Re: Is a 'neutral' metal-semiconductor junction possible?

olongdomango@xxxxxxxxxxx wrote: 
Hello,

While reviewing literature on metal-semiconductor junctions, I came
across a definition for a "neutral contact", which was described as the
junction that is formed when two materials with identical work
functions are joined, so that no band bending occurs. This makes sense
for two identical materials, but if for example, an n-type
semiconductor with a wide forbidden band gap and a metal somehow had
the same work functions and were joined, I appreciate that there would
be no built-in potential barrier, but wouldn't a Schottky barrier be
formed due to the position of the conduction band edge over the common
Fermi level? Would electrons still not have to overcome this barrier if
the metal was biased negative wrt the semiconductor? Is it that metal
electrons simply tunnel into the forbidden gap, where they are swept to
the conduction band under the influence of the external bias? If so,
wouldn't there not be a minimum bias required to achieve this, which
would be proportional to the size of the energy gap between the
Fermi-level and the conduction band? Sorry for all the questions, but I
would be glad if someone could help me to clarify my understanding.

This sort of "neutral contact" doesn't say anything about the transport
properties [I(V) curve].  If you have, say a typical semiconductor like
Si lightly doped p-type or GaAs very lightly doped n-type, you can get
a flat energy-band diagram.  The Fermi level lies near the middle
of the band gap, and with very precise control might be made to line up
with the pinned value of the metal Fermi level at the semiconductor surface.
In principle, these junctions should behave like normal Schottky diodes,
turning on with a voltage that pushes the semiconductor majority carriers
toward the junction.  In practice, the resistivity of the semiconductor
will be so high that all you are likely to see is leakage currents.

For a junction on a highly ionic (wide gap) material, where the electron
affinity rule actually works, the situation is the same, except that the
Schottky barrier depends on the metal work function, and has to be taken
into account.  But here the equilibrium carrier concentration is so low in the
neutral case that you probably won't have a carrier in any reasonable device
volume.

A very interesting case is InAs, whose bands lie so low in energy that the
conduction band edge is a bit below the Fermi level of metal contacts made
to it.  The neutral junction condition thus requires degenerate n-type doping,
and there is no energy barrier at all.  Since any reasonable carrier
concentration in InAs is degenerate (because the effective mass is very low)
putting just about any metal on n-type InAs gives an excellent Ohmic contact.

- Bill Frensley

.

Relevant Pages

  • Re: Is a neutral metal-semiconductor junction possible?
    ... only a true neutral junction can be ... >> semiconductor with a wide forbidden band gap and a metal somehow had ... >> formed due to the position of the conduction band edge over the common ... The neutral junction condition thus requires degenerate n-type doping, ...
    (sci.engr.semiconductors)
  • Re: Logarithmic response of PN junctions
    ... voltage of a PN junction. ... Is that why they're called "natural logs"? ... hex, octal, whatever, the curve and the math is the same. ...
    (sci.electronics.basics)
  • NPN diamond transistor
    ... semiconductor. ... P junction diamond has not been discovered. ... a silicon P junction sandwitched between two N doped diamond ...
    (sci.materials)