Searching for new Earths with micro-electromechanical systems
Far, far across the Milky Way galaxy is a small planet, much like Earth. It orbits close to a Sun-like star and may even be a water world like our own. At billions of years old, it has cooled and no longer glows with thermal radiation. Perhaps over those years, it even has developed life.
This Earth 2.0 hasn’t been found yet, but scientists are, statistically speaking, confident such a planet exists. In fact, there are probably millions of them. But these small, dark objects, obscured by blindingly bright stars, are nearly impossible to see; an Earth-sized planet would be about 10 billion times fainter than the star it orbits.
Scientists can image planets only up to 10 million times fainter than their host stars. Over the past two decades, they have imaged dozens of young, wide-orbit, Jupiter-sized planets still aglow with the heat of formation a few millions of years ago. However, it remains a challenge to image ones like our own Earth.
To reach the one-in-10-billion level of contrast requires a significant improvement to current technologies and ideally a platform in space, beyond the limiting turbulence of the atmosphere. In 2020, the ‘US National Academies of Sciences, Engineering, and Medicine Decadal Survey on Astronomy and Astrophysics recommended just that—a space telescope called the Habitable Worlds Observatory (HWO) that would be the first-ever designed to search for earthy counterparts that could host life. The National Aeronautics and Space Administration (NASA) took up the challenge and is now working with astronomers, engineers, and industry partners to improve key technologies to make HWO a reality. Of those technologies, some believe that one stands apart in its importance: deformable mirrors.
“The deformable mirror is the most critical component,” says Eduardo Bendek, a research scientist at NASA’s Ames Research Center. “And the deformable mirror needs to perform at an incredible level of accuracy to see a small planet.”
The bulk of the work in removing a star’s light to reveal a planet is done by a stellar coronograph—essentially an internal shade that blocks unwanted starlight. But stray light can sneak around a coronograph, and it alone is not enough to see an Earth or even a Jupiter. To compensate, astronomers use small deformable mirrors, which are placed behind the coronograph. With hundreds of moveable pistons called actuators under the mirror’s surface, they can level imperfections in the surface of the mirror, which is small enough to fit in the palm of your hand. Paired with a device called a wavefront sensor, which detects wiggles in incoming light, the deformable mirror can also continuously adjust its surface to compensate for changes in temperature or air turbulence that can cause stray beams of light that would otherwise hide the light of a faint planet.
There are a few technologies underlying these small deformable mirrors, but none have achieved contrasts of one in 10 billion. However, one technology showing promise for mirrors to achieve this ultra-high contrast level in space is micro-electromechanical systems (MEMS). Using components just micrometers in size, MEMS technology weds electronic and moving parts to create very small, but very capable machines.
A 2,000-actuator MEMS deformable mirror on the Subaru Telescope. Photo credit: Boston Micromachines Corporation.
Several groups of scientists and engineers are working on the next generation of MEMS deformable mirrors that could fly on NASA’s HWO. But to get there, they have to overcome two large remaining hurdles to MEMS deformable mirror technology—size and surface quality.
MEMS deformable mirrors are not a new technology. For more than two decades, they have been employed in the business of imaging exoplanets across the galaxy and can be credited with dozens of discoveries. Much of this has been possible due to industry leader Boston Micromachines Corporation, which has produced most of the MEMS deformable mirrors currently in use for astronomy.
“We started with a NASA Small Business Technology Transfer program [in 1999] to make deformable mirrors for space applications,” says President and CEO Paul Bierden.
While other companies have dabbled in MEMS deformable mirrors in the past, Boston Micromachines is, to his knowledge, the only one still producing MEMS deformable mirrors that can reach the contrast levels needed to see other worlds. The company fills a niche market and can create incredibly demanding one-off deformable mirrors, as is needed for NASA’s HWO.
Deformable mirrors come in two main classes, contact and contactless, separated by the style of mechanism that drives the tiny actuators used to deform the surface. Boston Micromachines’ deformable mirrors are contactless MEMS deformable mirrors, which use an electrostatic force between the mirror and an electrode below to shape the surface. This contactless method helps the MEMS deformable mirrors be less affected by changes in temperature and humidity. MEMS deformable mirrors are also lighter, quicker, and more compact than other contact class deformable mirrors, which is especially important for space telescopes. They have also shown exceptional durability over extended and repeated use.
Despite these traits, MEMS technology for deformable mirrors is still not ready for imaging an Earth-like planet. One remaining hurdle is size. The largest in-use MEMS deformable mirror, used by the Gemini Planet Imager at the Gemini Observatory, is one-inch square and has 4,096-actuators. But for HWO, scientists will need a deformable mirror with around 10,000 actuators. Fortunately, another key advantage of electrostatic MEMS deformable mirrors is their scalability.
“Since our devices are made in a semiconductor manufacturing process, we can replicate these patterns and then scale up in the two dimensions to get a larger number of actuators,” Bierden says. “So, the 10,000-actuator device is done with the same process as a 140-actuator device.”
