Effect of Curing Temperature and Architectural (Monolayer and Bilayer) of Hybrid Films Modified with Polyethylene Glycol for the Corrosion Protection on Tinplate

Hybrid films obtained by a sol-gel process are based on two important reactions: hydrolysis and condensation. The condensation reaction is influenced by the curing temperature, since it induces the development of the intrinsic properties of the gel. The aim of this work is to coat tinplate, a substrate widely used in the packaging industry, with monolayered and bilayered hybrid films modified with polyethylene glycol and obtained through dip-coating. The results showed that the bilayered hybrid film obtained at 60°C had a higher layer thickness, and the best performance in the electrochemical assays, as well as the most hydrophobic character, in relation to the other samples. For the monolayered systems, the 90°C-cured system showed a lower layer thickness; however, this system showed a more compact, uniform and less porous layer, and presented better electrochemical impedance results, in comparison with the 60°C-cured samples.


Introduction
Tinplate is a thin low-carbon mild steel sheet, cold reduced, coated on both sides with pure electrolytic tin, linking the strength and formability of the steel with the corrosion resistance and weldability of the tin 1,2 .Tinplate is largely used in food packaging, being ideally suited for this purpose since it is non-toxic, light in weight, strong, and corrosion-resistant 3 .
The resulting coating has the following stratified structure from bottom to top: an FeSn 2 alloy developed from flow-brightening which protects the steel against galvanic corrosion by oxidant species; and a free tin layer that determines the durability when acting as a sacrificial anode 4 .
Regarding the tinplate quality, the presence of tin oxides on the surface is remarkable, since they affect the metallic sheet appearance, weldability and capacity of being coated with organic films 5 .Currently, packaging is submitted to a chromate-based surface treatment, for it offers excellent corrosion resistance 6,7 .Nevertheless, nontoxic pre-treatments have been developed to substitute the chromatization process [8][9][10] .Among these alternatives, hybrid organic-inorganic films are found 11 .
Hybrid films are constituted by the combination of two compounds, organic and inorganic, which, normally, have complementary properties and give raise to materials with very different properties from those of their precursors 12 .Those materials are extremely homogeneous, since their mixture occurs at a molecular level 6 .Organic-inorganic hybrid films obtained through a sol-gel process have the advantage of being processed at low temperatures.For this reason there is neither solvent volatilization nor chemical species degradation 13,14 .
On the other hand, hybrid films obtained through the sol-gel method are not completely efficient as the films are not thick enough and tend to be cracked.In this sense, plasticizers, like polyethylene glycol 15 or polymethyl methacrylate 16 , are used, with the purpose of increasing the layer thickness and the sol viscosity, or of enhancing the molecular flexibility and mobility to obtain uncracked films which can be submitted to mechanical deformation without failure.Additionally, it is recommended to obtain coatings with a high thickness layer in order to promote efficient barrier properties between the substrate and the environment.The increase in the thickness can be achieved in two ways: first, it is possible to increase the number of layers, but a limit must be observed in order to avoid delamination problems 17 ; secondly, the sol viscosity can be increased, via temperature modification, however, in this case, the kinetics of the hydrolysis and condensation reactions varies and could induce modifications in the intrinsic properties of the gel 18 .
Pan et al. 19 evaluated the performance of silane films in different curing temperatures and concluded that a higher curing temperature accelerates the hydrolysis and condensation, leading to denser films with a better water barrier performance.
The aim of this work is to coat tinplate with a hybrid film obtained through a sol process, constituted of the following alkoxide precursors: 3-(trimethoxysilylpropyl) methacrylate (TMSPMA) and tetraethoxysilane (TEOS), with addition of 0.01 M cerium nitrate and the plasticizer polyethylene glycol at a concentration of 20 g.L -1 .The films were obtained by dip-coating process, applying monolayer and bilayer.The films were cured at different temperatures (60°C and 90°C) for 20 minutes.

