Something strange about doping semiconductors

[DrDu]

TL;DR

Doping leads to a reversal of electronegativity

Some post involving diodes made me reconsider pn junctions, with I first learned almost 40 years ago. Coming from a chemistry background, something always felt strange, but I could not tell what. Now I realized:

While e.g. Gallium arsenide is a semiconductor itself and can be described roughly as an ionic compound made of Ga 3+ and As 3- ions, if we dissolve a small amount of it in Silicium, something strange happens. Gallium acts as a source of holes, while Arsenic acts as a source of electrons. If both dopants are present in the same material, or in close proximity like in a PN junction, the electron and holes recombine and we are formally left with Ga- and As+ ions.

So somehow dissolution of GaAs in Si resulted in a reversal of electronegativity. It would be interesting to see when and how this change of oxidation states occurs as a ratio of mixing of GaAs and Si, assuming that the crystal structure does not change.

 

[DrDu]

In fact, the oxidation states of Ga and As remain +III and -III, respectively, Si has a higher electronegativity than Ga, but a lower one than As. As both Ga and As form 4 polar bonds with Si, the resulting oxidation states are +III and -III. We may compare this to the reaction of ammonia with hydrogen: 2 NH3+H2 -> 2 NH4+ + 2e- (in real life, you need some electron acceptor to stabilize the free electrons. In semiconductors, the high dielectric constant of the host may be sufficient to stabilize the electrons in addition to the entropy gain at low doping levels). So, as in chemistry, the energy gained in forming additional bonds with the host material overrules the electronegativity differences.