*2.2. Apparatus*

1H NMR spectra were recorded on a Bruker AV 300 MHz (Billerica, MA, USA) instrument using CDCl3 as solvent. Chemical shifts (δ) for 1H spectra are expressed in ppm relative to internal Me4Si as standard. Signals were abbreviated as s, singlet; d, doublet; t, triplet; m, multiplet.

Elemental analysis was performed by the Microanalytic Laboratory of Organic Chemistry (LOC) at ETH Zürich.

Gel permeation chromatography (GPC) analysis were performed by using Viscotek GPC system (Malvern, Worcs, UK) equipped with a pump and degasser (GPCmax VE2001, 1.0 mL min−<sup>1</sup> flow rate), a detector module (Viscotek 302 TDA) and three columns (2× PLGel Mix-C and 1× ViscoGEL GMHHRN 18055, dimensions 7.5 × 300 mm for each column) using tetrahydrofuran as an eluent.

Rheology measurements were carried out using an Anton-Paar MCR-302 rheometer with parallel plates (Graz, Austria), at a temperature of 120 ◦C. Complex viscosity was measured collecting 10 points at angular frequencies starting from 0.23–22 rad s<sup>−</sup><sup>1</sup> applying 5% of strain.

For thermogravimetric analysis (TGA), a Mettler Toledo TGA/DSC 3+ STARe system instrument (Mettler-Toledo, Schwerzenbach, Switzerland) was used heating the samples from 25 to 1000 ◦C with a heating rate of 10 ◦C min−<sup>1</sup> under nitrogen atmosphere.

Di fferential scanning calorimetry (DSC) was carried out with a Mettler Toledo DSC822e instrument (Columbus, OH, USA) using a cooling and heating rate of 10 ◦C min<sup>1</sup> under nitrogen atmosphere.

Pictures of polymer-coated AA2024 samples were obtained with an optical microscope (Wild Photomakroskop M400, Switzerland) equipped with a digital camera, combined with a portable UV lamp emitting at 395 nm (4 W, Lighting EVER).

Scanning Electron Microscopy (SEM)was performed with a LEO Gemini 1530.

#### *2.3. Synthesis of 4-Octyloxybenzyl Chloride*

4-Octyloxybenzyl chloride was synthesized according to the literature [15,16] as follows: 4-hydroxybenzaldehyde (0.16 mol, 20 g) was dissolved together with potassium hydroxide (0.2 mol, 11.12 g) in acetonitrile (270 mL) in a three-necked flask. The reaction was stirred at room temperature and heated to reflux at 90 ◦C. 1-Bromooctane (0.15 mol, 27 mL) was then added during 1 h and the reaction mixture was vigorously stirred overnight. Thereafter, the mixture was cooled to room temperature and quenched with 250 mL of water, transferred to a separatory funnel and extracted with hexane (200 mL). The organic layer was washed twice with 30 mL of NaOH solution (10% in mass) and finally with water (three times 30 mL). The organic phase was dried over anhydrous sodium

sulfate and concentrated in a rotary vapor at 300 mbar (40 ◦C) giving 4-octyloxy benzaldehyde as a pale-yellow oil (25.1 g, 0.11 mol, yield: 70%). 1H NMR (CDCl3): δ = 0.88 (t, 3H, CH3), 1.19–1.55 (m, 10 H, 5 CH2), 1.74–1.84 (m, 2H, CH2), 4.01 (t, 2H, CH2O), 6.65–6.98 (d, 2H, Ar), 7.79–7.82 (d, 2H, Ar), 9.85 (s, 1H, COH) ppm.

Afterwards, 4-octyloxy benzaldehyde (25.1 g, 0.107 mol) was dissolved in tetrahydrofuran (THF) (20 mL) and added dropwise to a stirred suspension of NaBH4 (8.1 g, 0.2 mol) in dry THF (140 mL) at 0 ◦C. The reaction mixture was then warmed up to room temperature and stirred overnight. The reaction mixture was quenched with 50 mL of water and the organic layer was separated, washed with water (three times) and dried over anhydrous sodium sulfate [17]. The solvent was then removed by evaporation at 11 mbar and 40 ◦C, giving a white solid that was dissolved in 5 mL of THF and poured into 400 mL of water under stirring. The white precipitate was filtered (cellulose filter) and the cake was dried under vacuum (0.7 mbar) to give 4-octyloxybenzyl alcohol as a white powder (0.08 mol, 20.14 g, yield: 80%). 1H NMR (CDCl3): δ = 0.86–0.91 (t, 3H, CH3), 1.20–1.49 (m, 10 H, 5 CH2), 1.73–1.82 (m, 2H, CH2), 3.93–3.95 (t, 2H, CH2OAr), 4.61 (s, 2H, CH2OH) 6.87–6.90 (d, 2H, Ar), 7.26–7.29 (d, 2H, Ar) ppm. C15H24O2 (236.35 g mol−1): calcd. C 76.23, H 10.23, found C 76.18, H 10.12.

