Introduction

Urushiol formaldehyde polymers (UFPs) are usually used in corrosion protection.1 However, their low alkali resistance greatly limits their application scope, and it is thus necessary to enhance the alkali resistance of UFP.2 Incorporation of dispersed graphene oxide (GO)–nano oxide can efficiently obstruct micropores by reducing internal stresses, thus enhancing the barrier properties of UFP-containing film as a nanofiller.3,4,5

GO has been successfully used to improve the formulation and anticorrosive properties of polymer coatings.1,2,6 The development of composite coatings functionalized by GO-polymers could effectively enhance the anticorrosion and physico-mechanical properties of anticorrosive coatings used for industrial equipment.7 Currently, GO is mainly fabricated from natural graphite via the modified Hummers method,8,9 which produces a nonuniform distribution of oxygen functional groups (carboxylic at the edges, hydroxyl on the surface and epoxide in the basal plane) on the graphene sheets. The preparation of GO-polymer composite coatings for various fields is strongly limited by the low dispersibility of GO. However, a variety of processing methods for improved dispersion of GO in polymer composites have been recently developed, including surface grafting and decorating with nanoparticles, in situ polymerization and organic modification.8,10,11 In particular, some studies have reported that the properties of GO can be improved by decorating it with nanoparticles such as oxides of titanium, aluminum, yttrium and zinc, as well as silica.12,13,14,15,16,17 At the same time, organic coatings comprising blended GO decorated with nanoparticles were introduced to achieve better dispersion of GO by preventing it from aggregating within the polymer coating.12

Nano-TiO2 has been widely applied as a polymer filler in organic coatings because of its high resistance to weathering and corrosion. Nevertheless, the high polarity and large surface area of nano-TiO2 particles make them prone to aggregation in polymer coatings.13,18,19 In addition, nano-silica (SiO2) was introduced into polymeric composite coatings which showed high corrosion resistance and mechanical properties. These were characterized by nano-SiO2 chain-like aggregates characterized by hydrogen bonds between the polymer and the surface hydroxyls of the nanoparticles.20,21,22,23,24 Similarly, nano-yttrium oxide has been exploited as a very promising additive to polymeric organic coatings on metal substrates, owing to its high thermal and chemical stability,25,26 as well as good mechanical and barrier properties.27,28 An as–fabricated GO–oxide dispersed composite was added into a polymeric organic coating, which not only greatly improved the corrosion resistance of the polymer coating but also improved the dispersion of nanoparticles within the matrix via a loose structure of GO.29

With potentially important applications and theoretical implications, the GO–metal oxide composites were prepared via a novel route in this study, which not only combined their inherent advantages but also improved the dispersion of GO within UFP. The synthesis of GO–TiO2 was conducted using 3-aminopropyltriethoxysilane (APTES), GO–SiO2 was fabricated via a sol-gel method using APTES and TEOS, while GO–Y2O3 was prepared via a chemical reaction between GPS–Y2O3 and APS–GO. Afterward, the GO–TiO2, GO–SiO2 and GO–Y2O3 particles were dispersed in UFP as nanofillers, and the corresponding UFP composite coatings were obtained via a solution-blending process. Finally, differences in the resistance of the resulting composite coatings against corrosion and alkali were investigated.

Materials and methods

Materials

Graphite powder with < 20 μm particle size was sourced from the Chinese Academy of Sciences Chengdu Organic Chemical Co., Ltd. (China). Nano-TiO2 (about 30 nm) was supplied by Guangzhou Metal Metallurgy Co., Ltd. (China). Nano-Y2O3 (about 20 nm) was purchased from Ji’nan Huasen Metal Products Co., Ltd. (China). Tetraethyl orthosilicate (TEOS) was sourced from Tianjin Kermel Chemical Reagent Co., Ltd. (China), and the silane coupling agents (SCA): (3-amino-propyl)-trimethoxysilane (APS), γ-(2,3-epoxypropoxy)propyl-trimethoxy silane (GPTMS) and 3-aminopropyltriethoxysilane (APTES) were obtained from Nanjin Chuanshi Chem Co., Ltd. (China). Pure urushiol was obtained via xylene extraction of raw lacquer sap (Shaanxi Lacquer Culture Industry Co., Ltd. China), as described previously.30 Analytical-grade potassium permanganate, hydrogen peroxide, aqueous ammonia solution, hydrochloric acid, sulfuric acid, xylene, methanol, formaldehyde, ethanol and acetone were commercially available.

