Why does thermite burn pretty slowly? Can it burn much faster if the components were ground to much smaller grain sizes? Can it be made to actually detonate--that is, the entire pile turning into liquid metal at the speed of sound, not necessarily exploding?

I suppose that it might require something like dissolving both the aluminum and iron oxide in something, then evaporating the solvent to achieve a properly fine mixture. Would that work? If not, why?

  • $\begingroup$ "Solid state explosives are very rare." - What did you mean by that? TNT, ANFO, dynamite are solid. $\endgroup$
    – John Smith
    Feb 22 at 0:10
  • $\begingroup$ Solid state explosives making only solid state products - that's the problem with thermite. There's no gas produced. $\endgroup$
    – Mithoron
    Feb 22 at 2:00
  • $\begingroup$ Dissolving aluminium and iron oxide means reactions, not true dissolution. You would not get back by evaporation. $\endgroup$
    – Poutnik
    Feb 22 at 7:43

2 Answers 2


There are two types of explosive: those that detonate and those that deflagrate (to use technical terms with a specific meaning). I suspect the question would accept either definition though it uses the word "detonate".

The difference between the two terms is the speed of propagation of the reaction in the explosive material: in detonation the reaction travels faster than sound; in deflagration it travels more slowly. Most would say both "explode".

TNT detonates. Guncotton deflagrates. TNT doesn't explode easily and needs to be triggered by a small very rapid detonation of something like a metal fulminate. In TNT the pressure wave from the initial trigger is sufficient to trigger an explosion in the bulk TNT and this propagates through the material much faster than sound. Guncotton is triggered in a similar way but the reaction propagation is slower than sound (which is valuable if you want a "gentle" bang that pushes a bullet or shell from the barrel of a weapon rather than bursting the barrel).

Those mechanisms are relevant to the question about thermite. The reaction in thermite is very hard to initiate. It requires a great deal of heat not pressure. This simply cannot propagate as fast as a pressure wave, so thermite cannot be turned into an explosive without some way to raise the temperature of the entire mixture to perhaps 2,000 °C very rapidly.

So thermite cannot "explode" with detonation though it might be termed deflagration though such a slow one it would not normally be called an explosion at all.

  • 1
    $\begingroup$ To be more specific, it is meant faster than speed of sound within the solid explosive. $\endgroup$
    – Poutnik
    Feb 22 at 13:06
  • $\begingroup$ Faster than sound? That's quite a shock! ;-) BTW, there's a schlieren video of shock waves from canned air and compressed air at youtube.com/watch?v=DfYlLns0els $\endgroup$ Feb 22 at 16:37

Heat moves in three general ways:

  • Conduction, mediated by phonons, lattice vibrations, that, in a crystal (e.g., diamond or graphene), or in a superfluid (He II), propagates at the speed of sound. However, thermite is a mixture, not crystalline, and phonons scatter, making heat travel far slower -- as in pitch.

  • Convection, where there is movement of material. When a brisant explosive detonates, material is ejected faster than the speed of sound, forming a shock-wave. However, there is no violent ejection of nitrogen, as in the detonation of TNT or C-4, in placid (if a bit warm) thermite. That displacement reaction offers no great volume change.

  • Radiation, where photons carry energy. Though thermite emits a great many photons (wear welders goggles while watching), the photons don't propagate inside the mixture, which is opaque.

In brief, then, the answer is "No, thermite cannot detonate."


In a transparent medium, it might be possible that a detonation could proceed at the speed of light! Perhaps someone can try this: fill two clear glass tubes, separated by a few cm, with nitroglycerin, and see if a detonation in one propagates to the second at the speed of light, initiated by photons.

If someone plans to do this, please let me know, so I can distance myself strategically.

  • $\begingroup$ While lattice vibrations might move at the speed of sound, heat clearly does not. If it did you would burn your hand every immediately on putting a metal poker into a hot fire. $\endgroup$
    – matt_black
    Feb 22 at 10:40
  • 1
    $\begingroup$ @matt_black, the speed of sound is the limit, but crystal defects prevent transfer at that speed. However, in crystalline carbon (diamond and graphene), and in liquid helium, there is little scattering and heat can propagate that fast. $\endgroup$ Feb 22 at 16:56

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