3.2.2. Synthesis of MMA\_α-Methylstyrene\_POMA Ter-Polymer

The reaction was performed under inert atmosphere in a 100 mL two-necked round bottom flask equipped with a nitrogen inlet adapter, a reflux condenser and an overhead magnetic stirrer. 40 mL of cyclohexane was mixed with MMA (14.6 g) and α-methylstyrene (4.3 g) POMA (0.8 g) and AIBN (0.3 g), used as free radical initiator. The resulting molar ratios were 8:2 MMA: <sup>α</sup>-methylstyrene, 1% mol·mol−<sup>1</sup> POMA: (MMA and <sup>α</sup>-methylstyrene), and 1% mol·mol−<sup>1</sup> AIBN: (MMA, <sup>α</sup>-methylstyrene and POMA). The solution heated in an oil bath at 70 ◦C for 24 h, then gradually cooled down to room temperature. A white solid was precipitated by addition of a large excess of methanol. After recovering the solid by filtration, the polymer was washed with methanol for several days under stirring to remove unreacted methacrylic monomers. After washing, the polymer was dried in a vacuum oven (ca. 4 mbar) at 40 ◦C for 48 h. The structure of the product was confirmed via 1H NMR spectroscopy (Figure S2).

### 3.2.3. Silica and Titania Sol Preparation

The silica sol was prepared by a modification of the procedure reported by Soliveri et al. [34]. Firstly, 10 g of tetraethyl orthosilicate (TEOS, Sigma-Aldrich, St. Louis, MI, USA) was added to a solution of 4.5 g of 0.1 M HCl and 25 g of ethanol. The mixture was stirred at room temperature for 120 min and then refluxed at 60 ◦C for 60 min. After cooling down, a solution prepared by dissolving 2 g of Lutensol ON 70 (BASF, Ludwigshafen, Germany) in 25 g of ethanol (>99.8%, Sigma-Aldrich) was added to the reaction mixture and stirred for 1 h. A stable, transparent sol was obtained.

For the preparation of the titania sol [48], 28.4 g of titanium isopropoxide (97%, Sigma-Aldrich) was dissolved in 79 g of ethanol and 0.9 mL of HCl 37% was added while stirring. Then, a solution of 0.47 g of Lutensol ON 70 in 79 g of ethanol was added. The resulting stable and transparent sol was stirred at 1 h at room temperature.

#### 3.2.4. Device Preparation

Polymer films were prepared via solution casting: 1.8 g of polymer was dissolved in 10 mL of CH2Cl2 and the resulting solution was cast onto a PTFE mould (7 cm in diameter). The films were dried overnight at 25 ◦C and atmospheric pressure. The air side of the polymer film was corona treated (Aslan Machinery, Germantown, MD, USA; voltage: 0.43 kW; exposition time: 5 min) in order to promote the adhesion of the oxide layers. Then, the SiO2 layer was deposited by spray coating the silica sol (nozzle diameter: 0.5 mm, target-nozzle distance ca. 30 cm; spraying time ca. 1 s, 3 layers). After

drying at room temperature, the TiO2 layer was deposited by spray coating a suspension of Hombikat UV100 (Sachtleben Chemie GmbH, Duisburg, Germany) (pure anatase, average crystallite size 10 nm, specific surface area ca. 350 m2 g−1) in the prepared titania sol (2 mg of Hombikat UV100 + 1 mL of ethanol + 0.3 mL titania sol), in the same conditions adopted for the silica layer. The prepared device was immediately dried in oven at 90 ◦C for 20 h. A further treatment was performed by immersing the device in water for 10 min at 70 ◦C and finally in 10−<sup>4</sup> M HNO3 for 20 min at 70 ◦C, according to a previously reported procedure [23]. Finally, the device was dried in an oven and irradiated for 3 h under UV light.

#### *3.3. Characterization Methods*

Nuclear magnetic resonance spectroscopy (NMR). 1H NMR spectra were collected at 25 ◦C with a Bruker 400 MHz spectrometer (Bruker, Billerica, MA, USA). The samples for the analyses were prepared dissolving 10–15 mg of POMA/ter-polymer in 1 mL of CDCl3.

