Waterborne Eco-Sustainable Sol–Gel Coatings Based on Phytic Acid Intercalated Graphene Oxide for Corrosion Protection of Metallic Surfaces

In the past few years, corrosion protection of metal materials has become a global challenge, due to its great economic importance. For this reason, various methods have been developed to inhibit the corrosion process, such as surface treatment approaches, by employing corrosion inhibitors through the deposition of opportunely designed functional coatings, employed to preserve from corrosion damages metallic substrates. Recently, among these techniques and in order to avoid the toxic chromate-based pre-treatment coatings, silane-based coatings and films loaded with organic and inorganic corrosion inhibitors have been widely used in corrosion mitigation water-based surface treatment. In this study, the synthetic approach was devoted to create an embedded, hosted, waterborne, and eco-friendly matrix, obtained by use of the sol–gel technique, through the reaction of functional alkoxysilane cross-linking precursors, namely (3-glycidyloxypropyl)trimethoxysilane (GPTMS) and (3-aminopropyl)triethoxysilane (APTES), in the presence of graphene oxide (GO) intercalated with natural and non-toxic phytic acid (PA) molecules. As a matter of fact, all experimental results from FT-IR spectroscopy, UV–Vis analysis, and SEM confirmed that PA molecules were successfully decorated on GO. Furthermore, polarization measurements and a neutral salt spray test were used to evaluate the anticorrosive performance on aluminum and steel substrates, thus showing that the GO-PA nanofiller improved the barrier and corrosion protection properties of the developed functional silane-based coatings.

Anticorrosion coatings, in particular, polymeric hybrid or nanostructured coatings, have up, until now, emerged as the most well-liked and successful methods of protecting metal from corrosion, due to their clear benefits, including their wide adaptability, economic savings, simple manufacturing, and routine maintenance. Enhancing the barrier property is of primary importance, in order to improve the anticorrosion capability of the developed Table 1. Recent sol-gel-based functional coatings for anticorrosion treatments of different surfaces.

Functional Sol
Anticorrosion Agent Treated Surface Ref.

Synthesis of the Waterborne Multicomponent Coatings
The four functional molecules shown in Figure 1 were opportunely selected and employed as synthons for the design and development of a multicomponent anticorrosive waterborne sol-gel coating featuring their synergistic effect.
sol-gel coating, based on phytic acid intercalated graphene oxide, was significa hanced, thus placing itself on a path towards a valuable, eco-friendly, and sim proach for obtaining efficient anti-corrosive and protective coatings for differen based application fields.

Synthesis of the Waterborne Multicomponent Coatings
The four functional molecules shown in Figure 1 were opportunely selected ployed as synthons for the design and development of a multicomponent antic waterborne sol-gel coating featuring their synergistic effect. In particular, the anticorrosive performances of graphene oxide intercalate acid are well-known in the literature [79][80][81][82]. Moreover, the sol-gel process is wid ployed for the preparation of thin films, thanks to their chemical inertness, outs resistance, and mechanical properties [83][84][85][86].
In this work, two alkoxysilanes (3-aminopropyl)triethoxysilane (hereafter A A) and (3-Glycidyloxypropyl)trimethoxysilane (hereafter GPTMS or G)) were c chosen as cross-linkers, in order to achieve the preparation of two types of functio ings [87,88]. In detail, GPTMS has been widely employed for coating applications to its two different functional ends [89][90][91][92], i.e., a binding trimethoxysilyl group choring epoxy ring; the latter may undergo an epoxy-ring opening reaction, du nucleophilic substitution reaction of the carboxylic groups, either in the GO-PA na while the silane end will lead through the usual silane stages of hydrolysis and In particular, the anticorrosive performances of graphene oxide intercalated phytic acid are well-known in the literature [79][80][81][82]. Moreover, the sol-gel process is widely employed for the preparation of thin films, thanks to their chemical inertness, outstanding resistance, and mechanical properties [83][84][85][86].
In this work, two alkoxysilanes (3-aminopropyl)triethoxysilane (hereafter APTES or A) and (3-Glycidyloxypropyl)trimethoxysilane (hereafter GPTMS or G)) were carefully chosen as cross-linkers, in order to achieve the preparation of two types of functional coatings [87,88]. In detail, GPTMS has been widely employed for coating applications, thanks to its two different functional ends [89][90][91][92], i.e., a binding trimethoxysilyl group and anchoring epoxy ring; the latter may undergo an epoxy-ring opening reaction, due to the nucleophilic substitution reaction of the carboxylic groups, either in the GO-PA nanofiller, while the silane end will lead through the usual silane stages of hydrolysis and, subsequently, condensation to the expected alkoxysilane polymerization towards the final development of a polyethylene oxide network (PEO) (see Figure 2b) [93,94]. Meanwhile, nucleophilic substitution reactions among the primary amino-silane APTES lead to the preparation of a reactive sol that can covalently bond the functional carboxylic nanofillers. The obtained coating may be used as a primer layer between the metal surface and an external top paint (Figure 2a).
The hydrolysis and condensations steps, characterizing the sol-gel processes, could promote their anchoring to specific substrates. In particular, covalent connections can be formed between the-OH groups of the metallic substrates (obtained by alkaline or acid surface-pretreatment) and hydrolyzed -OCH 2 CH 3 or -OCH 3 groups of the two functional silanes. Additionally, when metal-siloxane linkages are formed in a temperature-driven hydrolysis/condensation event, an enhanced adhesion may be observed. The formation of additional silanol groups (Si-OH) into -Si-O-Si-siloxane chains could also establish a strong network layer that serves as a powerful barrier towards aggressive species [95], as well as as potential sites for the adhesion of a subsequent layer of top paint. quently, condensation to the expected alkoxysilane polymerization towards the final development of a polyethylene oxide network (PEO) (see Figure 2b) [93,94]. Meanwhile, nucleophilic substitution reactions among the primary amino-silane APTES lead to the preparation of a reactive sol that can covalently bond the functional carboxylic nanofillers. The obtained coating may be used as a primer layer between the metal surface and an external top paint (Figure 2a). The hydrolysis and condensations steps, characterizing the sol-gel processes, could promote their anchoring to specific substrates. In particular, covalent connections can be formed between the-OH groups of the metallic substrates (obtained by alkaline or acid surface-pretreatment) and hydrolyzed -OCH2CH3 or -OCH3 groups of the two functional silanes. Additionally, when metal-siloxane linkages are formed in a temperature-driven hydrolysis/condensation event, an enhanced adhesion may be observed. The formation of additional silanol groups (Si-OH) into -Si-O-Si-siloxane chains could also establish a strong network layer that serves as a powerful barrier towards aggressive species [95], as well as as potential sites for the adhesion of a subsequent layer of top paint.

