In order to study the optimal electropolymerization concentration of AF17N, the atomic concentrations of PAF17 nanofilm were investigated by XPS. Figure 1 shows the XPS data of PAF17 nanofilm polymerized under different AF17N concentrations. It can be seen that the atomic concentrations of C1s, S2p, N1s and F1s on aluminum surface increase with increasing monomer concentration, while the atomic concentrations of Al2p and O1s decrease. The migration rate of ions to the electrode will slow down when the monomer concentration is low. With the monomer concentration increasing, the anions near the aluminum electrode become sufficient for electropolymerization to form the thicker PAF17 nanofilm. However, the micelles will form as the monomer concentration is over 5 mM due to higher molecular weight of the monomer, leading to a decrease of the reactional anions. The result preliminarily suggests that the optimal concentration of AF17N is 5 mM.

FT-IR spectra of PAF17 nanofilm obtained using different concentrations are shown in Figure 2. It is noted that when the AF17N concentration is up to 5 mM, the >C=N- and C-F absorption peaks reach the maximum intensities, which also indicates that the optimal concentration of AF17N is 5 mM. Based on the above XPS analysis, the electropolymerization rate of AF17N was slower while the oxidation reaction of aluminum was faster due to the lower monomer concentration. Therefore, the PAF17 nanofilm on an aluminum surface was so thin that the detected >C=N- and C-F absorption in the FT-IR spectra was very weak.

Figure 3 shows FT-IR spectra of PAF17 nanofilm obtained by using different electropolymerization times. With electropolymerization times ranging from 2 to 6 min, the intensities of >C=N- and C-F absorption peaks become stronger. When the electropolymerization time is more than 6 min, the >C=N- and C-F absorption peak intensities vary slightly. It can be speculated that the PAF17 nanofilm obtained by electropolymerization for 6 min is more compact and insulated, which could depress the further formation of PAF17 nanofilm and the depolymerization reaction could happen.

The effect of electropolymerization time on the three-dimensional morphology of PAF17 nanofilm is shown in Figure 4. The roughness (Ra) of PAF17 nanofilm decreases when the electropolymerization time is prolonged. However, when the electropolymerization time is over 6 min, the value of Ra shows little increase, which indicates that the PAF17 nanofilm obtained at 6 min is the most uniform. With increasing electropolymerization time, the peeling off of PAF17 nanofilm happens and the nanofilm becomes loose due to the depolymerization reaction. Based on the above analysis, we can conclude that the optimal electropolymerization time is 6 min.

The effect of electropolymerization temperature on the thickness and weight of PAF17 nanofilm is shown in Figure 5. When the electropolymerization temperature is 5 °C, the value of thickness and weight is low. It was presumed that the anion of triazinedithiol monosodium in the solution could diffuse to the aluminum surface very slowly, which influenced the rate of the electropolymerization reaction. According to law of thermodynamics, we know that the diffusion rate of a monomer anion depends on temperature. Therefore, the thickness and weight of PAF17 film becomes larger when the electropolymerization temperature is increased from 5 to 15 °C. However, an increase in temperature from 15 to 35 °C, the thickness and weight of PAF17 film appears to decrease. The oxidative reaction and stripping of aluminum electrode could be accelerated, which led to the formation of an alumina layer and pinhole. It was assumed that the alumina layer was insulated, hindering the polymerization of triazinedithiol monosodium.

FT-IR spectra measurements were performed by reflection absorption to study the chemical structure of PAF17 nanofilm. Figure 2 and Figure 3 show FT-IR spectra of polymeric film obtained by electropolymerization of AF17N under different conditions. The presence of the triazine ring is confirmed by absorption peaks at 1481 cm⁻¹, 1536 cm⁻¹ and 1560 cm⁻¹ due to >C=N- bonds [9,15,16]. Allyl perfluorodecyl amino groups are confirmed by absorption peaks at 1226 cm⁻¹, 1250 cm⁻¹, 1331 cm⁻¹, 1155 cm⁻¹ and 1143 cm⁻¹ due to C-F stretching vibrations of CF₃- groups and >CF₂- groups [12]. However, absorption bands centered at 1645 cm⁻¹ and 3090 cm⁻¹, 2975 cm⁻¹ which are assigned to C=C and =CH₂ of AF17N monomer, does not appear. It was supposed that allyl groups participated in the electropolymerization reaction [17]. At the same time, the peak of alumina at 956 cm⁻¹ was also observed. It was suggested that the electropolymerization of AF17N and oxidation of aluminum occurred simultaneously.

