We know that many distant stars have planets orbiting them because the planet's orbit jiggles the star and the oscillating Doppler shift of the starlight can be detected. But actually imaging the planet is a real challenge. The star is eight to ten orders of magnitude brighter than the planet, so to directly image the planet the glare from the host star's light needs to be suppressed by a factor of about a billion. If you try to do this in a telescope by blocking the star's light in an intermediate image plane and letting the planet's light pass, you find that diffraction from the edge of the "stop" blocking the star's light still obscures the planet. So we are working on fabricating a "soft-edged occulter" which will have minimal diffraction from the soft edges, which are continuously graded transmittance vs. radial coordinate in the focal plane. The required transmission vs. radial coordinate for minimizing diffraction is shown in Fig. 1, along with a cross section of the shape the stop would have to have if it were made of a material with spatially uniform volumetric absorption coefficient.
Fig. 1. Right: theoretical transmission vs. radial coordinate for an optimal soft-edged stop that blocks the star light (stop is centered over image of star) while minimizing diffraction so as not to obscure the planet. Left: thickness of uniformly absorbing material vs. radial coordinate required to produce the desired transmission vs. position profile.
This structure is too small and the tolerances too exacting to be fabricated by mechanical machining or laser machining. We are working toward making it using a Focused Ion Beam (FIB). This requires the development of an understanding ion sputter morphology evolution of tall, steep features.
Ruth Schlitz, as an undergraduate researcher working on this project, developed methodologies to make cones and fine points by FIB machining of silicon - the material about which the most is known regarding FIB machining. A couple examples of her work are shown in Fig. 2.
Fig. 2. Cone-like shapes machined in silicon. Top: comparison of (a) programmed ion beam parameters to (b) resulting physical shape of cone in silicon. (c) and (d) are two pillars with smooth surfaces and areas of high slope milled in silicon using an overlapping annulus technique.
We are collaborating with Dr. Volker Tolls of the Harvard Smithsonian Center for Astrophysics and with Dr. James W. Foley, a dye expert, at the Rowland Institute at Harvard. We are using the photoresist PMMA (poly methyl methacrylate) which we can dope with a variety of dyes to give uniform absorption vs. wavelength over a range of wavelengths. A schematic procedure is outlined in Fig. 3.
Fig. 3. Schematic of fabrication process for absorbing material. In (a), dye-doped PMMA (black) is deposited as a layer of known thickness onto a clear substrate such as glass (light blue), (b) then the mask shape (height function) is milled, and (c) the mask is embedded in clear, dye-free PMMA (also light blue) to minimize refraction.
Undergraduate researcher Shilpa Raja has been learning to dope PMMA and measure its optical constants. She has found that FIB milling of PMMA is much more complicated than milling of silicon, as shown by the image in Fig. 4, but under some circumstances there are encouraging signs for the prospect of milling the required shape, as shown by the image in Fig. 5.
Fig. 4. Unexpected surprises await those who try to machine polymers with a FIB!
Fig. 5. Small smooth cone machined from dye-doped PMMA is an encouraging sign.