So we have the theory regarding the dual nature of photons being both waves and particles. We know photons often interact with matter as a particle (ie photoelectric effect). But do photons ever interact with each other as particles?
I have an experiment which I am sure has been done. Take two laser beams and point them directly into one another. Will the EMW pass right through each other just like waves or will there be any “collision” of photons and scatter?
IANAQuantum Physicist, but I recall from reading Schroedinger’s Cat that the particles behave like waves when it suits them (?) - I’d imagine that they go all wavelike when meeting a load of other photons, and just pass straight through.
A particle/anti-particle reaction will result in the formation of two photons. The reaction should be reversible, but the cross-section (something like the probability) will be very small.
Under normal circumstances, photons don’t interact with each other. This is related to the usual assumption that electric-magnetic fields behave linearly in free space.
Simple answer: yes, they collide, but since they have no mass and no momentum, no one notices or cares.
Other answer: Depends on how you define “colliding”; does passing through each other count as a collision?
Photons are pretty weird.
I once saw the Feynmann diagrams for photon-photon interaction. The summed terms all seemed to involve creation and destruction of virtual particles (which interacted via the usual forces). The probabilities for photon-photon interactions must be really low, so you wouldn’t see it unless you had massively powerful lasers, or something.
“since they have no mass and no momentum”–Photons definitely have momentum. Whether or not they have mass depends on if you’re talking about rest mass (no) or relativistic mass (yes).
I am sure Chronos will be by shortly and help me out with this one.
But, yes they do interact. Higher energy photons are more likely to interact. Photons display more and more particle like behaviour as their energy (frequency) increases.
My understanding is that if the energy of two photons are high enough, they are virtually guarenteed to interact if they become co-incident in space.
Just to expand on CalMeacham’s response: The only way for photons to scatter off each other is for one photon to split into a virtual electron-positron pair, then the other photon scatters off one of the virtual particles, then the electron-positron pair recombines to make another photon. Or else even more complicated processes. This scattering is quite small but I think it has been measured.
Photons don’t interact directly with each other because the photons themselves have no electric charge.
Can constructive or destructive interferrence occur with light? I.e. Two waves that coincide reinforce and add to each other (constructive interferrence) but if a trough and a crest occur inthe same place they cancel each other out (destructive interferrence).
If we think of the photons in the OP laser experiment as waves could they be tuned to add to each other or cancel each other or would this not work at all?
Interferrence is a wave behaviour. This would not count as a “collision”.
A good example of a collision would be two gamma rays meeting to form a matter/anti-matter particle pair.
I have a sense that you are right but permit me to be stupid a second in search of a better understanding of what is happening.
Light can be considered both a wave and a particle. Does it matter for the experiment in the OP which it is? Does science consider light to be either a wave or a particle or both at the same time?
Your answer seems to suggest the constructive/destructive interferrence thing only occurs in waves and what we have here is particles. Who decides which way it is? Besides, doesn’t two waves existing in the same place at the same time constitute some sort of collision?
[sub]NOTE: This all assumes the interferrence I am talking about can even occur in light which I don’t know is the case.[/sub]
Light waves can interfere with each other, producing “ripples” in the intensity pattern. Particles like electrons can interfere with each other, too, producing ripples in their wave functions. An electron bouncing off a potential barrier will interfere with itself (just as a light wave reflecting from a mirror will interfere with itself). There are some very coool pictures of a calculated reflected wavefunction in Schiff’s book Quantum Mechanics.
There are differences, though. Two particles that run into each other will generally interact and deflect each other. Photons that run into each other do not, in general – they keep on moving as if the other wasn’t there. In fact, they probably won’t interfere with each other – their phases have likely got no relationship to each other.
Can anyone point me to documentation on either Feyman’s diagrams for photon photon interactions or on experiments like the one i described in the OP?
Thanks everyone for all your help! I love intelligent people 
Light and matter both display behaviour that we would associate with classical wave and particle nature.
Just looking at light. The energy (frequency) of light determines which behaviour is the most appearent.
The OP asks about collisions. Collisions are a particle behaviour. In our everyday lifes, light displays very little of this.
But, you can do experiments that will give conflicting results about the nature of light. (conflicting in a classical sense).
