The PRad Experiment

Maxime Levillain

September 1, 2017

Outline

The Proton Radius Puzzle

PRad Setup

Detectors Performance

Analysis

Summary

Outline

The Proton Radius PuzzleDifferent Methods of MeasurementElastic ep ScatteringNew Experiment Needed

PRad Setup

Detectors Performance

Analysis

Summary

Measurements of Form Factors

I First measurement at SLAC in 1961 through ep scattering

I 60 years of measurements, 4 possible different methods

Atomic HydrogenSpectroscopy

Lamb shift measurements byMPQ and LKB

Muonic HydrogenSpectroscopy

Lamb shift measurements byCREMA

ep Scattering

Accelerator based experiments atMainz, SLAC, JLab, etc

µp Scattering

Future experiment PSI/MUSE

Maxime Levillain The Proton Radius Puzzle 4 / 32

Spectroscopy Results

I Lamb shift measurements

I atomic hydrogen spectroscopy results compatible with epscattering results

I muonic hydrogen spectroscopy results at 0.84 fm

Maxime Levillain The Proton Radius Puzzle 5 / 32

Elastic ep Scattering

I Elastic cross-section in the limit of the firstBorn approximation:

dσ

dΩ=

(dσ

dΩ

)Mott

·E ′

E·

1

1 + τ· (Gn

E2(Q2) +

τ

εGnM

2(Q2))

with:Q2 = 4EE ′sin2θ/2 τ = Q2

4M2p

ε = 1/(1 + 2(1 + τ)tan2θ/2)

I Structureless proton:(dσ

dΩ

)Mott

=α2(1 − β2sin2θ/2)

4k2sin4θ/2

I GE can be expressed using a Taylorexpansion at low Q2:

GE = 1 −Q2

6< r2 > +

Q4

120< r4 > +...

which gives:< r2 >= −6 ·

dGpE

dQ2

∣∣∣Q2=0

e− e−

γ∗

p pGE ,GM

Phys. Rev. C 93, 065207

Maxime Levillain The Proton Radius Puzzle 6 / 32

The Proton Radius Puzzle

rp(e−) = 0.8770± 0.0045fm rp(µ−) = 0.8409± 0.0004fm

I Discrepancy between muonic hydrogen spectroscopy andatomic hydrogen (spectroscopy and scattering) measurements

Maxime Levillain The Proton Radius Puzzle 7 / 32

The PRad Experiment

I Previous measurements have largesystematic uncertainties and a limitedcoverage at small Q2

I Requirements for PRad Experiment:I large Q2 rangeI extend to very low Q2

I controlled systematics at sub-percentprecision

I Choices:I Non magnetic spectrometer methodI No target windowsI high resolution high acceptance

spectrometerI Normalization by Møller cross-section

Phys. Rev. C 93, 065207

Maxime Levillain The Proton Radius Puzzle 8 / 32

PRad Timeline

• 2011 - 2012 Initial proposal• 2012 Approved by JLab PAC39

• 2012 Funding proposal for windowless H2 gas flow target

• 2012 - 2015 Development, construction of the target

• 2013 Funding proposals for the GEM detectors

• 2013 - 2015 Development, construction of the GEM detectors

• 2015, 2016 Experiment readiness reviews

• January/April 2016 Beam line installation

• May 2016 Beam commissioning

• May 24 - May 31 Detectors calibration

• June 4 - June 15 1.1 GeV data taking

• June 15 - June 22 2.2 GeV data taking

Maxime Levillain The Proton Radius Puzzle 9 / 32

Outline

The Proton Radius Puzzle

PRad SetupJLab FacilityPRad SetupWindowless Gas Flow TargetHybrid CalorimeterGEM detectors

Detectors Performance

Analysis

Summary

JLab Facility

Maxime Levillain PRad Setup 11 / 32

JLab 12GeV Upgrade

I First experiment finished using 12 GeV accelerator (not at fullbeam energy)

