In late 2008 two teams made waves with the simultaneous announcement that they had managed to directly photograph planets in orbit around distant stars, also known as exoplanets. Although hundreds of exoplanets had already been found orbiting sun-like stars throughout the Milky Way, they had been discovered by indirect means—astronomers had inferred the presence of a planet by observing the dimming effects or gravitational wobble an orbiting companion induces on its parent star. In a few other cases a candidate planet had been observed near a star but had not been proved to follow a planet-like orbit.

The groundbreaking 2008 observations of the two planetary systems orbiting the stars Fomalhaut and HR 8799 relied on some of the field's most powerful observatories—the Fomalhaut group used the orbiting Hubble Space Telescope; HR 8799 was spotted by the Keck and Gemini North telescopes in Hawaii, which rank among the largest observatories on Earth. Now, a paper in the April 15 issue of Nature shows that, with a technological assist, smaller telescopes on which observing time is more plentiful can also spot systems such as HR 8799. (Scientific American is part of Nature Publishing Group.) What is more, improved technology should also allow larger observatories such as Keck to move from the few giant planets already imaged—all of which orbit their host stars at relatively large distances—to closer-in worlds more like our own. (Many worlds that reside in extremely tight orbits have been found by indirect means; it is the temperate, so-called Goldilocks zone where Earth analogues might be found that has proved most elusive.)

The challenge in directly imaging a planet is that the minimal amount of light reflected or emitted by a planet is dwarfed by the fiery output from its star, and the two appear very close together at interstellar distances. "The idea is to get very high contrast near a bright star," explains study co-author Eugene Serabyn, an astrophysicist at the NASA Jet Propulsion Laboratory in Pasadena, Calif. "You really have to get everything right, because you're looking for something that's one part in 10 billion of the star if you're looking for an Earth."

So astronomers use adaptive optics, in which deformable mirrors correct for atmospheric distortion, and instruments called coronagraphs to blot out the starlight. But at their crudest coronagraphs can bring complications. "Normally people put a black dot, for a lack of a better term, where the star lands," Serabyn says. "So you block the central piece of the star, and the stuff that's around the star gets by the black dot." But that dot obscures the star's immediate vicinity, where close-orbiting planets would reside, and introduces diffraction effects. So Serabyn's group installed a relatively new kind of instrument called a vortex coronagraph on a small-scale segment, called a sub-aperture, of the Hale telescope at the Palomar Observatory in California.

Rather than relying on an opaque spot aligned with the target star, a vortex coronagraph shifts the phase of the incoming light waves so as to create a create a singularity, or a unique point, in the center of the image, where the starlight falls. That central point of phase-shifted light is then redirected by a lens to the outside of the image, where it can be screened out, leaving only planet light in the center of the final image.

With a vortex coronagraph on the 1.5-meter Hale aperture, the researchers were able to detect the three previously known planets of the HR 8799 system. Those planets make for good imaging targets thanks to their great distance from their host star. The closest of the three known planets in the HR 8799 system orbits at a distance that, in our own solar system, would wedge it between Uranus and Neptune; the farthest orbits at a remove that would place it well beyond Pluto. Having demonstrated the efficacy of the approach on wide-orbit planets, Serabyn says he would like to apply the concept at larger telescopes to seek out worlds closer to their stars. "I think we'll do a lot better because we'll get the best of both worlds," he adds.

Christian Marois, an astronomer at the National Research Council Canada's Herzberg Institute of Astrophysics who led the 2008 study of HR 8799 on the 10-meter Keck and the 8.1-meter Gemini North telescopes, says that "being able to see all three of these planets with only a 1.5-meter aperture is pretty amazing." Marois is part of a group working on building a similar high-tech planet imager at the 8.1-meter Gemini South telescope in Chile, and was pleased to see the Serabyn team's proof-of-concept. "I believe they are probably the first ones to put all the pieces together for very high-contrast imaging instrument," he adds.