XPost: sci.physics   
   From: ram@zedat.fu-berlin.de   
      
   john larkin  wrote or quoted:   
   >I recall someone making a laser sort of thing that dispenses a single   
   >photon periodically.   
      
     So, the Heisenberg uncertainty principle basically says you can't set   
     up a state that's got a perfectly fixed number of photons at a su-   
     per well-defined frequency. Still, you can study what are called Fock   
     states, or number states, if the light field's spread over a pret-   
     ty narrow range of frequencies. In that case, a single-photon emitter   
     ends up acting like it's kicking out just one photon.   
      
     One of the old-school and simplest tricks to get something close to   
     single-photon light was just to tone down a normal laser until the   
     average number of photons per pulse was tiny. Since laser light fol-   
     lows a Poisson distribution, you can actually play with the odds of   
     getting one photon versus more than one. Like, if the average pho-   
     ton number mu=0.1, about 90% of the pulses have none, roughly 9% have   
     one, and around 1% spit out multiple photons.   
      
     Light from an ideal single-photon source acts totally quantum - it   
     shows antibunching, meaning there's always some nonzero delay between   
     photon hits. A basic laser doesn't do that; its second-order corre-   
     lation function equals one, so no antibunching there. But for stuff   
     like quantum cryptography, you need those antibunched photons.   
      
     A perfect single-photon source still doesn't exist, though. Right now,   
     people use things like single molecules, Rydberg atoms, diamond color   
     centers, or quantum dots to make single photons, and you can actually   
     see their antibunching if you set up the experiment right.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)