PR-09-018: SSA with SBS

28 Jan 2009 PR-09-018 (SSA with SBS) / PAC34 / JLab 1

PR-09-018: SSA with SBS

Measurement of the Semi-Inclusive and K electro-production in DIS regime from transversely polarized 3He target with the

SBS & BB spectrometers in Hall A

G. Cates, E. Cisbani, G.B. Franklin, B. Wojtsekhowski

and the SBS Collaborationhttp://hallaweb.jlab.org/12GeV/SuperBigBite

Evaristo Cisbani

Shall I mention why it is not an Hall A Collaboration ?


Propose to measure the SSA of SIDIS processes n(e,e’)X and n(e,e’K)X

– Extract Sivers and Collins (and Pretzelosity) asymmetries on and K with high statistics

– Provide 2D binning (at least) on the relevant variables: x, P and z, for

both hadrons

– Provide Q2 dependence

– Explore for the first time the high x valence region (with overlap to HERMES, COMPASS, JLab6 data)

• Understanding of QCD dynamics in the nucleon by the Sivers effect

• Improve knowledge of the nucleon structure in terms of parton distribution functions

• Shed more light on the origin of the nucleon spin

Evaristo Cisbani

Change title in Executive SummaryRemove the "general goals" and write the appaeatus we intend to use, correlation with other experiment and proposal ...


Method: from SIDIS to SSA to DFFF

Pretzelosity

Evaristo Cisbani

Deltaq - deltaq = k2/M2 * Pretzelosity Pretzelosity difference between elicity and transversity: a measure of the relativistic effects!


Hep

ex-0

802.

2160

Fro

m L

evor

ato

/ Tra

nsve

rsity

200

8F

rom

Pap

pala

rdo

/ Tra

nsve

rsity

200

8

Collins Moments on proton/deuteronCOMPASS/deuteron

proton

Proton (left)• K- and - with opposite sign, strong flavor

dependence (but K errors significant)!• K+ and + consistent with u dominance• COMPASS compatible with HERMES

Deuteron (up): consistent with 0, expected asymmetry on neutron as large as on proton

Larger effects at larger x


Sivers Moments on proton/deuteron

Hep

ex-0

802.

2160

Fro

m L

evor

ato

/ Tra

nsve

rsity

200

8F

rom

Pap

pala

rdo

/ Tra

nsve

rsity

200

8

Proton (left)• K+ twice + conflict with u dominance

expectation• K- and - consistent with 0• COMPASS smaller than HERMES

Deuteron (up): consistent with 0, expected asymmetry on neutron as large as on proton

COMPASS/deuteron


Extraction of DF and FF from a Global FitTransversity first moment (assume sea = 0)

Collins Fragmentation Function

Ans

elm

ino

et a

l 12/

08

Ans

elm

ino

et a

l 05/

08

Sivers function

e+e- from BELLE (Q2=110 GeV2)

HERMES/p + COMPASS/d (Q2~2.5 GeV2)

Sivers:large s-bar

sea small

Collins: Unfav/Fav FF puzzle(not seen in Unpol FF)


Theory Report for PR-09-018M. Burkardt, H. Kamano

“ … A measurement of the Sivers function at JLab would be very interesting for

several reasons. Most importantly, the statistics would be much higher… Since

there is a strong overlap in the x range with the HERMES data, it would be

possible to verify/falsify the “kaon puzzle”… The JLab data would not only

extend the Q2 leverage of the HERMES/COMPASS experiments but at the same

time provide results for two different value of Q2, and thus can shed some light

on the Q2 evolution of the Sivers function…

The HERMES and COMPASS results are more consistent with each other in

the case of the Collins function. However, also here a very interesting results

needs confirmation: it appears that the Collins function for “unfavored”

fragmentation is about as large (and of opposite sign) than that of the “favored”

fragmentation… A JLab experiment, with significantly higher statistics than the

HERMES data, could verify/falsify this result, and also help to study the Q2

evolution of the Collins function…”


Experimental Setup and parameters

Beam: 50 A, E=8.8 and 11 GeV (80% long. Pol.)Target: 65% polarized 3He GEn(2)/PR-09-016 Luminosity: 1.4×1037 cm-2s-1 , 0.05 sr

BB: e-arm at 30o

= 45 msrGEM TrackerGas CherenkovShower GMn/PR-09-019

SBS:h-arm at 14o

= 50 msrGEM trackerexcellent PID / RICHHadron CALO

e+3He→e’+(K)+X

Event rate: ~104×HERMES60 days of production expected stat. accuracy:

1/10 of proton HERMES

Evaristo Cisbani

cell: 75 mg of gas = 75 * 6.02 10^20current: 50 uAlumi = 75 * 6.02 * 10^20 * 50 * 10^-6 / (1.6 * 10^-19 = 1.4 10^37In background we use 4 10^37Deatail on calorimenter