Dispersibility of GO-PA
The potential responses to the PA intercalation of GO are depicted in Figure 3a. The epoxy group on the GO surface and PA establish a covalent link, leading to the formation of the functional nanofiller. The sedimentation test in Figure 3 illustrates the GO-PA and GO ability to disperse in aqueous solutions. In this investigation, GO and GO-PA were ultrasonically dispersed in water for half an hour, and the stable dispersion was subsequently kept in storage for several days without disturbance. It can be easily observed that GO-PA is better dispersed than GO in water. The GO-PA solution did not clearly stratify after 5 days, as seen in Figure 3b, demonstrating that the PA-modified GO has improved the water dispersion capacities.

Dispersibility of GO-PA
The potential responses to the PA intercalation of GO are depicted in Figure 3a. The epoxy group on the GO surface and PA establish a covalent link, leading to the formation of the functional nanofiller. The sedimentation test in Figure 3 illustrates the GO-PA and GO ability to disperse in aqueous solutions. In this investigation, GO and GO-PA were ultrasonically dispersed in water for half an hour, and the stable dispersion was subsequently kept in storage for several days without disturbance. It can be easily observed that GO-PA is better dispersed than GO in water. The GO-PA solution did not clearly stratify after 5 days, as seen in Figure 3b, demonstrating that the PA-modified GO has improved the water dispersion capacities.
The enhanced water dispersion capacity of GO-PA should be attributed to the successful modification of the GO surface with PA. Therefore, this evidence also provides optimistic conditions for GO-PA applications in the development of coatings featuring anticorrosive performances.
The analysis of the zeta potentials of GO and GO-PA in water dispersion was performed with the zeta potential test (both are 5 mg mL −1 ). To ensure that the results were accurate, two tests for each sample were performed. The findings (see Figure 4) indicate a −30.6 mV average zeta potential for GO. Meanwhile, for GO-PA, a notable negatively shifted average zeta potential, from −30.6 to −37.5 mV, was revealed. This result, together with the dispersion test, suggests better GO dispersion stability, due to PA modification, which is explained by the grafting of negatively-charged phosphate groups from PA molecules onto the GO surface [81].  The enhanced water dispersion capacity of GO-PA should be attributed to the successful modification of the GO surface with PA. Therefore, this evidence also provides optimistic conditions for GO-PA applications in the development of coatings featuring anticorrosive performances.
The analysis of the zeta potentials of GO and GO-PA in water dispersion was performed with the zeta potential test (both are 5 mg mL -1 ). To ensure that the results were accurate, two tests for each sample were performed. The findings (see Figure 4) indicate a -30.6 mV average zeta potential for GO. Meanwhile, for GO-PA, a notable negatively shifted average zeta potential, from -30.6 to -37.5 mV, was revealed. This result, together with the dispersion test, suggests better GO dispersion stability, due to PA modification, which is explained by the grafting of negatively-charged phosphate groups from PA molecules onto the GO surface [81].

FT-IR Analysis
The chemical structure of PA, GO, and GO-PA have been studied by FT-IR ( Figure  5) spectroscopy, in order to demonstrate the successful preparation of the GO-PA nanofiller. In the spectrum of PA (green line), its characteristic peaks appear at 976, 1126, 2354-1634, and 3363 cm -1 , and they were attributed to the P-O-C, P=O, P-OH, and -OH stretching vibrations, respectively [96,97].  The enhanced water dispersion capacity of GO-PA should be attributed to the successful modification of the GO surface with PA. Therefore, this evidence also provides optimistic conditions for GO-PA applications in the development of coatings featuring anticorrosive performances.
The analysis of the zeta potentials of GO and GO-PA in water dispersion was performed with the zeta potential test (both are 5 mg mL -1 ). To ensure that the results were accurate, two tests for each sample were performed. The findings (see Figure 4) indicate a -30.6 mV average zeta potential for GO. Meanwhile, for GO-PA, a notable negatively shifted average zeta potential, from -30.6 to -37.5 mV, was revealed. This result, together with the dispersion test, suggests better GO dispersion stability, due to PA modification, which is explained by the grafting of negatively-charged phosphate groups from PA molecules onto the GO surface [81].

FT-IR Analysis
The chemical structure of PA, GO, and GO-PA have been studied by FT-IR ( Figure  5) spectroscopy, in order to demonstrate the successful preparation of the GO-PA nanofiller. In the spectrum of PA (green line), its characteristic peaks appear at 976, 1126, 2354-1634, and 3363 cm -1 , and they were attributed to the P-O-C, P=O, P-OH, and -OH stretching vibrations, respectively [96,97].

