(This question has been migrated from the SciFi StackExchange.)

From Cixin Liu's 2010 sci-fi novel Death's End (pages 68–69):

"A radiation sail can be made very thin and light. Based on the current state of material sciences, we can make a sail of about fifty square kilometers and limit the mass to about fifty kilograms. That should be big enough." The speaker was a Russian expert who had once directed a failed solar sail experiment.


The good news was that the area of the radiation sail could be shrunk to twenty-five square kilometers, and with even more advanced materials, the mass of the sail could be reduced to twenty kilograms.

That's a mass of just one kilogram per square kilometer, or one gram per 1000 square meters, or one milligram per square meter. This immediately struck me as implausible... but I'm not sure.

I googled up this student report from circa 2014, which says:

For far-term missions, a value of 1 g/m² is required.

There is also ongoing research on [...] nanotubes [...] it could produce sails of ADD less than 0.1 g/m², 50 times smaller than that of a Mylar sail.

But that's still 100 times more massive than the 0.001 g/m² quoted by Death's End's "Russian expert."

Is a 0.001 g/m² solar sail

  • plausible by "near future" standards? (This part of Death's End is set in Year 1 of the Crisis Era, i.e. "201X".)

  • at least possible for a spacefaring civilization to achieve (in say the 24th century)?

  • unlikely or impossible, for fundamental physical reasons?

  • a translator's error? (If someone could show that the original Chinese said "fifty thousand square meters" instead of "fifty square kilometers," or something like that, then that would be a perfectly relevant and upvoted, if out-of-universe, answer.)


3 Answers 3


A solar sail with an areal density of $1~\mathrm{kg}/\mathrm{km}^2 = 1~\mathrm{mg}/\mathrm{m}^2 =0.001~\mathrm{g}/\mathrm{m}^2$ is impossible by known materials science because graphene has an areal density of $0.77~\mathrm{mg}/\mathrm{m}^2$.

Being a single atomic layer of a light atom, graphene is the absolute lower bound for the areal density of pretty much anything that can be constructed, including solar sails. A single atomic layer like graphene is unlikely to work well as a solar sail. Here are the issues:

  1. Graphene is nearly transparent: (only 2% of visible light is blocked by a single layer)
  2. A single layer of graphene would rapidly be degraded by radiation in space
  3. Large graphene sheets would collapse in on themselves without support. No technology currently exists for building such supports and they may not be possible for this areal density.
  • 19
    $\begingroup$ A radiation sail does not have to be continuous sheet. A grid of fine metal wires a fraction of a wavelength of light apart (a Faraday cage for optical radiation) would be enough. It could be as much as 80% empty space (30nm wide strips of metal a few nm thick could be 300nm apart). $\endgroup$ Commented Jun 7, 2021 at 16:36
  • 3
    $\begingroup$ @SteveLinton That is a good point, but I think the degradation problem would be even worse for a nanoscale mesh of graphene. The mesh would also be more flexible and require greater mechanical support. $\endgroup$ Commented Jun 7, 2021 at 16:46
  • 5
    $\begingroup$ Just want to mention the heliogyro design, which rolls out long sheets of solar sail and spins the craft for stiffness. That still leaves all the other reasons that the abovementioned device is impossible. $\endgroup$
    – Greg
    Commented Jun 7, 2021 at 17:09
  • 3
    $\begingroup$ @SteveLinton it is a good point, but the mesh needs to be made of highly conductive "wires" at optical frequencies, meaning it would have to have a sufficiently high density of conduction electrons. It takes many monolayers of even the most reflective metals before they can efficiently stop light from passing through and either absorb or reflect the incident photons' momentum. We still need some magic molecules to make sparse mesh with this areal density work for visible light. Now if it were some source of radio waves instead of light waves, a "radio sail" might be easier... hmm :-) $\endgroup$
    – uhoh
    Commented Jun 8, 2021 at 0:11
  • 1
    $\begingroup$ @SteveLinton so I've just asked How thin can the wires of a sub-wavelength gold mesh be and still be half as reflective as a solid sheet? What physics dictates this? $\endgroup$
    – uhoh
    Commented Jun 8, 2021 at 3:04

There is a extensive summary report on possible improvements of solar sail materials:

"Ultra-Thin Solar Sails for Interstellar Travel - Phase I Final Report"
December 1999, Dean Spieth, Dr. Robert Zubrin

When reading this report one has to keep in mind that they only look for the properties of the sail itself, not taking into account structural elements nor any payload. Nevertheless, it's a nice summary how the current state-of-the-art mylar foil could be improved.

