Casting Giants

When astronomer John M. Hill arrived as a new graduate student at the University of Arizona, Tucson, in 1979, he was greeted with some sage advice: Before choosing an advisor from among department faculty, he should wait to meet a professor named J. Roger Angel, who hadn't yet arrived on campus for the ritual grad student-professor matchmaking.

Hill took the advice and waited. He didn't know it yet, but when he did meet Angel, the two men would embark on a 40-years-and-counting professional relationship. Their work, along with the work of many other researchers and technicians, helped spur a revolution in the size and capabilities of ground-based optical telescopes and, in turn, shaped decades of research and discovery in astronomy and astrophysics.

Mirrors and mirror-making technology were fundamental to this revolution. In the 1970s, some 1960s-era surplus satellite mirrors from the US Air Force found a new home with a joint Smithsonian Institution/University of Arizona project called the Multiple Mirror Telescope (MMT), which would become the world's third-largest ground-based telescope.

Because the mirrors were originally built for outer space, they were made to be lightweight: Each low-expansion glass mirror had a honeycomb-like backing of hollow pockets. The six primary mirrors of the MMT were to function as separate parallel telescopes. Images from each mirror were then combined with other optical components to form a single, focused image. The effect was that the six 1.8-meter mirrors had the same light-gathering power as a single 4.5-meter mirror.

Despite the spare parts approach, MMT's multiple mirrors and lightweight honeycomb glass proved a historic milestone in the goal to build larger ground-based optical telescopes.

From its dedication in 1979, MMT was able to capture and combine light for high-resolution imaging both at optical and infrared wavelengths. The images it returned were better and less blurry than anyone had expected, says Hill, who today is technical director of the Steward Observatory's Large Binocular Telescope (LBT) on Arizona's Mt. Graham, as well as a research astronomer at the University of Arizona.

The mirrors for the MMT were initially chosen "mostly because they were available," says Hill. But Angel and others soon realized that, more than any other factor, it was the mirrors' stiff, lightweight quality—about 80% lighter than conventional mirrors of the same size—that so effectively reduced blurriness. Glass will deform under its own weight, and this had been one of the limiting factors in making any ground-based optical telescope larger than 5.1 meters.

Bigger, better, lighter

With inspiration from the MMT, Angel and Hill began experimenting to find a way to build bigger, lighter, better telescope mirrors. First, they tried cutting borosilicate glass into precise shapes that could be fitted together to make a honeycomb structure for the mirror interior. Then, plates of glass could be fused to the top and bottom of the structure to make the mirror surface. But the work was laborious. "We also could not purchase sheets of borosilicate glass larger than two meters across," adds Hill. They needed another approach.

Blocks of Ohara low-expansion E6 glass are carefully placed onto the mold for GMT mirror 5. Credit: Giant Magellan Telescope

They decided that it would be easier to build a mold with the inverse honeycomb structure inside, add the glass in broken chunks, and then melt it all in a furnace. The ceramic fiber used to make the honeycomb could be power washed out of the mirror once everything cooled, leaving the mirror 80 percent hollow.

But to get a "fast" mirror—with a focal ratio shorter than two—"you would have to grind out tons of glass," Hill says. "Grinding glass is routine, but not trivial."

To solve this problem, they decided to mount the furnace on a giant turntable that spins at a certain rate as the glass cools. The centrifugal and gravitational forces from spinning impart a parabolic shape close enough to the finished optical quality that less glass can be used at the outset and less glass needs to be ground out to reach the finished product.

The plan from the very beginning was to make eight-meter mirrors," says Hubert "Buddy" Martin, project scientist at University of Arizona's Richard F. Caris Mirror Lab, of which Angel is both the founder and scientific director. The lab started by fabricating 1.8-meter mirrors, then 3.5, 6.5, and up to today's maximum size of 8.4-meter mirrors.

They would eventually perfect the process. It takes the Mirror Lab about a year to make the lightweight ceramic honeycomb structure for a large mirror. It then takes a week to melt the glass into the mold, and another three months to cool, while spinning, from a temperature of 1200°C to 650°C, which is when the glass solidifies. When the temperature is within a few degrees of room temperature, the mirror can be lifted off the turntable and the mold removed. Finally, the mirror is ground and polished to an accuracy of 20 nm. This last stage is the "hard part" according to Martin, which might be an understatement since the surface generation and polishing can take more than two years to complete. After the polishing is finished and the mirror is shipped to the telecope, a thin layer of aluminum is applied so the mirror will reflect 90% at the incident light.

UA Mirror Lab staff review the glass placed in the mold, checking for space for the last few pieces of glass for GMT mirror 5. Credit: Giant Magellan Telescope

The competition

While the spin-casting approach to building giant mirrors has proven a success, there is an alternative approach to building large telescopes that involves using many smaller mirror segments fitted together to create a large one, with each mirror controlled by actuators to maintain their optics as a whole. This segmented mirror technology will be used for the Extremely Large Telescope being planned for the European Southern Observatory in Chile.

So which is better: monolithic spun-cast mirrors built on a lightweight honeycomb structure, or segmented mirrors? "They're two different approaches to achieve the same goal," says Hill, "which is a telescope that regularly produces images as sharp as nature allows. With large honeycomb mirrors, you invest a lot in the mirror and your reward is a guaranteed smooth wavefront over the largest possible aperture and the smallest possible number of elements to control."