FT-IR Analysis
The chemical structure of PA, GO, and GO-PA have been studied by FT-IR ( Figure 5) spectroscopy, in order to demonstrate the successful preparation of the GO-PA nanofiller. In the spectrum of PA (green line), its characteristic peaks appear at 976, 1126, 2354-1634, and 3363 cm −1 , and they were attributed to the P-O-C, P=O, P-OH, and -OH stretching vibrations, respectively [96,97].
Meanwhile, the absorption peaks appearing at 1720 and 1617 cm −1 correspond to the C=O stretching vibration and C=C from the benzene ring skeleton structure, indicating the reserved unoxidized graphitic domains [101].
The red line in Figure 5 represents the FT-IR GO-PA spectrum, where the characteristic absorption signals of both GO and PA can be observed. In particular, the peaks located at 977, 1115, and 1605-2341 cm −1 are assigned to the P-O-C, P=O, and P-OH groups of PA. At 1720 cm −1 , it is also possible to notice a stronger C=O stretching vibration band, compared to GO spectra, provided that the GO was successfully functionalized with PA. Moreover, the shift of -OH peak at 3096 cm −1 can be attributed to the formation of the association links between many PA hydroxyl groups [102,103]. It is also important to notice that, in the GO-PA spectrum, the addition of several hydroxyl groups from the PA molecules to the GO surface is responsible for the shifted signal of the OH signal. Additionally, it can be attributed to the increased electrostatic interactions and hydrogen bonding between hydroxyl units of the GO and PA molecules [104]. Therefore, these results confirm the synthesis of the GO-PA functional nanofiller. The pristine GO nanosheets (black line) showed peaks located at 1059, 1404, and 3395 cm -1 , which can be assigned to the stretching vibration of the epoxide C-O-C, hydroxyl C-O, and -OH groups, respectively [98][99][100]. Meanwhile, the absorption peaks appearing at 1720 and 1617 cm -1 correspond to the C=O stretching vibration and C=C from the benzene ring skeleton structure, indicating the reserved unoxidized graphitic domains [101].
The red line in Figure 5 represents the FT-IR GO-PA spectrum, where the characteristic absorption signals of both GO and PA can be observed. In particular, the peaks located at 977, 1115, and 1605-2341 cm -1 are assigned to the P-O-C, P=O, and P-OH groups of PA. At 1720 cm -1 , it is also possible to notice a stronger C=O stretching vibration band, compared to GO spectra, provided that the GO was successfully functionalized with PA. Moreover, the shift of -OH peak at 3096 cm -1 can be attributed to the formation of the association links between many PA hydroxyl groups [102,103]. It is also important to notice that, in the GO-PA spectrum, the addition of several hydroxyl groups from the PA molecules to the GO surface is responsible for the shifted signal of the OH signal. Additionally, it can be attributed to the increased electrostatic interactions and hydrogen bonding between hydroxyl units of the GO and PA molecules [104]. Therefore, these results confirm the synthesis of the GO-PA functional nanofiller. Figure 6 shows the comparison of the UV-Vis spectra of PA, GO, and GO-PA, in order to support the successful obtaining of the functional GO-PA nanofiller. The green line represents the PA spectra, showing its characteristic peak at 275 nm. The GO spectra (black line) is characterized by a relevant peak at 232 nm and shoulder-like peak at 280-330 nm, indicating the aromatic C=C π  π * and C=O n  π * transitions, respectively [105,106].  Figure 6 shows the comparison of the UV-Vis spectra of PA, GO, and GO-PA, in order to support the successful obtaining of the functional GO-PA nanofiller. The green line represents the PA spectra, showing its characteristic peak at 275 nm. The GO spectra (black line) is characterized by a relevant peak at 232 nm and shoulder-like peak at 280-330 nm, indicating the aromatic C=C π → π * and C=O n → π * transitions, respectively [105,106].

UV-Vis Spectral Changes
After the functionalization of GO with PA, a new characteristic peak was visible at 267 nm (red line), which is attributed to the absorption of the PA molecule and intercalation of the GO nanosheets. Therefore, the successful functionalization of GO with PA, is again confirmed. Additionally, Figure 6 shows the visual appearance of the water suspensions of GO, PA, and GO-PA (inset). It is possible to observe a change in their color from the brown of GO to the black of GO-PA, thus proving the effective grafting of PA on the GO surface [105].

Raman Spectroscopy
Raman spectroscopy was used to characterize the surface defect degree and degree of alteration of GO-PA, as well as to explore structural variation of carbon-based nanomaterials. Figure 7 displays the Raman spectra of GO and GO-PA, focusing on the D and After the functionalization of GO with PA, a new characteristic peak was visible at 267 nm (red line), which is attributed to the absorption of the PA molecule and intercalation of the GO nanosheets. Therefore, the successful functionalization of GO with PA, is again confirmed. Additionally, Figure 6 shows the visual appearance of the water suspensions of GO, PA, and GO-PA (inset). It is possible to observe a change in their color from the brown of GO to the black of GO-PA, thus proving the effective grafting of PA on the GO surface [105].

Raman Spectroscopy
Raman spectroscopy was used to characterize the surface defect degree and degree of alteration of GO-PA, as well as to explore structural variation of carbon-based nanomaterials. Figure 7 displays the Raman spectra of GO and GO-PA, focusing on the D and G peaks (typical characteristic reference peaks) that reflect the structural changes of graphene. Since defects and distortions of the sp 2 domains are common to all sp 2 carbon lattices and result from the stretching of the C-C bond, the broadened D peak, which is related, indicates the size reduction of the in-plane sp 2 domains of the graphite and prove the successful obtaining of GO [107].

X-ray Diffraction Spectroscopy (XRD)
The XRD spectra of GO and GO-PA are shown in Figure 8. The GO XRD pat showed a strong diffraction peak of the (001) plane of GO at 2θ = 13.2°, revealing graphene oxide basal planes were oriented most favorably parallel to the sample pl therefore demonstrating that the crystal structure of GO was complete and ordered [1 An interlayer distance of 6.7 Å was also observed, corresponding to the interlamellar s ing of the GO. This value of interlamellar spacing is attributable to GO in its dry stat the completely hydrated GO's layer distance can vary by up to 12° [113]. Subsequent the modification, the diffraction peak of GO-PA is approximately at 12.8°, with an in layer of 6.9 Å, which can be attributed to the intercalation of the PA molecules thro the GO nanosheets, thus indicating that PA molecules diffusion into the GO nanosh leads to the partial exfoliation of the nanosheets. It is, in fact, reported that PA grafte the GO surface may reduce the π-π stacking interactions between the sp 2 structure of nanosheets [82].
According to these findings, with the higher interlamellar spacing between the layers, it is possible to assess whether PA was successfully incorporated into In the GO spectrum, at 1338 and 1606 cm −1 , it is possible to observe the D and G band characteristics of GO (black line), referred to the sp 3 carbon atom vibration deriving from the functional groups and sp 2 carbon atoms of the in-plane vibration, respectively. A shift of the G peak from 1606 (GO) to 1602 cm −1 (GO-PA), is visible, indicating the effective functionalization of GO with PA [99].
The I D /I G ratio for GO, which measures the degree of disorder and is inversely correlated with the average size of the sp 2 clusters, was also determined. In this regard, the calculation results show that, compared with GO, the intensity ratio between the D and G peaks of GO-PA reveals a slight improvement from 0.70 (GO) to 1.32 (GO-PA), due to the increase of the surface defect density of GO, for the PA grafting [108][109][110][111].

