Courtesy of NASA.

Three images of T Tauri stars. Left: glowing dust and gas surround a pair of young stars (blue is not a true color). Middle: the shield of dust and gas that frequently hides the young stars. Right: an infrared view of a T Tauri star that pierces the dust veil.

Weintraub and Bary are pursuing a different possibility. They propose that most older T Tauri stars haven't lost their disks at all: The disk material has simply changed into a form that is virtually invisible to Earth-based telescopes. They published a key observation supporting their hypothesis in the September 1 issue of the Astrophysical Journal Letter that was highlighted by the editors of Science magazine as particularly noteworthy. The two researchers currently are preparing to publish additional evidence that supports their out-of-the-mainstream contention.

T Tauri stars range from about one-fifth to twice the mass of the Sun. They have been objects of scientific interest since the discovery of T Tauri in 1852. Initially, the reason for the interest was that their brightness varies dramatically. More recently astronomers have been studying them because they can provide important insights into how the Sun and solar system evolved.

Twenty years ago, scientists thought that T Tauri stars had extremely strong solar winds blowing outward at velocities of tens to hundreds of kilometers per second. The theory was based on a spectral analysis of the light coming from these stars. One of the emission lines in their spectrum, called the hydrogen-alpha line that is produced when protons and electrons combine to form hydrogen atoms, is unusually strong. Astronomers figured that it must be produced by exceptionally strong solar winds.

In the 1990's, however, astronomers were forced to re-evaluate this interpretation. Although T Tauri stars may have strong stellar winds, scientists now consider the strongest source of hydrogen-alpha emission to be hydrogen gas spiraling in from the surrounding disk to fall onto the star. As a result, strong hydrogen-alpha lines are now considered evidence that stars possess protoplanetary disks.

The dense disks of dust and gas surrounding classical T Tauri stars are easily visible because dust glows brightly in the infrared region of the spectrum. Although infrared light is invisible to the naked eye, it is readily detectably with specially equipped telescopes. The classical stars also possess the strong hydrogen-alpha line. But there is a second group of T Tauri stars that tend to be somewhat older - between three to six billion years - for which the hydrogen-alpha line is either very weak or absent and which show no evidence of disks. These have been labeled "naked" or "weak line" T Tauri stars.

The T Tauri stars also turn out to be strong X-ray sources. Naked T Tauri stars produce more X-ray emissions than their dustier, classical cousins. So in recent years, astronomers have been using X-ray telescopes orbiting Earth to search for them, and they've found hundreds.

Because the "naked" T Tauri stars do not have strong hydrogen-alpha lines and there is no visible evidence that they possess protoplanetary disks, astronomers have concluded that they must have lost the disk of dust and gas that they had when they were younger. The scientists argue that this material might have been absorbed by the star or blown out into interplanetary space or pulled away by the gravitational attraction of a nearby star.

The loss of disk material in less than 10 million years has serious consequences for planet formation. According to current theories, it takes about 10 million years to form a Jupiter-type planet and even longer to form a planet like Earth. If the planet-formation models are correct and if most Sun-like stars loose their protoplanetary disks in the first few million years, then very few stars like the Sun possess planetary systems.

This picture didn't sit well with Weintraub, however. "Approaching it from a planetary evolution point of view, I have not been comfortable with some of the underlying assumptions," he says.
The focus of his dissatisfaction has been that current models do not take into account the natural evolution that protoplanetary disks should go through as the planet-building process proceeds. Over time, the disk material should begin agglomerating into solid objects called planetesimals. As the planetesimals grow, an increasing amount of the mass in the disk becomes trapped inside these solid objects where it cannot emit light directly into space. As a result, the disk material should get progressively dimmer and more difficult to detect from a distance.

"Rather than the disk material dissipating," says Bary, "It may simply become invisible to our instruments."
For his doctoral dissertation, Bary has been working with Weintraub to find ways to determine if such "invisible disks" actually exist and can be detected even though standard methods have failed to find them. They realized that the constituents of the disk that astronomers knew how to detect - small grains of dust and carbon monoxide molecules - should quickly disappear during the first steps in planet building. But the disk's main constituent, molecular hydrogen, should stay around much longer. The hydrogen, which makes up the bulk of the mass of giant planets like Jupiter and Saturn, isn't vacuumed up into the planets until rocky planetesimal cores about ten times the size of Earth are formed.

