Taming the Tarantula: How Stellar Wind Feedback Shapes Gas and Dust in 30 Doradus

Abstract

Observations of massive star-forming regions show that classical stellar wind models over-predict the luminosity of the X-ray emitting gas, indicating a significant fraction of wind energy is lost. In this paper, we present a multi-wavelength analysis of the giant HII region 30 Doradus and its central star cluster R136 using 2 Ms of Chandra X-ray Observatory data, combined with James Webb Space Telescope and Hubble Space Telescope imaging and Spitzer spectral-energy distributions, to investigate how the hot gas energy is lost through turbulent mixing, radiative cooling, and physical leakage. We compare the spatial and spectral properties of the hot gas with those of the warm ionized gas and dust. We find no significant correlation between the dust and hot gas temperatures, suggesting they are not directly coupled and that the dust resides in the swept-up shells where it is heated radiatively. H$α$ and X-ray surface brightness profiles show that the X-rays peak interior to the H$α$ shells, demonstrating partial confinement of the hot gas. The fragmented shell structure and bright X-ray interior that declines near the H$α$ shell reflect efficient cooling from turbulent mixing at the hot-cold interface. We compare against recent simulations of stellar-feedback driven bubbles which have broad agreement with the morphology of the X-ray and H$α$ emission, but the simulations produce a dip in the interior X-ray surface brightness and a lack of hard X-rays compared to the observations. These differences may suggest thermal conduction is important as mass-loading of the hot bubble could reproduce the X-ray observables.

Figures (12)

• Figure 1: Composite X-ray images from 2 Ms of Chandra observations of 30 Doradus showing the complex structure of at least five superbubbles. a) Soft X-ray (0.5$-$1.2 keV) image of 30 Doradus. The white 50$^{\prime\prime}$$\times$50$^{\prime\prime}$ box was used for background-subtraction. b) Medium X-ray (1.2$-$2 keV) image of 30 Doradus. c) Hard X-ray (2$-$7 keV) image of 30 Doradus. d) Three-color X-ray image of 30 Doradus, where soft X-rays (0.5$-$1.2 keV) are in red, medium X-rays (1.2$-$2 keV) are in green, and hard X-rays (2$-$7 keV) are in blue. The cyan X marks the location of R136, and the cluster is removed from the images. At a distance of 50 kpc, 1$^\prime$ is about 14.6 pc, as shown by the scale bar. The image is 22$^\prime$$\times$22$^\prime$. North is up, and East is left.
• Figure 2: a) JWST NIRCam F335M observation of R136, tracing PAH emission. b) Spitzer 24$\mu$m observation, tracing warm dust heated by star formation. c) Chandra broad-band (0.5$-$7 keV) X-rays, tracing the hot gas, zoomed in on R136. The 84 (42$^{\prime\prime}$$\times$42$^{\prime\prime}$) regions used for X-ray spectral extraction are shown in white. d) 3-color image of 30 Doradus, with JWST NIRCam F335M in red, Spitzer MIPS 24$\mu$m in green, and broad-band X-rays (0.5$-$7 keV) in blue. The cyan X marks the location of R136. At a distance of 50 kpc, 1$^\prime$ is about 14.6 pc as shown by the scale bar. Images are 7$^\prime$$\times$9$^\prime$. North is up, and East is left. The hot gas fills the IR cavities, and the dust overlaps with the X-ray emission, suggesting that the dust may either occupy the same volume as the hot gas or remained confined to the shell.
• Figure 3: a) HST F658N image, corresponding to H$\alpha$. b) Chandra broad (0.5--7 keV) X-ray image with the regions used for the X-ray spectral extraction for the cavities seen in H$\alpha$ around R136. c) Composite image of the HST H$\alpha$ in green and Chandra broad X-ray in blue. The location of R136 is denoted by the cyan X and the magenta crosses mark the locations where we performed our surface brightness analysis (see Figures \ref{['fig:SB_R136']}--\ref{['fig:SB_smallcavity']}). At a distance of 50 kpc, 1$^\prime$ is about 14.6 pc as shown by the scale bar. Images are 5.5$^\prime$$\times$5.5$^\prime$. North is up, and East is left. The hot gas fills the H$\alpha$ cavities, suggesting the hot gas may be confined by a shell of cooler gas.
• Figure 4: Top: Map of the X-ray temperature $kT$ (keV) of 30 Dor derived from the spectral fits of the 84 regions (see Figure \ref{['fig:84regions']}) selected to explore the relationship between the hot gas and dust. Typical errors of hot gas temperature were $\sim$10%. Middle: Map of the column density $N_{\rm H}$ ($\times10^{22}$ cm$^{-2}$) of 30 Dor, with $\sim$10% error. Bottom: Map of the X-ray electron density $n_{\rm X}$ (cm$^{-3}$) of 30 Dor. The white contour shows the JWST NIRCam F335M observation. Each region is 42$^{\prime\prime}$$\times$42$^{\prime\prime}$. The magenta X marks the location of R136. North is up, and East is left. The X-ray temperature peaks at R136, and higher X-ray temperatures coincide with the JWST NIRCam F335M contours, suggesting that the hot gas from stellar feedback may influence the dust through heating or destruction.
• Figure 5: Top: Map of the X-ray temperature $kT$ in keV from the spectral fits of the 60 regions surrounding R136 and nearby cavities to explore relationship between hot gas and warm gas (see Figure \ref{['fig:hst_chandra']}), with $\sim$10%--20% error per region. Bottom: Map of the column density $N_{\rm H}$ ($\times10^{22}$ cm$^{-2}$), with $\sim$10% error per region. Region sizes vary and several are combined to account for low signal. The magenta X marks the location of R136. North is up, and East is left. HST H$\alpha$ contours are shown in white, highlighting the cavities east and north of R136. The values for each region correspond to the best-fit temperature from the 0.5$-$3 keV spectral analysis. The regions overlapping with the cavities east and north of R136 generally have low column densities and higher hot gas temperatures, though the eastern cavity shows relatively lower temperatures, suggesting a complex interplay of stellar feedback on the surrounding ISM.