ABSTRACT
The isoelectronic traplike behavior of substitutional nitrogen impurities in GaP drew a great deal of attention several decades ago [11] because it enabled the early design of visible-light-emitting diodes (LEDs). Due to the steep central-cell potential of the impurity and its consequent localization and trapping of electrons, the states Nx (generated by the isolated nitrogen impurity) and NN„, (generated by pairs of impurities) are delocalized in k-space, allowing for efficient photo-and electroluminescence from the otherwise indirect gap material GaP. More recently, success in increasing the nitrogen doping level from 1017 to 1021 cm-3 has led to excitement about the possibility of fabricating ultrabright LEDs from this material, which is closely lattice-matched to silicon. At these high N doping levels, the photoluminescence transitions from the Nx level (referred to as the Aline) and the various NNm levels broaden and merge due to formation of impurity bands [12]. The apparent lowering of this band-edge transition energy [13] is reminiscent of the abnormally large band-gap bowing recently observed for GaAsN [1].
