≡ Menu

Laser Travel by Photonic Thruster

A recent article in Popular Mechanics about Young K. Bae’s work on laser propulsion is generating some buzz, enough that I went back to look at the 2008 report Bae did for the NASA Institute for Advanced Concepts. Actually there are two reports, a Phase I and a later Phase II (with additional funding) on the topic of “A Contamination-Free Ultrahigh Precision Formation Flight Method Based on Intracavity Photon Thrusters and Tethers.” Bae was interested in spacecraft formation flight down to precisions of mere nanometers.


The idea relates to missions that would use multiple spacecraft to study astronomical phenomena. In fact, setting up a proper configuration would allow the study of biosignatures in the atmospheres of terrestrial exoplanets, using the technique of interferometry. Here the image produced has a resolution equal to that of a telescope as large as the maximum antenna separation, but the spacecraft involved have to maintain their alignment. Both ESA (with the Darwin mission) and JPL (with one configuration of Terrestrial Planet Finder) have studied interferometry in this context, and Bae’s work would assist such formation flying.

Image: Physicist Young K. Bae. Credit: Y. K. Bae Corporation.

Making it possible, in Bae’s view, was the use of laser methods in combination with tethers, but the work has grown beyond that original concept into what is being described as a ‘photonic laser thrust system,’ in which a laser is generated in a cavity between two mirrors. The resultant beam is fired from a space-based launch platform stabilized by conventional thrusters, and now we’re in territory made familiar by Robert Forward, who envisioned enormous laser installations firing beams to departing ‘lightcraft,’ sails pushed by laser light all the way to nearby stars.

The photonic laser thruster (PLT) is described on Bae’s website as:

…an innovative photon thruster that amplifies photon thrust by orders of magnitude by exploiting an active resonant optical cavity formed between two mirrors on paired spacecraft. PLT is predicted to be able to provide the thrust to power ratio (a measure of how efficient a thruster is in terms of converting power to thrust) approaching that of conventional thrusters, such as laser ablation thrusters and electrical thrusters. Yet, PLT has the highest specific impulse (a measure of how fast the fuel can propel spacecraft) orders of magnitude larger than that of other conventional thrusters.


We’re very early in this game, although a small amount of research on beamed propulsion to sails has been performed in the laboratory by Jim and Gregory Benford. A space-based experiment using what was to have been the Planetary Society’s COSMOS sail never occurred after the sail’s launch failure in 2005, but even then the plan had been to use the Goldstone dish of the Deep Space Network to bounce a microwave beam off the sail to measure the effects. Laser experiments are sure to follow as we gain expertise at operating sails in space.

Forward’s laser installations would use abundant solar power from orbits near that of Mercury, firing out a laser beam that would have to be kept narrow — collimated — by a vast lens built in the outer Solar System. The same problem occurs with Bae’s lasers, but at least in the early days, he’s talking about near-Earth and lunar missions, and eventual flights to Mars. The diffusion problem occurs as we start contemplating significantly longer missions. And note this, from the article: “One compensation is the idea of doing smaller platforms to create a ‘photonic railway,’ each acting as a sort of refueling station in between to get the craft where it needs to be.” Here Charles Quarra’s idea of using multiple lenses for beam refocusing comes to mind.

Popular Mechanics doesn’t mention it, but Quarra’s ‘starway’ concept grows out of work first performed by Geoffrey Landis back in the 1990s and posits an eventual ‘light bridge’ connecting nearby solar systems (see A Light Bridge to Nearby Stars for more). Bae is evidently studying beam collimation in terms of ‘Bessel beams’ as well, beams that do not diffract like a conventional laser. What both these ideas get us away from is the need Forward saw of building a lens somewhere beyond the orbit of Saturn that would focus the laser light tightly on the departing spacecraft. Such lenses would demand building space artifacts the size of Texas.

Photonic laser thrusters have now received a Phase II grant from the resurrected NIAC, with the focus on the near-term development of the technology for spacecraft maneuvering and formation flying. For the longer view, see Dr. Bae’s “Prospective of Photon Propulsion for Interstellar Flight,” in Physics Procedia Vol. 38 (2012), pp. 253-279 (abstract), which describes a ‘photonic railway’ that would, in Bae’s words, “…bring about a quantum leap in the human economic and social interests in space from explorations to terraforming, mining, colonization, and permanent habitation in exoplanets.” That’s an energizing goal, but let’s get formation flying using photonic thrusters demonstrated first to see how all this scales up.


Comments on this entry are closed.

  • Michael Fisher October 21, 2013, 11:55

    I’m not a physicist, but I assume the laser beam engine & the lens would require thrusters to balance the push of the laser beam and if the laser beam engine is using solar energy the collectors would need to be balanced by an opposing thrust too?

    Could this balancing act be performed easily without resorting to chemical or mass thrusters that need to be resupplied?

  • Rob Flores October 21, 2013, 14:02

    Sure for a civilization, with an established colony or two, I can see it.
    And using Mercury as laser base for solar system propulsion will probably be cost effective provided your lasers are reliable and cheap.