Subsequently, 4-octcyloxybenzyl alcohol (0.03 mol, 7.08 g) was dissolved in dichloromethane (200 mL) and thionyl chloride (0.036 mol, 4.26 g) was added to the solution dropwise at 0 ◦C. The mixture was then stirred at the same temperature during 2 h and the reaction was subsequently quenched with water (100 mL). The organic layer was extracted, washed with water (40 mL) and saturated aqueous sodium hydrogen carbonate (140 mL) and dried over anhydrous sodium sulfate. The dichloromethane was evaporated at reduced pressure (100 mbar at 40 ◦C) obtaining 4-octyloxybenzyl chloride as a yellow pale oil (0.016 mol, 4.16 g, yield: 54%). 1H NMR (CDCl3): δ = 0.89–0.93 (t, 3H, CH3), 1.31–1.49 (m, 10 H, 5 CH2), 1.75–1.84 (m, 2H, CH2), 3.94–3.98 (t, 2H, CH2OAr), 4.57 (s, 2H, CH2Cl) 6.87–6.90 (d, 2H, Ar), 7.29–7.32 (d, 2H, Ar) ppm. C15H24OCl (254.80 g): calcd. C 70.71, H 9.10, found C 70.92, H 9.22.

#### *2.4. Synthesis of the Octyloxy-Containing PPM Derivatives*

Copolymerization of benzyl chloride in presence of 4-octyloxybenzyl chloride was performed with two ratios of the comonomers, 5.3% mol/mol and 11.2% mol/mol of 4-octyloxybenzyl chloride, respectively. First, propylene oxide stabilizer (0.25% w/v) was removed exposing benzyl chloride to vacuum (0.7 mbar) overnight. The removal of propylene oxide stabilizer was verified by 1H NMR spectroscopy by disappearance of signals of propylene oxide at 2.96, 2.73, 2.41, and 1.31 ppm. In the case of 5.3% mol/mol, 1.58 g (1.6 mL, 6.2 mmol) of 4-octyloxybenzyl chloride was added under nitrogen atmosphere to 14.9 g (13.7 mL, 118 mmol) of destabilized benzyl chloride in a 100 mL three-neck flask equipped with a mechanical stirrer. In the case of 11.2% mol/mol, 1.49 g (1.50 mL 5.84 mmol) of 4-octyloxybenzyl chloride were added under nitrogen atmosphere to 6.1 mL (52 mmol) of destabilized benzyl chloride in a 100 mL three-neck flask equipped with a mechanical stirrer. Thereafter, the respective mixtures were heated up to 60 ◦C and 0.05 mL (0.46 mmol) of SnCl4 was added. The copolymerizations were carried out under nitrogen flow of 0.4–0.5 mL min−<sup>1</sup> to allow the produced HCl to leave the reaction environment. After 3 h, due to the increase of viscosity, the temperature was risen to 120 ◦C for 3 h and subsequently to 180 ◦C for 17 h. During the reaction, the color shifted from deep red after the addition of SnCl4 to clear amber at the end of the reaction. Afterwards, the mixture was cooled down to room temperature and the product was solubilized in 10 mL of THF. This solution was then poured into 400 mL of methanol under vigorous stirring, and after 4 h, the obtained powder was filtered through a cellulose filter and dried under vacuum (10−<sup>2</sup> mbar) for 12 h, yielding 3.87 g and 13.26 g of product, respectively, containing 13.4% and 6.1% of the octyloxy repeat units, respectively, as obtained from 1H NMR spectra (see below). Accordingly, yields of 63% and 75%, respectively, were calculated in the case of 13.4% and 6.1% octyloxy repeat unit.

#### *2.5. Preparation of Coated AA2024*

Sheets of 12 cm in length, 3 cm in width and 4 mm in thickness of high strength aluminum alloy AA2024 (4.3%–4.5% copper, 1.3%–1.5% magnesium, 0.5%–0.6% manganese and less than 0.5% of other elements) were provided by Aviometal s.p.a (Varese, Italy) and used as substrate.

Samples of 4 cm in length were cut and subsequently polished with abrasive papers of 300, 500, 800, 1200, and 4000 grit. Immediately after polishing, the samples were cleaned by immersion in ethanol in an ultrasonic bath (Banderlin, Berlin, Germany) for 5 min. Then AA2024 samples were removed from the ethanol bath and the residual alcohol at the surface was evaporated by means of a flush of nitrogen.

A layer of benzyltriethoxysilane was applied by spin coating (3500 rpm, 30 s) on freshly cleaned AA2024 samples and subsequently heated up to 100 ◦C for 1 min, whereupon condensation of benzyltriethoxysilane to respective polysiloxanes proceeded [1]. These samples were finally coated with substituted PPM as described in the section Results and Discussion, using about 100 mg of polymer.

#### *2.6. Electrochemical Characterization of Coated AA2024*

The anticorrosion ability of the PPM derivatives as thin protective films was studied by means of electrochemistry techniques, carrying out tests on AA2024 samples coated with the two copolymers (6.1% mol/mol and 13.4% mol/mol). The protocol for the coating deposition is described in the Results and Discussion section.