The spun-cast borosilicate telescope mirrors pioneered by Angel, Hill, and colleagues have led to some of the biggest ground-based telescopes ever built, such as the 6.5 meter Baade and Clay telescopes built by the Carnegie Institution at Las Campanas, and the twin 8.4-meter mirrors of the LBT on Mount Graham that have the combined observing power of a single 11.8-meter mirror.

The Mirror Lab's large, lightweight mirrors are also part of a novel concept to create so-called turnkey observatories, which are high-powered scientific instruments that are affordable to single research institutions or small groups of institutions.

A scientist "shouldn't spend their life building a research instrument," says Steward Observatory Associate Director, Jeffrey Kingsley. Given the capabilities of the Mirror Lab and possibilities for siting at existing observatories, he and colleagues began to ask whether it would be possible to put together, within five years of an order, a 6.5-meter mirror optical-infrared observatory.

The answer appears to be yes. They have plans to build the first turnkey observatory on Arizona's Mt. Lemmon, which has superior observing conditions and existing infrastructure. A fixed price of about $60 million, and $1 million per year in operating costs, would be affordable to many research institutions, with the help of government and philanthropic funding, says Kingsley.

The mirror spins while it cools, which is where the technique gets its name. Credit: Giant Magellan Telescope

Science goals for large mirror telescopes

As for the science, mirrors from the mirror lab are expected to help discover habitable planets around nearby stars, capture unparalleled images of star formation, help determine what dark matter exists in the universe, and power sky surveys that will reveal more about galaxy formation and the large-scale structure of the Universe. In combination with specialized adaptive optics systems and cameras, future telescopes—notably the Giant Magellan Telescope (GMT) and the Vera C. Rubin Observatory—will capture information about the Universe that can't be obtained with existing ground-based systems. Both are scheduled to come online in the 2020s.

In a finished telescope, "you want to collect light as efficiently as you can," says GMT Project Manager James Fanson. "Monolithic mirrors reduce the number of edges, so you get a very clear pupil on the telescope and good, high-contrast images as a result."

The GMT will consist of six off-axis 8.4-meter mirrors that surround a central on-axis mirror to form a single optical surface 24.5 meters in diameter with a total collecting area of 368 square meters. In fact, the GMT mirrors will collect more light than any telescope ever built, with a resolving power 10 times greater than the Hubble Space Telescope. Light from the Universe will first reflect off GMT's seven primary mirrors, then reflect again off seven smaller secondary mirrors, and finally, down through the center primary mirror to the telescope's charge-coupled device imaging cameras.

In addition to reducing weight, Fanson says that the honeycomb structure on the back of the GMT's mirrors will help prevent visual distortion from temperature differentials, such as might happen when a mirror is opened to the night sky with much lower ambient temperatures than inside the telescope dome. With the spin-cast mirrors, air can be circulated throughout the night through the back of the mirror and "you can keep the mirror at the same temperature as the atmosphere," he says.

One of the telescope's many science objectives, Fanson says, "is to study the very early universe, for example, galaxy formation 100 to 500 million years after the Big Bang." The light from that early in the universe is shifted to the near infrared, and the GMT's light-gathering ability will be critical to capturing those wavelengths. "It's a key aspect of what we want this scope to do."

Other work by the GMT will include the search for planets outside our solar system that could support life. Because the reflected light from these planets is faint and often drowned out by their host stars, the GMT's light-collecting power will be critical in finding and learning more about these other worlds.

The science goals of the Rubin Observatory, whose primary and tertiary mirrors were created by the Mirror Lab, are quite different. Steven M. Kahn, director of the Rubin Observatory, describes his facility as a "unique telescope designed to survey the sky with a wide field of view."

Indeed, starting sometime in 2021, it will commence imaging the entire Southern Hemisphere of sky every three nights for ten years to deliver a 200 petabyte set of images and data products to answer some of the most pressing questions about the structure and evolution of the Universe. Among them: understanding dark matter and dark energy, hazardous asteroids and the remote solar system, the transient optical sky, and the formation and structure of the Milky Way Galaxy.

Kahn says the Mirror Labs' stiff, lightweight, and highly accurate mirrors are essential to the observatory's mission and ability to operate. Because the mirror has such a low focal ratio of 1.2 and a very wild field of view, eliminating possible distortions on the mirror is critical. The survey nature of the Rubin Observatory mission means the telescope will move every few seconds whenever it is operating and, from its perch atop Chile's Cerro Pachon, its mirrors will have to be continuously monitored and kept near the very cold ambient outside temperatures. Chromatic aberrations, distortions caused by temperature differences, telescope movement, and more can all impact image quality.

As "first light" approaches for the Rubin Observatory—and later, the GMT—the Mirror Lab will have provided the largest individual mirrors possible for two of the largest ground-based telescopes ever built. After that, Mirror Lab scientists say it's hard to make any prediction about where giant telescopes and mirror-making technology may go. But it's safe to say that the engineering process designed by the Mirror Lab will influence the direction.

William G. Schulz is a freelance science journalist in Washington, DC, who regularly covers optics and photonics research.