How can it be explained that the carrier distribution is different in case of doped semiconductors? I understand it naturally, that the distribution of electrons is higher for n-type semiconductors. But is it also possible to explain it via the density of states or fermi level or so? I guess it should as the second column represents the density of states and there you can see that the number of electrons is rising with with the density of states which is rising in return with the energy.

I know the DoS for electrons is:

$ g_c(E) = \frac{2m^*}{h^2}^{\frac{3}{2}} \cdot \sqrt{E} $

in case of electrons, $ \sqrt{E} $ is $ \sqrt{E - E_C} $

and then? What is E (exactly)?

doped tpyes

  • $\begingroup$ The density of states in the conduction/valence bands is the same (for reasonable levels of doping). The occupation of those states differs greatly because of the doping. Why that is will be explained earlier in the chapter most likely... $\endgroup$ – Jon Custer Aug 16 '19 at 19:17
  • $\begingroup$ I'm really unclear on what the question is. $\endgroup$ – Gilbert Aug 17 '19 at 12:25

In your formula, $g(E)$ is the density of states for an electron gas and $E$ is the energy of a state. An undoped semi-conductor cannot be approximated as an electron gas. To use this DOS, one needs to dope it and the extra holes/electrons in the valence/conduction bands could be considered as an electron gas. Then the formula can be applied.

For low doping, the DOS does not usually depend of the doping, only the occupations. For large doping, one might need to consider interactions and then, maybe the electron gas approximation is not valid anymore.

| cite | improve this answer | |
  • $\begingroup$ Thank you! The DoS for the carriers is constant inside the Semiconductor, or? $\endgroup$ – Ben Aug 17 '19 at 12:34
  • $\begingroup$ I'm not sure I understand your question, but the actual shape of the DoS is material dependant. For a doped semi-conductor, the DoS in the conduction band could be approximated as the one for an electron-gas where $g(E)\propto \sqrt{E-E_C}$. But this woule be only valid close to the band edge. Elsewhere it might be very different. $\endgroup$ – fgoudra Aug 17 '19 at 15:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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