Can you create a black hole from light? Also, Whiteholes vs. Blackholes

[bernsten69]

I recall reading about creating a black hole from having an area of space-time occupied by a sufficiently high enough number of high-energy photons, so as to create a black hole. I believe this has some German name, but I can't recall it at the moment.

Does anybody have an opinion on this? Could, for example, 2 or more sufficiently high-energy gamma bursts "collide," creating a zone of space-time with sufficient 'effective mass,' so as to create a black hole? If so, what would happen (ie, how much of the energy would be converted into momentum, resulting in a high-speed black hole flying through the universe). If none of it is converted into momentum, doesn't this imply a preferred rest frame?

Also, is there any difference between a Blackhole (made of regular matter) and a Whitehole (made of anti-matter)? It seems to me that a Blackhole violates CP symmetry, and thus a whitehole is no different from a blackhole (since all of the matter is compressed to nothingness and loses all of its properties besides charge, angular momentum and mass).

If they are the same, then can you ever preserve CPT symmetry with a black hole? Attempting time reversals of various particles and forces being absorbed by the black hole (with their various carrier-virtual-particles) seems to imply CPT symmetry violation. . . Is that correct?

[DaleSpam]

If so, what would happen (ie, how much of the energy would be converted into momentum, resulting in a high-speed black hole flying through the universe). If none of it is converted into momentum, doesn't this imply a preferred rest frame?Momentum would in principle be conserved. So the black hole would have the same net momentum as the pair of gamma rays. That would not in any way imply a preferred frame.

[PAllen]

This made me wonder about the likelihood of such an interaction. I seem to conclude if a whole bunch of prior absurdities are assumed, it must be a very likely result; and the black hole could be long lived.

Consider a photon whose energy/c^2 is equivalent to the mass of an asteroid. As ridiculous as that is, if it didn't have anything to interact with (even protected from CMB photons), it could not produce any matter (or black hole) without violating conservation of energy or momentum. I guess it could instantly decay into a mass of comoving less energetic photons (they are bosons after all), but I've never heard of that reaction, so lets pretend it doesn't happen.

Now imagine its interaction with an identical photon moving in the opposite direction (head on, within the limits of QFT). You would now have two asteroids worth of available mass/energy localized incomprehensibly (wavelength of such photons). It would thus seem, if you got this far, a black hole formation would have to be exceedingly likely.

Comments?

[bcrowell]

Yes, it's definitely possible to make a black hole out of colliding EM waves. There are exact electrovac solutions that show this process happening. See Griffiths, Colliding Plane Waves in General Relativity, available here http://www-staff.lboro.ac.uk/~majbg/ , e.g., ch. 8.

This is all classical physics, so you don't need to worry about photons or QFT. I guess QFT does allow light waves to interact in ways that are not possible according to Maxwell's equations, which are linear. However, GR is nonlinear, so you don't need to add a quantum ingredient to get the nonlinearity.

Singularities are a generic feature of GR, and the no-hair theorems tell us that black holes aren't picky about what they eat.