Fabrication process of the composite coatings

Synthesis of APTES-GO

The modified Hummers method was used to synthesize graphene oxide (GO) from natural graphite.31 Briefly, 1 g of APTES and 0.05 g of GO were first suspended in 50 g of anhydrous ethanol ultrasonication at 50°C for 60 min. Subsequently, the resulting solution was stirred in a water bath at 80°C for 4 h. Then, 5 g distilled water was dropped into the mixed solution, and the as-prepared solution filtrated and washed several times with a 50% (v/v) aqueous ethanol solution to remove the unreacted coupling agents. The obtained solid was dried for 24 h at 60°C to obtain APTES-GO powder. The synthesis of APS–GO powder using the second silane coupling agent APS was done in analogy to that of APTES-GO.

Synthesis of GO–SiO2

A one-step sol-gel process was used to prepare the nano-GO–SiO2.32,33 A schematic of the GO–SiO2 fabrication process is shown in Fig. 1. Specifically, 0.05 g of APTES-GO was added to 288 mL of an aqueous ethanol solution (240 mL ethanol and 40 mL DI water) and dispersed by ultrasonication at 50°C for 60 min to form a homogenous solution. Subsequently, 360 μL of a 28% (w/v) aqueous ammonia solution and 0.6 mL of TEOS were dissolved in the APTES-GO suspension in a 500-mL three-necked condenser flask with a magnetic stirrer. The reaction was performed at 60°C in a water bath for 4 h to ensure the complete hydrolysis of the mixture, followed by incubation of 75°C for another 1 h. The as-prepared solution was filtrated and washed several times with a 50% (v/v) aqueous ethanol solution to remove the unreacted coupling agents, and the solid was dried for 24 h at 60°C to obtain GO–SiO2 powder.

Fig. 1
figure 1

A schematic illustration of the GO–SiO2 synthesis procedure

Full size image

Synthesis of GO–TiO2

As illustrated in Fig. 2, the preparation of GO–TiO2 followed a two-step method as follows:

  1. (1)

    6 g of APTES and 0.2 g of nano–TiO2 were suspended in 150 g of anhydrous ethanol by ultrasonication at 50°C for 30 min. The resulting solution was stirred for 6 h in a water bath at 80°C, after which 44 g of a 15/7 (w/w) aqueous/ethanol solution were dropped into the mixed solution. Finally, the as-prepared solution was filtrated and washed several times with the aqueous ethanol solution to remove the unreacted coupling agents. The remaining solid was dried for 24 h at 60°C to obtain APTES–TiO2.

  2. (2)

    0.3 g of GO and 0.2 g of APTES–TiO2 were added to 100 mL of xylene and uniformly dispersed through ultrasonication for 30 min. The reaction mixture was stirred for 3 h at 105°C, and the resulting GO–TiO2 was obtained by filtrating, washing five times with anhydrous ethanol, and drying for 24 h at 60°C, followed by repeated grinding to obtain a powder.

Fig. 2
figure 2

A schematic illustration of the GO–TiO2 synthesis procedure

Full size image

Synthesis of GPS–Y2O3 and GO–Y2O3

The amino-functionalized Y2O3 was obtained using GPS to modify nano-Y2O3. Briefly, 0.1 g of nano–Y2O3 and 2 g of GPS were homogeneously dispersed in 100 g of anhydrous ethanol by ultrasonication at 50°C for 30 min. The resulting solution was stirred for 4 h in a water bath at 78°C with drop-wise addition of DI water (10 g). Finally, the as-prepared solution was filtered and washed several times with a 50% (v/v) aqueous ethanol solution to remove the unreacted chemicals. The remaining solid was dried for 24 h at 60°C to obtain GPS–Y2O3.

A schematic representation of the reaction between APS–GO and GPS–Y2O3 is shown in Fig. 3. An aliquot comprising 0.25 g of APS–GO was added to 100 mL of xylene and dispersed through ultrasonication for 30 min. Subsequently, 0.05 g of GPS–Y2O3 was added into the suspension and dispersed via ultrasonication for 30 min. The resulting mixed solution was stirred for 5 h at 105°C, after which the GO–Y2O3 solid was filtered, washed several times with a 50% (v/v) aqueous ethanol to remove the unreacted coupling agents, and dried at 60°C for 24 h.

Fig. 3
figure 3

A schematic illustration of the preparation of GO–Y2O3 and the corresponding UFP coatings

Full size image

Preparation of the UFP composite coatings

A homogeneous distribution of nanofiller was ensured by using a low content of GO–TiO2, GO–SiO2 and GO–Y2O3 in the organic coating to prevent them from agglomerating within the UFP matrix. For this purpose, the modified GO dispersion in xylene (1:30, w/w) was sonicated for 30 min and then added to a stoichiometric amount of UFP resin at room temperature using an S312-250H high shear mechanical mixer (SENCO Technology Co., Ltd., China). The mixing ratio of UFP to modified GO was 100:1 by weight. The prepared coatings were applied to JDMK-50 tinplates (Tianjin Jingkelian Material Testing Machine Co., Ltd., China) using the roll coating technique.33 The UFP composite coated tinplates were sequentially baked at 120°C for 1 h and 220°C for 2 h in an oven to achieve complete curing. In addition, the 30 μm thickness GO-metal oxide/UFP composite coatings were prepared on the tinplates and glass sheets as a reference sample.

