Exclusive Physics with the HERMES Recoil Detector Erik Etzelmüller on behalf of the HERMES Collaboration!!! workshop on! Exploring Hadron Structure with Tagged Structure Functions! Thomas Jefferson National Accelerator Facility,18.1.214 1
Content Motivation (GPDs - Generalized Parton Distributions) DVCS (Deeply Virtual Compton Scattering) The HERMES experiment and the Recoil detector Overview of the analysis and the results Associated DVCS/ΔVCS (ep eγnπ) 2
GPDs Some milestones 1988 EMC publication triggers spin crisis 1995 HERMES experiment at DESY starts data taking. Mission: spin puzzle 1997 Ji finds a way to access the total angular momentum of a nucleon 21 First HERMES DVCS paper published 26 HERMES Recoil detector starts taking data 3
GPDs The proton spin puzzle contribution from quarks lower than expected from naive quark model Ji found a way to access total angular momentum through GPDs [X. Ji, Phys. Rev. Lett. 78 (1997) 61]: GPDs 4
GPDs reduced Wigner distribution (GTMDs) TMDs PDFs FFs 5
GPDs Experimentally GPDs can be accessed through measurements of hard exclusive lepton-nucleon scattering processes. e e e γ* γ hard scattering process (QED & QCD) e * x+ ξ x ξ soft non-perturbative part (GPDs) x+ x N N N N DVCS deeply virtual Compton scattering HEMP hard exclusive meson production 6
DVCS Theoretically cleanest access to GPDs through DVCS But: final and initial state is equal to Bethe-Heitler scattering (BH) e γ* e γ virtual photon interacts with quark inside the nucleon e γ e x+ ξ x ξ real photon is emitted γ* N N N N DVCS BH 7
DVCS d 4 dq 2 dx B dtd = x B e 6 32(2 ) 4 Q 4p 1+ 2 T ep!ep 2 Amplitude of Bethe-Heitler scattering is dominant at HERMES kinematics DVCS amplitude is amplified by BH in the interference term 8
DVCS at HERMES Access through the measurement of asymmetries: longitudinally polarized beam, unpolarized target Fourier coefficients comprise Compton-Form factors (CFFs) which are convolutions of the GPDs with the hard scattering kernel. Example: 9
The HERMES experiment m 2 FIELD CLAMPS DRIFT CHAMBERS TRIGGER HODOSCOPE H1 27 mrad PRESHOWER (H2) 1 Recoil Detector DRIFT CHAMBERS e+ 14 mrad p FC 1/2 PROP. CHAMBERS MC 1 3 LUMINOSITY e+ MONITOR27.5 GeV 1 TARGET CELL DVC SILICON HODOSCOPE H STEEL PLATE BC 1/2 BC 3/4 TRD γ CALORIMETER 14 mrad 2 RICH 27 mrad MAGNET 1 2 3 4 5 6 7 8 9 Recoil detector upgrade for 26/27 running 1
The HERMES Recoil detector Enables the measurement of the recoiling charged particle and therefore full ep epγ event reconstruction 11
The HERMES Recoil detector Sketch of front- and backside of a silicon strip detector module (SSD) Schematic design of the scintillating fibre tracker (SFT) 12
The HERMES Recoil detector The silicon strip detector (SSD) The scintillating fibre tracker (SFT) 13
The HERMES Recoil detector Kinematic coverage MC MC SFT SSD Scintillating fibre tracker (SFT) and silicon strip detector (SSD) complement each other 14
The HERMES Recoil detector Recoil tracking taking energy loss into account improves momentum resolution for low p azimuthal-angle resolution: 4 mrad, polar-angle resolution: 1 mrad (for p>.5 GeV) 15
Improvement through the Recoil detector 1 (N/N DIS ).2.15.1 forward spectrometer only measured s i m i l a r b a c k g r o u n d proton s a m e k i n e m a t i c a c c e p t a n c e unresolved sample unresolved-reference sample pure sample experimental data simulation (sum) ep epγ ep e + γ semi-inclusive.5 5 1 15 5 1 15 2 M X [GeV 2 2 ] M X [GeV 2 ] associated processes (ep eγδ+) 5 1 15 M X 2 [GeV 2 ] pure ep eγp Missing mass: 16
Improvement through the Recoil detector Elastic BH Resonant Kinematic event fitting: χ²-value of interest penalty term constraints 4-momentum conservation as constraints lowest χ²-value in case of multiple recoil tracks per event minimum of 1 % fit probability required, which corresponds to χ² < 13.7 17
Beam-spin asymmetries ep eγp A. Airapetian et al, JHEP1 (212) 42 sin A LU.2 -.2 unresolved unresolved-reference pure -.4 sin(2) A LU.2 -.2 overall 1-1 -t [GeV 2 ] 1-1 x B 1 1 Q 2 [GeV 2 ] Magnitude of the leading asymmetry has increased by.54 ±.16 All sets are strongly correlated but the unresolved samples contain an average contribution of 12-14 % of associated processes 18