To confirm the chemical structure of PAF17 nanofilm prepared on an aluminum plate, the S2p fitted curve of XPS spectra was investigated (Figure 6). The S2p spectra consists of peaks assigned to S-Al groups at 160.3 eV, C-S-C groups at 163.5 eV, C-SS-C groups at 163.6 eV and SO₄²⁻ groups at 167.7 eV. Peak based on S-Al groups indicates the reaction between SH groups and the aluminum substrate during electropolymerization. Peaks based on C-S-C groups suggest the reaction between SH and allylic groups in AF17N during electropolymerization. Peak for the C-SS-C group reveals the electrochemical reaction of thiols. From the S2p XPS data, the PAF17 nanofilm prepared on aluminum plate are confirmed to consist of poly(6-(Nallyl-1,1,2,2-tetrahydroperfluorodecyl)amino-1,3,5-triazine-2,4-disufide). It is assumed that AF17N monomer is dissolved as dithiolate anions in the electrolyte aqueous solution and the dithiolate anions transfer two electrons to the anode (aluminum plate) to change to bisthiyl radicals .Then the bisthiyl radicals cause coupling with each other to yield PAF17 film on the aluminum surface. Besides, lots of bubbles could be observed on the aluminum surface and irritant gas could be smelt simultaneously, which suggested that NaNO₂ also participated in the polymerization reaction. Based on the above experimental facts, the reaction mechanism is proposed as follows (Equations 1–6):

Test specimens (50 × 30 × 0.1 mm) of pure aluminum (purity no less than 99.9995%) were prepared by cutting a large plate into pieces. All test plates were degreased by ultrasonic washing in acetone for 15 min, then were dried in nitrogen air. 6-(N-allyl-1,1,2,2-tetrahydroperfluorodecyl) amino-1,3,5-triazine-2,4-dithiol monosodium (AF17N) was prepared by reaction between 6-(N-allyl-1,1,2,2-tetrahydroperfluorodecyl)-amine-1,3,5-triazine-2,4-dichloride and NaSH, according to the method described previously [18]. The structure of AF17N is shown in Figure 7. All of the chemicals were employed as analytical reagents (AR) without further purification. Distilled water was used as solvent, and sodium nitrite (NaNO₂) was applied as supporting electrolyte. The concentration of supporting electrolyte was kept at 0.15 M.

The electropolymerization of AF17N was performed by using an electrochemical measurement apparatus (Hokuto Denkou Co. Ltd., HD-3000). The electrolytic cell was equipped with working electrode (aluminum surfaces), Pt counter electrode and reference electrode (saturated calomel electrode, SCE), then was filled with electrolytic solution containing AF17N. Electropolymerization of the AF17N monomer was performed galvanostatically with a current density of 0.30 mA/cm² [13]. Concentrations of AF17N monomer at 1, 3, 5 mM were investigated. Electropolymerization time was 2, 4, 6, 8 min according to our previous study [12].

An electronic balance with measurement accuracy of 0.01 mg (CP225D, Sartorius) was used to investigate the polymeric film weight. Film thickness was measured by using a JASCO M-150i ellipsometer (Jasco Tokyo Japan). FT-IR spectra were carried out using JASCO IR-5500 (Jasco Tokyo Japan) by high-performance reflection absorption spectroscopy (RAS). A reflection attachment was used at an incident angle of 80° together with a wire grid polarizer. X-ray photoelectron spectroscopy (XPS) was performed to investigate the elemental composition of aluminum surface. Spectra were obtained by using a ULVAC PHI-5600 spectrometer equipped with monochrome Al Kα radiation (1,486.6 eV). The pressure in the preparation chamber was less than 10⁻⁷ Torr and less than 4 × 10⁻¹⁰ Torr in the analysis chamber. Samples were examined over an area of 800 × 2000 μm, and photoelectron spectra were recorded with a take-off angle of 45°. The treated substrate was observed by atomic force microscopy (AFM) (Nanoscope III Scanning Probe Microscope, Digital Instruments, Veeco Metrology Group) with contact mode to characterize the three-dimensional morphology.