X-ray Diffraction Spectroscopy (XRD)
The XRD spectra of GO and GO-PA are shown in Figure 8. The GO XRD pattern showed a strong diffraction peak of the (001) plane of GO at 2θ = 13.2 • , revealing that graphene oxide basal planes were oriented most favorably parallel to the sample plane, therefore demonstrating that the crystal structure of GO was complete and ordered [112]. An interlayer distance of 6.7 Å was also observed, corresponding to the interlamellar spacing of the GO. This value of interlamellar spacing is attributable to GO in its dry state, as the completely hydrated GO's layer distance can vary by up to 12 • [113]. Subsequently to the modification, the diffraction peak of GO-PA is approximately at 12.8 • , with an interlayer of 6.9 Å, which can be attributed to the intercalation of the PA molecules through the GO nanosheets, thus indicating that PA molecules diffusion into the GO nanosheets leads to the partial exfoliation of the nanosheets. It is, in fact, reported that PA grafted on the GO surface may reduce the π-π stacking interactions between the sp 2 structure of GO nanosheets [82]. therefore demonstrating that the crystal structure of GO was complete and order An interlayer distance of 6.7 Å was also observed, corresponding to the interlamel ing of the GO. This value of interlamellar spacing is attributable to GO in its dry the completely hydrated GO's layer distance can vary by up to 12° [113]. Subsequ the modification, the diffraction peak of GO-PA is approximately at 12.8°, with a layer of 6.9 Å, which can be attributed to the intercalation of the PA molecules the GO nanosheets, thus indicating that PA molecules diffusion into the GO nan leads to the partial exfoliation of the nanosheets. It is, in fact, reported that PA gr the GO surface may reduce the π-π stacking interactions between the sp 2 structu nanosheets [82].
According to these findings, with the higher interlamellar spacing between layers, it is possible to assess whether PA was successfully incorporated i nanosheets. According to these findings, with the higher interlamellar spacing between the GO layers, it is possible to assess whether PA was successfully incorporated into GO nanosheets.

Scanning Electron Microscopy (SEM)
The morphology of the GO and GO-PA was characterized by scanning electron microscopy (SEM), in order to gain insight regarding the effects of the PA intercalation on GO. SEM images for the GO and GO-PA were displayed in Figure 9a-d. Figure 9a, and the relative magnification in Figure 9b, with micrographs of the GO powder, show the characteristic sheet structure of the few-layer graphene oxide nanomaterial, distinguished by its smooth surface and wrinkled texture [114]. Additionally, the borders of the sheets, including the kinked and wrinkled portions, can be seen on such few-layer GO films, which are occasionally folded or continuous. However, Figure 9c, and the relative magnification in Figure 9d of the modified GO powder (GO-PA), demonstrate that the GO intercalated PA nanofiller is characterized by a more granular morphology than pure GO. Moreover, in the GO-PA matrix, the micrometric agglomerates of the rough folds are visible in the structure of the single sheet.
These characteristics are further proof of the successful decoration of GO surface with PA molecules [115].
few-layer GO films, which are occasionally folded or continuous. However, Figure 9c, an the relative magnification in Figure 9d of the modified GO powder (GO-PA), demonstra that the GO intercalated PA nanofiller is characterized by a more granular morpholog than pure GO. Moreover, in the GO-PA matrix, the micrometric agglomerates of the roug folds are visible in the structure of the single sheet.
These characteristics are further proof of the successful decoration of GO surface wi PA molecules [115].

Optical Microscopy and Roughness Measurement
In order to investigate the morphology of the obtained sol-gel nanohybrid coatin with GO-PA nanofillers, the optical microscopy was employed, to observe in particul the variations in terms of roughness (see Figure 10a-f).
The images show that, in the case of aluminum (see Figure 11a,b), GO-PA coatin treatment has no apparent effect on the surface morphology of the substrates, and th roughness increases only for the GO-PA/APTES coating, as proven by the roughness pr file values collected in Table 2. This increased roughness may enhance the subseque attachment of the paint to the substrate, as demonstrated by adhesion tests, thus provi ing promising benefits, in terms of protective activity. For the GO-PA/GPTMS coating, th average roughness remains almost constant. On the other hand, in the case of steel (s    The images show that, in the case of aluminum (see Figure 11a,b), GO-PA coating treatment has no apparent effect on the surface morphology of the substrates, and the roughness increases only for the GO-PA/APTES coating, as proven by the roughness profile values collected in Table 2. This increased roughness may enhance the subsequent attachment of the paint to the substrate, as demonstrated by adhesion tests, thus providing promising benefits, in terms of protective activity. For the GO-PA/GPTMS coating, the average roughness remains almost constant. On the other hand, in the case of steel (see Figure 11d-f), the GO-PA coating treatment seems to alter the surface morphology of the substrates, with the formation of circular pores for the GO-PA/APTES coating absent in the starting sample. Indeed, from the data shown in Table 2 Figure 12a-c) was used to additionally investigate   Figure 12a-c) was used to additionally investigate the dispersion of the GO-PA nanofiller in GPTMS-and APTES-based matrices on aluminum and steel substrates. According to the EDX mappings, which are displayed in Figure 12 d-f, the C, O, and P elements are evenly distributed across the surface of GO-PA. The fact that the P element is evenly distributed across the entire surface demonstrates that the PA is distributed uniformly. In conclusion, the GO-PA nanofiller exhibits well-dispersed performance in both silane-based coating matrices, indicating that it can greatly improve the coatings' barrier properties. This effect will be further examined in the characterization of the anti-corrosion performances.