What is special in this experiment

• High Luminosity:– 105 larger than in HERMES– High target polarization (65%)– Fast target polarization switch (120 seconds)– 4 (8) transverse polarization directions

• Use of SBS (and BB):– Large solid angle (50 msr), very good angular and

vertex resolutions– Large momentum coverage (2-7 GeV/c)– Excellent hadron PID

• Reuse equipment from GEp(5) (approved), GEn(2) and GMn (proposals)


Hadron Arm: SBS

• Angular Resolution: (p = 4 GeV)_h = 0.09+0.59/p [mrad] (0.3 mrad)

_v = 0.14+1.34/p [mrad] (0.4 mrad)

• Vertex Resolution: 0.53+4.49/p [mm] (0.2 cm/sincentral)• Momentum resolution p/p = 0.03 p+0.29 % (0.4 %)• CALO Trigger Threshold: 1.5 GeV (online), 2.0 (offline)

- Magnet: 48D48 - 46 cm gap2 Tm field integral -100

tonInsert for beam pipe

• GEM chambers for tracking with 70 m resolution

• HERMES RICH for hadron-ID• Segmented Hadron CALO

(15x15 cm2 blocks)


Hadron PID: HERMES RICH on SBS

Very stable performance (n/(naerogel-1)= 1%, 9 years)Stored at UVa under safe/controlled conditions(also additional wall of spare Aerogel)

5.5 GeV K+

14.6 GeV e-1.5 GeV -

REAL DATA from NIMA 479 (2002) 511

C4F10 gas


Background Rate and Trigger LogicBB-Shower (1 GeV):

200 kHzBB-Cherenkov (3-4 pe):

2.5 MHz

40 ns coinc./segmentedaccidental: 4 kHz

DIS-e: 1 kHz

BB-Offline / Tracking0 suppression: 1 kHz

SBS Backgroundon HCALO

SBS-HCalo (1.5 GeV):3 MHz

Online 50 ns coinc.real coinc: 100 Hz

acc: 750 Hz

4 ns time gateTotal: 12 Hz

Vertex Correlation(6=3 cm): 0.6 Hz

Momentum cut > 2GeV

Bck = 0.3 HzSignal = 67 Hz

Evaristo Cisbani

Wiser code for calorimeter counting ratesCharge particle below 1 GeV swept out of the CALOe-e rate at the level of 0.01 Hz


Challenges in large acceptance/high luminosity

• SBS Tracker rate 60 kHz/cm2; 3xGEM support rate >10 MHz/cm2

• Track reconstruction: BB first,SBS from vertex to segmented HCALO hit

• RICH PID: high segmentation of photon detector (2000 PMT) is the optimal solution:

–Expected 35 extra hits/event from:soft photons → Compton electronsin aerogel (50 ns gate width)

2-5% occupancy

~20% of the HERMES RICH PMT array


Q2 coverage

Current Transversity Exp. E06-010

Prop. Exp. Ebeam = 8.8 GeV

Prop. Exp. Ebeam = 11.0 GeV

We will investigate the Q2 dependence of the Sivers and Collins functions, with overlap in the region of HERMES; reveal higher twist effects.

Analysis of the Q2 effect will use also the results of presently running 6 GeV E06-010 Transversity experiment


Azimuthal Coverage

0.15 < x <0.65h_cent. = 14°

Ebeam = 11 GeV4 target spin

directions

Complete coverage of the Collins, Sivers and “Pretzelosity” azimuthal angles with 4 target spin directions

(with 8 target spin directions even better uniformity)

Partial coverage of h but in sin/cos sensible regions

h


Figure of Merit

Two beam energy runs for Q2 dependence studies:


Expected Statistical Accuracy on

High x region, with partial overlap with HERMES2D binning in (x,z), (x,P) and (z,P) for and Kand Q2 dependence

5×5 bins (0.15<x<0.65, 0.2<z<0.7) (only one shown)

DF from CTEQ5MFF from DSS

Con

talb

rigo

/SP

IN08


Expected Statistical Accuracy on K

• Superior quality of Kaon data

• Extend at higher x with partial overlap with existing data on proton, deuteron and expected results of HallA Transversity 6 GeV DF from CTEQ5M

FF from DSSRate normalized to HERMES/p+d K production


Systematics• Physics Effects:

– FSI on nuclei

• 3He: Pp~2%, |d|2~10%, Presc~10-20%

• D: Pp~85%, |d|2~6%, Presc~5-10%

– Higher Twist Terms of SIDIS asymmetries

• Experimental/Analysis:– Random background– Vector Meson– Particle ID– Acceptance Effects– Radiative Corrections

