In calculating the transition levels of defects in semiconductor, we may meet with the issue of so-called negative-U center.

For example, in Prof. Chris G. Van de Walle's paper: Role of nitrogen vacancies in the luminescence of Mg-doped GaN ( Appl. Phys. Lett. 100, 142110 (2012) ), for the N-vacancy in GaN they mentioned:

" The 2+ charge state is always higher in energy than the + and 3+ charge states, forming a negative-U center. "

" The negative-U character, in which the 2+ charge state is always higher in energy than the 3+ and + states, is due to large and charge-state dependent relaxations around VN. "

So here is my question:

1. How to understand the negative-U center

2. Compared with the general defect transition level, does it feature any special characteristics and impacts to the semiconductor (or device) performance?

3. Can the negative-U center be experimentally measured and how?

Thank you for sharing.