Characterization of composite coatings

The FTIR test was performed in a wavelength range of 400–4000 cm−1 using a NICOLET IS10 FTIR spectroscopy instrument (PerkinElmer, USA). A K-Alpha X-ray photoelectron spectrometer (XPS; Thermo Fisher Scientific, USA) was used as follows: Al Ka was used as the radiation source at 10-7 mbar and a voltage of 5 kV. An S-4800 field emission scanning electron microscope (FE-SEM, Hitachi Ltd., Japan) was used to examine the morphology of the prepared coatings. XRD (D8 ADVANCE A25, Bruker, Germany) was used to measure the phase crystallinity and composition of the prepared coatings, with Cu Ka as the radiation source at 40 kV and 30 mA.

A PGSTAT 30 electrochemical workstation was used for electrochemical impedance spectroscopy (EIS) and potentiodynamic polarization curve measurements, with a three-electrode cell configuration for both EIS tests and Tafel plots.29 The measurement apparatus consisted of a 1 cm2 tinplate sample coated with the modified GO–metal oxide/UFP as the working electrode, a platinum counter electrode and a saturated calomel electrode (SCE) as the reference electrode.12,37 An applied potential range from − 0.25 to + 0.25 V and relative to open circuit potential (OCP) were used for the polarization curves test at a scan rate of 1 mV/s.37 A frequency range from 0.01 Hz to 100 kHz, with a sinusoidal voltage of 10 mV, was used for the EIS test. All EIS measurements were implemented on three different replicates for checking the repeatability of data.

Analysis of chemical resistance properties

The anticorrosion properties of the UFP composite coatings were investigated using an induced corrosion experiment in chemical solutions. Individual coated glass sheets were separately immersed in 3% NaCl, 30% H2SO4 and 10% NaOH for 14 days at room temperature. Loss of color, brightness or the formation of foam on the surface of the coating were visually followed. Every 24 h, the samples were checked and photographed.

Results and discussion

FTIR analysis

A goal of the structural analysis was to investigate the completion of the reaction between GO and APTES–TiO2. The FTIR spectra of GO–TiO2/UFP and GO–TiO2 are shown in Fig. 4a. The absorption peaks at 2920, 2843, 1038 and 1109 cm−1 could be attributed to the stretching vibrations of –CH3 and –CH2, as well as Si–O–C and Si–O–Si, respectively, most likely corresponding to the functional groups of APTES. The –NH2 absorption band at 1640 cm−1 and the Ti–O–Ti stretching band at 765–620 cm−1 of GO–TiO2 were observed. These results indicated that nano-TiO2 had been successfully modified using silane molecules (APTES) through chemical reactions with functional groups.

Fig. 4
figure 4

FTIR spectra of UFP films and pure urushiol (a), GO–TiO2/UFP films and GO–TiO2 (b), GO–Y2O3/UFP films, APS–GO, GPS–Y2O3 and GO–Y2O3 (c), APTES–GO and GO–SiO2 (d)

Full size image

When comparing typical spectra of GO–TiO2 and GO, it can be seen that the stretching vibration of C–O–C (GO) at 1120 cm−1 disappeared along with appearance of the stronger –CN and –NH absorptions at 1545, 801 and 1545 cm−1, respectively, which demonstrated the successful chemical reaction between epoxide groups on GO surface and amino groups from the modified nano–TiO2 in APTES.15

The FTIR peaks of APS–GO, GPS–Y2O3, GO–Y2O3 and GO–Y2O3/UFP are shown in Fig. 4b. The distinctive absorption peaks of GO generally appear at 3447, 1220, 1120 and 1741 cm−1 corresponding to –OH, C–O, C–O–C and C=O, respectively, indicating the presence of hydroxyl, epoxide and carboxyl groups on the GO surface. Accordingly, the feature in the range of 2800–3000 cm−1 was the result of the –CH stretching modes in the FTIR spectra of APS–GO and GPS–Y2O3. The strong peaks at 1040 cm−1 were associated with Si–O–Si/Si–O–C bonding from APS and GPS,34 with the appearance of new peaks at 1385 and 1508 cm−1 corresponding to Si–O–Y,35 indicating the Y2O3 and GO have been successfully modified by the silane coupling agents APS and GPS through chemical bonding, accordingly. In the FTIR spectra of the GO–Y2O3 example, the absorption peaks of C–O–C bonding became obscured by the appearance of a stronger absorption at 1525 and 1410 cm−1 attributed to the bending of the secondary amide N–H and the C–N stretching vibration, while Si–O–Si/Si–O–C and Si–O–Y bonding was still observed (Fig. 4b). These results implied that reactions took place between the epoxide and amino groups to some extent.36

