On Friday, June 14, 2019 at 4:27:19 PM UTC-5, Jos Bergervoet wrote:   
   > Consider a Stern-Gerlach device using a field in the z-direction.   
   >   
   > Entering is a single-electron Gaussian wave packet with spin   
   >   Psi(x) ( |u> + |d> ) / sqrt(2)   
   > where the initial spin is clearly in x-direction.   
   >   
   > We know the apparatus will split the trajectory of the components   
   > and due to the non-uniform grading of the field two separate wave   
   > packets will come out, moving in slightly different directions.   
   >   
   > But before the separation is complete, isn't there fast precession   
   > of the spin around the z-axis? And then wouldn't the electron radiate   
   > energy, so fall down to the energy level of the |d> spin which in   
   > the field is lower than the |u> spin?   
   >   
   > So my question is: do we expect more down-spin to come out than up-   
   > spin? (When the input components both have exactly equal amplitude,   
   > I mean!) Or is there a reason for this mechanism to be insignificant?   
   >   
   > --   
   > Jos   
      
   A practical answer is that the only radiation would be via the   
   magnetic moment of the electron, which is so weak that it doesn't   
   couple strongly and thus the radiation would be very weak.  i.e.   
   it would be a rare event.  I haven't calculated the energy for   
   reasonable magnetic fields, but I rather suspect it would correspond   
   to a very low frequency radio wave, which again would couple very   
   poorly to the electron magnetic moment.   
      
   Even if this did happen, what would be observed (if you didn't   
   detect the radiation) is a slight increase in the statistics of the   
   |d> state over the |u> state.  Given the (i expect) rare transition   
   probability it might be very hard to discern the difference.   
      
   Rich L.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)