some stars blow themselves a cocoon
In spectroscopic notation, the Roman numeral after an atomic symbol refers to its ionization state. O VI is 5-times ionized: it’s missing 5 electrons.
Astronomers have observed light with wavelength around 1030Å from many stars. This doesn’t quite match hydrogen or H₂ absorption lines, but it does fit the profile of 5-times-ionized oxygen.
But how could this happen? Most stars have effective temperatures of 30,000-50,000 Kelvin, which is way too cold for O VI.
One exciting answer proposed by Castor, McCray, and Weaver is that stars live in bubbles. These bubbles, formed by solar winds, have outer edges dense enough to support oxygen ions through O VI and potentially, molecules like H₂.
Stars produce heat through nuclear fusion. The atoms on the surface vibrate faster (remember: heat makes atoms vibrate), some of which escape the Sun’s gravity.
Note: some winds are formed by radiation pressure, where light literally pushes particles outwards.
At first, these atoms freely expand into the surrounding space. This is the wind! They quickly become supersonic (see Parker’s model), sending a shock wave into any interstellar gas they meet.
This eventually causes the wind itself to be shocked, heating it up and sending a negative-pressure blast wind back into the star. But overall, the wind keeps moving outwards, now transferring its energy into the shocked-gas (similar to a supernova remnant; see the Sedov-Taylor solution for more details).
At some point, enough energy is lost from radiation (i.e. light emission), and the wind and gas continue to expand, “snowplowing” their way through more interstellar gas. This is often called the “momentum-conserving” phase, and we can measure the photon energies from this radiation.
In the snowplowing phase, radiative cooling dominates energy loss from the shocked gas, collapsing the region into a thin bubble’s edge. This dramatically increases density. The super-hot shocked wind still conducts with this edge, causing super-ionized atoms to slam into other atoms and neutralize.
Note: when ions decelerate, they emit light, in this case measured in the soft X-ray spectrum.
Castor, McCray, and Weaver calculate that the expected densities of O VI match up with observations from the Copernicus satellite. Also, they calculate that the sphere of ionized hydrogen likely ends within the shell. This makes it possible for neutral H and thus H₂ to form in the edge of the bubble.
This complicates some understanding of Wolf-Rayet stars, where the heavier element spectra may come from the bubble itself, not the dying star’s fusion. Stars also can be moving, which stretch the bubble from a perfect sphere. For example, the Sun’s motion morphs the heliosphere (its bubble) into a comet-like shape.
Supernova also form bubbles from their explosions, often much bigger. We live inside a bubble likely formed from a supernova.
These bubbles may be hard to distinguish from solar wind bubbles, for example in RCW 114. Some blurring in the spectra comes from Doppler broadening, which means the star is still blowing a wind, as opposed to a leftover gas cloud from the supernova.
Although there are massive objects in relatively few places in the universe, gas and plasma may have blown from these objects through much of the universe. This would make interstellar space heterogeneous and clumpy, especially when bubbles collide into superbubbles.
The Universe may look very different from bubble to bubble!
I gave this as a talk in May 2019 as part of Eugene’s Astrophysical Fluid Dynamics course. Please send me a note if you have thoughts or corrections.