In its infancy, before its planets formed, our sun was ringed by a disk of gas and dust. Astronomers estimate that the dust in this preplanetary disk coagulated and began forming structures in less than about five million years, but the stages of planetary birth are still very uncertain. Similar preplanetary disks are seen around many young stars because infrared observations can detect the emission from the warm dust. In many of the newly discovered disks, the dust appears still to be distributed very uniformly, but in some slightly older stars the disks show signs of having begun to evolve.
As the dust in a preplanetary disk begins to coagulate, and as the plasma wind and radiation from the evolving young star begin to strengthen, one of the first structural features to develop in the disk is a gap around the star - a donut hole. The presence of a developing planet inside the gap could further accelerate the clearing out of the inner regions. Such gaps have usually been inferred from the shape of the infrared spectrum of the source, because dust near the star is the hottest dust in the disk and its absence means its characteristic features are absent.
Still, the inference of a gap is a notoriously indirect one since other possible scenarios, such as the presence of a binary star, could yield the same spectrum.
The CfA's Submillimeter Array (SMA) is uniquely suited to probe the very small distance scales associated with dust gaps in preplanetary disks. Two SAO astronomers, Charlie Qi and David Wilner, together with three colleagues, used the SMA to obtain some of the first direct images of a donut hole: the cleared annulus in a dust disk around one suspicious disk with a suggestive infrared spectrum. Writing in this week's Astrophysical Journal Letters, the astronomers report that the SMA was clearly able to image a 40 AU gap (one AU is the distance corresponding to the average distance of the earth from the sun) around the star, which itself is thought to be only about three million years old. The result is significant because it substantiates and refines the use of infrared spectra to infer the existence of gaps, and proves that at least sometimes the inner part of a disk can develop first.