*3.4. Photocatalytic Activity*

The photocatalytic activity of the device was tested under both UV (Jelosil HG500, effective power density: 17 mW cm−<sup>2</sup> in the 280–400 nm) and simulated solar irradiation (UltraVitalux lamp, Osram, Munich, Germany, effective power density: 4.5 mW cm−<sup>2</sup> in the 280–400 nm range and 14 mW cm−<sup>2</sup> in the 400–800 nm range).

Tests were carried out both in the gas phase and in water. The gas phase degradation of ethanol was carried out using a previously reported setup [37]. An active surface area of 38 cm2 and an initial pollutant concentration of 198 ppm were employed; a gas chromatographic system was adopted for monitoring the pollutant disappearance, and the formation of acetaldehyde (main reaction intermediate) and CO2 (complete degradation product). Three consecutive photocatalytic tests were performed in the same conditions to test the stability and the reusability of the device. The degradation of a tetracycline hydrochloride (TC) in water was conducted in an open reactor, with a total active surface of 60 cm2 and an initial pollutant concentration of 12 ppm (V = 300 mL). The reaction was carried out at spontaneous pH and oxygen saturation was maintained via air bubbling. Before irradiation, the device was kept in the dark for 30 min in order to achieve adsorption-desorption equilibrium. The molecule disappearance was monitored by UV-vis spectrophotometry, by recording the intensity of the characteristic absorption peak of tetracycline at 357 nm, in agreement with previous literature reports [41,49,50].

#### **4. Conclusions**

In this work, we presented a floating photocatalytic device based on TiO2 photocatalyst immobilized over an ad hoc synthesized ter-polymer. The developed methylmethacrylate, α-methylstyrene and perfluoroctyl methacrylate ter-polymer is characterized by a highly porous morphology and inherent hydrophobicity, which enable a stable buoyancy. Furthermore, the support, displaying the characteristic optical transparency and oxygen permeability of PMMA, was engineered to possess enhanced thermal stability, mechanical resistance and photostability, in order to promote the device durability. The compatibility of the inorganic top coating was enhanced by a series of strategies: (1) an ad hoc polymer casting method leading to reorganization of the hydrophobic chains and dual wetting features of the opposite film sides; (2) the corona treatment of the polymer surface aimed at increasing hydrophilicity and creating surface pitting to bolster adhesion of the inorganic coating; (3) the deposition of an intermediate SiO2 layer, which improves the adhesion of the TiO2 top layer and protects the polymer support from the radical species generated by photocatalytic oxidation. The adopted TiO2 layer contains commercial nanoparticles with high photocatalytic activity bound together by a titania sol promoting adhesion. The final device showed promising results in photocatalytic degradation tests of both water and gas phase pollutants, also in recycle tests. Tests were carried out under both UV and simulated solar irradiation, with either front or back irradiation, confirming the ability of the device to work also when capsized. Future work will further optimize the TiO2 amount to boost the photocatalytic activity of the device by both increasing the nanoparticles content in the top layer and using fibers and sponge architectures as an alternative to polymer films.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2073-4344/8/11/ 568/s1, Figure S1. 1H NMR spectrum of POMA monomer, Figure Figure S2. 1H NMR spectrum of MMA\_ α-methylstyrene\_POMA polymer, Figure S3. O2 transfer with respect to time of the synthesized MMA\_α-methylstyrene\_POMA, Figure S4. FT-IR spectra of MMA\_α-methylstyrene\_POMA collected before and after the UV stability test at the air (i) and PTFE (ii) side, Figure S5. Ethanol disappearance and acetaldehyde and CO2 formation during the photocatalytic test under UV irradiation, Figure S6. Cross-sectional SEM images after UV irradiation for over 15 hours in working conditions.

**Author Contributions:** Conceptualization, V.S., D.M., S.A.; methodology, L.T., V.S., L.R.; formal analysis and data curation, L.R., V.S., D.M.; writing—original draft preparation, D.M., V.S., L.R.; writing—review and editing, D.M., S.A., V.S., L.R.; resources, M.A.O., H.F., S.A.; supervision, M.A.O., H.F., S.A.

**Funding:** This research received no external funding.

**Acknowledgments:** The authors wish to thank Stefano Farris and Riccardo Rampazzo of the Dipartimento di Scienze per gli Alimenti, la Nutrizione e l'Ambiente (Defens) at the Università degli Studi di Milano, for assistance during corona treatment and gas permeability measurements.

**Conflicts of Interest:** The authors declare no conflict of interest.