Surface preparation
Tinplates with dimensions of 20×40 mm were obtained from an industrial sheet, rinsed with acetone and dried.Then, the plates were degreased by a 10 min immersion in neutral detergent at 70°C, washed and dried.

Elaboration of hybrid films
The hydrolysis reactions were conducted with hybrid film silane precursors (TMSPMA) 3-(trimethoxysilylpropyl) methacrylate (C 10 H 20 SiO 5 ) and (TEOS) tetraethoxysilane (C 8 H 20 SiO 4 ) with the addition of cerium nitrate in a concentration of 0.01 M. Ethanol and water were used as solvents.PEG 1500 was added in the formulation of the sol, at a concentration of 20 g.L -1 .In addition, a sample without plasticizer was studied.The hydrolysis time was 24 hours.The application of hydrolyzed hybrid solutions was performed by dip-coating process, with a removal rate of 10 cm.min -1 and a permanence of 5 minutes in the solution.The films were obtained by dip-coating process, applying one or two layers to obtain monolayered and bilayered systems, and cured at two different temperatures, 60°C and 90°C, for 20 minutes.A description of the samples used is presented on Table 1, and the protocol for the sol preparation as well as for the coatings is detailed on the flow chart (Figure 1).

Experimental techniques
The morphological characterization was carried out by using Scanning Electron Microscopy (SEM), performed in a JEOL 6060 device, with an acceleration voltage of 20 kV.The samples were observed in top view and in cross-section to determine the layer thickness.
The wettability of the hybrid films was determined by the contact angle measurement from the sessile drop method using equipment developed by the Research Laboratory of Corrosion (LAPEC) of UFRGS.The contact angle was determined by image analysis software.
The infrared spectroscopy (FTIR) measurements were performed on a spectrophotometer Spectrum ASCII 100.The measurements were performed with the beam in the mid-infrared (4000-500 cm -1 ).The spectra were obtained from the films without a substrate (free-standing films).For this characterization, two different curing temperatures (60°C and 90°C) for 20 minutes were used.
The corrosion performance of the coatings was evaluated by open circuit potential (E cor ) monitoring and Electrochemical Impedance Spectroscopy (EIS) measurements in a 0.05 M NaCl solution.This concentration is sufficiently high to activate corrosion in a relatively short exposure time, but low enough to enable the effects of various acids to be determined.Kozhukharov et al. 20 also used 0.05 M NaCl to ensure a concentration low enough to allow for the observation of inhibition effects.A three-electrode cell was used to perform the analysis, with a platinum wire as the counter electrode and SCE as the reference electrode.The working electrode area was 0.626 cm 2 .For the EIS measurements, the systems were previously monitored for up to 96 hours.The amplitude of the EIS perturbation signal was 10 mV, and the frequency studied ranged from 100 kHz to 10 mHz using an frequency response analyzer model NOVA system and a potentiostat AUTOLAB PGSTAT 30.

