2016MNRAS.461.2309R

This paper presents a component-level comparison of the polarized v=1 J=1-0, v=2 J=1-0 and v=1 J=2-1 SiO maser emission towards the supergiant star VY CMa at milliarcsecond-scale, as observed using the Very Long Baseline Array at λ=7 and 3 mm. An earlier paper considered overall maser morphology and constraints on SiO maser excitation and pumping derived from these data. The goal of the current paper is to use the measured polarization properties of individual co-spatial components detected across multiple transitions to provide constraints on several competing theories for the transport of polarized maser emission. This approach minimizes the significant effects of spatial blending. We present several diagnostic tests designed to distinguish key features of competing theoretical models for maser polarization. The number of coincident features is limited by sensitivity however, particularly in the v=1 J=2-1 transition at 86 GHz, and deeper observations are needed. Preliminary conclusions based on the current data provide some support for: (i) spin-independent solutions for linear polarization; (ii) the influence of geometry on the distribution of fractional linear polarization with intensity; and, (iii) π/2 rotations in linear polarization position angle arising from transitions across the Van Vleck angle (sin ²θ=2/3) between the maser line of sight and magnetic field. There is weaker evidence for several enumerated non-Zeeman explanations for circular polarization. The expected 2:1 ratio in circular polarization between J=1-0 and J=2-1 predicted by standard Zeeman theory cannot unfortunately be tested conclusively due to insufficient coincident components.