Maxime Levillain PRad Setup 12 / 32

PRad Setup

I Electron beam or tagged photon beam at ∼ 1 GeV and ∼ 2 GeV

I Windowless H2 gas flow target

I Vacuum box

I GEM detectors

I Primex HyCal

Maxime Levillain PRad Setup 13 / 32

Windowless H2 Gas Flow Target

I gas target of cryogenically cooledhydrogen at 19.5 K

I beam opening: 2 mm, length: 4 cm

I cell density: ∼ 2 · 1018 H atoms/cm2

I pressures:I cell pressure: 471 mTorrI chamber pressure: 2.34 mTorrI vacuum chamber pressure: 0.3 mTorr

Developed and build by JLab target group

Maxime Levillain PRad Setup 14 / 32

Vacuum Box

I 1.7 m diameter, 2 mm aluminum vacuum window

→ Limited background

Maxime Levillain PRad Setup 15 / 32

Primex HyCal

Hybrid detector:

I Central part:I 34 x 34 matrix of PbWO4 detectorsI dimension of block: 2 x 2 x 18 cm3

I 2 x 2 blocks removed from the centerfor beam line to pass through

I Peripheral part:I 576 lead glass detectorsI dimension of block: 4 x 4 x 45 cm3

I Successfully used for Primex experiments

Maxime Levillain PRad Setup 16 / 32

GEM Detectors

I Two large area GEM detectors: 55 cm x 123 cmI Purpose:

I improve spatial resolution by a factor 20 to 40 → 100 µm→ to reduce uncertainties on θ and Q2

I Central overlap between the 2 planes and central hole for thebeam line

Developed and build by UVA

Maxime Levillain PRad Setup 17 / 32

Outline

The Proton Radius Puzzle

PRad Setup

Detectors PerformanceHyCal Energy ResolutionTrigger EfficiencyGEM Matching EfficiencyGEM Spatial Resolution

Analysis

Summary

HyCal Energy Resolution

I Crystal energy resolution with statistical uncertainties

E (MeV)400 600 800

(E)/

Eσ

0.02

0.03

0.04

0.05

0.06

E GeV0.026 =

E(E)σ

I Achieved expected energy resolution:I 2.5% at 1 GeV for crystal partI 6.1% at 1 GeV for lead glass part

Maxime Levillain Detectors Performance 19 / 32

Trigger Efficiency

(GeV)γE0.4 0.6 0.8 1

Trig

ger

Effi

cien

cy

0.992

0.994

0.996

4PbWO

PbGlass

I Plateau from 500 MeV with an efficiency of 0.995

I Good uniformity

Maxime Levillain Detectors Performance 20 / 32

GEM Matching Efficiency

0.5 1 1.5 2 2.5 3 3.50.8

0.85

0.9

0.95

1

1.05

theta (degree)

effic

ienc

y

I GEM detection efficiency measured in both photon beamcalibration (pair production) and production runs (ep and ee)

I Almost flat efficiency > 97% after removal of spacers anddead zones

Maxime Levillain Detectors Performance 21 / 32

GEM Spatial Resolution

I Really good spatial resolution ∼ 74 µm

I 20 to 40 times better than HyCal spatial resolution

Maxime Levillain Detectors Performance 22 / 32

Outline

The Proton Radius Puzzle

PRad Setup

Detectors Performance

AnalysisStabilityYieldsCross-sections

Summary

Data Collected

I Calibration with tagged photon beamI Every calorimeter module moved into the beamI Allows study of resolution, linearity, trigger efficiency

I 1.1 GeV electron beamI 4.2 mCI 604 M events with targetI 53 M events with “empty target”I 25 M events with 12C target for calibration

I 2.2 GeV electron beamI 14.3 mCI 756 M events with targetI 38 M events with “empty target”I 10.5 M events with 12C target for calibration

