Magnitude of the electric dipole moment in semiconductors

[ppdbacke]

When using k.p theory to calculate the magnitude of the electric dipole moment in semiconductors for band-to-band transitions, the value of the dipole moment is basically fitted to the effective mass of the electron in the conduction band. However, the values turn out to be larger than the dimension of the elementary unit cell, which does not make that much sense. We are basically calculating an overlap integral between two Bloch functions which have unit cell dimensions and a position operator. How can that ever be larger than the unit cell?

 

[ppdbacke]

As a follow up to the previous question. While deriving optical processes in semiconductors practically all textbooks use the vector potential and the electron momentum (A.p). I haven;t found a single reference that actually uses the electric dipole moment to do this and yet this is the value all books state at the end of the calculation. Does anybody have a solid reference that simply uses the electric dipole moment from the beginning to the end of the calculation?