Experiments that show interferrence demonstrate a wave nature. That is the nature of the experiment. Waves do not collide, they pass through each other. If two passing waves are observed passing through each other, there will reinforcement and its converse at their intersection, but if you let them continue on past each other, they will be unchanged.
Experiments that show a collision demonstrate a particle behaviour.
In the history of science the status of light as wave or particle was hotly contested for a long time. This is because different experiments pointed to difference conclusions.
Today it is accepted that light behaves both ways. Its energy(frequency) determining which behaviour is the most obvious.
With this in mind, I read the OP as asking if light displays a particle nature. The answer is yes. If the question was asked if it displays a wave nature, the answer would still be yes.
Since the OP was asking about collision (a particle behaviour), experiments that show wave nature (interference) just don’t speak to it.
I hope that makes sense.
Any book on Quantum Field Theory will probably have this, though at an advanced level. E.g. the one by Ryder, p. 343, which, BTW, lets me know I got it wrong last post. The Feynman diagram looks like, well, draw a square. Now put a wiggly line going out from each corner. Those are photons. The sides of the square are electron/positrons (no distinction in Feynman diagrams). So what’s happening is:
-one photon splits into an electron-positron pair
If you were keeping track, we now have two brand-new photons, but we can consider them to be the old photons that have uncergone scattering. This process has FOUR interactions (the corners of the square), and so occurs only rarely.
Experiments? Don’t ask me, I’m a theorist… 
Thanks scotth and Cal…that does help.
So, for the sake of the OP, the light would be considered as a particle but being a wave doesn’t change anything anyway.
Would it be possible to tune the lasers such that they destructively interferred with each other?
If so would the beams seemingly disappear? If they do ‘disappear’ where is the energy represented in the beams hiding?
Easy experiment: I have done this myself. Take a single laser. run the beam through a splitter. Direct both beams at the same target. As the beams will be coming from not quite the same direction, there will be rings of constructive and destructive inteference. It is plain to see. The tinest movement in the setup will cause the rings to shift around and the beams wavelength line up different due to the length of one path or the other changing. Getting someone to breath warm air across one of the beams is enough to see a disturbance.
The energy isn’t “hiding” anywhere. If you saw the same bahaviour with water waves coming up against a beach it would look perfectly natural.
It is possible to get beams to completely cancel and or reinforce. It is a tougher setup. After splitting, the beams would have to be “unsplit” and hit the target from the exact same angle. The adjustment to line the waves up will have to be extremely fine. You will need to be able to adjust the length of one of the legs at interval shorter than the wavelength of light being used. It can be done. In fact it has been done. I have never done it personally.
Although it’s easy to make sound waves (forr instance) from different sources interfer with each other, it’s really hard to do this with light waves. I’ve heard about one experiment where they actually successfully interfered light waves from two different lasers with each other. I’ve been out of this loop for years, so maybe it’s become less unlikely. But it’s still not an easy thing to do. In principle you could tune two different lasers to interfere with each other. But in any Real World situation, I’d just sp[lit one beam up into two parts and keep the path length differences less than a coherence length.
But, as I say, you can easily demonstrate destructive interference in this way with a HeNe laser, but you probably won’t see the kind of photon-photin scattering you’re lookibng for. I think you’d want a couple of mega-laser, like the ones they use for Laser Fusion – Shiva or Omega.
Interference is done regularly with two lasers. One of the great experiments in General Relativity was done by interfering two lasers. By putting putting one laser up on an upper floor and bouncing its beam down elevator shaft to merge with a beam generated on the ground floor, it was shown that time runs faster at the top of the building than at the bottom. Because time ran faster at the top, the interference pattern continuously cycled through from constructive to destructive inteference.
As far as the “scattering” idea, it would be hard to demonstate with any laser we have ever made. A gamma ray laser would probably do the job, but we don’t have any of those. It is not the intensity of the light, but the frequency that makes it likely.
Assume I manage this experiment with a single laser beam split (as both you and Cal mention it would be far easier this way) and manage to get them to destructively interfere at a given focal point. If I place a piece of paper at that focal point will the laser still burn through it (assume the lasers have more than sufficient power to put a hole in paper or whatever material we place there and assume you are capable of the precision necessary to perform this experiment)?
What I am getting at here with the energy ‘hiding’ is how much energy is actually present where the two waves cancel each other. Is it all still really there such that I still zap whatever is there or do they really zero out?
Sorry for the hijack but I’m curious.