Maxime Levillain Analysis 24 / 32

Target Stability

I Control of target properties (pressure, temperature, position)via EPICS

Run number0 20 40 60 80 100 120 140 160

2N

/cm

800

850

900

950

1000

10501510×

1.1GeV

2.2GeV

3%±~

Target Thickness

Weizhi Xiong→ Less than 3% deviation

Maxime Levillain Analysis 25 / 32

Phase Space

I Phase space after background subtraction

I Separation of ep and Møller phase space (for θ > 0.85deg for1 GeV)

Maxime Levillain Analysis 26 / 32

ep Yields

2 (GeV)2Q-410 -310 -210 -110

epN

210

310

410

510

610

710

~10% of data

Very preliminary

1.1 GeV

2.2 GeV

2Q vs. ep→epN

I 1.1 GeV data set: Q2 ∈ [2 · 10−4, 1.3 · 10−2] GeV2

I 2.2 GeV data set: Q2 ∈ [8 · 10−4, 6 · 10−2] GeV2

Maxime Levillain Analysis 27 / 32

Yields Stability

I Stability of ratio ep/ee after background subtraction fordifferent beam intensity

#run1100 1200 1300

ep/e

e

0.98

1

1.02

5 nA 10 nA 15 nA

#run1350 1400 1450 1500

ep/e

e0.98

0.99

1

1.01

1.02

15 nA 25 nA 55 nA 40 nA 55 nA

I Good stability for the 2GeV period

Maxime Levillain Analysis 28 / 32

Extraction of Cross-section

I Normalization of ep cross-section by Møller cross-section:(dσ

dΩ

)ep

=Nexp(ep → ep in θi ±∆θ)

Nexp(ee → ee)·εeegeomεepgeom

·εeedetεepdet·(dσ

dΩ

)ee

I Several event generators have been developped for ep andMøller scattering taking into account complete calculations ofradiative corrections beyond ultra relativistic approximations

I A. V. Gramolin et al., J. Phys. G Nucl. Part. Phys.41(2014)115001

I I. Akushevich et al., Eur. Phys. J. A 51(2015)1

I Geant4 is used to take into account all external radiativeeffects

σBornep =

(σepσee

)exp

/

(σepσee

)sim

· σBornee

Maxime Levillain Analysis 29 / 32

Preliminary ep Cross-section

I Preliminary ep cross-section for the 2.2 GeV data setI Statistical uncertainties at ∼ 0.2% per pointI Conservative point-to-point systematic uncertainties at ∼ 2%

(deg)θ0.5 1 1.5 2 2.5 3

(m

b/sr

)ep

→ep

Ω/dσd

0

100

200

300

400

500

600

700

800

~50% 2.2 GeV data

Very Preliminary

elastic scattering cross sectionep

Maxime Levillain Analysis 30 / 32

Preliminary ep Cross-section

I Preliminary ep cross-section for the 2.2 GeV data setI Statistical uncertainties at ∼ 0.2% per pointI Conservative point-to-point systematic uncertainties at ∼ 2%

) 2 (GeV2Q-410×6 -310 -310×2 -210 -210×2

(m

b/sr

)ep

→ep

Ω/dσd

-110

1

10

210

310~50% 2.2 GeV data

Syst. Uncertainty

Very Preliminary

01020

elastic scattering cross sectionep

Maxime Levillain Analysis 31 / 32

Summary

I The PRad experiment was uniquely designed to address theProton Radius Puzzle

I The experiment was successfully performed in May-June 2016

I Wide range of Q2 without normalization on more than twoorders of magnitude (2 · 10−4 GeV2 to 6 · 10−2 GeV2)

I Lowest Q2 data set of ep elastic scattering (2 · 10−4 GeV2)

I First preliminary extraction of the proton radius expected atthe end of October

Thanks to JLab, Hall B, Accelerator Division and Target Group

PRad is supported in part by NSF MRI award PHY-1229153, as well as DOE awards for GEM;

my research work is supported by NSF awards: PHY-1506388 and PHY-0855543

Maxime Levillain Summary 32 / 32