Figure 4c shows the FTIR spectra of APTES–GO and GO–SiO2. In the spectrum of APTES–GO, many new characteristic absorption peaks were clearly observable, including strong bands at 465 and 1100 cm−1 assigned to the Si–O–Si bending vibration and the Si–O–C/Si–O–Si asymmetric stretching vibration, respectively, which could be attributed to the chemical reaction between GO and APTES. After APTES-GO reacted with TEOS, three new characteristic peaks appeared at 465 (1090), 694 (1124) and 3260 cm−1. These were attributed to the Si–O–Si bending vibration (asymmetric vibration), Si–O–C stretching and NH2, respectively.12,37 This suggests that SiO2 was formed on the GO surface through chemical reactions between epoxy and hydrogen groups of GO and the silane coupling agent APTES,38,39 which was in good accordance with the XPS analysis.

Microstructure and chemical composition of GO–metal oxide in UFP

To evaluate the distribution of GO–Y2O3, GO–TiO2 and GO–SiO2 composites in the UFP matrix, SEM images of the cured UFP composite coatings were also recorded. The SEM micrographs in Fig. 5 show the bulk structural information on the GO–TiO2/UFP, GO–Y2O3/UFP and GO–SiO2/UFP composite films. Figs. 5a and 5b show the fracture morphology of the GO–TiO2/UFP composite coating. As can be seen in Fig. 5a, GO–TiO2 had a better dispersion performance in UFP at 2 wt%, arising from the composite blending process of GO–TiO2 in UFP, which has been observed to result in thin layers. By contrast, the GO–Y2O3/UFP showed an entirely different fracture morphology. As shown in Figs. 5c and 5d, GO–Y2O3 exhibited a good affinity and compatibility with UFP. There was no obvious aggregation into clusters, and a thin-layered structure was observed instead, which improved the poor dispersion of GO–Y2O3 nanoparticles or the laminar pattern within the UFP polymer. The fractured morphology of the GO–SiO2/UFP composite (Figs. 5e and 5f) showed no visible agglomerations or clusters. In particular, GO–SiO2 was well dispersed in the UFP matrix with a multilayered structure, implying that covalent bonds were formed between the APTES and the GO surface, which was consistent with the structure deduced from the FTIR results. The excellent dispersion could enhance the anticorrosion properties of the UFP coatings, which should be verified in further measurements.

Fig. 5
figure 5

SEM images of (a) the GO–TiO2/UFP composite coatings, (b) high magnification image of the GO–TiO2/UFP coating, (c) the GO–Y2O3/UFP composite coating, (d) high magnification image of the GO–Y2O3/UFP coating and (e) the GO–SiO2/UFP composite coating, (f) high magnification image of the GO–SiO2/UFP coating on silicon chip substrates

Full size image

XRD analysis

XRD spectra of GO–TiO2 composite and the GO–TiO2/UFP composite coating are shown in Fig. 6a. In the as-prepared GO–TiO2 samples, a new diminished protuberance was visible at 2θ = 12.5° corresponding to the structure of GO (Fig. 6a), which suggested that the introduction of nano–TiO2 did not destroy the closely packed layered and highly ordered structure of GO.7,33 The GO–TiO2 wide-angle X-ray diffraction showed a very broad peak at 21.7°, which could indicate that the amorphous GO–TiO2 structure was transformed through APTES.40 When comparing the pattern of GO–TiO2 with GO–TiO2/UFP, no obvious characteristic diffraction peaks of GO–TiO2 were observed in the XRD pattern of the GO–TiO2/UFP because the TiO2–GO hybrid nanoparticles were completely dispersed in the UFP matrix.

Fig. 6
figure 6

XRD spectra of GO–TiO2 and GO–TiO2/UFP films (a), APS–GO, GPS–Y2O3, GO–Y2O3 and GO–Y2O3/UFP films (b), APTES–GO, GO–SiO2 and GO–SiO2/UFP films (c)

Full size image

Figure 6b shows the XRD patterns of APS–GO, GPS–Y2O3, GO–Y2O3 and GO–Y2O3/UFP. According to the Bragg equation [2dsinθ=λ (λ=0.154 nm)],33 the interlayer distance of GO (0.80 nm) was larger than that of pristine graphite (0.34 nm) owing to the presence of oxygenated functional groups on the carbon sheets. When comparing the XRD pattern of GO–Y2O3 with that of APS–GO, a small peak is visible around 2θ = 7.8° (d = 1.13 nm) in the curve of GO–Y2O3, which can be attributed to GO. The expansion of the interlayer distance of GO confirmed the successful blending process between APS–GO and GPS–Y2O3.14

The XRD spectra of APTES–GO, GO–SiO2 and GO–SiO2/UFP are shown in Fig. 6c. Wide-angle X-ray diffraction displayed a very broad peak at 19.9° due to the presence of amorphous structure domains in the UFP, which was characteristic of UFP.41 Figure 6c also shows that the intrinsic diffraction peak of GO was apparently weakened in the GO–SiO2 and disappeared in the XRD spectrum of the GO–SiO2/UFP. This means that better dispersion of GO–SiO2 nanohybrids was achieved in this case.12

XPS analysis

To investigate interactions between GO and the metal oxide nanoparticles (TiO2, Y2O3 and SiO2) as well as the chemical state of the GO-metal oxide composites, XPS was applied to identify the changes in the surface elemental composition of the GO-metal oxide.

