1. 1 Polarimeter for dEDM experiment G. Venanzoni Laboratori Nazionali di Frascati for the dEDM collaboration Workshop on Flavour in the era of LHC – Cern 9-11 Oct 2006

2. 2 A typical experimental layout contains: beam direction scattering plane y x z left detector β right detector target (φ=0) A polarization of the beam (p) causes a difference in the rates for scattering to the left compared to the right. analyzing power (determined by nuclear effects in scattering) governs spin sensitivity unpolarized cross section (determined by nuclear effects in scattering) governs efficiency φ θ left and right detectors useful for vector polarization vertical component of the polarization For protons (S=1/2):

3. 3 A typical experimental layout contains: beam direction scattering plane y x z left detector β right detector target (φ=0) A polarization of the beam (p) causes a difference in the rates for scattering to the left compared to the right. φ θ left and right detectors useful for vector polarization For deuterons (S=1): analyzing power (determined by nuclear effects in scattering) governs spin sensitivity unpolarized cross section (determined by nuclear effects in scattering) governs efficiency

4. 4 Left-Right and Down-Up Asymmetries  In dEDM experiment we are looking for an increase with time of the vertical polarization  p V,y  over the length of a beam store  By making measurement at  1 and  2 =  1 + , we look at the Left-Right (LR) asymmetry, which is sensitive to p V,Y :  In the same way, the Down-Up (DU) asymmetry is sensitive to the component in the plane p V,x  LR Asymmetry due to tensor polarization will have other distinctive signature (phase and oscillation), and is not expected to increase with time  LR will increase with time  DU will oscillate with g-2 frequency

5. 5 Up Down Right Left Carbon target Iron absorber Segmented scintillator Functional polarimeter elements (must run continuously) Left-right asymmetry carries EDM information. Down-up asymmetry carries information on g-2 precession. Carbon target is small. Two possibilities for the polarimeter: extract using jet in ring onto annular target tune beam onto annular target slow extraction into second beam Angle range: 4° - 15°

6. 6 angle D L U R R D Δ “extraction” target - gas “defining aperture” primary target detector system Target could be Ar gas (higher Z). Target “extracts” by Coulomb scattering deuterons onto thick main target. There’s not enough good events here to warrant detectors. Events must imbed far enough from hole to not multiple scatter out of primary target, thus Δ << D. Δ, which is a large fraction of the deuteron range, sets scale for polarimeter. Hole is large compared to beam. Every- thing that goes through hole stays in the ring. (It may take several orbits to stop scattered particle.) Detector is far enough away that doughnut illumination is not an acceptance issue: Δ < R. Primary target may need to be iris to allow adjustment of position and inner radius. It may also need to be removed during injection. Layout of the Polarimeter (if it were situated on the ring)

7. 7 Polarimeter optimization ingredients   unp (  )   unp (  ) (largest at forward angle) iT 11 (  )  Vector analyzing power iT 11 (  ) (usually increases at larger angles)  unp (  ) and iT 11 (  ). (FOM)  The solution is often a compromise between  unp (  ) and iT 11 (  ). It’s common to define a figure of merit (FOM) :  = error in the measurement of the p V components

8. 8 Current design: p d = 1.0 -1.5 GeV/c  T d = 250 – 525 MeV Deuteron data Previous design: p d = 0.7 GeV/c  T d = 126 MeV Data: POMME at Laboratoire National Saturne (France) (B.Bonin et al. (1990), V.P. Ladygin et al. (1998) ) T d in (0.175 – 1.8) GeV θ in: [4°, 15°] per T d < 300 MeV [2°, 20°] per T d > 300 MeV SMART at RIKEN (Japan) (Y. Satou et al.): T d in (200 – 300) MeV Data: S. Kato et al., NIM A 238 (1985) 453-462 T d in (35 – 70) MeV θ in: [30°, 65°] Data also from Test beam at KVI ( Groningen, NL)

9. 9 Data from Pomme polarimeter at energies > 200 MeV Work here. efficiency (%) average iT 11 momentum (GeV/c) iT 11 (  )  unp (  ) An iron absorber was placed to remove non elastic particles from the scattered flux. At about 700 MeV it loses its effectiveness and iT 11 starts to decline. wire chambers Carbon target Iron absorber

10. 10 Data from 270-MeV deuteron elastic scattering – RIKEN Optimize here: favor larger analyzing power, leverage against systematics Optimize here: favor statistical precision At this energy (p~1 GeV/c) these two choices lead to different angle covarage. together But as momentum rises FOM and analyzing power peak together. iT 11 5o5o 10 o 14 o 18 o 24 o iT 11 Lab angle (deg) 20 o 30 o 10 o Lab angle (deg) 20 o 30 o FOM  700 MeV Two possibilities:

11. 11 Momentum dependence of FOM, iT 11 and efficiency The SOLID dots and lines follow the FORWARD peak in the FOM curve. The open/dashed dots and lines follow the analyzing power peak (where there is enough data to use). The AVERAGES shown here integrate over some angle range that covers the relevant feature in FOM or iT 11. Satou gets even larger analyzing powers by cutting out more protons and losing “efficiency”. The FOM is down about 30% from the open dots.

12. 12 run time 6·10 7 s polarimeter efficiency 2% particles per fill 10 12 spin coherence time 1000 s 5.7·10 -10 rad/s p = 1.5 GeV/c 0.36x0.6 Polarimeter statistical error At d=10 -29 e  cm effective beam use fraction 0.5 2 years to arrive to  d =10 -29 e  cm

13. 13 1 Displacement / angle errors detectors θ angle shift θ x position shift Remedy: measure on both sides (L/R) flip initial spin opposite sign for EDM accumulation spin detector left/right efficiency differences cancel +/ luminosity differences cancel Polarimeter systematic errors (examples of issues) (What other sources arise for a left/right asymmetry?)

14. 14 Errors that are second-order in θ and u=p + +p - Example: With  max  10 -7 and requiring  < 10 -5   <0.02 o A is the analyzing power  is the difference in the mean acceptance angle of the detector btw the + a – polarization states. For LR asymmetry  is the angular shift of the detector/target.

15. 15 2 Polarimeter rotation L R U D φ target We are helped by the time dependence of ε DU and φ-dependence in analysis of segmented detector. 3 Parity violation Effects start to appear at ε < 10 -6 associated with p x. We are helped by time dependence. Polarimeter systematic errors (cont’d)

16. 16 4 For spin = 1, tensor contributions (t 21 ) detectors The left/right asymmetry is maximal along 45 ° but reverses sign in the perpendicular direction. Tensor polarization requires and equal population of m=1 and m=-1 deuterons that is different from m=0. The left/right asymmetries oscillate as the spin rotates in the ring plane. They do not grow with time. The effect appears at 10 -4 with ~1% tensor contamination of the polarized beam. Polarimeter systematic errors (cont’d)

17. 17 Conclusion and outlook  Considerable effort in the past to develop a set of data on which to base the design for the polarimeter (at p=0.7 GeV/c)  At the current design (1.0-1.5 GeV/c) we need to define/study:  “extraction” of the beam  operating momentum of the ring  sensitivity of the polarimeter to different error sources  thickness of the target  segmentation of the detector  Readout and DAQ  Test beam-polarimeter interactions (at COSY), in the case we decide to have the polarimeter in the ring  Prepare (and test) prototype