It has been about two years since Yuri Milner announced his most audacious piece of science-focused philanthropy: Breakthrough Starshot, an attempt to send hardware to Alpha Centauri by mid-century. Although the technology involved is a reasonable extrapolation of things we already know how to make, being able to create materials and technology that create that extrapolation is a serious challenge. So much of Breakthrough Starshot's early funding has gone to figuring out what improvements on current technology are needed.

Perhaps the least well-understood developments we need come in the form of the light sail that will be needed to accelerate the starshots to 20 percent of the speed of light. We've only put two examples of light-driven sails into space, and they aren't anything close to what is necessary for Breakthrough Starshot. So, in this week's edition of Nature Materials, a team of Caltech scientists looks at what we'd need to do to go from those examples to something capable of interstellar travel.

The size of the problem

One of our best examples of a light sail was put into space on the IKAROS craft, which was capable of accelerating up to speeds of 400 meters/second. Breakthrough Starshot's craft are expected to travel in the area of 60,000 kilometers/second and accelerate to that speed before leaving the Solar System. So the amount we can learn from the existing craft is fairly limited.

Those speeds—and the acceleration needed to get there—provide a rough idea of the size of the sail we'd need, and the dimensions are pretty impressive: 10 square meters but weighing less than a gram. That, the Caltech researchers calculate, means that the sail will have to average out to being 100 atoms thick yet still be able to transmit the force of acceleration. Graphene is one of the strongest materials we're aware of, so it might work for Breakthrough, but it's also transparent, so it can only act as structural support.

There is some good news in that 100-atom-thick figure. The sail will run into a variety of energetic particles in the interplanetary and interstellar medium, but most of these will be hydrogen and helium. And, based on how deeply those particles penetrate into other materials, there's a good chance that the hydrogen and helium will pass straight through the solar sail. Dust particles create more of a problem, but the authors estimate that they would obliterate about 0.1 percent of the total sail area, and most of that would occur after the acceleration has been completed.

So what we're left with is primarily the challenge of building the material that will reflect light from the solar sail. Or, as we'll see, the collection of interlocking challenges.

Sailing

Light sails work because photons carry momentum, and they'll impart a bit of that to reflective surfaces as they bounce off. Starshot proposes to sync up a large collection of lasers on Earth and focus this on the light scale in order to accelerate it rapidly. That means the lasers will have to be at a wavelength that can pass through the atmosphere without being absorbed or scattered; the authors suggest that the near-infrared (with a one- or two-micrometer wavelength) would be a good choice.

So the sail surface would have to reflect in those wavelengths. Gold and silver, among other metals, are already known to do so and can be made in thin films. Problem solved, right?

Not exactly. As the authors note, managing light isn't just a matter of tackling the problem of reflecting light; it's a series of interlinked problems. Even if we could make thin films of these metals a few atoms thick on a sail, the relatively high weight per-atom means that we could easily blow past the sail's weight limit. And, even more problematic, while these metals reflect most of the light in these wavelengths, a lot of what doesn't get reflected is absorbed. At the intensity of the lasers involved in accelerating the craft, the heat from those absorbed photons would quickly destroy the sail.

Instead, the authors look into partly reflective materials that have a high refractive index and low absorption. The high refraction allows for the possibility of making light-manipulating structures on the surface of the sail that contribute to its reflectivity. The researchers consider a handful of semiconducting materials that fit the bill, rejecting a number of them because their component atoms weigh too much (like tin). The reflection/refraction of these materials also have to cover a broad range of wavelengths since, once the craft is moving fast enough, the incoming photons will be red-shifted.

In the end, nothing meets the researchers' full list of desires. So they settled for two out of three and focus on silicon, diamond, and molybdenum disulfide.

There are other issues, however. These materials also have to efficiently radiate away any heat they do wind up with—technically, they have to have a high emissivity. This will obviously cause problems with, as the researchers put it, "melting or other thermal failure modes." But it can cause problems even below that. As silicon heats up, the amount of light it absorbs will increase, creating the prospect of a thermal runaway.

Right now, the team suggests that we don't even have good measurements of some of the material properties that we'd need to fully evaluate some of these details. We've got measurements of bulk material, but not thin films. So there's an entire research field that has to be advanced for us to fully understand the tradeoffs we face.

Structure

While materials like silicon won't reflect much of the light, they have a high refractive index, which means they will bend a lot of it. If you structure the surfaces of these materials on scales similar to the wavelength of the incoming light, they can potentially bend it in a way that is functionally equivalent to reflecting it. So, the researchers also considered a variety of designs for the surface.

And here they found yet another trade-off. While the best light reflection (nearly perfect reflection) came with a multi-layered, 3D structure, that added significantly to the weight. By contrast, they weren't able to reflect as much light using a pattern of holes drilled into the silicon, but this significantly lowered the weight. When they compared the acceleration of the two options, they were able to get a noticeably larger acceleration from the less-efficient solution purely due to the weight.

Adding to the challenges they face, the sail itself won't be a simple flat sheet. That's because even the slightest imperfections on the surface or tilt of a sheet could send it off in unexpected directions during acceleration. Instead, current thinking focuses on a spherical sail and a donut-shaped light field, which could self-correct for small perturbations in the sail. That design, however, probably isn't conducive to making the sail from single sheets of material (assuming we could figure out how to grow them at scales of 10 square meters). So we're going to have to bond many individual panels of these materials together, something that we don't currently know how to do.

Overall, the paper does a good job of laying out what we'd need to know to start choosing materials for a Breakthrough light sail. But it also highlights that this isn't a matter of finding the one perfect solution; instead, it's about managing multiple, sometimes conflicting priorities and engineering a solution that partially satisfies all of them. "We argue that a successful design of the light sail will require synergetic engineering," the authors conclude, "simultaneous optimization and consideration of all of the parameters described above."