Adhesion Measurements: Pull-Off and Cross-Cut Test
Evidence of excellent adhesion of paint after the treatment of aluminum and steel substrates with the GO-PA/GPTMS and GO-PA/APTES coatings was attained through coatings' thickness and adhesion strength evaluations.
Pull-off adhesion and cross-cut tests were performed to investigate the effect of GO-PA-based coatings (GO-PA/APTES and GO-PA/GPTMS) on the adhesion strength of a commercial paint on the surface of aluminum and steel substrates. Figure 13 reveals that treatment with GO-PA coatings can effectively improve the adhesion strength between the paint and aluminum or steel substrates. The phosphate groups of PA have very good metal ion chelation capacity, according to the literature [116]. Therefore, GO-PA-based coatings are able to interact directly with the substrate.

Adhesion Measurements: Pull-Off and Cross-Cut Test
Evidence of excellent adhesion of paint after the treatment of aluminum and steel substrates with the GO-PA/GPTMS and GO-PA/APTES coatings was attained through coatings' thickness and adhesion strength evaluations.
Pull-off adhesion and cross-cut tests were performed to investigate the effect of GO-PA-based coatings (GO-PA/APTES and GO-PA/GPTMS) on the adhesion strength of a commercial paint on the surface of aluminum and steel substrates. Figure 13 reveals that treatment with GO-PA coatings can effectively improve the adhesion strength between the paint and aluminum or steel substrates. The phosphate groups of PA have very good metal ion chelation capacity, according to the literature [116]. Therefore, GO-PA-based coatings are able to interact directly with the substrate.
The maximum adhesion values for both substrates, ISO 0 and ASTM 5B, were obtained via the cross-cut adhesion test, which provides a visual comparison technique for measuring paint adhesion integrity, as verified against ISO 2409 and ASTM D 3359 standards. Additionally, an ASTM 5B value is provided to a coating if all of the edges of the cut are entirely smooth and none of the lattice squares that were generated as a result of the cuts are detached [117]. According to Valli, who was discussing thin, durable coatings for steel protection, adherence is a coating's most crucial quality for future effective use [118]. According to the reference standards, none of the squares in the lattice are detached, and the edges of the cuts in Figure 14a-d are entirely flat.
Pull-off adhesion and cross-cut tests were performed to investigate the effect of GO-PA-based coatings (GO-PA/APTES and GO-PA/GPTMS) on the adhesion strength of a commercial paint on the surface of aluminum and steel substrates. Figure 13 reveals that treatment with GO-PA coatings can effectively improve the adhesion strength between the paint and aluminum or steel substrates. The phosphate groups of PA have very good metal ion chelation capacity, according to the literature [116]. Therefore, GO-PA-based coatings are able to interact directly with the substrate.  measuring paint adhesion integrity, as verified against ISO 2409 and ASTM D 3359 standards. Additionally, an ASTM 5B value is provided to a coating if all of the edges of the cut are entirely smooth and none of the lattice squares that were generated as a result of the cuts are detached [117]. According to Valli, who was discussing thin, durable coatings for steel protection, adherence is a coating's most crucial quality for future effective use [118]. According to the reference standards, none of the squares in the lattice are detached, and the edges of the cuts in Figure 14a-d are entirely flat. This coatings' remarkable qualities, which cannot interfere with the adhesion of the following paint treatment, are demonstrated by their exceptionally excellent high scratch resistance. As a result, a notable improvement is made to the coatings' overall adhesion capacity.
For AQ-36 aluminum substrates, the chromatic map in Figure 15a-c indicates average thicknesses of about 102 and 160 µm for the GO-PA/APTES and GO-PA/GPTMS coatings, respectively. As for the QD-36 steel substrates, the chromatic map in Figure 15d Table 3. Table 3. Thickness values of the uncoated and coated samples obtained from the chromatic maps. This coatings' remarkable qualities, which cannot interfere with the adhesion of the following paint treatment, are demonstrated by their exceptionally excellent high scratch resistance. As a result, a notable improvement is made to the coatings' overall adhesion capacity.

Sample Code T (μm) Sample Code T (μm)
For AQ-36 aluminum substrates, the chromatic map in Figure 15a-c indicates average thicknesses of about 102 and 160 µm for the GO-PA/APTES and GO-PA/GPTMS coatings, respectively. As for the QD-36 steel substrates, the chromatic map in Figure 15d Table 3. Table 3. Thickness values of the uncoated and coated samples obtained from the chromatic maps.

Evaluation of Anticorrosive Performance
Polarization measurements and salt spray test for painted steel and aluminum specimens were performed.
One approach for evaluating coating anticorrosion performance is the polarization test, which measures the coating's corrosive potential (Ecorr), corrosion current density (icorr), polarization resistance (Rp), and corrosion rate (CR). Generally, the shift through a more positive value for Ecorr and lowering of icorr, compared to the untreated sample, indicates improved anticorrosion capabilities [119]. The corrosion protective efficiency (PE), based on the corrosion current density of bare AQ-36 aluminum and QD-36 steel, quantifies the corrosion protective ability of the different coatings, according to the following equation [120,121]: where icorr represents the corrosion current density of the coating, and i represents the corrosion current density of bare AQ-36 aluminum and DQ-36 steel. Tafel polarization curves for the GO-PA-based coating samples were obtained in 3.5 wt.% NaCl aqueous solution, and they are shown in Figure 16a