The XPS survey results of GO–TiO2 (a), C 1s (b), O 1s (c), and Ti 2p (d) are shown in Fig. 7. In the XPS survey spectrum of the composite comprising GO–TiO2 (Fig. 7a), the C 1s spectrum exhibited four peaks located at 284.4, 284.9, 286.6 and 287.9 eV (Fig. 7b), which could be assigned to Ti–O–C, C–C, C–O and C=O, respectively. Observation of Ti–O–C bonding revealed that nano-TiO2 has been modified by APTES to some extent. The O 1s signals at, respectively, 530.6 and 532.3 eV were associated with two Gauss-Lorentzian peaks related to Ti–O and C–O (Fig. 7c).42 The Ti 2p, which consisted of peaks at 458.9 and 456.8 eV (Fig. 7d), was related to the Ti 2p1/2 and Ti 2p3/2 core levels of TiO2, respectively. Taken together, the XPS results indicated that GO was effectively reacted with APTES–TiO2 to form the surface coating.

Fig. 7
figure 7

(a) XPS survey spectrum and high-resolution spectra of GO–TiO2, (b) C 1s, (c) O 1s, and (d) Ti 2p

Full size image

The main structural information, including GO–Y2O3, N 1s and O 1s, is shown in the XPS survey spectrum (Fig. 8). As can be seen in Fig. 8a, C 1s, O 1s, N 1s, Si 2p and Y 3d were the main signals observed in the XPS spectrum of the GO–Y2O3. The O 1s spectrum (Fig. 8b) was separated into five peaks attributed to the binding energies of C–O–C/OH/Si–O–Si (532.8 eV), O=C–O (532.7 eV), C=O/C–O/C–O–Si (532.1 eV), Y–O–Si (531.7 eV) and Y–O (530.6 eV).14,43 These results indicated that GO and Y2O3 were modified by APS and GPS to some degree. The N 1s spectrum of GO–Y2O3 in Fig. 8c could be decomposed into two peaks, in which C–N was assigned to 399.1 eV and N–H to 400.4 eV. Based on the results obtained from the XPS analysis, it could be clearly seen that the GPS–Y2O3 were grafted onto the APS–GO surface via chemical bonding with two types of functional groups (hydroxyl and carboxylic).

Fig. 8
figure 8

(a) XPS survey spectrum and high-resolution spectra of GO–Y2O3, (b) O 1s, and (c) N 1s

Full size image

Figure 9 shows wide-scan XPS spectra of GO–SiO2, whereby peaks in the C 1s, O 1s and Si 2p spectra were detected at binding energies of 283.8, 526.8 and 101.2 eV, respectively. Figure 9b displays the C1s XPS spectrum of GO–SiO2 with seven peaks appearing at 284.1, 284.7, 285.3, 285.5, 286.7, 288.0 and 289.1 eV. These peaks correspond to the C–Si, C–C/C=C, C–N, C–O/C–O–Si, C–O–C, C=O and O–C=O bonds, respectively, indicating that APTES and TEOS successfully reacted with GO forming chemical bonds with two types of functional groups.44,45 In addition, the O 1s signals at 530.9, 531.7, 532.3, 532.9 and 533.4 eV were, respectively, related to the five Gauss-Lorentzian peaks of O–C=O, C=O, C–O/C–O–Si, Si–O–Si and C–O–C/OH (Fig. 9c).46,47 It could be observed that two peaks at 102.5 and 103.4 eV in the Si 2s spectrum (Fig. 9c) were assigned to C–Si and Si–O–Si, respectively, indicating that the GO successfully reacted with silane groups, mainly in the form of SiO2 nanoparticles.44

Fig. 9
figure 9

(a) XPS survey spectrum and high-resolution spectra of GO–SiO2, (b) C 1s, (c) O 1s, and (d) Si 2s

Full size image

As can be seen from Table 1, the Ti 2p, Y 3d and Si 2s signals included 37.12% O–Ti–C, 27.11% Y–O–Si, 23.66% Si–C and 76.34% Si–O–Si peaks, which indicated that most of the silane moieties could be successfully grafted onto the GO surface via chemical bonding in the formation of the UFP composite coatings. Importantly, the results were in good agreement with the FTIR analysis described above.33,48

Table 1 The XPS results for the GO–TiO2, GO–Y2O3 and GO–SiO2 sample
Full size table

Electrochemical tests

For better understanding and to judge the corrosion stage, the Nyquist and Bode plots were fitted with proper equivalent electrical circuits as shown in Fig. 10. The EIS fitted results were indicative of the two different stages of the corrosion process (the initial and terminal stages). In the equivalent electrical circuits, the elements contain the electrolyte resistance (Rs), coating capacitance (Cc), and coating resistance (Rc), respectively.

