Bandgap energy modeling of the deformed ternary GaAs(1-u)N(u) by artificial neural networks.

Appraising the bandgap energy of materials is a major issue in the field of band engineering. To better understand the behavior of GaAs(1-u)N(u) material, it is necessary to improve the applied calculation methodologies. The band anticrossing model (BAC) allows modeling of the bandgap energy when diluted nitrogen is incorporated into the material. The model can be improved using artificial neural networks (ANN) as an alternative solution, which is rarely applied. Our goal is to study the efficiency of the (ANN) method to gauge the bandgap energy of the material from experimental measurements, considering the extensive strain due to the lattice mismatch between the substrate and the material. This makes the GaAsN material controllable with (ANN) method, and is a potential candidate for the fabrication of ultrafast optical sensors.