ABSTRACT
In Figure 5 we give an example. By slightly shifting the objective away from the plane of the nanocrystal core towards the bottom SiO2 layer, strikingly different PL spectral shapes are observed. In Figure 5a when the objective is aligned with the waveguide core, a smooth Gaussian curve is detected, which is characteristic of nanocrystal emission and very little difference is noticed between the normalized TE and TM spectra. However, if the objective is slightly off alignment, as shown in graph (b) (displacement exaggerated for demonstration), the spectrum is drastically altered. If we use a polarizer to separate the TE and TM polarized light, regularly distributed peaks are observed in the spectra with the peaks in one polarization corresponding to the valleys in
the other. This phenomenon was observed before and was attributed to modes of the waveguiding layer.30 However, we believe this is simply caused by light interference due to the layered structure. These fringes are equally spaced in energy and fitting shows their spacing corresponds well with the film thickness. This strongly suggests that the signal detected in this case is not from the waveguiding layer, but dominated by stray PL.
