Effects and implications of photon reabsorption phenomena on confocal micro‐photoluminescence measurements in crystalline Si

Abstract

Confocal micro‐photoluminescence (PL) spectroscopy has become during the last years one of the most important tools for carrying out novel studies of advanced solar cell concepts with micron resolution. This work presents a comprehensive study about the effect of photon reabsorption in confocal micro‐PL spectroscopy measurements. First, supported by theoretical calculations, we study the dependence of reabsorption phenomena on the different setups and experimental parameters, i.e. excitation wavelength, pinhole aperture, numerical aperture (NA) of focusing lenses. Second, we analyze the effects of reabsorption on the emission line‐shape of the resulting micro‐PL spectra. Finally, in order to prove the importance and implications of this study, we present a current and relevant application, namely the estimation of doping densities in crystalline Si (c‐Si) via micro‐PL, where reabsorption processes must be taken into account to avoid the misinterpretation and misquantification of the obtained micro‐PL data. Figure 1(a) shows the normalized PL emission spectra calculated with our theoretical model (based on the generalized Plank's law) for photons spontaneously emitted at different X distances from the front surface (see inset diagram). The variation in the reabsorption level is evidenced by a decrease in the PL signal between 1000 and 1100 nm. This is correlated to the increase of Si absorption coefficient at lower wavelengths (see the black dashed curve). Figure 1(b) shows confocal micro‐PL spectra recorded from a p‐type c‐Si wafer by changing the collection point distance from the front surface, i.e. 0 (on focus), 100 and 200 µm. The measured PL spectra of Figure 1(b) are qualitatively the same than the calculated spectra of Figure 1(a), proving the consistency of our theoretical model. When the collection point is positioned at the surface (common conditions in micro‐PL measurements), the reabsorption level is dependent on the selected experimental parameters, which define the extension of the collection volume, and hence, the available photon travelling distance. This is observed in Figure 2(a) for different NA values, and in Figure 2(b) for different pinhole apertures. The higher the confocality, the lower the reabsorption. The effect of reabsorption on the PL spectra line‐shape can have important implications in some type of applications. One of these examples is the estimation of doping densities in c‐Si via micro‐PL spectroscopy, where the band gap shift that takes place in heavily doped c‐Si can be studied at room‐temperature by monitoring the center of mass of the PL spectra. Figure 3 shows the study and quantification of doping densities in a laser‐doped region (LDR) in c‐Si. We present data for two lenses with different NA. A picture of the studied LDR is shown in Figure 3(a), and the change in the center of mass of the PL spectra recorded along the LDR is depicted in Figure 3(b). The large shift between the two curves is linked to a change in reabsorption level. As expected, the two curves are constant outside the LDR, but the center of mass increases remarkably inside the LDR, showing, apparently, an increase in doping density. By means of the calibration curves obtained from c‐Si wafers with different doping densities (see Figure 3(c)), we can transform the center of mass profiles depicted in Figure 7(b) into the doping density profiles of the LDR under study (see Figure 7(d)). Now, the two curves show a very similar and reliable doping density profile across the LDR, reaching doping levels around 7x10 19 cm ‐3 at the very center of the processed region. This example proves the importance of considering reabsorption effects in all those micro‐PL studies that are based on the analysis of the PL spectra line‐shape and position.