In order to clarify the origin of the phosphorus-related hyperfine-electron-spin-resonance (ESR) signal we study the properties of differently P-doped amorphous hydrogenated silicon (a-Si:H) prepared at low substrate temperature (${\mathit{T}}_{\mathit{s}}$=50 °C) and the changes induced by subsequent annealing at temperatures up to ${\mathit{T}}_{\mathit{A}}$=250 °C. In addition to the standard ESR and electrical conductivity we use light-induced ESR, subgap absorption, and photomodulation spectroscopy to characterize the samples. We have found that part of the controversy concerning the origin of the phosphorus hyperfine (hf) signal is related to the fact that the usual assumptions, namely that the sample is homogeneous, the conduction-band edge energy ${\mathit{E}}_{\mathit{C}}$ is fixed, and only the Fermi level ${\mathit{E}}_{\mathit{f}}$ moves are not satisfied. When the substrate temperature decreases from 250 to 50 °C the hydrogen content and the optical gap of P-doped a-Si:H increases and the anti-Meyer-Neldel behavior indicates the shift of the transport path (${\mathit{E}}_{\mathit{C}}$). Although the total density of deep defects changes only slightly by annealing, their character changes substantially. To explain the details of the ESR results heterogeneity given by long-range potential fluctuations must be introduced.

• Received 21 August 1992

DOI:https://doi.org/10.1103/PhysRevB.47.13283