Beam-spin asymmetries ep eγp P. Kroll, H. Moutarde, F. Sabatié, Eur. Phys. J. C (213) 73:2278 HERMES (without Recoil) HERMES (with Recoil) Derived from GPDs, which are extracted from HEMP Recoil data leads to a significantly better overlap with HEMP data 19
Beam-spin asymmetries ep eγnπ Besides a better understanding of the unresolved sample, associated DVCS in principle also allows further access to GPDs. In the large Nc-limit the remaining N Δ GPDs can be related to the N N isovector GPDs: 2
Selection of associated events ep eγpπ⁰ ep eγnπ+ Uncharged particle remains undetected Kinematic fitting in case of ep eγnπ hypothesis therefore not as strong Additional selection criteria:! Recoil PID information Lower-cut on ep eγp hypothesis E. Etzelmüller, EHS-TSF at JLab, 18.1.214 21
Beam-spin asymmetries ep eγpπ⁰ A. Airapetian et al, arxiv:131.581; JHEP 14 (in press) E e = 27 GeV, Q 2 = 2.5 GeV 2, x B =.15, t = -.25 GeV 2, = 9 o BSA d/(dq 2 dx B dt ddw N ) (nb/gev 5 rad).35 e - p e - + n.35 e - p e - p.3.3.25.25.2.2.15.15.1.1.5.5 e - p e - + n.3.3 e - p e - p.25.25 Shown amplitudes are corrected for background (only overall fractions are listed here):.2.15.1.5.2.15.1.5.15 Associated DVCS/BH (ep eγpπ⁰) 85 ± 1 Elastic DVCS/BH (ep eγp) 4.6 ±.1 SIDIS (ep exπ⁰) 11 ± 1 1.1 1.2 1.3 W N ( GeV ) 1.1 1.2 1.3 W N ( GeV ) P. A. M. Guichon, L Mossé, M. Vanderhaeghen, Phys. Rev. D 68 (23) 22
Beam-spin asymmetries ep eγnπ+ A. Airapetian et al, arxiv:131.581; JHEP 14 (in press) E e = 27 GeV, Q 2 = 2.5 GeV 2, x B =.15, t = -.25 GeV 2, = 9 o BSA d/(dq 2 dx B dt ddw N ) (nb/gev 5 rad).35 e - p e - + n.35 e - p e - p.3.3.25.25.2.2.15.15.1.1.5.5 e - p e - + n.3.3 e - p e - p.25.25.2.2 Shown amplitudes are corrected for background (only overall fractions are listed here):.15.1.5.1.15.1.5 Associated DVCS/BH (ep eγnπ+) 77 ± 2 Elastic DVCS/BH (ep eγp).2 ±.1 SIDIS (ep exπ⁰) 23 ± 3 1.1 1.2 1.3 W N ( GeV ) 1.1 1.2 1.3 W N ( GeV ) P. A. M. Guichon, L Mossé, M. Vanderhaeghen, Phys. Rev. D 68 (23) 23
HERMES DVCS Summary cos(φ) A C cos φ A C cos(2φ) A C cos(3φ) A C sin φ A LU sin (2φ) A LU sin φ A LU,I sin φ A LU,DVCS sin(2φ) A LU,I HERMES DVCS Hydrogen Deuterium Hydrogen Pure CFFs Re(H) Im(H) The HERMES Recoil detector allows complete event reconstruction for ep eγp and reduces the resonant background to a negligible level sin(φ - φ ) s A UT,I sin(φ - φ ) s A UT,DVCS sin(φ - φ ) cos φ s A UT,I cos(φ - φ ) sin φ s A UT,I cos(φ - φ ) s A LT,I cos(φ - φ ) s A LT,BH+DVCS sin(φ - φ ) sin φ s A LT,I Im(H-E) Re(H+E) First measurement of asymmetry amplitudes in associated DVCS (ep eγπn) cos(φ - φ ) cos φ s A LT,I sin φ A UL sin(2φ) A UL cos(φ) A LL cos φ A LL cos(2φ) A LL -.4 -.3 -.2 -.1.1.2.3 Amplitude Value ~ Im(H) ~ Re(H) A. Airapetian et al, JHEP 6 (28) 66, Nucl. Phys. B 829 (21) 1-27, JHEP 6 (21) 19, Nucl. Phys. B 842 (211) 265-298, JHEP 7 (212) 32, Phys. Lett. B 74 (211) 15-23, JHEP 1 (212) 42, arxiv:131.581; JHEP 14 (in press), JINST 8 P512 (213) 24
Backup 25
Backup Recoil magnet r [cm] 2 15 1 5-1 1 2 3 z [cm] 1.2 1.8.6.4 [T] B z r [cm] 2 15 1 5-1 1 2 3 z [cm].5.4.3.2.1 -.1 -.2 -.3 -.4 [T] B r NbTi-wires with.85 mm diameter mounted on copper rings 166 A operating current 1 T at center 26
Backup SSD (silicon strip detector) helix charge divider HLCU ACC p-side hybrid vacuum feed through sensor HADC HV vacuum n-side hybrid LV pitch adapter 5.8 cm away from lepton beam, 1.5 cm gap sensor thickness 295 um - 315 um thickness of target cell 75 um 27
Backup SFT (scintillating fibre tracker) stereo parallel SFO particle track 1 outer barrel stereo parallel SFO4 SFO3 SFO2 SFO1 stereo parallel SFI 1 inner barrel stereo parallel SFI4 SFI3 SFI2 SFI1 beam 11.5 cm (18.5 cm) inner (outer) radius 1318+132 (2198+218) fibres with a diameter of 1 mm each readout by 64-channel Hamamatsu H7546B MAPMTs 28
Backup Recoil PID Energy Loss [MeV] 14 12 1 8 3 25 2 P + 6 15 4 2-1 -.5.5 1 Momentum [GeV/c] 1 5 2 15 1 PID ( SSD + SFT ) 5-5 -1-15 -2 1.8.6.4.2 Momentum / [GeV/c] discrimination between protons and positively charged pions parent distributions were crucial and determined experimentally 29
Backup Kinematic fitting for ep eγpπ⁰ ep eγpπ⁰ ep eγp SIDIS ep eγpπ⁰ hypothesis ep eγp hypothesis χ²ep eγpπ⁰ < 4.6 χ²ep eγp > 5 Using powerful kinematic fitting of ep eγp hypothesis is crucial for the ep eγnπ analysis 3