    But this idea as early turn of the century interstellar project For humanity? Well, I honestly think that we’ll have a practical interstellar propulsion system designed and tested ( I grant that a laser propulsion can get us to very high end of the speed of light barrier) Way before we will be able to manipulate and assemble an object about the mass of a moon about 200-250KM diameter, with it’s attendant support system. I am assuming your
    mirror is solid.

    Would love to peek at the future and see if the early interstellar vechicle solution turns out to be either 2,000 ton behemoths or lithe and nimble 80 tons for a 6 person crew.

  • CZ October 21, 2013, 17:19

    Looking on an intrastellar scale this tech looks perfect for propelling a small sail-based civilization around its own solar system. Sending a craft/probe/human to mars or the moon would have much lower power requirements and allow the opportunity to practice focusing the beams at ever increasing scales.

  • Brett Bellmore October 21, 2013, 21:04

    “beams that do not diffract like a conventional laser. ”

    As near as I can tell, based on an educated layman’s knowledge of optics and EM theory, this is not possible. Aside from some (Low rate) interactions at extremely high energies, photons do not interact in the absence of charge carriers, which means you can do anything you want with your beam, but it’s going to diffract anyway.

    A brief reading on “Bessel beams” suggests that what is really going on is that you simulate non diffraction for a finite distance along the beam center by replacing the photons that diverge from the beam center with other photons from the beam periphery. But that in the far field the result is going to be worse than a beam using the same aperture and simply going for minimized diffraction.

    However, it has occurred to me that photons do interact gravitationally, and that it’s possible a sufficiently well focused and intense laser beam would be genuinely non-diffracting based on photons on the edge of the beam being bent back into the beam by gravity. Any idea how much power this would require.

  • James M. Essig October 22, 2013, 1:27

    Hi Folks;

    I cover staged beam bounce relativistic rocket type craft in my second book, The Galactic Explorer: Advanced Concepts In Relativistic Rocket Flight.

    I consider remote beamed sources for which imput energy does not originate in a spacecraft Mass Ratio.

    I also have three Volumes of my series by the primary title, Sailing Into Cosmic Destinations and am working on Volume 4 of the same series.

    My seventh book by the primary title Pin Wheel Sails Of Cosmic Fortune should publish within a month. This will be Volume 1 of a continuing series.

    Beam sails are just plain fun and as I like to say, are somewhere in the exotic between reactionary thrust via rocket exhaust and the proposed general relativistic methods of antigravity, wormhole transport, and warp drive.

    I share Paul Gilsters interest in beamed energy concepts.

    Either way, the crew of beam ships can effectively travel arbitrary many unitary multiples of light speed in the ship frame, as well as time travel into the future as it will actually turn out, instead of some parallel history time travel invoked as a means to remedy otherwise dangerous causal issues assoicated with forward then backward time travel.

  • Eniac October 22, 2013, 22:57

    Brett Bellmore:

    However, it has occurred to me that photons do interact gravitationally, and that it’s possible a sufficiently well focused and intense laser beam would be genuinely non-diffracting based on photons on the edge of the beam being bent back into the beam by gravity. Any idea how much power this would require.

    Gravitation would probably not work for light beams.

    There is a better option, though: self-focussing beams. Believe it or not, the vacuum is optically non-linear at sufficiently high intensities. This follows from the laws of quantum electrodynamics (QED), part of the standard model. I once researched this, there is a number of publications out there on this subject.

    Self-focussing is a well known phenomenon in non-linear optics, where light creates it’s own channel to be confined by. Such self-focussing beams narrow down to very high power densities and easily destroy the material that makes them possible. They are a common nuisance in high power laser installations.

    If self-focussing beams in vacuum could be generated and maintained, you could thread them between stars like railway tracks, and fairly simple devices could couple to them and derive energy as well as momentum. The main problem with this extremely attractive idea is that there is at least a dozen orders of magnitude between the power levels needed for vacuum self-focussing beam and those we could realistically hope to achieve in practice. There goes another grand idea…

  • Brett Bellmore October 23, 2013, 15:20

    A quick search says the optical non-linearity of vacuum at high intensities is scattering, not self-focusing, so maybe we should be glad the required intensity is absurdly high.

    As for gravitation, ask Google, and you get an answer: Gravitational self-confinement of a photon beam

    Alas, I don’t have a subscription. Can anybody here read this?

  • Eniac October 23, 2013, 23:02


    A quick search says the optical non-linearity of vacuum at high intensities is scattering, not self-focusing

    Could you point me to a reference on that? I found this
    http://www.munich-photonics.de/fileadmin/media/Toshis_Seite/NL5011.pdf, for example, where is shows a formula on page 20 that indicates increased refractive index with power. Increased refractive index would mean self-focussing. Or am I missing something?

    Also, let me know if you ever obtain a copy of this very interesting paper on gravitational self-focussing.

  • Brett Bellmore October 24, 2013, 12:58

    I was looking at page 22 of the same pdf; At really high energies, photons can undergo scattering interactions.