Morphological characterization
Figure 2 shows the SEM micrographs for the hybrid films before the electrochemical tests.The samples F1P60D (Figure 2b) and F1P90M (Figure 2c) show the formation of a homogeneous and uncracked film.Furthermore, the curing temperature showed to be adequate for the monolayered system.
It was observed that the 60°C-cured monolayered hybrid film was not cracked (Figure 2a).This is associated with a lower curing temperature which did not cause cracking; however, dark spots on the film can be observed, formed by non-evaporated retained solvent during the curing process.This is related to the addition of the plasticizer PEG 1500, due to its high molecular weight and increased viscosity, which limits the evaporation of volatile organic compounds during curing after application of the films 16 .
In the system F1P90D (Figure 2d) formation of cracks can be observed on the film.These results are in agreement with some researchers 21,22 who report that a further increase of thickness, or an addition of a higher number of layers, would promote an improvement of the system against corrosion.However, the morphological analysis shows microcracks for the films exposed to high curing temperature, due to the fragile state of thicker layers, and a consequent decrease in corrosion resistance.Additionally, it is generally accepted that the cracked microstructure of the bilayered hybrid films is due to the drying stage 23 , because of the application of compressive stress during the crosslinking of the silane layer in the curing step 24 .
The layer thicknesses of the films determined by the analisys of cross section images obtained through SEM (Figure 3), as well as the results are shown in Table 2.The F1P60D and F1P90D bilayered samples had higher layer thickness values than monolayered systems, as it was expected.This result is associated to the fact that the first layer anchors the second one due to the presence of the PEG plasticizer in the sol formulation, since the organic phase allows better anchoring after the coating application 25 .
The results reported on Table 2 denote that the systems cured at 60°C had higher layer thickness values than the systems cured at 90°C for the same number of applied layers.This is in agreement with the results reported by  other authors 26,27 whose works show that the curing process contributes to the good formation and performance of the silane film, as well as to the reduction of the layer thickness.Besides, sample F1P60D (Figure 3b) had a denser, thicker and more compact film with perfect adhesion of the two layers, than the F1P90D system (Figure 3d).This is associated to the fact that the excess of heat in the silane curing process causes an increase in the crosslinking reactions, reducing their reactivity, which is due to the silane-siloxane conversion.Consequently, the film gets denser and the reaction and/or adhesion with subsequent layers gets harder 28 .
Despite the fact that the F1P90M sample has shown a lower thickness value than the F1P60M sample, the former showed a more compact, uniform and probably denser layer.These results are consome with other researchers 29,30 , in regards to the fact that silane layers improves the barrier properties by the dense reticulation, which forms a less porous layer, thus enhancing the corrosion protection.

Contact angle
From Table 3 it is possible to observe the contact angle measurements obtained for the studied systems and for the tinplate substrates (FL) The F1P60D sample showed the highest contact angle value for the studied systems.This result proves that the higher the thickness of the layer is, the more difficult is the water penetration in denser layers which are completely polymerized, as it has been previously observed in relation to to the perfect adhesion of the two layers.Thus, the pore volume being smaller, water cannot permeate the structure and the formation of more hydrophobic films occurs 15 .Moreover, the 90°C-cured bilayered system did not show a high contact angle value, perhaps due to the fact that the system was cracked, as observed in the SEM micrographs (Figure 2d) and so water could have penetrated in the cracks.
Comparing the monolayered systems, the 90°C-cured system showed a higher contact angle value than the F1P60M sample.This can be associated with the fact that a higher curing temperature not only enhances the barrier properties, due to the formation of a less porous layer given the dense reticulation, but also obstructs the passage of water, and makes the film more hydrophobic.
For tinplate substrates (FL), the observed hydrophobicity is attributed to the presence of tin oxides (SnOx), such as  SnO, SnO 2, and due to the fact that their hydrated forms provide complete or partial surface coverage.

Fourier Transform Infrared Spectroscopy Analysis (FTIR)
The FTIR spectra of the hybrid films prepared with different temperatures (60°C and 90°C) are presented on Figure 4 and Table 4.As seen in Figure 4, the strong bands, between 1000 and 1200 cm -1 , are attributed to Si-O-Si, the structural backbone of the hybrid material.The formation of the Si-O-Si layer protects the tinplate substrate.The bands at intermediate frequencies, between 900 and 960 cm -1 , arise from the SiOCH 2 CH 3 group due to the incomplete hydrolysis of the TEOS.Peaks at approximately 2900 cm -1 were identified and associated with symmetrical and asymmetrical CH stretching (CH 2 and CH 3 ) in the aliphatic chain of the organosilane (TMSPMA).These peaks are present in the totality of the prepared hybrids films, suggesting that the temperature does not influence the formation of the hybrid film, indicating full hydrolysis and a subsequent crosslinking of these samples.The broad absorption, between 3200 cm -1 and 3700 cm -1 , is characteristic of the axial deformation of OH, which could arise from silanol groups (Si-OH) that were not condensed during the synthesis.In this case, the intensity of the peaks on the two samples was very similar, showing that a temperature of 60°C was sufficient for crosslinking reactions.The bands close to 1728 and 1622 cm -1 are associated with the stretching vibrations of C=O and C=C groups, respectively.

