusion

• Projected uncertainties from proposal

• More Q2 points– Shifted range down– Better linearity tests– Slightly smaller range

• Analysis underway