    I really need to settle down and do the math, but my reasoning is that as the amount of power in the beam rises, so does the associated gravitational field due to that energy density. This will give rise to an ‘escape velocity’ for the beam. It’s likely to be quite low, of course. But the radial component of a photon’s velocity if extremely well focused is going to be very low, too, and that’s the only part of it’s velocity that’s relevant. Say you’ve got a diffraction limited focus of 0.1 arcsecond; I figure that’s a radial component of under 200mps. So if you had enough energy in the beam that it’s escape velocity was over 200mps, it should self-focus.

    That’s a LOT of energy, of course.

  • stephen October 25, 2013, 10:43

    I hope these will be of interest…

    Slow Train to Arcturus, by Eric Flint and Dave Freer…
    In this one, the starship doesn’t slow down, but habitats separate from the ship to enter a star system, while the starship keeps going.

    And Edward M. Lerner has a novel, “Dark Secret”, anthologized in Analog, April – July/August 2013; one of the ideas is that a cosmic string will be a region where lightspeed is faster, so a vessel following the string can go FTL.

    If a cosmic string is lightyears long, maybe different sections will have different physical laws?

  • Brett Bellmore October 25, 2013, 16:59

    There’s an older SF novel, I really hate that I can’t recall the title or author, just that the cover was by Kelly Freas, and it was premised on exactly this idea: An interstellar “railroad” based on self-focusing laser rails.

    I may have to dig out my Freas art books to track it down.

  • Eniac October 25, 2013, 19:48


    I was looking at page 22 of the same pdf; At really high energies, photons can undergo scattering interactions.

    You may not have noticed that the title on page 22 says “Beyond QED” and the formulas refer to the Riemann Tensor and are thus about gravity. Quantum gravity, in this case. Such effects are bound to be many orders of magnitude weaker even than the QED effects. That also goes for your gravitational self-focussing idea, I suspect.

    I remain convinced that QED vacuum non-linearity leads to self focussing of intense light beams, and that it is a much lower-hanging fruit than light/gravity interaction of any sort. Still VERY high up, though….

  • Andrew Palfreyman October 26, 2013, 15:26

    I’d be interested in seeing a figure for this high energy density regime (J/m^3) such that photon-scattering and self-focus occurs. I presume one can be induced to win out over the other via a careful choice of beam geometries.

    As the energy density increases further, we end up in black hole territory of course. If we’re talking about a pencil beam of light, then I assume the line degenerates into a string of black hole beads in the limit. Depending on their relative velocity and relative size and separation, these may or may not coalesce into larger black holes. I suspect this entire process is stochastic via quantum noise, and therefore inherently uncontrollable. The beam has now disappeared, and we are left with a rather useful artifact – a chain of small black holes , which would likely come in handy for maintaining heading for any craft intent on following a similar route.

    It looks like a rich topic for a Type III civilisation, but perhaps even ourselves.

  • Brett Bellmore October 26, 2013, 18:37

    “then I assume the line degenerates into a string of black hole beads in the limit.”

    Not necessarily: Beam divergence in a diffraction limited beam is inversely proportional to the aperture diameter. It’s possible that self-focusing reaches a balancing point where the increasing divergence due to the size of the beam relative to a light wavelength compensates for self-focusing, and causes the size of the beam to stabilize.

    Or maybe not. I’d really like to see a good, thorough analysis of this.

  • Eniac October 28, 2013, 0:22


    It’s possible that self-focusing reaches a balancing point where the increasing divergence due to the size of the beam relative to a light wavelength compensates for self-focusing, and causes the size of the beam to stabilize.

    This is almost certainly correct. The beam cannot become narrower than the wavelength of the light, I suppose, and that will happen well before self-gravity plays any role at all.

    This final diameter is going to be very small, which will help with the amount of power needed, as much of the extreme intensity is provided simply by the very low cross-section. IIRC I looked up work on self-focussing in non-linear optical media and plugged in the QED vacuum parameters, and got extremely high power levels even taking the self-narrowing into account. Oh well.

    I still think this is worth following up on, though, I might easily have gotten something wrong.

  • CharlesJQuarra October 28, 2013, 15:25

    gravitational self-focusing is largely suppressed to first order corrections. Parallel photons exert essentially no mutual gravitational interaction. Gravitational interaction only exists if there exists one frame where the net momentum of the photons is zero. If the photons are 100% parallel, there is no such frame. Even if the photons are ultra-energetic along the almost parallel direction, the energy that matters for gravitational coupling purposes is the difference, not the sum


    I’m not sure, but I suspect that the configuration of maximal gravitational self-interaction for a flux of photons is achieved when they form a spherical wavefront.

  • Eniac October 28, 2013, 21:22


    Parallel photons exert essentially no mutual gravitational interaction.

    Good point. You could imagine self-focussing beams to be bidirectional, though, with equal and opposite fluxes going through the same self-induced waveguide. Not sure what difference that would make with respect to self-gravitation, but I think it would be beneficial for other reasons to have neutral overall momentum of the beam.

    In any case, I fully concur that gravitational interactions will be negligible under any circumstances, compared with QED photon-photon effects.

  • CharlesJQuarra November 13, 2013, 12:03

    QED photon-photon interactions are way more feasible, and it has been discussed in the literature:


    But the article is a bit lacking on a estimation of beam divergence. More work on this are is needed