Photoluminescence and electroluminescence characterization of a-Si: H/c-Ci interfaces

Abstract

Devices based on silicon heterojunctions (HT) formed between hydrogenated amorphous silicon (a-Si:H) and crystalline silicon (c-Si) are very attractive for solar cell applications; efficiencies of 24.7% have been reached (Terakawa et al., MRS Proc 1536:17–26, 2014), with relative low cost of fabrication. The interface of the a-Si:H/c-Si HT solar cell and its characterization remain critical issues in order to achieve optimal efficiencies. Luminescence spectroscopy techniques are powerful approaches that are nowadays currently used for the characterization of solar cells. More specifically, photoluminescence (PL) and electroluminescence (EL) are increasingly carried out under different intensityfrequency regimes to investigate interfaces, surface passivation and bulk properties at the different steps of the a Si:H/c Si HT solar cell fabrication. In the first part of this study, PL and EL measurements have been performed under different excitation regimes, namely steady-state and variable frequency small signal modulated conditions, and conducted on different HT configurations: a Si:H(p)/c-Si(n) and a-Si:H(p)/a Si:H(i)/c-Si(n). In modulated conditions, experimental results show that the light intensity dependence of the effective lifetime extracted by PL is in good agreement with the one determined from EL technique (Fig. 36.1). The effective lifetime estimated for the different HT configurations points out higher values for the HTs with an intrinsic layer. This confirms that passivation with an intrinsic layer improves solar cells performances. The second part of the study concerns a multiscalar PL analysis of c-Si through two complementary experimental setups. A confocal microscope has been used for the microscopic scale approach, and a non-confocal system for the macroscopic scale. The same excitation wavelength (785 nm) was used in both experiments. Depending on the PL setup we observed different PL spectra behavior. At shorter wavelengths, a shoulder is evidenced for the confocal PL system, which does not exist for the non-confocal PL system (Fig. 36.2). The difference between both spectra is explained by the depth scan capabilities of the confocal system that allow one to discriminate between a volume element of photons emitted from the near-surface and the bulk. On the other hand, the non-confocal PL system integrates a PL signal from different depths where reabsorption has a high probability to take place for the more energetic photons; the relative weight of these high energy photons in the measured PL is thus reduced, making the short wavelength shoulder disappear. The following interpretation has been discussed and supported by numerical modeling.