Electrochemical characterization
Open circuit potential (E OC ) monitoring was conducted using an NaCl 0.05 M solution in order to verify the potential variation with time.The curves are shown in Figure 5.In comparison to the uncoated sample (FL), all hybrid films shifted their potential towards the positive direction, which is due to the formation of a barrier between the substrate and the electrolyte.
As shown on Figure 5, the potential of the F1P60D hybrid film shifted to the positive direction compared to uncoated tinplate, or these samples presented a value of corrosion potential more noble than the substrate, indicating an improvement in corrosion resistance 31 .This result is related to the fact that the first layer has enabled a better adhesion of the second layer due to the presence of the PEG plasticizer in the sol formulation, which allows better anchoring of the coating to be applied later, and the temperature of 60°C, which is low enough to enable a complete crosslinking, as shown in the FTIR results (Figure 4), allowing the adhesion of subsequent layers.
On the other hand, sample F1P90D did not show such nice performance.Despite being formed by two layers, these did not interact to one another, which resulted in cracking when the second layer was being formed (Figure 2d).Table 4. FTIR peaks assignment as function of peak position 28 .

Peak position (cm -1 )
Peak assignment Consequently, cracks became preferential pathways for the electrolyte, thus reducing the film corrosion resistance.Furthermore, it is remarkable that the F1P60M sample presented a potential close to that of uncoated tinplate, matching the behavior of these hydrophilic coatings with their contact angle value (Table 3).Also, formed by nonevaporated retained solvent during the curing process were present in the formation of these hybrid films, as shown by SEM micrographs (Figure 2a), that constitute a preferential pathway to electrolyte permeation.
Figure 6 presents the Nyquist diagrams, obtained by electrochemical impedance assay for the F1P60M, F1P60D, F1P90M and F1P90D films, as well as for the uncoated tinplate, performed in 24 and 96 hours in a solution of 0.05M NaCl.
From the mentioned diagrams (Figure 6), it is possible to observe that although the curves are similar in shape, they differ in magnitude.Therefore, the same fundamental phenomenon could have occurred in all these coatings but over a different effective area in each case.The diameter of the arc can be regarded as the polarization resistance (R p ) of the sample 32 .
All hybrid films coated samples showed a larger semicircle related to uncoated tinplate, demonstrating the protective action of these hybrids films.The performance of the samples is associated with the fact that a small increase in the branching of the ethylene oxide (20 g.L -1 ) radical was sufficient to decrease the rate of the condensation reaction in the triorganosilane groups.The condensation of the tetrafunctional alkoxides was also retarded in the presence of this monomer (PEG) due to the increased steric hindrance in the transition state.Therefore, a more flexible film was obtained with better adhesion to the substrate and hence an improvement in the anticorrosive properties of this coating 33 .
The high resistance of the F1P90M sample, after 24 hours of immersion (Figure 6a), is most likely due to the fact that this system had a more compact, uniform and less porous layer, causing an improvement in the corrosion protection.However, the resistance, the phase angle, and the impedance modulus of this system lowered significantly from 24 to 96 hours of immersion (Figure 6b and Figure 7b).This result reveals the fragility of these coatings, allowing the permeation of the electrolyte through the film, perhaps because of the smaller layer thickness (Figure 3 and Table 2)  between the studied films, thus not resisting to long periods of immersion.
As shown on the F1P60D sample, a bilayered coating helped to improve the protective performance of the hybrid film, as suggested by the highest resistance and impedance modulus values after 96 hour of testing (Figure 6b and Figure 7b).There was an excellent adhesion between the layers, producing a denser, more compact and uncracked film, as per the SEM micrographs (Figure 2b), with a high layer thickness value (Figure 3b).These results are in agreement with studies conducted by Van Ooij et al. 17 , who demonstrated that a monolayered silane film is not enough to produce a good corrosion protection to a metallic substrate.Hence, a bilayered film coating on the metallic substrate is suggested, involving a two-step treatment.
The bilayered sample cured 90°C had the worst electrochemical impedance results after 96 hours of immersion.This may be associated to the fact that the F1P90D sample developed cracks on the film, as seen on the SEM images (Figure 2d).Therefore, the number of layers must be adjusted to the film curing temperature, for the increase of the film thickness (Figure 3 and Table 2) was not related to an improvement against corrosion.However, the chosen temperature which lead to the formation of cracks due to the embrittlement of the thickest layers, hence the reduction in corrosion resistance 34 .