That realization led Bary and Weintraub to search for evidence of molecular hydrogen. Unfortunately, this form of hydrogen is notoriously difficult to stimulate into emitting light, so astronomers had not previously tried to look for it in the spectra from T Tauri stars. The fact that T Tauri stars also produce X-rays gave them an idea. What if some of these X-rays were striking the hydrogen molecules in the disk? X-rays are energetic enough to split the hydrogen molecules into atoms, protons and electrons.

Under the proper conditions, these particles in turn could heat up the surrounding hydrogen gas to the point that it would emit infrared radiation of a distinctive wavelength that could be detected from Earth. Studying various theories of planet formation, they concluded that hydrogen molecules should be present in appropriate conditions in a "flare region" near the outer edge of the protoplanetary disk.

Molecular hydrogen emission lines In four T Tauri stars The images and infrared spectral lines of four T Tauri stars demonstrate that they are surrounded by disk of molecular hydrogen. Molecular hydrogen emits light at a characteristic wavelength of 2.1218 microns (0.00008 inches). In the star LkCa 15, the disk is edge-on. The double peak in its spectrum is caused by the Doppler effect. The right-hand peak is produced by light emitted by hydrogen molecules on the side of the disk that is traveling away from Earth so it is shifted slightly to a longer wavelength. The peak on the left is caused by light coming from the other side of the disk, where the molecules are traveling toward the Earth, causing their wavelength to shorten slightly. This allowed the astronomers to determine that the region where the light is being produced is at about the same distance from the star as Saturn is from the Sun.

The next step was to get observation time on a big telescope to put their theory to the test. "That was the hardest part," says Weintraub, looking back. Their proposals were turned down for several years, but they were finally allocated viewing time on the four-meter telescope at the National Optical Astronomical Observatory in Kitt Peak, Arizona. When they finally took control of the telescope and pointed it toward one of their prime targets - a naked, apparently diskless T Tauri star named DoAr21 - they found the faint signal for which they were searching.

"We found evidence for hydrogen molecules where no hydrogen molecules were thought to exist," says Weintraub.

When Bary calculated the amount of hydrogen involved in producing this signal, however, he came up with about a billionth of the mass of the Sun, not even enough to make the Moon. As they argued in their Astrophysical Journal Letter article, they believe that what they have detected is only the tip of the iceberg since most of the hydrogen gas will not radiate in the infrared. The question that remains is whether the iceberg constitutes a complete protoplanetary disk or just its shadowy remains.

Since this first discovery, Bary and Weintraub have detected the same hydrogen emission line around three classical T Tauri stars with visible protoplanetary disks. They have found that the strength of the hydrogen emission lines in the three is comparable to that measured at DoAr21.They have used these results to obtain the ratio between the mass of hydrogen molecules that are producing the infrared emissions and the mass of the entire disk in each of the three systems. For all three they calculate that this ratio is about one in 100 million.

"If the ratio between the amount of hydrogen emitting in the infrared and the total amount of hydrogen in the disk is about the same in the two types of T Tauri stars, which is not an unreasonable assumption, this suggests the naked T Tauri star has a sizable but hard-to-detect disk," says Bary.

In one of the stars, the disk is edge-on. That allowed the researchers to measure the Doppler shift of the region producing the infrared emissions. The shift corresponds to an orbital velocity comparable to that of Saturn, which clearly places the location of the emitting region within the protoplanetary disk, right where they expected it.

Weintraub and Bary admit that they have more work to do prove their theory. They have been allocated timeon a larger telescope, the eight-meter Gemini South in Chile, in order to search for a second, fainter hydrogen emission line. If they find it, comparison of the strength of the two lines will provide additional insights into the process that is exciting the hydrogen gas. To determine if the hydrogen emissions that they have discovered are caused by a general mechanism involved in the planetary formation process, the researchers also plan to survey about 50 more naked T Tauri stars for molecular hydrogen emission lines.

Currently, the number of naked T Tauri stars that have been discovered is much greater than the number of known classical T Tauri stars. If a significant proportion of them have kept their protoplanetary disk, it could mean that solar systems similar to our own are a common sight in the universe.

David Weintraub's home page