Fig. 10
figure 10

The proposed equivalent circuit used to fit the EIS experimental data

Full size image

Potentiodynamic analysis

The anticorrosive properties of the three UFP composite coatings were estimated using the potentiodynamic polarization analysis.37,49 Figure 11 shows the Tafel plots of tinplates with and without coatings (GO–SiO2/UFP, GO–TiO2/UFP and GO–Y2O3/UFP) in 3.5% NaCl after soaking for 18 d. The electrochemical parameters from the polarization curves were obtained and listed in Table 2. According to the polarization curves, the samples coated with GO–TiO2/UFP displayed more negative corrosion potentials (Ec, −461.48 mV) and significantly lower corrosion current densities (ic, 1.86×10-9 A/cm2). By contrast, the GO–Y2O3/UFP-coated tinplate substrate exhibited a more positive Ec (−374.82 mV vs SCE), with a significant increase of ic (1.29×10-8 A/cm2). However, the corrosion rate (C.R., 2.18×10-5 mm/a) and corrosion protection efficiency (P.E., 99.59%) of the GO–TiO2/UFP composite coatings were significantly changed by the reduction of ic accompanied by a more positive Ec due to the addition of GO–TiO2.49 These results indicated that the GO–TiO2/UFP coating provided better corrosion protection to the metal substrate than other UFP composite coatings and effectively decreased the corrosion rate of the tinplate. Thus, the anticorrosion properties of the GO–TiO2/UFP composite coating were improved by adding GO–TiO2, especially due to its good distribution within the UFP matrix,50 which further reinforced the composite’s anticorrosion performance. Well-dispersed GO–TiO2 embedded in UFP matrix provided a superior barrier effect due to its two-dimensional sheet structure and plugging inherent micropores properties, thereby exhibiting superior corrosion protection to the metals.

Fig. 11
figure 11

Potentiodynamic polarization curves for the GO–TiO2/UFP, GO–Y2O3/UFP and GO–SiO2/UFP composite coatings on tinplate substrates after immersion in 3.5% NaCl solution for 18 days

Full size image
Table 2 Electrochemical parameters obtained from the polarization curves of the UFP coatings containing different contents of GO–TiO2, GO–Y2O3, GO–SiO2 and MGO
Full size table

In this study, the protection efficiency (P.E.) and corrosion rate (C.R.) of composite coatings were used to better understand the corrosion process of metal substrates coated with the UFP composite coatings. The corresponding values were calculated based on equations (1)53 and (2),51,52 respectively.

$${\rm{C.R.}} = 3.27 \times {10^{ - 3}} \times \frac{{{i_{corr}}}}{\rho } \times {E_w}$$
(1)
$${\rm{P.E.}}(\% ) = ({I_0} - {I_c}/{I_0}) \times 100$$
(2)

In equation (1), EW is the equivalent weight of the tinplate (4.01 g) and ρ is the density of the tinplate (7.85 g/cm3). In equation (2), I0 (μA/cm2) is the corrosion current of the neat tinplate, while Ic (μA/cm2) is the corrosion current of the tinplate coated with the UFP composite coating. The parameters extracted from the polarization curves were in good agreement with the electrochemical tests (Table 2).

EIS measurements of the composite coatings

The electrochemical impedance spectroscopy (EIS) measurements of the different GO-metal oxide/UFP composite coatings were displayed in the form of the Nyquist plots and Bode diagrams. In addition, the EIS fitted data with suitable electrically equivalent circuits were also shown in the figures for comparison (solid lines). The determination of the corrosion stage plays a key role in the study of corrosion processes. In the initial stage of the corrosion process, the electrolyte solution penetrated into the coating through inherent micropores. In the middle stage, the corrosion reaction occurred. The terminal stage comprises a diffusion process, but UFP polymer wearing is nonconspicuous.53,54,55,56

Figures 12 and 14 show the Bode and Nyquist plots of GO–SiO2/UFP, GO–TiO2/UFP and GO–Y2O3/UFP composite coatings after immersion for 2 and 16 days, respectively.

