Astronomers have discovered almost 300 planets (exoplanets) around other stars.1 Although current detection techniques strongly favor relatively massive bodies, smaller Earth-like rocky planets are now believed to be abundant. Such planets, if located within the ‘habitable zone’ of their parent star, could potentially harbor life. Exoplanet discoveries are currently being carried out with indirect detection techniques: the planet light itself is not imaged; instead, the planet's gravitational influence on its parent star is detected as a periodic velocity variation or periodic light dimming of the star if the planet happens to pass in front of the star. Direct imaging of planets—isolating planet light from starlight—is ultimately necessary to characterize exoplanets by spectroscopy and to find out whether habitable (and possibly inhabited) planets are abundant.
Figure 1. The phase-induced amplitude apodization (PIAA) coronagraph uses aspheric beam-shaping optics to ‘apodize‘ a telescope beam. The conventional non-apodized telescope beam (left) would produce an image with strong Airy diffraction rings. PIAA optics (right) transform this beam into an apodized, Gaussian-like beam that yields a stellar image free of Airy rings. Since the transformation is done without losing light or angular resolution, the PIAA coronagraph is extremely efficient.
Imaging planets like our own is extremely challenging: the light from an Earth-like planet is approximately 10 billion times fainter than the parent starlight, and the planet-to-star angular separation is very small (typically 0.1arcsec for nearby stars). A conventional telescope, even if free from optical aberrations, cannot image exoplanets as they would be lost within the much brighter Airy diffraction rings of the central star's image. To overcome this challenge, a space telescope would need to be equipped with a ‘coronagraph,’ an optical device that can efficiently remove starlight while preserving light from any orbiting planets. Producing diffraction-free star images
Because the Airy diffraction rings are due to the sharp edges of the telescope's aperture, analogous to the pupil of an eye, a possible solution to the problem is to place a mask in the telescope beam to produce a modified or ‘apodized‘ Gaussian-like pupil that yields images free of diffraction rings. Apodization purposely changes the shape of a mathematical function to suit certain optical and laser systems. To reach the 1010 contrast required for imaging Earth-like planets, this mask would unfortunately need to remove most of the light gathered by the telescope. Since most of the light removed is at the edges, this would be equivalent to reducing the telescope diameter by two to three times. A large space telescope, at least 4m diameter, would then be required to image Earth-like planets.
The phase-induced amplitude apodization (PIAA) approach we have developed uses high precision beam-shaping optics to do exactly the same job: producing an apodized Gaussian-like beam.2–4 However, the advantage is that no light is lost. Figure 1 shows that instead of removing light, PIAA optics geometrically steer light to form the desired beam. An image of the star can be formed without diffraction rings, and a small opaque focal plane mask can then physically remove starlight. The precious unchanged planet light is left for the science camera. The PIAA coronagraph design also includes a more conventional apodizer mask that eases manufacture of the beam-shaping optics with little cost in performance. The very strong field aberrations introduced by the beam-shaping optics are managed by small corrective optics after starlight has been removed.
The PIAA system removes starlight without losing sensitivity or angular resolution, and planets as close as 2λ/D from the star can be imaged at 1010 contrast. It can also work in polychromatic light of many frequencies—a significant challenge for a number of other coronagraph concepts—and accounts for the fact that stars are partially resolved small disks, not points. PIAA combines, for the first time, all of the characteristics needed for a coronagraph to image Earth-like planets with a small telescope in space. We are currently researching a space telescope concept with a 1.4m primary mirror equipped with a PIAA system for NASA, which would be able to detect an Earth analog in the proximity of 20 nearby stars.
Thanks to recent advances in aspheric optics manufacturing, the PIAA system can now be made to a high degree of accuracy with mirrors or lenses, and we have assembled several PIAA systems at the Subaru Telescope laboratory.5 Our main prototype PIAA system, cofunded by Japan (Subaru Telescope) and the United States (NASA), also includes a deformable mirror to correct for small errors in the optical elements. We have demonstrated with this system that a PIAA coronagraph can deliver high-contrast images. Figure 2 shows a contrast better than 1,000,000:1 at 1.2 λ/D from the optical axis and was achieved in air.
Figure 2. Comparison between the simulated image of a unresolved star with a conventional telescope (left) and an image obtained with a PIAA coronagraph prototype at the Subaru Telescope laboratory (right). Both images are shown at the same brightness and angular scales. In the PIAA image, the D-shaped dark area immediately on the right of the star's position was made extremely dark (>1,000,000:1 contrast) by the coronagraph to allow imaging of exoplanets. The diffraction ring at ≈4.5λ/D from the optical axis is due to the particular design used for this experiment, and will be removed in future PIAA coronagraphs.
With the participation of NASA's Jet Propulsion Laboratory and Ames Research Center, we are now planning to demonstrate the level of performance required for imaging Earth-like planets. This new effort will include operating a PIAA system in a vacuum with higher-quality optics and improved hardware and software for controlling light wavelengths with the same phase (wavefront control). A lens-based PIAA coronagraph system is also under assembly for the Subaru Telescope and will be ready to start observations in approximately 18 months. This system includes a few specific innovations. For instance, the remapping performed by the PIAA optics removes the central obstruction and spider pattern, another diffraction effect, from the Subaru Telescope's pupil.
The PIAA coronagraph's ability to deliver high-contrast images at no cost in sensitivity or angular resolution can lead to a telescope as small as ≈1.5m diameter that could discover and characterize Earth-like planets around nearby stars within the next decade.
National Astronomical Observatory of Japan Subaru Telescope
University of Arizona
Olivier Guyon is an astronomer at Subaru Telescope and the University of Arizona. His research interests include adaptive optics for astronomy and exoplanet discovery. He is developing new concepts for wavefront control and coronagraphy to enable direct imaging of exoplanets and disks from both ground-based and space telescopes.