Evaluation of Anticorrosive Performance
Polarization measurements and salt spray test for painted steel and aluminum specimens were performed.
One approach for evaluating coating anticorrosion performance is the polarization test, which measures the coating's corrosive potential (E corr ), corrosion current density (i corr ), polarization resistance (R p ), and corrosion rate (CR). Generally, the shift through a more positive value for E corr and lowering of i corr, compared to the untreated sample, indicates improved anticorrosion capabilities [119]. The corrosion protective efficiency (PE), based on the corrosion current density of bare AQ-36 aluminum and QD-36 steel, quantifies the corrosion protective ability of the different coatings, according to the following equation [120,121]: where i corr represents the corrosion current density of the coating, and i 0 corr represents the corrosion current density of bare AQ-36 aluminum and DQ-36 steel. Tafel polarization curves for the GO-PA-based coating samples were obtained in 3.5 wt.% NaCl aqueous solution, and they are shown in Figure 16a        As shown in Table 5 The salt spray test was performed to further analyze the corrosion protection capability of the GO-PA-based coating samples. Figure 17 shows visual pictures of the painted aluminum (a) and steel (b) panels treated with PA-GO based coating (GO-PA/GPTMS and GO-PA/APTES) and untreated samples after exposure in neutral salt spray environment. The figure shows that different substrates exhibited varying extents of corrosion areas, following the application of equivalent GO-PA coatings, in agreement with the polarization measurement indicated above. For the steel QD-36 substrates, the corrosion product accumulation and paint peeling were visible around the scratch zone, thus showing the poor anti-corrosion ability of the GO-PA-based coatings. Among them, the corrosion of the St + GO-PA/GPTMS was the most serious, while the corrosion spots of St + GO-PA/APTES was relatively minimal. This could be due to the incompatibility of the coating with the steel substrate or with the paint, probably due to the acidic nature of the nanofiller, thus reducing the anti-corrosion performance of the coatings. This behavior could be counteracted in the future by buffering the coating solution, thus increasing the pH value and improving corrosion inhibition. In contrast, the GO-PA coatings exhibit improved corrosion resistance for aluminum substrate, which does not lead to any accumulation products and coating delamination among all coating samples, increasing the specimen's resistance to NSS from 900 h of the bare aluminum to 1300 h of the coated ones.

Synthesis of Graphene Oxide
The GO nanosheets were made using a modified version of Hummer's method, as follows: (i) 5 g of graphite powder was added in 120 mL of concentrated H2SO4 and stirred for 2 h; (ii) 15 g of KMnO4 and 2.5 g of NaNO3 were added slowly into the mixture and stirred for 72 h; (iii) the obtained mixture was diluted with 600 mL of deionized water and H2O2, added dropwise after 30 min, obtaining a yellow-brown solution; (iv) subsequently, the raw product was firstly washed with a 1:10 HCl solution and deionized water, and then it was centrifugated for 3 min at 3000 rpm, until reaching neutral pH, in order to obtain the final GO. The figure shows that different substrates exhibited varying extents of corrosion areas, following the application of equivalent GO-PA coatings, in agreement with the polarization measurement indicated above. For the steel QD-36 substrates, the corrosion product accumulation and paint peeling were visible around the scratch zone, thus showing the poor anti-corrosion ability of the GO-PA-based coatings. Among them, the corrosion of the St + GO-PA/GPTMS was the most serious, while the corrosion spots of St + GO-PA/APTES was relatively minimal. This could be due to the incompatibility of the coating with the steel substrate or with the paint, probably due to the acidic nature of the nanofiller, thus reducing the anti-corrosion performance of the coatings. This behavior could be counteracted in the future by buffering the coating solution, thus increasing the pH value and improving corrosion inhibition. In contrast, the GO-PA coatings exhibit improved corrosion resistance for aluminum substrate, which does not lead to any accumulation products and coating delamination among all coating samples, increasing the specimen's resistance to NSS from 900 h of the bare aluminum to 1300 h of the coated ones.

Synthesis of Graphene Oxide
The GO nanosheets were made using a modified version of Hummer's method, as follows: (i) 5 g of graphite powder was added in 120 mL of concentrated H 2 SO 4 and stirred for 2 h; (ii) 15 g of KMnO 4 and 2.5 g of NaNO 3 were added slowly into the mixture and stirred for 72 h; (iii) the obtained mixture was diluted with 600 mL of deionized water and H 2 O 2 , added dropwise after 30 min, obtaining a yellow-brown solution; (iv) subsequently, the raw product was firstly washed with a 1:10 HCl solution and deionized water, and then it was centrifugated for 3 min at 3000 rpm, until reaching neutral pH, in order to obtain the final GO.

Functionalization of GO with PA
Firstly, GO nanosheets (0.3 g) were poured into deionized water (200 mL) and then placed in an ice bath sonicator for 30 min to obtain a uniform dispersion of GO. Subsequently, a mixture of GO and PA (6 g) was prepared and sonicated for 2 h under stirring and at room temperature. Therefore, after the conclusion of the reaction, PA-modified GO (GO-PA) was obtained, and the excessive non-reacted PA was removed by several washings using deionized water. Finally, when the neutral pH was reached, the cleaned PA-modified GO was uniformly dispersed in deionized water by sonication, dried in oven at 70 • C for 12 h, and cell pulverized for storage.

Preparation of Nanohybrid Coatings/Hybrid Sol
The nanohybrid coatings were obtained through the following steps. Aqueous suspension of GO-PA was mixed with an alkoxysilane water-based solution, prepared by using GPTMS or APTES as precursors. The resultant solution was vigorously stirred at 25 • C for 24 h to obtain homogeneous dispersion. This sol was used to coat aluminum and steel substrates specimens.

Sample Preparation and Coating Method
In this study, Q-PANEL standard test substrates (AQ-36 aluminum and QD-36 carbon steel), whose dimensions are shown in the Table 6, were purchased from Q-Lab Corporation. In order to prepare the samples for dip-coating, they were cleaned by ultrasonic degreasing with 10% detergent solutions, DEGRIX L 420 ® for the steel and DEGRIX L 321 ® for the aluminum provided by NoxorSokemGroup ® , and rinsed by demineralized water. This cleaning process thoroughly cleans the panels and removes any contaminants that might be on the surface due to preservation before use of the specimens. The preparation of the sol-gel films involved the dip-coating of the prepared specimens in the sol solution for five minutes, then rinsing them with deionized water. Following coating application, the specimens were dried for 1 h in an oven at 60 • C before being cured for 1 h in an oven at 130 • C. A powder paint was applied to the prepared substrates and then cured at 180 • C in an industrial line. The nanohybrid coating behaves similarly to a conventional sol-gel polymeric matrix: the alkoxysilane hydrolysis and condensation reactions firstly bring the synthesis of a functional sol, which, by a dip-coating process, is employed to coat different metallic substrates; this will also lead to the formation of a xerogel film that, after a curing phase, will finally allow us to obtain the desired functional coated surface, as shown in Figure 18. Firstly, GO nanosheets (0.3 g) were poured into deionized water (200 mL) and then placed in an ice bath sonicator for 30 min to obtain a uniform dispersion of GO. Subsequently, a mixture of GO and PA (6 g) was prepared and sonicated for 2 h under stirring and at room temperature. Therefore, after the conclusion of the reaction, PA-modified GO (GO-PA) was obtained, and the excessive non-reacted PA was removed by several washings using deionized water. Finally, when the neutral pH was reached, the cleaned PAmodified GO was uniformly dispersed in deionized water by sonication, dried in oven at 70 °C for 12 h, and cell pulverized for storage.

