Speaker
Description
We present the transverse single spin asymmetries $A_N$ for very forward neutrons measured by the PHENIX zero-degree calorimeters (ZDCs) in high-energy polarized proton-nucleus (p-A) collisions at RHIC. First-ever p-A data taken in the RHIC-2015 run exhibit positive and remarkably large $A_N\sim0.18$ only in p-Au collisions whereas nearly zero $A_N$ in p-Al collisions. Comparisons of these new results with $A_N=-0.05$ from the previous PHENIX-ZDC measurement in p-p collisions raised the following question: what mechanisms do produce such a large $A_N$ in only p-Au collisions and smaller or close to zero $A_N$ in other collisions? One attempt is a single-Reggeon exchange model considering the interference of pion (spin flip) and $a_1$-Reggeon (spin nonflip). Single-Reggeon exchange well explains $A_N$ in p-p collisions and can be extended to p-A collisions by taking into account strong nuclear absorption correction as well as nuclear breakup. However, the predicted $A_N$ by the extended single-Reggeon exchange model is far smaller than the PHENIX data in p-Au collisions. In this talk, we will discuss an alternative and rather unknown mechanism: ultra-peripheral p-A collisions (UPCs, also known as Primakoff effects). UPCs lead to very large $A_N$ of about 0.35 and have cross sections proportional to $Z^2$ of the nuclei. UPCs contribute to inclusively measured $A_N$ modestly in p-Al collisions and significantly in p-Au collisions. To discuss quantitatively, we will show that the Monte Carlo simulations incorporating both the Reggeon exchange and UPCs, where virtual photon--polarized proton interactions follow the MAID 2007 isobar model, successfully describe the PHENIX data in p-Al and p-Au collisions simultaneously.