Size Exclusion Chromatography (SEC). The polymer molecular weight before and after accelerated ageing was investigated by SEC using a Waters 1515 Isocratic HPLC pump (Waters, Milford, MA, USA) and a four Phenomenex Phenogel (5 × <sup>10</sup>−<sup>3</sup> Å–5 × <sup>10</sup>−<sup>4</sup> Å–5 × <sup>10</sup>−<sup>5</sup> Å–5 × 500 Å) column set with a RI detector (Waters 2487, Milford, MA, USA) using a flow rate of 1 mL/min and 40 μL as injection volume. Samples were prepared dissolving 40 mg of polymer in 1 mL of anhydrous THF; before the analysis, the solution was filtered with 0.45 μm filters. Molecular weight data were expressed in polystyrene (PS) equivalents. The calibration was built using monodispersed PS standards having the following nominal peak molecular weight (Mp) and molecular weight distribution (D): Mp = 1,600,000 Da (D ≤ 1.13), Mp = 1,150,000 Da (D ≤ 1.09), Mp = 900,000 Da (D ≤ 1.06), Mp = 400,000 Da (D ≤ 1.06), Mp = 200,000 Da (D ≤ 1.05), Mp =90,000 Da (D ≤ 1.04), Mp = 50,400 Da (D = 1.03), Mp = 30,000 Da (D = 1.06), Mp = 17,800 Da (D = 1.03), Mp = 9730 Da (D = 1.03), Mp = 5460 Da (D = 1.03), Mp = 2032 Da (D = 1.06), Mp = 1241 Da (D = 1.07), Mp = 906 Da (D = 1.12), Mp = 478 Da (D = 1.22); Ethyl benzene (molecular weight = 106 g/mol). For all analyses, 1,2-dichlorobenzene was used as internal reference. The molecular weights of the samples obtained after the UV exposure test were also determined.

Differential Scanning Calorimetry (DSC). The polymer glass transition temperature (Tg) was measured by DSC analyses on a Mettler Toledo DSC1 (Zurich, Switzerland), using a 10 ◦C/min heating rate and under nitrogen atmosphere. Before measurement, samples were heated at 90 ◦C to eliminate residual internal stresses from the synthesis.

Fourier Transform Infrared Spectroscopy (FT-IR). FT-IR spectra were collected on a Spectrum 100 spectrophotometer (Perkin Elmer, Waltham, MA, USA) working in attenuated total reflection (ATR) mode. A single-bounce diamond crystal was used with an incidence angle of 45◦.

Water Contact Angle (WCA) analyses. Water contact angle measurements were carried out using a Krüss EasyDrop (Krüss, Hamburg, Germany), on at least ten independent measurements using 5 μL water droplets.

UV-vis spectroscopy. UV-vis transmittance spectra were acquired in the 250–800 nm range by using a Shimadzu UV2600 spectrophotometer (Tokyo, Japan).

Scanning electron microscopy (SEM). Top and cross-sectional SEM images were collected on a Zeiss LEO-1430 (Oberkochern, Germany). Samples were sputter coated with Au before measurements.

Mechanical properties. The polymer sample preparation and the determination of their elastic modulus, tensile strength and elongation at break were performed in agreement with the ISO 527-1/2 standard method using a Kistler 9273 dynamometer (Winterthur, Switzerland).

Permeability test. The permeability test was conducted in oxygen atmosphere, at 23 ◦C and with a relative humidity of 40%, in accordance to the ASTM D3985 standard test method for oxygen gas transmission rate through plastic films and sheeting using a colorimetric sensor.

UV stability test. An accelerated aging test was conducted according to the UNI10925:2001 standard method to evaluate the stability under UV radiation of the prepared polymer foils. The test was conducted for 100 h (T = 25 ◦C and *p* = 1 atm), with a Jelosil HG500 lamp (Milan, Italy; 45 mW cm−<sup>2</sup> in the 280–400 nm range).
