To look for DIBs, an astronomer points the telescope at a star and scans through a rainbow made up of thousands of wavelengths of light. This rainbow, or spectrum, is extended a bit beyond visible light, into the UV at the blue end and into the infrared at the red end.

DIBs are not defined by what astronomers see while doing this, but by what they don't see. The colors missing from the rainbow, marked by black stripes, are the ones of interest. Each one is a wavelength being absorbed by some kind of atom or molecule.

A DIB is one of these regions where the color is missing. But compared to the nice, neat "absorption lines" that are identified with atoms or simple molecules, a DIB is not well-behaved, which is why it stands out.

"Astronomers were used to seeing quite sharp, narrow bands where typical atoms and molecules absorb," says Cordiner. "But DIBs are broad; that's why they are called 'diffuse.' Some DIBs have simple shapes and are quite smooth, but others have bumps and wiggles and may even be lopsided."

The mystery deepens

Over time, astronomers have been building up catalogs to show exactly which wavelengths are absorbed by all kinds of atoms and molecules. Each molecule has its own unique pattern, which can be used like a fingerprint: if a pattern found during an astronomical observation matches a pattern in one of the catalogs, the molecule can be identified.

It's a pretty straightforward concept. So, early researchers "would surely not have thought that the solution to the diffuse band problem would still be so elusive," wrote Peter Sarre in a 2006 review article about DIBs. Sarre, a professor of chemistry and molecular astrophysics at the University of Nottingham, U.K., supervised Cordiner's graduate-school work on DIBs.

The significance of the first DIBs, recorded in 1922 in Mary Lea Heger's Ph.D. thesis, was not immediately recognized. But once astronomers began systematic studies, starting with a 1934 paper by P. W. Merrill, they had every reason to believe the problem could be solved within a decade or two.

No such luck

More than 400 DIBs have been documented since then. But not one has been identified with enough certainty for astronomers to consider its case closed.

"With this many diffuse bands, you'd think we astronomers would have enough clues to solve this problem," muses Joseph Nuth, a senior scientist with the Goddard Center for Astrobiology who was not involved in this work. "Instead, it's getting more mysterious as more data is gathered."

Detailed analyses of the bumps and wiggles of the DIBs, suggest that the molecules which give rise to DIBs—called "carriers"—are probably large.

But like beauty, "large" is in the eye of the beholder. In this case, it means the molecule has at least 20 atoms or more. This is quite small compared to, say, a protein but huge compared to a molecule of carbon monoxide, a very common molecule in space.

Recently, though, more interest has been focused on at least one small molecule, a chain made from three carbon atoms and two hydrogen atoms (C3H2). This was tentatively identified with a pattern of DIBs.

Tenacious D

On the list of DIB-related suspects, all molecules have one thing in common: they are organic, which means they are built largely from carbon.

Carbon is great for building large numbers of molecules because it is available almost everywhere. In space, only hydrogen, helium and oxygen are more plentiful. Here on Earth, we find carbon in the planet's crust, the oceans, the atmosphere and all forms of life.