BARELY 700 million years after the big bang, the universe was filled with dust that went on to form stars and planets 鈥� and eventually us too. This dust is thought to have been produced by supernovae that signalled the end of the first generation of stars.
But there has been a problem with this theory: the remnants of more recent supernovae appeared to have far less dust than would have been necessary for supernovae to be the 鈥渄ust factories鈥� of the universe.
Now a pair of studies may have solved the mystery of the missing dust. Both teams used NASA鈥檚 Spitzer infrared telescope to estimate the amount of dust around type II supernovae, which occur when stars bigger than 10 times the mass of the sun are destroyed in a cataclysmic blast.
Advertisement
Ben Sugarman of the Space Telescope Science Institute in Baltimore, Maryland, and colleagues looked at the remnants of supernova 2003gd in the spiral galaxy NGC628, about 35 million light years away. They estimate that in the two years since 2003gd lit up its galaxy, enough dust to make 7000 Earth-mass planets has condensed around it. That鈥檚 10 times the quantity found in any previous supernova observation, and suggests that such supernovae could indeed have created the dust observed in the early universe (Science, DOI: 10.1126/science.1128131).
So why have previous observations seen so little dust around supernovae? The key may be in the timing. Snezana Stanimirovic of the University of California, Berkeley, announced last week at the meeting of the American Astronomical Society in Calgary, Alberta, that her team has seen only the usual, paltry amount of dust around E0102, a 1000-year-old supernova remnant. In that time the dust could have cooled enough to evade detection or may even have been destroyed. 鈥淪hock waves created when the expanding supernova remnant strikes the surrounding environment can destroy dust,鈥� Stanimirovic says. But she argues that supernovae may not be the only answer. 鈥淲e need another source of dust in the early universe,鈥� she says.
Both sides plan to use the Spitzer satellite to see more supernovae. It is sensitive enough to pick out warm dust out to a distance of around 60 million light years, and this should bring enough supernovae into view to settle the issue of how much dust they produce.
If supernovae are the major source of dust in the early universe, it raises the possibility that the first planets formed around the dying embers of the first stars 鈥� unlike the planets in our own solar system, which formed from the dust left over from the material that coalesced to form the sun.
The quest for dust
Lyman Spitzer of Princeton University, after whom the NASA Spitzer telescope is posthumously named, developed a model of dust formation in the 1960s. He suggested that dust forms over hundreds of millions of years as individual atoms occasionally collide in interstellar space.
In the 1980s, the IRAS satellite detected more infrared radiation than expected coming from some stars. This was interpreted by astronomers as evidence that the dust grains grew much more quickly 鈥� over timescales of just a few decades 鈥� in the cooler atmospheric gases of the stars, rather like soot forming in the cooler parts of a flame.
Also around this time, theorists began to wonder if dust could form in supernova debris. In 1988, Eli Dwek of NASA estimated that dust with a mass equal to the mass of the sun might be formed in a supernova blast, and that this might almost blot out our view of the recently exploded supernova 1987a. However, observers detected only a thousandth of that quantity.
This may be explained by the dust being clumpy, rather than uniformly spread. Estimates of the amount of dust around supernovae are very sensitive to assumptions about clumping, according to Ben Sugarman of the Space Telescope Science Institute in Baltimore, Maryland. If the dust is assumed to be a uniform cocoon, estimates of its mass can be up to ten times lower than if the distribution is assumed to be clumpy.
