C-Si hybrid photonic structures by full infiltration of conjugated polymers into porous silicon rugate filters

Loading of one-dimensional (1-D) porous silicon photonic crystals (PS-PhCs), known as rugate filters, with luminescent materials is generally limited by the potential for (undesired) “pore clogging,” in relation to the size of the nanoparticles (e.g. quantum dots) or molecular species, and so far mainly restricted to small molecular weight materials or small nanocrystals, or in situ polymerized dyes. Here we report the infiltration 1-D PS-PhCs with a green-emitting commercial luminescent polymer (F8BT, poly[(9,9-di-n-octylfluorenyl-2,7-diyl)-alt-(benzo[2,1,3]thiadiazol-4,8-diyl)]), with a molecular weight of approximately 46 kDa across their whole depth (approximately 7.5 μm), thereby showing that pore clogging is not a concern for these structures. We also characterize the modification of the photoluminescence (PL) and decay rates, and investigate the detailed inner morphology of the filters with the help of (scanning) transmission electron microscopy. We observe both suppression (in the stop-band) and enhancement (at the high-energy band-edge) of the PL. We also find that the photonic stop-band is red-shifted after polymer infiltration, due to the increased effective refractive index of the polymer-infiltrated nanostructured system. The presence of just one unbroadened peak in the reflectance spectra after infiltration confirms that infiltration extends for the whole depth of the rugate filters.


I. DEPTH PROFILE OF THE RUGATE FILTER
shows a profile of the TEM micrograph shown in the main text in figure 1(b). Here, the maxima correspond to the brighter layer, whereas the minima are the darker layers. In this way it is therefore possible to evaluate the interplanar distance as the spacing between two minima (or two maxima). This average spacing is found to be ~150 nm, in excellent agreement with the two-dimensional fast-Fourier-transform (2D-FFT) data. To provide statistical information on the distribution of periods etc., we have also carried out a full fast-Fouriertransform (FFT) analysis of 5 depth profiles taken from figure 1b. Figure S1. Profile of the internal structure extracted from figure 1(b) of the main text.
We have also tried to fit with a sinusoid a section of the profile number #5. The poor fitting (R 2 =0.7) highlights the differences with respect to a pure sinusoidal wave.

II. ATOMIC FORCE MICROSCOPY (AFM) CROSS-SECTION IMAGE
We show in figure S2 an AFM cross section of the rugate filter at the interface between untreated silicon and the pores. In the bulk silicon we can see the cleavage plane. From observation the size of pores is not uniform, but mostly in the range of 40-100 nm.  Figure S4 Contour plot of the angle resolved reflectance spectra for the tuned PhC before a) and after (b) polymer infiltration. The angular dispersion and the width of the stop -band are not affected by the presence of the polymer. After polymer infiltration the stop -band is red-shifted of about 12 nm. Figure S5 Contour plot of the angle resolved reflectance spectra for the detuned PhC before a) and after (b) polymer infiltration. The angular dispersion and the width of the stop -band are not affected by the presence of the polymer. After polymer infiltration the stop -band is red-shifted of about 16 nm

III. ANGLE-RESOLVED SPECTROSCOPY
The apparent oscillating behaviour of the secondary reflectance fringes in Fig. S4 or the non-dispersive one on Fig. S5 is actually a visual/optical effect due to the fact that there are many (similar) fringes shifting with angle, and that at some point it is hard to follow the fringe of the same order when passing from one spectrum to the next.
We illustrate this "trompe l'oeil" effect by looking at the actual spectra taken (and from which the contour plots were derived) and reported below in Fig. S6 for both the tuned (a) and detuned (b) sample before polymer infiltration. Here, the fringes have a normal/expected blue-shift as a function of angle. To make it more obvious we have now highlighted the evolution with angle by focusing on the 10 th fringe and the 3 rd for from the main stop-band peak for the tuned and detuned sample, respectively (marked with a square as far as possible, as beyond a certain angle some of the intermediate fringes cannot be identified). To sum up, the visual effect in the contour plot is due to the shift of 1 entire period, and thus this would appear as a shift in the opposite direction, instead of a real blue-shift. Figure S6 Angle resolved reflectance spectra for the tuned (a) and detuned (b) PhC before polymer infiltration. In both cases there is a blue shift of both the main reflectance peaks and the secondary fringes. Figure S7 Contour plots of the PL spectra as a function of the incidence angle for the F8BT infiltrated into the tuned (a) and detuned (b) PhC.

High-energy edge
Film on Si 1150±57 1.496 Stop-band Film on Si 1180±57 1.800 Figure S1 Temporal decay constant (τ) extracted from mono-exponential fits of the PL decay curves of the F8BT neat film with their reduced  2 .