Fig. 12
figure 12

Nyquist (a) and Bode (b) plots of UFP coatings corresponding to GO–TiO2/UFP, GO–Y2O3/UFP and GO–SiO2/UFP after 2 days of immersion in 3.5% NaCl solution; the marker points and solid lines show the experimental and fitted data, respectively

Full size image

Figures 12a and 12b, respectively, showed the Nyquist and Bode plots of the GO–SiO2/UFP, GO–TiO2/UFP and GO–Y2O3/UFP coatings after 2 days of immersion in 3.5% NaCl solution. The radius of the impedance arc of the GO–TiO2/UFP composite coating was significantly higher than that of the other two composite coatings; the impedance modulus at the low frequency for GO–SiO2/UFP, GO–TiO2/UFP and GO–Y2O3/UFP coatings was 1.62 × 108, 1.27 × 1010 and 4.14 × 108 Ω cm2, respectively. Signals at the larger radius of the impedance arc correspond to greater anticorrosion resistance and more difficult electron transfer,37,57 whereby the GO–TiO2/UFP composite coating exhibited a significantly higher value than both GO–SiO2/UFP and GO–Y2O3/UFP. According to Figs. 12a and 12b, one time constant appeared in initial state, which indicated that the corrosion process did not occur, and these three coatings displayed excellent barrier properties. In addition, in Bode plots of the three composite coatings, the phase angle of the GO–TiO2/UFP composite coatings could reach more − 90° of the highest value (Fig. 12b) even after immersing for 2 days, which indicated that the GO–TiO2/UFP composite coatings had pronounced anticorrosive properties. This indicated that incorporation of GO–TiO2 is a promising approach for improving the barrier properties of the UFP coatings at initial immersion. The corresponding phase angle of the GO–TiO2/UFP composite coating was the highest among the three UFP composite coatings, which was another clear sign of the improved corrosion resistance of GO–TiO2/UFP composite coatings.

Figures 13a and 13b showed the Nyquist and Bode plots of the GO–SiO2/UFP, GO–TiO2/UFP and GO–Y2O3/UFP composite coatings. GO–TiO2/UFP maintained a relatively high modulus and phase angle, which may be attributed to the GO–TiO2 hybrid nanoparticles, including their two-dimensional structure and uniform dispersion properties. Accordingly, it was confirmed that the GO–TiO2 composite coating remarkably improved the corrosion resistance of the UFP coating.

Fig. 13
figure 13

Nyquist (a) and Bode (b) plots of UFP coatings corresponding to GO–TiO2/UFP, GO–Y2O3/UFP and GO–SiO2/UFP after 16 days of immersion in a 3.5% NaCl solution; the marker points and solid lines show the experimental and fitted data, respectively

Full size image

The variation of the |Z|0.01 Hz values of the GO–metal oxide/UFP samples vs immersion time is shown in Fig. 14. The |Z|0.01 Hz values of the GO–TiO2/UFP composite coating were higher than those of the two other samples during the immersion, indicating a better corrosion protection performance of the GO–TiO2/UFP composite coating due to the barrier effect of the GO–TiO2 in the UFP polymer. The impedance values decreased for all UFP composite coatings with the increase of immersion time. It suggested that the corrosive electrolyte gradually diffused into the coating, thereby increasing the porosity and presence of electrolyte pathways.12 Obviously, these results were in good accordance with the electrochemical tests.

Fig. 14
figure 14

The variation of the |Z|0.01 Hz value versus immersion time in a 3.5% NaCl solution for GO–TiO2/UFP, GO–Y2O3/UFP and GO–SiO2/UFP samples

Full size image

According to Table 3, the GO–TiO2/UFP coating had higher Rc than the other two samples. Indeed, the Rc values of the UFP composite coatings decreased with prolonged immersion time in the order of UFP < GO–SiO2/UFP < GO–Y2O3/UFP < GO–TiO2/UFP. The increase of the Rc value indicated the good dispersion of GO–TiO2 in the coating, which improved its resistance against the uptake of corrosive species by decreasing its porosity.58,59

Table 3 Electrochemical parameters obtained from the EIS spactra of the UFP coatings containing different contents of GO–TiO2, GO–Y2O3 and GO–SiO2
Full size table

Chemical resistance of the UFP composite coatings

To demonstrate the superior alkali resistance of the GO–SiO2/UFP, GO–TiO2/UFP and GO–Y2O3/UFP composite coatings, the chemical resistance of the coatings was examined by immersing coated samples in 10% NaOH, 30% H2SO4 and 3% NaCl for up to 14 days at room temperature, in conjunction with visual observations for cracking, blistering or other signs of solvent attack on the coatings surface. The results of the chemical resistance tests conducted on the composite coatings are listed in Table 4 as a reference for the electrochemical impedance spectroscopy (EIS) data. The GO–SiO2/UFP showed a better alkali resistance than the other UFP composite coatings, with no damage after 14 days of immersion in a 10% NaOH solution, which indicated that the alkali resistance could be improved via the chemical reaction between GO–SiO2 and UFP by decreasing the hydroxyl group content of pure UFP coating.1

Table 4 Resistance of the GO–TiO2/UFP, GO–Y2O3/UFP and GO–SiO2/UFP composite coatings against chemical attack after immersion in 30% H2SO4, 10% NaOH and 3% NaCl solution at room temperature
Full size table

Corrosion protection mechanism

An explanation of the corrosion-resistance mechanisms of the GO–metal oxide/UFP composite coating was proposed according to the preceding observations and analysis.

