Factorization breakdown of two-particle correlations in pPb and PbPb collisions at CMS

Technique of two-particle correlations has been widely used in studying azimuthal anisotropy flow in relativistic heavy-ion collisions. A key assumption imposed in this approach is the factorization of Fourier coefficients extracted from two-particle correlations into a product of single-particle anisotropies of trigger and associated particles. It is recently predicted by hydrodynamics that due to initial-state participant fluctuations, a transverse momentum ($p_{T}$) dependence of event-plane angle would be induced, leading to a breakdown of factorization, even if hydrodynamic flow is the only source of correlations. We present a systematic examination of the factorization assumption in 5.02~TeV pPb and 2.76~TeV PbPb collisions with the CMS experiment. Significant breakdown of factorization (up to $20\%$) is observed in a large sample of ultra-central ($0-0.2\%$) triggered PbPb events, where initial-state fluctuations play a dominant role. Comparison of data and viscous hydrodynamics predictions, as a function of $p_{T}$ and centrality, allows new constraints on the modeling of initial condition and shear viscosity to entropy density ($\eta/s$) ratio of the medium created in heavy-ion collisions. Furthermore, the measurement is also extended to high-multiplicity pPb collisions. As the initial-state geometry of a pPb collision is expected to be entirely a consequence of fluctuations, quantitative studies of factorization breakdown will help to investigate the nature of the observed long-range correlations in pPb collisions, particularly in the context of hydrodynamic models.