They summarize as follows:

• 25X improvement by removing the plastic substrate, leaving ~100 nm Al layer,
• 300X by reducing aluminum sail thickness to ~4 to 5 nm,
• 500-5,000X by perforating the aluminum sail, possible in the near term, and
• 10,000-100,000X by doping carbon nanotubes, well into the next century.

Along these lines, the first step is to get rid of the plastic support material that makes up the largest part of mylar. Obviously, a 100 nm thin aluminium layer itself is not stable enough to be transported and deployed, so they suggest to use a plastic material that gets torn away after deployment, e.g. destroyed by UV radiation. Part of the weight saved has to be added again in the form of stiffener materials - but current carbon fiber based materials should be able to do the job with less weight.

Second step is to reduce the thickness of aluminium - They found a 5 nm thick "foil" as the optimum in terms of acceleration per mass. It gets about 50% transparent at this point, but the reduced mass more than balances the reduced reflectivity. Again, additional material has to be added to keep this foil stable.

The third and radical step is to not use a foil at all, but a very sparse grid. Remember that structures well below the wavelength of light can't be resolved, and a grid of 5nm wires spaced 200nm apart has almost the same optical properties than a solid foil, at 1% the weight. Producing such a material on a small scale should be well possible today, but manufacturing a kilometer-sized sail that can be deployed in deep space will be a major challenge.

If these materials can be manufactured into a sail of the given weight remains an open question - but at least the foil part using technology available today on a small scale could possibly be in the 10 mg/m² range.

  • 1
    $\begingroup$ It seems that the perforated foil is effectively an aluminum fiber fabric. $\endgroup$ Commented Jun 8, 2021 at 17:07
  • $\begingroup$ 5 nanometers thick is astonishingly fine wire. The bond length in aluminum is 286 pm, so those wires would only be about 20 atoms across! The likely rate of wire-snapping damage, due to cosmic ray events or even just voltage spikes, seems extremely high. Even if the sail could be made in the first place, how long would it last once deployed? $\endgroup$
    – Ryan C
    Commented Jun 9, 2021 at 4:26
  • $\begingroup$ Technically any material is a "sparse grid", if you look at it closely enough. $\endgroup$
    – T.E.D.
    Commented Jun 9, 2021 at 13:40
  • 1
    $\begingroup$ The material in that article is made of handwavium. Then an unrealistic X25 modifier is applied. Then a unrealistic X300 modifier is applied on top of that. Then, just for good measure, further x5000 and x 100000 are applied. That is pure handwavium-impregnated unobtanium. then the author applies good, solid, reliable scientific calculation to figure out how good this miracle material will perform. $\endgroup$ Commented Jun 9, 2021 at 15:50
  • $\begingroup$ @PcMan The starting material is well known, and at least the first optimization steps are well described. There is nothing impossible mentioned. If and how something like this can be manufactured is an open question, but the design ideas themself are well within reach of current technology. $\endgroup$
    – asdfex
    Commented Jun 9, 2021 at 17:24

All of you have to remember that NIAC, at least at the time of the report, did not fund contracts that were easy to do nor simply an engineering challenge - they wanted far-out concepts, not achieveable today, yet not violating any known laws of physics.

Yes, we did look at structural mechanical properties of materials versus temperature, as well as payload mass in the report. Cindy, or Bob, before I joined the outfit around Y2K, came up with the brilliant idea to consider carbon nanogrids, which can be manufactured at a small scale, and one usually obtains extremely high mechanical properties using nano-technology or electroformed metals. Carbon and metals are conductive so an SGEMP effect is unlikely; similarly, cosmic rays and charged particles have little effect on conductive materials because they already have free electrons. And you would not assemble a carbon structure in LEO because of atomic oxygen erosion, as well as increasing orbital debris issues, but once above LEO (say MEO or GEO) it is not an issue (yet still a manufacturing challenge). A mesh will retain its shape even after a large number of micrometeroid hits. The big issue with a large solar sail, if it can be assembled in high orbit, is how to steer it once it leaves the influence of solar light pressure. It might be possible to steer it with a small plasma engine, or heat from a small RTG, as it appears RTGs affected a slight trajectory deviation of the Voyager spacecraft.

(fyi, on a different subject, it would require several "solar sails" the size of Alaska in MEO to temporarily block out the sun for about half of an orbit for a few minutes over any land mass, and thereby offset global warming, but it is probably impractical to build something of that large scale let alone keep it in orbit.) - Dean Spieth, retired from Ball Aerospace


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.