Evaristo Cisbani

Compromise between number of bins (in x,z,Pt, Q2) and systematics (acceptance and other detector related): larger the bins, better the systematicsP_resc = Probability of Rescattering


Summary• We propose to measure the SSA in the transversely

polarized SIDIS processes: n(e,e’)X and n(e,e’K)X at two Q2

• Experiment will re-use part of GEp(5) equipment and the HERMES RICH

• Will be ready to take data in 2014, with no significant extra costs respect to SBS apparatus for GEp(5)


Backup Slides


Collaboration ResponsibilitiesAll spokespersons are active members of

GEp(5) experiment


Experimental Setup with SciFi

SciFi/NSERC

A small part of the SBS-SciFi tracker can be used in front of the HCALO to constraint the track reconstruction


Polarized SI-DIS process


GEn(2) High Lumi Target

• 60% transverse polarization

• Support 60 uA beam current

• Extended target cell (60 cm)• Use proved alkali-hybrid mixture technique for polarization transfer • Improved diagnostics for optimal tuning of the potassium-rubidium ratio• Use line-narrowed high-power diode-laser arrays• Use convection for gas mixing between target and pumping chambers• Better description of the target behavior by the identification of the relaxation rate (X-

term)• Use of metal target cell


Electron Arm: GEn(2), GMn BigBite

• Use the current BB setup:– Gas Cherenkov– Two-Layer Electromagnetic Calo,– Scintillator Hodoscope – GEM based Tracker

Replace MWDC tracker by:

• GEM based tracker, from components of the GEp(5) front tracker


SBS Solid Angle

SBS can go very forward, at reduced solid angle.Moving from 14 to 10 degree, the solid angle goes from 50 to ~40 msr

Solid angle of SBS as a function of central angle (and distance from target)


ALICE/STAR RICH Reconstruction

Expected ~12% occupancy with 3 /K separation

Performance proved by MC and STAR (3--5% occupancy) real data.

D. Di Bari, RICH07/Trieste


Phase Space


Phase Spase of the Relevant Variables

Ebeam = 11 GeV


Azimuthal Angles Distributions

The coverage of can be extended moving forward SBS to central ~ 10 degree (with a decrease of SBS acceptance) and/or changing (not dramatically) the BB settings


Azimuthal angles coverage vs x

Azimuthal angles coverage does not very significantly depend on x,Q2

Number of target spin directions can be increased to have a better uniformity; 8 would be optimal in this respect.


h coverage vs x

h

Partial coverage of one (or both) azimuthal variables.

When sin or cos (or S) are close to 0 (red band) the AUT equation degenerates into the sum or difference of AC and AS. The variation of S is helpless.

In all other case, even with partial coverage of and S the extraction of the two modulated terms is possible.

Error on the extraction roughly goes like (coverage_fraction)-½

sin/cos functions


Background Rate and Trigger Logic

BB-Shower (1 GeV): 200 kHz

BB-Cherenkov (3-4 pe):2.5 MHz

40 ns coinc./segmentedaccidental: 4 kHz

DIS-e: 1 kHz

BB-Offline / Tracking0 suppression: 1 kHz

BigBite Backgroundon Shower

SBS Backgroundon HCALO

SBB-HCalo (1.5 GeV):3 MHz

SBB-RICH:5% occupancy

Online 50 ns coinc.real coinc: 100 Hz

acc: 750 Hz

4 ns time gateTotal: 12 Hz

Vertex Correlation(6=3 cm): 0.6 Hz

Momentum cut > 2GeV:

Bck = 0.3 Hz

Signal = 67 Hz

Evaristo Cisbani

Non chiaro differenza in pi0 dei due spettrometri; legato forse al differente angolo centrale ?Come distinguiamo gli elettroni su SBS ?


Terms entering the measured asymmetry


Systematics/Physics

• Target FSI relative error: expected <7% following the analysis of Scopetta/Transversity/2008

• Higher Twists:– we will study the Q2 dependence with high

statistics;– terms will be included into the fit whose

stability benefits again of the high statistics– Unpolarized analysis will also be carried on


Systematics/Exp. Apparatus• Random

– dilution factor expected to be small < 1% (S/N = 67/0.3)– Relative error well below 10%

• Vector Meson:– from PHYTIA prediction tuned on HERMES data below 2.5%; lower

at higher x– can be studied at higher z

• Acceptances:– azimuthal angles are well covered even with partial h coverage– X,z,P,Q2 effects suppressed by 2D (at least) binning

• Radiative QED Effects:– Influence x,Q2 and azimuthal angles; PHYTIA can be used to

estimate the correction factor (and systematic error)– According to HERMES we expect a systematic error <5%

• PID:– RICH detector with up to 5% occupancy expect to provide 4

separation at least

Documents

PR-09-018: SSA with SBS