This answer to Pre-mixing cryogenic fuels and using only one fuel tank quotes John D. Clark's Ignition! Chapter 11: The Hopeful Monoprops, and the quote includes the line:

How he avoided suicide (the first rule in handling liquid oxygen is that you never, never let it come in contact with a potential fuel) is an interesting question, particularly as JPL later demonstrated that you could make the mixture detonate merely by shining a bright light on it.

It is also quoted in this answer to Why did it take so long for methane to be used as a rocket propellant?

Question: How did JPL detonate a liquid oxygen methane mixture by shining light on it? Is there any information on the procedure and/or the intensity of the light?

CO2 lasers come in kilowatt+ varieties for example, and were around since about 1967. They can burn through just about anything, so just saying that they used light does not necessarily mean that they used "a light touch".

Was it a purely bulk thermal effect, or was this really a photo-initiated chemical reaction, i.e. individual photons stimulated chemical reactions directly?

  • $\begingroup$ A little preliminary poking around finds a lot of different papers in different places talking about methane photochemistry, but in all cases they talk about vacuum ultraviolet photodisassociation. I'm not necessarily sure I'd call a strong VUV source a "bright light", but I'm sure some sort of arc-lamp would do... $\endgroup$ Commented Dec 20, 2019 at 14:50
  • 1
    $\begingroup$ @StarfishPrime those may be about very low pressure gases rather than cryogenic liquid mixtures; Vacuum UltraViolet is called that for a reason, and photodissociation is a lot different than detonation, though maybe that's all it takes. $\endgroup$
    – uhoh
    Commented Dec 20, 2019 at 16:07
  • $\begingroup$ Gotta cleave some atomic bonds somehow, and those C-H and O-O bonds don't usually just fall apart when you point visible light at them. $\endgroup$ Commented Dec 20, 2019 at 16:14
  • $\begingroup$ Longer wavelength UV of <241nm wavelength is enough to disassociate oxygen, though I wouldn't know if the mere presence of oxygen radicals is enough to kickstart combustion. The Methyl-H bond is weaker than an O-O bond (photolysed by 272nm UV, I think), and so with a suitable UV flux you should be able to form enough of the right sort of radicals in close enough proximity to spark off a more convenient chemical reaction. If methane-oxygen mixtures were more troublesome than that, you'd expect someone to have a bit more to say about it than this one anecdote... $\endgroup$ Commented Dec 20, 2019 at 20:31
  • 2
    $\begingroup$ LOX-hydrocarbon mixtures can also explode from impact. $\endgroup$
    – ikrase
    Commented Dec 21, 2019 at 7:32

1 Answer 1


At low temperatures, the activation energy for pure CH4 O2 oxidization is about 170kj/mole. (See figure 1 here) That’s about 1.8eV per atomic reaction.

1.8eV can be provided by 688nm red light, or any shorter wavelength. So generally, visible light can initiate reactions.

I can’t quantify how many photons/cm2 it’ll take to start a runaway reaction from the liberated energy, but LMethane/LOX is an ideal mix for turning a few catalyzed interactions into runaway combustion: unlike air, it has no N2 to carry away energy or slow the kinetics. Mixed liquid reactants increase the rate of CH4 2O2 configurations. Etc.

(more detail added:)

Typically, a combustion reaction starts when heat is added, which provides activation energy to enough molecules for some to react. The energy (really, enthalpy) from those reactions provide the activation energy for the next, which provides energy for the next, etc. Once started, only more reactants is required if each time there's a surplus of energy after losses.

That question of "after losses" is significant. If you're burning with air instead of pure oxygen, some of the heat goes into heating (and perhaps disassociating) N2. Some gets conveyed away through the thin gas reactants, so remaining reactants don't get enough of the energy. Etc. Those are all losses. Ideally, you have pure (so other chemicals can't take heat or drive other reactions) fully mixed (so any molecule that picks up energy can react) liquid (ditto) reactants; those are easiest to ignite and will burn fastest.

A fully mixed LCH4 LO2 environment is about as good as you can get for those conditions. Once some activation energy is provided, the resulting reaction energy goes straight to nearby reactants that are ready to use it. It’s so favorable that the mixture actually detonates, with the combustion front moving as a shock wave at about 4600 m/sec.

The only countervailing term I can think of (I haven't done any experiments with this stuff!) for optical ignition is the long optical length of light in LCH4 and LOX: The incident light isn't absorbed right away, so spreads its energy along a path. (That can be useful if you want to ignite the while volume all at once, though)

  • $\begingroup$ Let us continue this discussion in chat. $\endgroup$
    – uhoh
    Commented Dec 20, 2019 at 5:30
  • $\begingroup$ Why doesn't carbon fiber overwrapping in LOX catch fire? (watch this video first) $\endgroup$
    – uhoh
    Commented Dec 20, 2019 at 6:34
  • $\begingroup$ This is helpful and thanks for the edit! I'm going to hold out for some information on how JPL actually did this, and if they in fact used a catalyst or not. $\endgroup$
    – uhoh
    Commented Dec 21, 2019 at 3:22
  • $\begingroup$ @WillCrawford I don't see any doubt expressed, only a carefully worded Stack Exchange question looking for a Stack Exchange answer that provides some "information on how JPL actually did this, and if they in fact used a catalyst or not." I'm pretty sure that the mechanism is related to the energy of the individual photons, not the intensity of a focused beam. but until we track down a reference we won't know. $\endgroup$
    – uhoh
    Commented Apr 24, 2020 at 2:14
  • 2
    $\begingroup$ Might I suggest asking Randall Munroe? I believe he worked at Langley ... $\endgroup$ Commented Apr 24, 2020 at 2:20

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.