Nowadays, sensor systems on the basis of porous matrices are of great interest for the detection of toxic and dangerous substances as well as for biosensing [1–3]. However, most of the sensors are arranged in such a way that the contact of the medium with the inner surface of porous matrixes is limited by the substrate. A freestanding porous matrix, as the active element of a sensor, significantly increases the detection efficiency due to the possibility of pumping the analyte through the porous structure. The latter is true both for sensors working on the principle of the reflection spectrum shift due to a local change in the refractive index, and for hybrid sensors, based on the registration of the selective fluorescent labels signal [1,4,5] or luminescence quenching  of luminophores embedded into the porous structure. In addition, freestanding porous matrices are able to increase the efficiency of solar cells  and can be used as transmission optical elements: narrowband and edge filters [8,9].
In our study, we have fabricated freestanding porous silicon microcavities and measured their reflection spectra before and after lifting them off from the substrate. The reflection spectra of the microcavities remain almost unchanged, which allows us to state that the used lift-off technique does not affect the integrity of the samples. Furthermore, we have shown that thermal oxidation could be used to stabilize optical properties of freestanding porous silicon photonic crystals and does not lead to the significant deformation of freestanding porous silicon structure.
2. Materials and methods
We have used electrochemical etching of highly doped monocrystalline silicon wafers in hydroalcoholic solutions of 48% aqueous hydrofluoric acid to produce porous silicon photonic crystals as described in details in [10–12]. Briefly, the etching current density was varied from 3 mA/cm to 30 mA/cm to create a porous silicon microcavity. The process has been carried out in home-built PTFE etching cell with etching surface area around 1 cm . Highly boron-doped p-type monocrystalline silicon wafers with (100) orientation and resistivity of 0.003 Ω cm have been used as a substrate. Once samples of photonic crystals have been fabricated, their reflection spectra were measured with the OceanOptics USB2000+ spectrometer, just before lifting-off from the substrate.
In order to lift-off porous structure from its substrate, we have applied an electropolishing method. During electropolishing, the layers of the silicon atoms were removed, in a layer-by-layer fashion, and the porous silicon was separated from the wafer, i.e. lifted-off. The switching from porous silicon formation to electropolishing was ensured by the increase of the current density to approximately 300 mA/cm (using an electrolyte with the same hydrofluoric acid concentration as that for the porous silicon formation mode) . However it was shown that in order to obtain a higher surface quality, it is more appropriate to use a diluted HF solution rather than HF solution used for porous silicon formation . In fluoride-depleted solution, the electropolishing process leads to the formation of oxygen compounds of silicon at the interface between the porous layer and the silicon substrate, where holes, provided by the substrate, are easily available to promote silicon oxidation. Subsequent attack by fluoride ion in the electrolyte dissolves the oxide layer, resulting in the lifting-off the porous silicon structure from the substrate. The lower HF concentration also makes it possible to reduce the current density necessary for the lift-off by 1-2 orders of magnitude. Thus, our electropolishing current density was 6 mA/cm . After all, freestanding porous silicon microcavities were placed in hexane to avoid adhesion to the silicon substrate and to minimize the possibility of shattering of the film due to the strong capillary forces and thermal stresses exerted when ethanol evaporates from the pores. The same monocrystalline silicon wafer can be etched multiple times. Finally, the reflection spectra of freestanding porous silicon microcavities were measured and compared with the reflection spectra of non-lifted-off samples.
3. Results and discussion
The comparison of the reflection spectra of porous silicon microcavities on the substrate with the freestanding is shown in Figure 1. It can be seen, that the spectra do not differ significantly: the form of the reflection spectra remains almost unchanged, its eigenmode and characteristic secondary peaks are preserved. There is also a 5-10 nm blue shift, which could be attributed to the partial oxidation and subsequent dissolution of the silicon on the surface of inner porous layers during the lift-off. The fact that the reflection spectra of microcavities, the eigenmode position and full width at half maximum in particular, remain almost the same allows us to state that the technique we used for lifting-off porous structures from the substrate does not lead to the collapse of multilayer structure and does not affect their morphology. Minor blue shift observed is due to the change of effective refractive indexes of porous layers.
In order to stabilize optical properties of freestanding microcavity samples we performed thermal oxidation. We used thermal oxidation in soft regimes to prevent freestanding porous films from crushing due to high mechanical stresses arising during the transition of silicon into silicon oxides. In details, after the lifting-off the photonic crystal samples have been oxidized in air atmosphere in tube furnace at 500 C during 3 hours and the reflection of the oxidized samples have been measured (Figure 2). One can see that the photonic bandgap was shifted by about 140 nm, while the shape of the spectrum did not change significantly (Figure 2). Narrowing of the photonic bandgap full-width-at-half-maximum was attributed to the decrease of the refractive index contrast between the layers with different porosity. Thus, we can conclude that the oxidation of the freestanding porous silicon microcavities does not lead to the destruction of multilayer structure and preserves the photonic properties of the microcavity.
In this study, we have investigated the technique of lift-off of the porous silicon photonic crystals from monocrystalline silicon substrates using electropolishing method. It has been demonstrated that there is no significant changes in the reflection spectra of freestanding porous structures. Photonic bandgap width and position, reflectance and eigenmode position and full-width-at-half-maximum remained the same, and only the small blue shift has been observed. Further, the thermal oxidation of the freestanding microcavities can be performed without any subsequent disruptions of the multilayer structure and morphology. The obtained results show that the lifting-off technique developed we used in our study is not destructive for the multilayer structure of the samples and morphology of the photonic crystal and could be used to increase the performance of many devices in photonics and sensing.