Based on the coating capacitance [equations (3) and (5)]60,61,62 and the Brasher–Kingsbury equation [equation (4)], it was assumed that the transport of water through the polymer coating follows Fick’s law:63,64

$$C_{c} = \varepsilon \varepsilon _{0} A/d$$
(3)
$$\phi _{w} = \frac{{\log \frac{{C_{c} }}{{C_{0} }}}}{{\log \varepsilon _{w} }}$$
(4)
$$C_{c} = \left( {Y_{0} } \right)^{\frac{1}{n}} \cdot \left( {R_{{{\text{pore}}}} } \right)^{{\frac{1 - n}{n}}}$$
(5)

In equation (3), A is the active area, d is the coating thickness, ε is the relative dielectric constant of the UFP composite coating and ε0 the permittivity of vacuum (8.85 × 10−14 F cm−1). In equation (4), εw is the dielectric constant of water (εw ≈ 80), Cc is the coating capacitance at immersion time t, and C0 is the initial capacitance of the dry coating. The volume fraction of water uptake (φw) in the GO–TiO2/UFP coatings can be calculated from their capacitance (Cc). In equation (5), the coating capacitance value was obtained using the EIS method including the admittance parameter (Y0) and an exponent (n). The constant phase element consisting of the two parameters represented the deviation from ideal capacitive behavior. According to equation (4), this showed that the effect of an increase of the dielectric constant with the permeation of water may increase the Cc value. That is, a change of the capacitance value indicates water uptake via the GO–metal oxide/UFP polymer film. In the case of the GO–TiO2/UFP sample, corrosive electrolytes diffused into the coating increasing the amount of water uptake. Subsequently, crosslinking density of the UFP composite coating contributed to the significant decrease in the amount of water uptake with increasing immersion time. This indicated that silanols reacted with UFP by hydrolysis and condensation. It could be confirmed that the electrochemical corrosion resistance of the GO–TiO2/UFP composite coating was superior to that of the other GO–metal oxide/UFP materials since GO–TiO2 served as a barrier that blocked the diffusion paths of corrosive agents.

It is not difficult to see that the protection efficiency (P.E.) of the GO–TiO2/UFP coatings increased remarkably by adding 3.0 wt% GO–TiO2, which serves to demonstrate that the GO–TiO2/UFP composite coating exhibited anticorrosive properties of metallic substrates by making the diffusion pathways of corrosive agents in the UFP more tortuous.65,66 According to a few reports, the atomic/molecular situation of the interface of metal and organic coatings was assessed based on QM (quantum mechanics) approaches.69 The QM computations also evidenced that with increasing the amino groups bound to GO surface, the interactions of UFP with modified GO surface strengthened, which could result in improved properties of prepared coatings of UFP and modified GO. Inclusion of GO–SiO2, GO–TiO2 and GO–Y2O3 into the UFP coating improved its anticorrosion properties. The highest improvement of the polymer barrier performance was seen in the case of GO–TiO2 composites. As a result, owing to stronger interfacial interactions, a metal substrate with higher OH coverage will be coated with a thicker cured-UFP layer, which leads to enhanced protection of carbon steel.67,68,69,70 Accordingly, a corrosion protection mechanism for GO–TiO2/UFP composite coatings could be proposed as shown in Fig. 15.

Fig. 15
figure 15

Schematic representation of the most likely mechanism of corrosion protection by the GO–metal oxide/UFP composite coatings on tinplate substrates

Full size image

Conclusions

Metal oxide nanoparticle-decorated graphene oxide composites (GO–SiO2, GO–TiO2 and GO–Y2O3) were synthesized via a simple two-step solution approach. Their structure and morphology were characterized and they were incorporated into UFP-based coatings. The corrosion-resistance properties of the GO–metal oxide/UFP composite coatings were investigated by various techniques. The main findings are as follows: (1) Compared with the two other GO–metal oxide composites, the dispersion, compatibility, barrier properties and corrosion resistance of the GO–TiO2/UFP composite coating were dramatically enhanced. (2) The addition of GO–SiO2 appears to be an appropriate approach to evidently improve the alkali resistance of UFP at low doses. At 0.1 wt% and above, the addition of GO–SiO2 drastically increased the alkali resistance of the resulting GO–SiO2/UFP composite coating due to the chemical reaction between GO–SiO2 and UFP, providing a more effective barrier against the ingress of corrosive ions.