Preparation of Nanohybrid Coatings/Hybrid Sol
The nanohybrid coatings were obtained through the following steps. Aqueous suspension of GO-PA was mixed with an alkoxysilane water-based solution, prepared by using GPTMS or APTES as precursors. The resultant solution was vigorously stirred at 25 °C for 24 h to obtain homogeneous dispersion. This sol was used to coat aluminum and steel substrates specimens.

Sample Preparation and Coating Method
In this study, Q-PANEL standard test substrates (AQ-36 aluminum and QD-36 carbon steel), whose dimensions are shown in the Table 6, were purchased from Q-Lab Corporation. In order to prepare the samples for dip-coating, they were cleaned by ultrasonic degreasing with 10% detergent solutions, DEGRIX L 420 ® for the steel and DEGRIX L 321 ® for the aluminum provided by NoxorSokemGroup ® , and rinsed by demineralized water. This cleaning process thoroughly cleans the panels and removes any contaminants that might be on the surface due to preservation before use of the specimens.
The preparation of the sol-gel films involved the dip-coating of the prepared specimens in the sol solution for five minutes, then rinsing them with deionized water. Following coating application, the specimens were dried for 1 h in an oven at 60 °C before being cured for 1 h in an oven at 130 °C. A powder paint was applied to the prepared substrates and then cured at 180°C in an industrial line. The nanohybrid coating behaves similarly to a conventional sol-gel polymeric matrix: the alkoxysilane hydrolysis and condensation reactions firstly bring the synthesis of a functional sol, which, by a dip-coating process, is employed to coat different metallic substrates; this will also lead to the formation of a xerogel film that, after a curing phase, will finally allow us to obtain the desired functional coated surface, as shown in Figure 18.
The final coated samples were stored until the tests ( Figure 19).   The final coated samples were stored until the tests (Figure 19).  Figure 19. Aluminum AQ-36 panels (a) and QD-36 steel panels (b) treated with GO-PA-based coatings.

Characterizations
Fourier transform infrared (ATR-FT-IR) spectroscopy, ultraviolet-visible (UV-vis) spectroscopy, Raman spectroscopy, and X-ray diffraction (XRD) were employed to characterize the structural modification of GO after PA functionalization.
For FT-IR analysis, a V-6600 Jasco spectrometer, including the intuitive Spectra Man-ager™ Suite with integrated search software solution, KnowItAll® Informatics and database JASCO Edition (JASCO Europe s.r.l., Cremella, LC, Italy), endowed with an attenuated total reflection (ATR) accessory, was employed, and the spectra of liquid samples recorded, with spectra range of 4000-500 cm −1 at room temperature.
UV-vis spectroscopy (UV-Vis V-770, Jasco equipped with Standard Measurement and Analysis Programs, and Spectra Manager™ Suite Spectroscopy Software, JASCO Europe s.r.l., Cremella, Italy) was performed at 25°C via a wide range from 200 to 800 nm.
Raman measurement was performed to gain the Raman spectra, employing a BRAVO (Bruker Optics, Billerica, Massachusetts, United States) spectrometer, operating in the 450-3200 cm -1 range. The source was constituted by two lasers operating at the wavelength of 785 and 1064 nm. The scanning spectral range was 1000-3200 cm -1 . The spot size was 10-15 micron at 10× lens.
XRD measurements were carried out by using a D8 Advance Bruker instrument (Bruker, Billerica, Massachusetts, United States) equipped with a monochromatic CuKα radiation source (40 kV, 40 mA). Bragg-Brentano theta-2theta configuration and a scanning speed of 0.1°/s were used to examine the samples in a wide range, from 10° to 80°. SEM micrographs were collected by a Quanta 450 FEI, with a large-field detector (LFD), on Cr-coated samples, with an accelerating voltage of 5 kV in high vacuum (10 −6 mbar).
The zeta potential test was performed to analyze the potential variation of the GO nanosheets after PA modification by using the Zeta sizer 3000 instrument (Malvern Panalytical, Malvern, UK. The coating surface investigation were performed by electron microscopy using a Zeiss Sigma VP (Zeiss, Oberkochen, Germany) field emission scanning electron microscope (FE-SEM), endowed with a Bruker Quantax energy dispersive X-ray (EDX) spectrometry microanalysis detector (Bruker, Billerica, Massachusetts, United States), which was used to investigate the surface morphologies and element content of different covered GO-PA panels.
Optical analysis of the samples was obtained by using a Hirox digital microscope KH-8700 (Hirox, Tokyo, Japan), with optical microscope in xyz (3D) and mapping mode. Further, MX(G)-5040Z lens was used to record the optical images.