The EIS was used in the present work to characterize the corrosion behavior of the developed coatings.A more detailed interpretation of the EIS measurements was performed by fitting the experimental plots using equivalent electrical circuit models which simulated the electrochemical behavior of the coatings (Figure 8).These models were based on the combination of resistances, capacitances, and other elements that should have a physical meaning, in relation to the electrochemical response of the system.In this work two equivalent electrical circuit models were used, and they were shown in Figure 8.
Tables 5-8 demonstrates the electrical parameter values obtained by fitting the equivalent electrical circuit from the experimental EIS data, obtained for the F1P60M,   In several circuits the capacitance was substituted by a CPE in order to take into account the non-ideality of the systems.In these circuits (Figure 8a), Re represents the electrolyte resistance, R HFM and CPE HFM represent, respectively, the resistance and a constant phase element associated to the electrolyte permeation resistance through the monolayered hybrid film.In the same circuit, R HFB and CPE HFB represent the resistance and a constant phase element indicating the effect of electrolyte permeation resistance through the bilayered hybrid film.
The same equivalent circuit model (Figure 8a) was used for the electrochemical behavior simulation in all of the studied immersion times (1, 24, 48, 72 and 96 hours) for the F1P60D sample, and just for the immersion time of 1 hour for the F1P90D sample.This behavior has been observed by the other authors 35,36 , and this model confirms that a monolayered film is not enough to produce a good corrosion protection to the substrate.Indeed, it is recommended that metallic substrates must be coated with bilayered films in order to enhance the corrosion protection, avoiding cracks and delamination problems.The F1P60D hybrid film resistance (R HFB ) could be estimated by the Nyquist diagram at the end of the experiment (Figure 6b), taking into account the highest semicircle diameter at high frequency ranges.The RT (monolayered system resistance + bilayered system resistance) values were found to be around 251 kΩ cm 2 for the 96 hours of immersion (Table 6).
In the circuits shown on Figure 8b, Re represents the electrolyte resistance, R HFM and CPEHFM represent, respectively, the resistance and a constant phase element associated to the effect of electrolyte permeation resistance through the monolayered hybrid film.The same equivalent circuit model (Figure 8b) was used for the electrochemical behavior simulation at all immersion times studied (1, 24, 48, 72 and 96 hours) for the F1P60M and F1P90M samples.Electrochemical behavior simulation for 24 hours of immersion was done in the F1P90D sample.This behavior was observed by other authors [37][38][39] indicating the hybrid films which retards the corrosion processes on the tinplate surface.
Figure 9 shows the evolution of the coating properties -i.e., the total resistance, as a function of immersion time.Generally, the high-frequency resistance values indicated a decrease during the first hours of immersion, due to the development of conductive pathways inside the film 40 ; occurring more significantly in the F1P60D, F1P90M and F1P90D hybrids films.It can be observed that the F1P60D sample presented the highest resistance values after 48 hours of immersion (Figure 9) in comparison to other samples (F1P60M, F1P90M and F1P90D).This can be associated to the high layer thickness value (Figure 3 and Table 2), and consequently, it can be said that the curing temperature of 60°C, was enough to obtain a dense, hydrophobic film (as shown in the contact angle results -Table 3 -and in the FTIR results -Figure 4 and Table 4).
The evolution of the coating resistance is a major characteristic of the barrier properties of a protective layer 41 .The F1P90M and F1P90D samples presented a sharp reduction of the resistance values, which may be associated to the presence of cracks and irregularities in the formation of F1P90D shown by the SEM micrographs (Figure 2).The case of the F1P90M sample is due to the fragility of these coatings, allowing the permeation of the electrolyte through the film.This can be associated to the lower layer thickness value (Figure 3 and Table 2) and, consequently, not resisting to long periods of immersion.Figure 10 shows the SEM micrographs for the samples after 96 hours of electrochemical impedance assays in a 0.05 M NaCl solution.It can be observed that the F1P60D sample (Figure 10b) had less corrosion products, as it was expected regarding the good electrochemical performance showed by this system.