However, this production style also creates small but unfavorable surface conditions on the mirror. In the manufacturing process, layers of silicon are deposited, patterned, and etched. This results in lower layers creating a quilting pattern on the surface of the mirror. While the surface variation is only on the nanometer level, it limits the ability of the system to function property when working at the 1-in-10-billion contrast level.
“We make what is considered to be a very good optic for everybody but people who are trying to get to 1-in-10-billion contrast,” Bierden says. “There are very few people in this world who need subnanometer surface figure errors.”
To combat these issues, Boston Micromachines is delving into fluid flow physics and developing new techniques to deposit the silicon in a way that can improve surface quality. They believe they’ll be able to resolve the issue before a design is selected for HWO in 2029.
While Boston Micromachines works to improve MEMS deformable mirror technology, other academic and government groups are ensuring it will work in space. The first space flight for a full MEMS deformable mirror was in 2015 aboard a sounding rocket. While the unit only spent a few minutes aloft, the flight showed the deformable mirror could function in space and that such a device could survive a sounding rocket launch, which is harsher than for larger rockets.
In a MEMS deformable mirror, small posts connect the mirror (yellow) to electrostatic actuators. Photo credit: Boston Micromachines Corporation.
In 2020, a DARPA-funded CubeSat called the Deformable Mirror Demonstration Mission, or DeMi, spent 18 months testing its postage stamp-sized MEMS deformable mirror. While issues with the ground-based telecommunication system prevented full testing, the mission was able to use an internal laser to test the mirror and show that it was able to iteratively correct for changes in light.
“Our goal was to get under 100 nanometer optical quality, and we exceeded that,” says Ewan Douglas, an astronomer and assistant professor at the University of Arizona and payload engineer and project scientist for DeMi at MIT in the Department of Aeronautics and Astronautics. “For a CubeSat that’s pretty notable, as CubeSats don’t have a super high success rate.”
While 100 nm was a notable accomplishment, HWO will need to reach 0.01 nm, which requires better sensors than could fit in DeMi. The University of Arizona recently completed construction of a vacuum test bed that will allow Douglas and others to more easily test the performance of components than launching into space.
“There are also some interesting effects that only show up when you go to vacuum, because there’s no air dampening the motion of the mirror,” Douglas says. “When the mirror is pushing against vacuum, it resonates a little bit more.”
These extra opportunities for testing are essential for the success of MEMS deformable mirrors. In addition to electrostatically forced MEMS, two other deformable mirror technologies—an electrostrictive design using a piezoelectric material and an electromagnetic one actuated by magnetic coils—have been shortlisted for HWO. All three technologies have limitations and all three have groups of researchers and companies working together to prove their technologies can ultimately be the most robust when the HWO working group selects a design.
“It’s in NASA’s best interest to have a couple of irons in the fire to make sure that they have the best technology at the end of the day,” Bierden says. “We’re fairly confident that after this four-year period of proving [our MEMS deformable mirrors], we will be demonstrably the best technology.”
DeMi’s deformable mirror. Photo credit: Ewan Douglas, STAR Lab, PI
Kerri Cahoy, MIT Department of Aeronautics and Astronautics.
A similar winnowing process was used for NASA’s Nancy Grace Roman Space Telescope, scheduled to launch in 2027. MEMS deformable mirrors were also shortlisted for that telescope, with which astronomers hope to image old Jupiter-sized planets. However, when it came time to select a front runner, the MEMS technology was not as well tested as an electrostrictive deformable mirror—a contact class deformable mirror that uses a mechanically connected actuator—developed by AOA Xinetics. Ultimately, MEMS technology wasn’t selected for flight.
At the time, in addition to surface quality issues, MEMS deformable mirrors were also struggling with issues during vibrational testing, one of the key tests to ensure a telescope component can hold up to the harshness of launch conditions. In a 2021 paper, scientists reported that a MEMS deformable mirror had failed a vibrational test, which came as a surprise. The test of the mirrors revealed that a few of the actuators showed anomalies after the shaking. The scientists suspected perhaps dust, picked up as the mirror was transported between several laboratories, had compromised the devices and decided to conduct further tests.
“The MEMS are microscopic devices, so if you have particles of dust, then they can get on different parts of the system and create a short circuit,” Bendek says.
In 2023, Bendek and his colleagues reported the results of the second test, which used a larger deformable mirror. This time, they made sure to keep it in a more controlled clean environment. Despite the fact that the deformable mirror had more connectors, wires, and actuators, it passed with flying colors.
While plenty of work remains to ensure that a MEMS deformable mirror can truly reach 1-in-10-billion contrast levels, researchers from academia, government, and industry agree that the technology has the potential to help us find habitable worlds like our own.
“All the evidence seems to be adding up to that it is a good technology for space,” Douglas says. “Everything seems to be pointing to it being a viable technology for high-contrast imaging. More work needs to be done to show we can meet the extreme requirements for imaging Earth-like exoplanets.”
Mara Johnson-Groh is an award-winning science writer and photographer who writes about everything under the Sun, and even things beyond it.