Characterizations
Fourier transform infrared (ATR-FT-IR) spectroscopy, ultraviolet-visible (UV-vis) spectroscopy, Raman spectroscopy, and X-ray diffraction (XRD) were employed to characterize the structural modification of GO after PA functionalization.
For FT-IR analysis, a V-6600 Jasco spectrometer, including the intuitive Spectra Man-ager™ Suite with integrated search software solution, KnowItAll®Informatics and database JASCO Edition (JASCO Europe s.r.l., Cremella, LC, Italy), endowed with an attenuated total reflection (ATR) accessory, was employed, and the spectra of liquid samples recorded, with spectra range of 4000-500 cm −1 at room temperature.
UV-vis spectroscopy (UV-Vis V-770, Jasco equipped with Standard Measurement and Analysis Programs, and Spectra Manager™ Suite Spectroscopy Software, JASCO Europe s.r.l., Cremella, Italy) was performed at 25 • C via a wide range from 200 to 800 nm.
Raman measurement was performed to gain the Raman spectra, employing a BRAVO (Bruker Optics, Billerica, MA, USA) spectrometer, operating in the 450-3200 cm -1 range. The source was constituted by two lasers operating at the wavelength of 785 and 1064 nm. The scanning spectral range was 1000-3200 cm −1 . The spot size was 10-15 micron at 10× lens.
XRD measurements were carried out by using a D8 Advance Bruker instrument (Bruker, Billerica, MA, USA) equipped with a monochromatic CuKα radiation source (40 kV, 40 mA). Bragg-Brentano theta-2theta configuration and a scanning speed of 0.1 • /s were used to examine the samples in a wide range, from 10 • to 80 • . SEM micrographs were collected by a Quanta 450 FEI, with a large-field detector (LFD), on Cr-coated samples, with an accelerating voltage of 5 kV in high vacuum (10 −6 mbar).
The zeta potential test was performed to analyze the potential variation of the GO nanosheets after PA modification by using the Zeta sizer 3000 instrument (Malvern Panalytical, Malvern, UK. The coating surface investigation were performed by electron microscopy using a Zeiss Sigma VP (Zeiss, Oberkochen, Germany) field emission scanning electron microscope (FE-SEM), endowed with a Bruker Quantax energy dispersive X-ray (EDX) spectrometry microanalysis detector (Bruker, Billerica, MA, USA), which was used to investigate the surface morphologies and element content of different covered GO-PA panels.
Optical analysis of the samples was obtained by using a Hirox digital microscope KH-8700 (Hirox, Tokyo, Japan), with optical microscope in xyz (3D) and mapping mode. Further, MX(G)-5040Z lens was used to record the optical images.
With the portable and compact roughness tester, Surftest SJ-210, series 178 (Mitutoyo S.r.l., Milan, Italy), the surface roughness (Ra) was calculated by the following, Equation (2), |Yi| (2) where Ra represents the arithmetic mean of the absolute values of the deviations of the evaluation profile (Yi) from the mean line. The JIS2001 roughness standard was employed for the measurement conditions, five sampling lengths, lengths of cut-off (λs = 2.5 mm, λc = 0.8 mm), and a stylus translation speed of 0.5 mm/sec. The average profile was obtained on n. 4 roughness profiles per type of sample. The pull-off adhesion test was performed by a Lloyd LR10K (AMETEK GmbH, Meerbusch, Germany) universal testing machine, according to the ASTM D4541 standard test method for pull-off strength of coatings. During the test, steel metal dolls were attached perpendicularly on the steel or aluminum metal samples, uncoated and coated with GO-PA/APTES or GO-PA/GPTMS. The test conditions were load cell 10 KN, with pre-load 1.00 N and speed 1 mm/min.
A commercial cross-hatch adhesion tester (SAMA Tools SADT502-5, SAMA Italia, Viareggio, Italy) was employed for the testing of the adherence of coating films to metallic substrates, according to ASTM D3359e2 standard test method for measuring adhesion by tape. An approximately 10 × 10 cm grid incision was made in a selected test area using an appropriate cutter, characterized by horizontally and vertically spaced (2 mm) incisions on the surface. By a soft brush, the excess particles produced in the region were removed. Moreover, a 3 M adhesive tape was applied on the cutting grid by a light pressure and then peeled off in an orderly motion. The ISO 2409:2013 reference images were used for comparison with the selected pictures, in order to visually evaluate the condition of the damage. Based on the number of flaked off squares and appearance, a cross-cut parameter, ranging from 0 (very good adhesive strength) to 5 (extremely poor adhesive strength), was assigned.
A high precision digital coating thickness gauge, SAMA TOOLS-SA8850 (S.A.M.A Italia S.r.l., Viareggio, Italy), was used to measure the thickness of the films on metal bases in a non-destructive way. A grid was created on the metal specimen to acquire, at the intersection points, the different thickness values for the entire sample.
A PalmSens2 potentiostat from PalmSens (Houten, The Netherlands) was used for the linear polarization measurements. In particular, the analyses were conducted in three electrodes cell, using NaCl 0.1 M solution (Ag/AgCl) (KCl 3M) reference electrode and Pt counter electrode, working electrode as the tested surfaces, with 1.77 cm 2 surface area. The linear polarization was performed by measuring open circuit potential (EOC) for 10 s and scanning potential from EOC ± 0.2 V, with 1 mV potential steps, 1.7 mV/s as scan rate, and the data analyzed by PSTrace software, in order to determine the Tafel slope of the polarization in the corrosion mode.

Conclusions
This work presents the efficient design and development of polymeric silane-based, multicomponent, water-based, and eco-friendly coatings, featuring cross-linked functionalized GO with PA nanofillers, as obtained by sol-gel technology. The synthetized nanostructured coatings were tested as an anticorrosive finishing for a metal surface. It has been demonstrated that the functionalization of GO nanosheets with PA is a key step for enhancing their dispersion in polymeric coatings. The results of the FT-IR spectroscopy, UV-Vis analysis, and SEM indicated the successful decoration of PA molecules on the surface of GO. Anticorrosive performance was tested by polarization measurements and neutral salt spray test, thus confirming that GO-PA nanofillers enhance the barrier and corrosion protection properties of the silane-based coatings. This behavior is attributable to the coatings' enhanced impermeability and adhesive power, as a result of the addition of uniformly dispersed GO-PA nanofillers. In this context, phytic acid is a type of natural organic coordination molecule that, because of its special properties, such as its non-toxicity, biocompatibility, and environmental friendliness, can serve as a safe and green inhibitor of corrosion. The developed functional waterborne multicomponent anticorrosive coatings are expected to pave the way for other efficient cross-linked coatings that may be used as a primer between metal surfaces and a top aesthetic paint, and they are able to: (i) improve the performances and lifetime of the painted metal surfaces against corrosion, due to chemical and environmental agents, (ii) reduce the maintenance phases of the metal surfaces, and (iii) lead to an overall great economic advantage for different metal-based application fields. Therefore, the waterborne, sol-gel-based hybrid coatings bearing GO intercalated PA anticorrosive agents, as developed in this study, could represent a useful, environmentally safe, and easy method of achieving good anti-corrosive and protective coatings for various metal-based application fields.