It is argued that (1) the so-called ‘‘DX center’’ is a simple substitutional donor which is displaced from its normal, centered, lattice position, (2) its deep state is derived from a triplet of symmetry $T_2$, not $A_1$ as is generally assumed, and (3) the donor displacement, driven by the energy of an occupied antibonding orbital, resembles a strong Jahn-Teller effect but differs because of the large potential barrier between the centered ($T_2$) donor state and the distorted deep state. This model provides a natural explanation for the bulk of the data published on DX centers, especially the large barrier for thermal capture and the far-infrared absorption spectra reported by Theis et al. for the 1s-2p transition in Si-doped ${\mathrm{Al}}_{\mathrm{x}}$${\mathrm{Ga}}_{1\mathrm{-}\mathrm{x}}$As. The far-infrared absorption spectra indicate that the lowest state of symmetry $A_1$ lies where it is expected, about 70 meV below the L conduction-band edge, or nearly 100 meV above the deep ‘‘DX’’ level.

• Received 2 April 1986

DOI:https://doi.org/10.1103/PhysRevB.34.2664