Electrochemical corrosion tests were conducted in a naturally aerated near-neutral simulated marine environment prepared by dissolving 0.6 mol L−<sup>1</sup> sodium chloride (≥99.0%, Sigma-Aldrich) in MilliQ ® water. The pH value was adjusted to 6.7 ± 0.1 by adding few drops of 0.2 mol L−<sup>1</sup> sodium hydroxide solution to the stock solutions. All the experiments, if not otherwise stated, were carried out at ambient temperature (24 ± 3 ◦C, with a variation lower than 2 ◦C during each single run). In all cases, the operative temperature was below the glass transition temperature of PPM derivatives (see below).

The apparatus used for the measurements consisted of a glass cell with a hole (1 cm in diameter) in the middle of the flat bottom part which assures the contact between the coated metallic plate (working electrode, exposed area 0.78 cm2) and the working solution (0.6 M NaCl). The sealing was guaranteed by a bi-adhesive layer (a2 Soluzioni Adesive, Italy) pressed between the sample and the bottom of the cell. The electrochemical setup also included a platinum coil as counter electrode and an aqueous saturated calomel electrode as reference one (*E*SCE = 0.242 V vs. SHE). The latter was inserted into a glass double bridge (filled with the same working solution) ending with a Luggin capillary aimed to minimize the ohmic drop between working and reference electrode. No instrumental compensation of the residual ohmic drop was performed.

The electrochemical characterization included both potentiodynamic and potentiostatic methods. The former consisted of a single anodic polarization scan, sweeping the potential from OCP to 2.5 V vs. SCE, at a scan rate of 10 mV min−<sup>1</sup> (each run lasting ca. 5.5 h). A limit current density of 4 mA cm<sup>−</sup><sup>2</sup> was imposed, thereafter the scan was automatically aborted independently by the achievement of the final potential. The second characterization implies the application of a constant potential to the metallic sample and the recording of the current flow between working and counter electrode. In our experiments, an oxidizing potential of 0 V vs. SCE was applied for 24 h. Potentiodynamic and potentiostatic curves were recorded after an initial delay time of 600 s for assuring the equilibration of the system at OCP.

Some potentiodynamic curves were recorded also at a fixed temperature of 35 ◦C, just above the glass transition temperature of the modified PPM (see Results and Discussion). For these experiments, a suitable cell surrounded by a jacket filled by a flux of water controlled by a thermostat (Haake CH Fisons coupled to a Haake F3 Fision) was adopted.

#### **3. Results and Discussion**

#### *3.1. Synthesis and Structural Characterization*

PPM-based copolymers containing *n*-octyloxy side chains were obtained by mixing the comonomers benzyl chloride and 4-octyloxybenzyl chloride at two ratios, 5.3% mol/mol and 11.2% mol/mol, respectively, and polymerization was subsequently carried out under the conditions reported in the Materials and Methods section. For both ratios the number-average molar masses of the resulting copolymers (Mn) amounted to about 2500 g mol−<sup>1</sup> and the weight-average molar masses (Mw) to about 5400 g mol−1, resulting in polydispersity indices (PDI= Mw/Mn) of about 2.2–2.3 (Table 1). Notably, the fluorescence observed for PPM itself also emerged in the copolymers (Figure 2).

**Table 1.** Molar masses of copolymers prepared from benzyl chloride and 4-octylbenzyl chloride (6.1% mol/mol and 13.4% mol/mol).


**Figure 2.** Photographs taken under UV-light illumination of PPM and its derivatives solutions in chloroform (≈0.5% m/m, 2 cm diameter of the vials): PPM (left), copolymer containing 13.4% (center) and 6.1% (right).

The presence of octyloxy groups in the copolymers was assessed with 1H NMR spectroscopy. With regard to the reported values of PPM [18,19], 1H NMR spectra of the copolymers showed the aromatic resonances (HAr) in the typical region of 6.8–7.2 ppm. The position of the bridging methylene signals (Hm) at 3.9 ppm differed from that of the corresponding signals of the CH2Cl group of the comonomers at 4.5 ppm (Figure 3). Notably, the presence of the signals corresponding to the alkoxy chains (HAlk) in the region of 0.5–1.5 ppm confirms the inclusion of octyloxy side chains into the polymers. Furthermore, the ratio between integrated peak areas of alkyl signals (HAlk) and those of the methylene bridges (Hm) were used to calculate the effective molar composition of the copolymers. The obtained values

showed somewhat higher molar ratios of constitutional repeat units of the octyloxy groups (13.4% and 6.1%) than the comonomer ratio employed for the reaction (11.2% and 5.3% respectively). Obviously, polymers with a higher fraction of unsubstituted phenylene methylene units were removed during sample workup with somewhat higher preference.

**Figure 3.** 1H NMR spectra. From the bottom to top: 4-octyloxybenzyl chloride (blue line), copolymer with 6.1% mol/mol (red line), and 13.4% mol/mol 4-octyloxybenzyl-containing constitutional repeat units (green line).
