The 'Scopes Have Eyes

January 05, 2009

We're back, fresh from a nice long holiday break and ready to sample the wonders the universe has to offer in 2009. It just so happens that this is the start of the official International Year of Astronomy, a joint effort by the International Astronomical Union and UNESCO around the theme, "The Universe: Yours to discover!" It's a huge endeavor, with thousands of sponsored events all year long, including "The Cosmic Diary" -- a group blog initiative with 50 bloggers from 35 countries -- and daily podcasts from "365 Days of Astronomy." You can find a complete list of event highlights here.

Sadly, I will not be flying to Paris later this month for the official Opening Ceremony, but I thought it might be nice to honor one critical development in particular: the rise of adaptive optics to correct various optical aberrations that typically plague even highly advanced telescopes. Telescopes gather light from the stars, but before it reaches the instrument (at least here on earth), the light first must pass through our atmosphere, and all those layers of air with varying temperatures and densities causes that light to become distorted. In fact, that's why stars seem to twinkle instead of just appearing as a sharp pinpoint of light.

An adaptive optics system corrects the "wavefronts" of light, basically straightening the paths to improve resolution and contrast. The result is brighter, clearer images. Current systems use a wavefront sensor to sample the light collected by a telescope's primary mirror and send that data to a computer. The computer in turn controls a deformable mirror that can be adjusted to cancer out any atmospheric distortions. SUre, you could send a telescope into space -- the Hubble Space Telescope hasn't done too badly -- but it's expensive to do so. And many scientists believe that ground-based telescopes with cutting-edge adaptive optics could very well produce images of even better resolution than Hubble (at least in the infrared region of the spectrum, when is where AO works best). The benefits of AO increase dramatically with telescope size, because more light can be collected. 

But adaptive optics aren't just for telescopes anymore. The human eye -- especially the cornea and lens -- can also distort wavefronts and give rise to those pesky aberrations that plague telescopes as well. Eye doctors use instruments called opthalmocopes to image their patients' eyes, but aberrations often make it more difficult to spot developing problems. But scientists have now figured out how to incorporate adaptive optics and micro-electro-mechanical systems (MEMS) technology to build a new kind of opthalmoscope capable of imaging individual retinal cells.

The system -- designed by a team at Lawrence Livermore National Laboratory in conjunction with other universities, medical centers and so forth -- is the first instrument that automatically performs aberration measurements, makes the necessary corrections, and enables the corrected image to be viewable immediately by your eye doctor. It has tiny telescopes inside that relay light to two deformable mirrors and into the patient's eyes, focusing the light beam onto the retina. A wavefront sensor measures any optical aberrations in the incoming and outgoing light, and a MEMS-based deformable mirror corrects the distortions. The rest is basic camerawork: pass the corrected light through a confocal pinhole and into a photomuliplier tube and voila! You have a lovely high-resoltion digital video of the retina.

And that's not all. Just as telescopes help you see things that are far away, microscopes help illuminate the world of the very small. So it's probably not all that surprising that adaptive optics are proving useful in that arena as well. Philip Batson, a physicist with IBM's T.J. Watson Research Center, has incorporated AO into his electron microscope, improving resolution to such a degree that he can now routinely image the positions of single atoms. He swears that "the signal from a single atom is strong enough to get good quality images in a few tens of milliseconds, allowing the taking of sequences of images to follow atomic processes.

Electron microscopes, for the curious, use magnetic lenses to focus electrons into very small beams so scientists can peer at atomic-scale details in very thin slices of materials -- like silicon and similar compounds commonly used in the electronics industry. But just as with any other lens, there are aberrations -- most notably the dreaded spherical aberrations -- blurring the images.

Batson and his colleagues found an ingenious solution: they combined seven sets of magnetic lenses with modern computers to actively correct the aberration in real time. So their souped-up electron microscope can produce an electron beam only three billionths of an inch wide -- smaller than a single hydrogen atom. And that means they can now see objects smaller than a single hydrogen atom. This, in turn, makes it easier to spot defects in the atomic structure of semiconductor materials, such as extra or missing atoms, and figure out how to correct those defects. As Batson succinctly put it, "We can't fix what we can't see."

So there you have it: an ingenious collection of tiny waveguide sensors, actuators, computers and MEMS-based deformable mirrors can shed light not just on the mysteries of our great expansive universe, but also on the realm of the very, very small -- a Grand Unified Technology, if you will. This may be the International Year of Astronomy, but other branches of science shouldn't feel left out: ultimately, it's all connected.

Photos: (top) Images of individual retinal cells. Source: Lawrence Livermore National Laboratory. (bottom) Image of a lattice crystal using IBM's scanning electron microscope. Source: Cornell University.

Comments

[wavefront control technology]

Hi,

thanks for your post.

Could you develop the passage about the electron microscope adaptive optics? For instance what is the active component? It looks as if it uses a succession of magnetic lenses, the focal length of which changes to correct for aberrations. But what are the aberrations due to in this context?

I know it is possible to find X-ray wavefront sensors out-there, but I've never heard about one for an electron beam. I guess it is then open-loop wavefront control, is that correct?