X-rays are emitted when high velocity electrons in an X-ray tube collide with a metal target. Some high velocity electrons on collision with the metal atoms cause an inner electron of an atom to be ejected, creating a vacancy in the inner shell of that atom. Outer electrons fill up this vacancy and emit an X-ray photon (corresponding to characteristic X-rays).

Most explanations I have found say so much. My question is, is it certain that the inner electron is ejected completely out of the atom, and not to another vacant, higher energy level? In calculation of characteristic K$_\alpha$ ray frequencies for instance, I have always taken the difference of the energies of an ionised atom in the two states when an electron is lost from a K shell, and when it is lost from the L shell of that atom. The only difference, of course, in the situation I propose is where the excited electron finally ends up - whether it becomes free of the atom, or makes a transition to a higher energy level. Will the prediction of characteristic X-ray frequencies differ if the electron is assumed to be excited to a higher level - or is the difference too minimal to matter (the energy difference between the K and L shell energy levels for an electron is hardly affected by outer electrons anyway)? And does that electron, in reality, actually become free of the atom, or can it also move to a higher level energy state?

  • 1
    $\begingroup$ So, how do you feel about Auger electrons - do you feel there are similar problems? $\endgroup$
    – Jon Custer
    Commented Nov 2, 2015 at 15:21
  • 2
    $\begingroup$ This is an interesting question. It really boils down to whether a material particle can kick a bound electron from a lower to a higher energy level. Of course we know photons can do it but can electrons? I don't really see where Auger electrons enter the picture though... $\endgroup$
    – Gert
    Commented Nov 2, 2015 at 16:13
  • $\begingroup$ I've read of the Franck-Hertz experiment. From what I can gather, the experiment passed a stream of energetic electrons through mercury vapour. The electrons collided with the atoms and caused electrons of mercury to be excited to the second quantum level. So electrons, it seems, can excite bound electrons. $\endgroup$
    – Charles
    Commented Nov 3, 2015 at 4:20
  • $\begingroup$ There's not much difference in the energy in the resulting photon if an inner electron is knocked free or knocked to an outer bound state with weak binding energy... by definition. The only way I can imagine a difference in the generated X-rays between those two cases is if there are selection rules when going to the outer shells that don't exist when going to an unbound state. $\endgroup$
    – DanielSank
    Commented Feb 23, 2016 at 18:42

1 Answer 1


Regarding the core question,

is it certain that the inner electron is ejected completely out of the atom, and not to another vacant, higher energy level?

this is the most likely scenario but it is not completely necessary. It is perfectly possible (at least in the sense that it's not forbidden) for the electronic collision to leave the atom in an excited state with a core hole and that electron occupying some valence or Rydberg orbital, and for that state to decay radiatively to the ground state of the neutral atom, or indeed to some slightly excited state that then decays by emitting a visible or IR photon.

If that is the case, then the frequency of the emitted photon will be slightly below the shell edge. This is good and bad: on one side, the core-valence energy difference tends to be big enough that the effect of where the original electron ends up will be minimal, but on the other side, any emission below the edge will stick out like a sore thumb.

However, I suspect that the probability of this process is essentially negligible, because it's competing with the Auger effect, in which the valence electron drops back into the core hole and uses the energy to ionize another electron from the valence band, with a very fast timescale.

It is hard to make general statements because there is a wide variety of possible situations covered by the question, but my impression is that pre-edge peaks tend to be somewhere between negligible and weak, though they can be observed if your apparatus is sensitive enough:

Image source


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.