One of the most widely known and repeated astrophysical facts is that stars produce all the heavy elements that eventually make planets, shrubberies, and the likes of us. It's absolutely true, but how exactly do they get those elements out into the universe to do all that?
A major route is stellar explosion. When supernovae go off they spew element-rich matter into the cosmos on a big scale, pushing it out for as much as a few hundred light years. But it's not the only way for a star to dig into its pockets to hand out loose change. In fact, all moderately massive stars - from roughly solar mass to several times larger - can go through a phase after they've exhausted the fusion of hydrogen in their cores where they expel huge amounts of material. They do this by periodically inflating their outskirts, and then blowing this matter out to interstellar space. As much as half the mass of the star can be cast off this way. This freshly produced star-stuff consists of both gas and microscopic dust grains that are produced as the gas cools down away from the star and quite literally condenses out, forming silicate or carbon particles (the latter from lower mass stars). This hazy outflow dumps new elements into space to produce some beautiful structures, known rather confusingly as 'planetary' nebula.
The problem is that we haven't fully understood how stars perform this trick. The only tool they have at their disposal is the pressure of stellar photons - light flooding from the star can push and accelerate material away from it. However, getting this light to push against the gas of the stellar atmosphere efficiently enough to set it in motion has seemed difficult. One option is that the tiny grains of dust act like miniature solar sails, that in turn snowplough through the gas to accelerate it along in front of them. However this theory has had some gaps in it; figuring out the necessary combination of dust grain composition, size, and location of formation has been tricky.
A new investigation recently published by Norris et al. in the journal Nature (and discussed in an excellent companion piece by Susanne Höfner) exploits some very clever astronomical observations of flatulent old stars to find a possible solution. By studying the polarized light from a number of these systems, and by using interferometric techniques, the authors were able to test the properties of dust that appears to be produced remarkably close to stellar surfaces, at barely a couple of stellar radii away.
The dust particles are surprisingly large, with diameters of about 600 nano-meters (0.0006 millimeters), and must be quite transparent to the stellar light or else they would be boiled away as they absorbed radiation. Yet this would seemingly make them poor solar sailors. The solution to this conundrum is that these are silicate grains (perhaps magnesium silicate) that scatter the starlight rather than absorb it, like rather rough mirrors. These 'big' grains can be readily pushed outwards at speeds of 20,000 miles an hour, and they will sweep up anything in their way.
Thus, the dispersal of elements into the cosmos may owe a lot to a most peculiar type of sandstorm, taking place in the messiness around dying stars. This remarkable process may be critically important to understand for cosmological reasons as well. Höfner points out that the more massive stars that go through this stage are also the likely progenitors of Type Ia supernovae, the explosions cosmologists use to track the changing expansion rate of the universe. Proper knowledge of the true, sandy environment of these vital yardsticks would be a very good thing.