Revealing the accelerating wind in the inner region of the colliding-wind binary WR 112

TL;DR

WR 112, a long-period, dust-producing WC+O binary, is studied with Chandra and Swift X-ray imaging, extensive radio monitoring, and high-resolution Keck infrared imaging to connect inner wind dynamics with the extended dust spiral. The authors test circular, eccentric, and accelerating wind models against multi-scale data and find that only an accelerating post-shock wind consistent with dust formation and radiative driving can reproduce the near-infrared dust morphology and the phase-locked radio variability. The resulting framework links hot-plasma physics, non-thermal radio emission, and dust production in a self-consistent picture, providing critical boundary conditions for future hydrodynamic simulations of colliding-wind binaries. WR 112 thus joins WR 140, WR 104, and Apep as a benchmark system for exploring the diversity of orbital architectures and wind physics in massive-star binaries.

Abstract

Colliding winds in massive binaries generate X-ray-bright shocks, synchrotron radio emission, and sometimes even dusty "pinwheel" spirals. We report the first X-ray detections of the dusty WC+O binary system WR 112 from Chandra and Swift, alongside 27 years of VLA/ATCA radio monitoring and new diffraction-limited Keck images. Because we view the nearly circular orbit almost edge-on, the colliding-wind zone alternates between heavy Wolf-Rayet wind self-absorption and a near-transparent O-star wind foreground each 20-yr orbit, producing phase-locked radio and X-ray variability. This scenario leads to a prediction that the radio spectral index is flatter from a larger non-thermal contribution around the radio intensity maximum, which is indeed observed. Existing models that assume a single dust-expansion speed fail to reproduce the combined infrared geometry and radio light curve. Instead, we require an accelerating post-shock flow that climbs from near-stationary to ~1350 km/s in about one orbital cycle, naturally matching the infrared spiral from about 5" down to within 0.1", while also fitting the phase of the radio brightening. These kinematic constraints supply critical boundary conditions for future hydrodynamic simulations, which can link hot-plasma cooling, non-thermal radio emission, X-ray spectra, and dust formation in a self-consistent framework. WR 112 thus joins WR 140, WR 104, and WR 70-16 (Apep) as a benchmark system for testing colliding-wind physics under an increasingly diverse range of orbital architectures and physical conditions.

Figures (12)

• Figure 1: WR 112 N-band ($\lambda\sim12\mu$m) images from 2 epochs lau2020. Note the different orientations of the U-shaped structures within $2"$ of the central binary, indicative of orbital motion. The rightmost figure shows a dust model applied to the 2016 N-band image. Orbital motion and the expansion of the dust arcs in the wind from WR 112 reproduces the observed spatial distribution of the inner and outer dust arcs projected on the sky.
• Figure 2: Left: False-color Broad-band (0.2 --10 keV) CHANDRA image of the WR 112 field from ObsID 25129; Center: CHANDRA image, ObsID 29739; Right: 2MASS J-band ($\lambda\sim1.23\mu$m) image of WR 112. Contours from the 2MASS image are overlaid on the CHANDRA images. Note that, aside from WR 112, there are no other J-band sources clearly detected in the CHANDRA X-ray images.
• Figure 3: The 0.1 -- 10 keV band CHANDRA X-ray lightcurve of WR 112, using 500 second bins.
• Figure 4: The summed Chandra ACIS X-ray spectrum of WR 112. The data are binned for visual clarity. The best-fit two-temperature plasma model is shown in red (tbabs$\times$vapec$+$vapec). The kT$=0.5$ keV and kT$=2$ keV components are shown in cyan and blue, respectively. The strongest lines in the model are He-like Si XIII, S XVI, Ar XVII, and Fe XXV (1.87, 2.46, 3.14, and 6.70 keV, respectively). The spectrum is obscured at low energy by the high line-of-sight column density. See Table 3 for model parameters and errors.
• Figure 5: Left: The 8 GHz light curve of WR112, including synthesized flux estimates when only nearby frequencies were available. Right: The spectral index correlation with flux, as WR 112 transitions from the "high state" dominated by non-thermal radio emissions to the "low state" where we only see thermal (free-free) emission from the ionized winds.