A1−T2splitting for substitutional nitrogen in diamond

The reason for which extended Hückel theory leads to an inverted order for the $A_1$ and $T_2$ levels of substitutional nitrogen in diamond is discussed in some detail. This situation is shown to originate from an overestimation of the interactions involving "$s$" states, which also leads to a bad description of the conduction band. The reduction of the corresponding parameters leads to a normal ordering $E_{A_1}<E_{T_2}$ which agrees with the local-density calculations. This finding has implications for the existence of Jahn-Teller distortions which are discussed.