Transverse Spin Asymmetries in the CNI region of elastic proton-proton scattering at √s=200 GeV

Precise measurements of transverse spin asymmetries in proton-proton elastic scattering at very small values of four-momentum transfer squared, t, have been performed using the Relativistic Heavy Ion Collider (RHIC) polarized proton beams. The measurements of both single and double spin asymmetries were made at the center-of-mass energy s = 200 GeV and in the region 0.003 ≤|t|≤ 0.035 (GeV/c)2, which was accessed using Roman Pot devices incorporated into the STAR experimental setup. The obtained set of asymmetries is sensitive to the poorly known hadronic contribution to the spin-flip amplitudes and provide significant constraints for the theoretical descriptions of the reaction mechanism of proton-proton elastic scattering at high energies.


Double spin asymmetries
A P P n A P n P n A P s P s

Luminosity monitors: BBC and ZDC
• Manifests as a shift in raw asymmetry, not a scaling factor ! • Main assumption: two different processes can have the same spin sensitivity only in the case if it is zero in both of them • Numerically, two (or more) processes can be considered free of spin effects if they give zero difference in corresponding independent normalization ratios R 2 , R B , R Y • Careful choice of STAR subsystems as luminosity monitorsbest consistency checks, high statistics, two arms each: • ZDCzero degree calorimeters • BBCbeam-beam counters • Difference in R 2 normalization ratio for BBC and ZDC is systematically shifted from zero at the level of 1.4•10 −3 , averages out with fake polarization pattern • Have to conclude: one of the two monitors feels double spin effects at this level => further investigation  A N result confirms conclusions at lower energies: r 5 =0, no hadronic single spin flip is seen  (A NN −A SS )/2 found to be close to 0, i.e.A NN ≈A SS and ϕ 4 ≈ 0 as required by angular momentum conservation  (A NN +A SS )/2 is significantly different from zero and have small negative values about 5•10 −3  r 2 is well constrained in its imaginary part and compatible with zero in real part  r 4 has large errors as expected and hypothesis r 4 =0 is not controversial  Hadronic double spin-flip amplitudes ϕ 2 and ϕ 4 behave differently at our kinematic range.This indicates that the exchange mechanism is more complex than an exchange of Regge poles only (factorization of amplitudes does not hold).
 Results are at variance with the latest models and may attract the attention of theorists

*
ITEP) for the STAR Collaboration  Roman Pots integrated with STAR detectorclosest proximity to the beam. CNI region : 0.003 < −t < 0.03. Ideal beam optics: β*= 21m and parallel to point focusing terms other than L EFF in the transport matrix very small. High transverse polarization of both beams ~60%. Excellent detector performancenearly 100% efficiency and only 5 dead/noisy strips per ~14000 active strips.Detector and experimental conditions Unique possibility to measure A N , A NN , A SS and, in general, A LL at collider energies Averaged for 4 fills from the official Run'09 RHIC polarimeter data: https://wiki.bnl.gov/rhicspin/Resultsx D  L x eff Θ x * y D  L y eff Θ y * Dmitry Svirida (ITEP) for the STAR Collaboration 2π acceptance in azimuthal angle. Exactly the same sample of elastic events for A N and A NN &A SS studies  2ITEP) for the STAR Collaboration Single spin asymmetry A N Re r 5 = 0.0017 ± 0.0063 Im r 5 = 0.007 ± 0.057  Statistical error of <3% in each of 5 points  Many systematics checks  Highest accuracy in extraction of r 5 Phys.Lett.B719 (2013) 62-69  -Square root formula‖no need for external luminosity normalization Re r 5 = -0.033±0.035Im r 5 = -0.43±0.56no hadronic spin-flip STAR Dmitry Svirida (ITEP) for the STAR Collaboration Cross section azimutual angular dependence for transversely polarized beams:

-•
is the vector in the scattering plane, normal to the initial momentum ; -polarization vectors of the two beams cos 2φ dependence for A NN A SS  NO angular dependence for A NN +A SS  Cannot use square root formula  Must use external normalization Luminosity monitors must not have double spin effects • For normalization studies use 3 independent luminosity ratios (Ntotal monitor count) : R 2 = (N ++ + N --)/Nfraction of parallel spin interactions (double spin) R B = (N ++ + N +-)/Nfraction of interactions with spin UP in Blue beam R Y = (N ++ + N -+ )/Nfraction of interactions with spin UP in Yellow beam • Uncertainties:


Difference in R 2 double spin normalization ratio for BBC and ZDC Dmitry Svirida (ITEP) for the STAR Collaboration Luminosity monitors: inside BBC • Three different subprocesses were compared to each other and to BBC as a whole in terms of R 2 normalization ratio:  High multiplicity: N>5 tiles in one arm and a hit in the opposite arm  Inner: single hit in one of the Inner tiles and a hit in the opposite arm  Outer: single hit in one of the Outer tiles and a hit in the opposite arm • Confirmed that the subprocesses have significantly different physics -angular dependence of single spin ratios R B , R Y is of opposite sign • East and West arms show extremely good consistencyaverage them • Though R 2 spread is relatively large fill by fill, the averages for our 4 fills are very close to zero at 10 −4 level • Averaged ΔR 2 of each subprocess and the whole BBC are added in quadratures to form the total uncertainty δR 2 Very clean data set of ~20 million pp-elastic events is taken with Roman Pots integrated into the STAR detector for low t studies  Both single and double spin asymmetries are extracted with unprecedented accuracy in 5 t-ranges at s=200 GeV.