237 words
1 minutes
MD Notes 2 - Distribution of Carriers

Distribution of Carriers#

Distribution of electrons at certain energy n(E)n(E), is the product of state density of that energy gc(E)g_c(E), and the possibility that the energy level is filled by electrons f(E)f(E).

(spacial) Density of electrons is the integral of n(E)n(E)

n=ECtopgc(E)f(E)dE=NceEcEFkTn = \int_{E_C}^{top} g_c(E)f(E){\rm{d}}E = N_ce^{-\frac{E_c-E_F}{kT}}

under Boltzmann approx. that requires:

Ev+3kTEFEc3kTE_v + 3kT \leq E_F \leq E_c - 3kT

in other words, the semiconductor is **non-degenerately doped**.
  • NcN_c: the effective density of states in the conduction band, integral of gc(E)g_c(E).
  • n+n^+: degenerately doped n-type semiconductor, where EFEcE_F \approxeq E_c.
  • p+p^+: degenerately doped p-type.

Similarly, p(E)=gv(E)(1f(E))p(E) = g_v(E) \cdot \left( 1-f(E) \right)

alt text

Derive EFE_F from nn:

EF=EckTln(Ncn)E_F = E_c - kTln(\frac{N_c}{n})

Intrinsic Fermi Level#

Using the fact that n=pn=p in intrinsic semiconductor:

EFEiE_F \equiv E_i

EiE_i is the intrinsic Fermi level.

Also, given that n=nin=n_i

Nc=nieEcEikTN_c = n_i e^{\frac{E_c-E_i}{kT}}

So carrier concentration could be written as a function of nin_i and EiE_i:

n=nieEFEikTn = n_i e^{\frac{E_F-E_i}{kT}}

p=nieEiEFkTp = n_i e^{\frac{E_i-E_F}{kT}}

Band Gap Narrowing#

If the dopant concentration is a significant fraction of the silicon atomic density, the energy-band structure is perturbed, and the band gap is reduced by ΔEG\Delta E_G.

Dopant Ionization#

  • Dopant compensation: the effect of one type of dopant is completely or partially cancelled by adding dopant of the opposite type.
  • Net dopant concentration: the difference between the concentration of donor and acceptor dopant.

At extreme high temperature, the intrinsic excitation dominates,

n=p=nin = p = n_i

and at extreme low temperature, the ionization becomes less significant,

n=NcNd2eEcEd2kTn = \sqrt{\frac{N_cN_d}{2}}e^{-\frac{E_c-E_d}{2kT}}