From: pireddag@hotmail.com   
      
   I posted this a few days ago in it.scienza.fisica (in Italian), but I   
   did not get any answer there. I try here.   
      
   I raed a few days ago  “The chemical potential of radiation,”, di  P.   
   Wurfel, J. Phys. C Solid State Phys., vol. 15, 3967 (1982).   
      
   It explains (I forgot many details but the synthesis here should be   
   enough) how to calculate radiation from a semiconductor in which some   
   electrons have been excited from te valence to the conduction band.   
      
   The steps are the following.   
      
   Excited electrons and the one in the valence band reach quickly and   
   separately an equilibrium distribution, while the two separate   
   populations are not in equilibrium with each other (so both   
   distributions are in quasi-equilibrium).   
      
   The two distributions (conduction band and valence band electrons) are   
   characterized by two different chemical potentials   
      
   The interaction with the electromagnetic field is given by absorption,   
   stimulated emission and spontaneous emission.   
      
   One then imposes that the sums of transitions per unit time and per unit   
   energy caused by stimulated emission, absorption and spontaneous   
   emission leaves the system in equilibrium:   
      
   r_sp + r_st + r_a = 0   
      
   I do not understand this. Why should radiation maintain the   
   quasi-equilibrium of the system? Is it approximately true? If so, why   
   and what is the approximation? Should I perhaps see the semiconductor as   
   under continuous excitation by a source which I am not counting in the   
   above equation?   
      
   The experimental data shown in the article (luminescence from a GaAs   
   diode) match well the calculation done in the article (perhaps there is   
   a fitting parameter, did not read again before posting).   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)