It is my luck that i am the first to read this question and give an answer to it.
At low doping concentration the doping level appears discrete near the band edge.
It is separated from it by the ionization energy Ei= Eedge-E doping
where Eedge is the edge of the intended band either the conduction band in case of donor doping and near to the valence band edge in case of acceptor doping.
I assume here you have acceptor doping if your host material is silicon.
As the doping concentration increases the discrete level broaden into band because of the increase in the interaction between the neighboring doping atoms.
At the same time the ionization energy decreases because of the increase in the electron concentration. This phenomenon is similar to the bandgap narrowing due to increases doping. It is due to heavy doping effects.
I treated this problem before and published some papers on it.
At certain doping concentration Ei tends to zero and the impurity band will be entered into the corresponding band.
I would like that you follow the two papers:
Article A New and Simple Model for Plasma and Doping induced Band Ga...
Article The dependence of diffusion length, lifetime and emitter Gum...
It is my luck that i am the first to read this question and give an answer to it.
At low doping concentration the doping level appears discrete near the band edge.
It is separated from it by the ionization energy Ei= Eedge-E doping
where Eedge is the edge of the intended band either the conduction band in case of donor doping and near to the valence band edge in case of acceptor doping.
I assume here you have acceptor doping if your host material is silicon.
As the doping concentration increases the discrete level broaden into band because of the increase in the interaction between the neighboring doping atoms.
At the same time the ionization energy decreases because of the increase in the electron concentration. This phenomenon is similar to the bandgap narrowing due to increases doping. It is due to heavy doping effects.
I treated this problem before and published some papers on it.
At certain doping concentration Ei tends to zero and the impurity band will be entered into the corresponding band.
I would like that you follow the two papers:
Article A New and Simple Model for Plasma and Doping induced Band Ga...
Article The dependence of diffusion length, lifetime and emitter Gum...
I am glad to meet you. For your question, I think when the doping content is low, the state of defect energy is discrete. However, once the content is higher than one value, the the state of defect energy can change to energy band. For the case of higher doping content, I think the band gap should be smaller.
The above is just my opinion. If helpful, I will be very happy.