Conclusion
The results showed that all the studied hybrid films: F1P60M, F1P60D, F1P90M and F1P90D, presented an increase in the corrosion resistance when compared to the uncoated tinplate.
For the monolayered samples, the 90°C-cured system showed a lower layer thickness, however, not only the layer was more compact, more uniform and less porous, but it also had a better electrochemical performance at the end of the electrochemical impedance test than the 60°C-cured one.
As for the bilayered samples, the 60°C-cured system had a better electrochemical performance in comparison to the 90°C-cured bilayer system.This behavior is associated to the high layer thickness obtained for this film, as well as to the formation of a homogeneous, uncracked, and continuous film, and a higher contact angle value.The temperature of 60°C was enough to obtain a dense, hydrophobic film and a higher corrosion resistance at the end of the electrochemical impedance test, in comparison with all of the other samples.
The 90°C-cured bilayer system showed a worse corrosion resistance when compared to the 90°C-cured monolayered system, to the 60°C-cured bilayer and also to the other monolayered samples.This is associated to the presence of cracks and irregularities in the formation of the hybrid film.

Figure 4 .
Figure 4. FTIR spectra of the hybrid films formulated with the different temperatures (F1P60 and F1P90).

Figure 5 .
Figure 5. Open circuit potential (OCP) monitoring for all the hybrid films and the uncoated tinplate.

Figure 6 .
Figure 6.Nyquist plots obtained for hybrid film coated and uncoated tinplate in 0.05 M NaCl solution: (a) 24 and (b) 96 hours of immersion.

Figure 7 .
Figure 7. Bode plots obtained for hybrid film coated and uncoated tinplate in 0.05 M NaCl solution.

Figure 9 .
Figure 9. Evolution of the hybrid films resistance for F1P60M, F1P60D, F1P90M and F1P90D in a 0.05M NaCl solution with the immersion time.

Table 1 .
Description of the samples.
Figure 1.Experimental parameters of synthesis of the sols and processing to deposit coatings.

Table 2 .
Layer thicknesses for the hybrid films.

Table 3 .
Contact angle values obtained by the sessile drop method.

Table 5 .
Electrical elements fitted values for F1P60M film up to 96 h of immersion in 0.05 M NaCl solution.The error % associated with each parameter value is given in parenthesis.

Table 6 .
Electrical elements fitted values for F1P60D film up to 96 h of immersion in 0.05 M NaCl solution.The error % associated with each parameter value is given in parenthesis.

Table 7 .
Electrical elements fitted values for F1P90M film up to 96 h of immersion in 0.05 M NaCl solution.The error % associated with each parameter value is given in parenthesis.

Table 8 .
Electrical elements fitted values for F1P90D film up to 96 h of immersion in 0.05 M NaCl solution.The error % associated with each parameter value is given in parenthesis.