Silicon doping of AlxGa1−xN layers with high aluminum mole fractions (0.8 < x < 1) was studied. The AlGaN:Si layers were pseudomorphically grown by metalorganic vapor phase epitaxy on low defect density epitaxially laterally overgrown AlN/sapphire templates. The effects of SiH4/III ratio and aluminum content on the resistivity, the carrier concentration, and the mobility have been investigated. By variation of the SiH4/III ratio during the growth of AlxGa1−xN:Si layers, a recorded low resistivity of Al0.81Ga0.19N:Si was obtained with 0.026 Ω cm. The resistivity increases exponentially with increasing aluminum content to 3.35 Ω cm for Al0.96Ga0.04N, and the optimum SiH4/III ratio is shifted towards lower values. Hall effect measurements show that the increase of the resistivity with increasing aluminum mole fraction is mainly caused by a decrease of the carrier density. The optimized Al0.81Ga0.19N:Si exhibits a carrier concentration of 1.5 × 1019 cm−3 and a mobility of the carriers of 16.5 cm2 V−1 s−1.

The dependence of the activation energy as well as the energetic levels of the neutral charge state and the DX center of the Si donor in AlxGa1−xN:Si samples on aluminum content and SiH4/III ratio were investigated by electron paramagnetic resonance (EPR) measurements, Van-der-Pauw resistivity measurements, and Hall-effect measurements. It was found by EPR measurements that the energy distance of the neutral charge state of the Si donor from the conduction band increases with increasing aluminum content from 61 meV for x = 0.82 to 106 meV for x = 0.89. Additionally, the formation of a DX center below the neutral charge state which is deepening from 6 meV for x = 0.82 to 58 meV for x = 0.89 is observed. This results in a linearly increasing activation energy with increasing aluminum content from 67 meV for x = 0.82 to 164 meV for x = 0.89. This is consistent with the activation energies as determined by Hall-effect measurements showing a linear increase from 24 meV for x = 0.85 to 211 meV for x = 0.96, as well as the activation energies as determined by Van-der-Pauw resistivity measurements. By varying the SiH4/III ratio we observed a formation of a minimum resistivity in accordance with the room temperature charge carrier density. However, no clear dependence of the activation energy was observed. EPR measurements of samples with a high SiH4/III ratio hint to an increased incorporation probability of a deep secondary donor species which might explain the increase in resistivity.