A Spectropolarimetric Study of Southern WR + O Binaries

The classical Wolf-Rayet (WR) state is the evolved stage of a massive star, post main-sequence. They are characterized by their strong emission line spectra and stellar winds that are often more than 10 times denser than that of their progenitor O-type stars, which have mass loss rates of 10-6 MΘyr-1. The evolution of WR stars and their connection to specific types of supernovae (SNe) is an open question. Current theory suggests that rapidly rotating massive stars may be the progenitors of SNe that produce long-duration gamma-ray bursts. The interaction between WR stars and their companion in binary systems may provide sufficient angular momentum to create such progenitors. Angular momentum (and therefore rotation) tends to create aspherical structures in astronomical objects (e.g. Be star disks, T Tauri jets caused by decretion and accretion respectively) that can be investigated using linear polarimetry, even for unresolved sources. I have investigated WR stars in detail to determine the geometric structure of their winds using spectropolarimetry. I began by using archival broadband polarimetric data to search for intrinsic polarization in a sample of more than 40 single and binary WR stars, finding that 12 of the stars exhibit intrinsic continuum polarization or line polarization effects that indicate aspherical or non-uniform winds. In the later stages of the project, I used the Southern African Large Telescope (SALT) to obtain time-dependent spectropolarimetric observations of 10 of the stars in that sample, along with 8 additional targets. These targets are all WR + O binary systems, whose complex winds are best observed over time with spectropolarimetry to determine the geometry of the wind across different emission lines and the continuum. I investigated two stars in the sample, WR 42 and WR 79, and found that they exhibit classic continuum polarization signatures of binary orbits, as well as intriguing orbital line polarization effects. I compared the line polarization behaviour with the predictions of existing spectrally-derived models of the systems to obtain new information about the structure of the colliding wind regions. Finally, I have modified an existing 3-D Monte Carlo radiative transfer code to include an additional source of photons that represents a companion star. This allows the code to treat the asymmetric structures seen in massive binary systems. I used this updated code to simulate the well-observed WR + O system V444 Cygni. I created a set of emission regions to simulate line emission from both the WR wind and wind-wind collision regions, finding that the wind-wind collision creates very strong polarimetric signals that appear similar to those in other systems in my SALT sample. The results shed new light on the relationships among WR + O binaries and yield clues to their subsequent evolution and potential roles as SN and GRB progenitors