| Popis: |
Two series of a-Si:H solar cells with varying absorber thickness $(t_{\mathrm{i}}= 10-500$ nm) and deposition temperature ( $T_{\text{dep}-\mathrm{i}}$ =200-350°C) were investigated by Fourier-transform photocurrent spectroscopy to determine the sub-bandgap absorption originating from the electronic localized states in the a-Si:H absorber i-layer. The thickness series showed that the bulk defect is increased by thickening $(t_{\mathrm{i}}{=}$ and by light soaking while no p-i or n-i interface defect prevails. It is demonstrated that the performance of the state-of-the-art a-Si:H p-i-n solar cell is dominated by the total number of (native and light-induced) midgap defects in the bulk i-layer. On the other hand, variation of $T_{\text{dep}-\mathrm{i}}$ has a different impact on the creation of localized states in the i-layer depending on the layer stack sequence (p-i-n or n-i-p). When depositing a-Si:H i-layer at high temperatures (>200°C), p-i-n cells showed a larger performance decrease than n-i-p cells, along with a broader bandtails and a larger number of defects in the i-layer. The results indicate that the creation of such electronic states is influenced by the local Fermi level position in the i-layer, i.e., the closer the Fermi level to the valence band (p-i interface), the more electronic states are created at high $T_{\text{dep}-\mathrm{i}}$ . These findings agree with the earlier report that p-doped a-Si:H and c-Si passivated by a p-i stack of a-Si:H layers exhibit a lower thermal stability. |
