Corrosion and Anti-Corrosion Mechanisms of Epoxy Resin/Graphene and Epoxy Resin/Graphene Oxide Composite Coatings on Magnesium Alloys

Abstract

Graphene and graphene oxide are potential anti-corrosion materials. In this study, epoxy resin/graphene and epoxy resin/graphene oxide composite coatings were succeed prepared. Hydrogen evolution and electrochemical experiments were conducted to determine key parameters—including hydrogen evolution rate, hydrogen evolution volume, corrosion current density, and corrosion potential—of the designed composites in a 3.5 wt.% NaCl solution. The sample with the highest graphene oxide content was 0.8295 μA/cm², representing a two-order-of-magnitude decrease compared to the matrix. Combined with scanning electron microscopy, the surface morphologies of various coatings after corrosion were observed, and the corrosion mechanisms of magnesium alloys with these coatings were carefully discussed. Based on electrochemical analysis, this study proposes and verifies that the working mechanism of the composite coatings relies on a physical barrier rather than a redox reaction.

1. Introduction

As a lightweight metallic material, magnesium alloy exhibits broad application prospects in multiple fields, owing to its low density, high specific strength, and excellent electrical/thermal conductivity [1,2,3,4]. However, magnesium has a standard electrode potential of −2.372 V and extremely high chemical activity, which severely restricts the popularization and application of magnesium alloys [5,6]. Under natural conditions, a protective oxide layer forms on the surface of magnesium alloys; however, this naturally generated oxide film has a loose, porous structure that cannot effectively block the intrusion of external corrosive media (e.g., oxygen, water, and chloride ions), failing to provide long-term protection for the magnesium matrix.

Therefore, blocking the contact between magnesium alloys and external corrosive media is a simple and effective anti-corrosion strategy. However, monolithic organic coatings suffer from drawbacks such as insufficient adhesion and poor aging resistance [7]. Epoxy resin (EP) coating, a green and environmentally friendly surface treatment technology developed in the last century, offers advantages including low cost and eco-friendliness [8]. Graphene is a two-dimensional material with excellent physical and chemical properties; when it is added to coatings, it not only enhances physical barrier performance but also effectively inhibits electrochemical corrosion [9]. Nevertheless, graphene exhibits poor dispersibility in coatings and weak interfacial cohesion strength. Simply adding graphene without considering the overall coating system will also reduce its protective effect [10,11].

Jin et al. [12] fabricated a new type of graphene oxide (GO), which formed a compact, smooth coating that significantly improved the corrosion resistance of the AZ31 magnesium alloy. Liu et al. [13] added acrylic resin to EP coating on the AZ31 magnesium alloy, enhancing both adhesion and corrosion resistance—improved adhesion helps the coating maintain optimal performance. Xue et al. [14] constructed a composite coating of graphene, EP, and Ce film, which remarkably enhanced the corrosion resistance of magnesium alloys. In this paper, two types of graphene-based materials were used to prepare epoxy resin-based composite coatings. The structural and performance differences between the two coatings were systematically investigated and compared, and the effects and distinctions of graphene/graphene oxide on the corrosion protection mechanism of epoxy coatings were elucidated through experimental methods.

2. Materials and Methods

Graphene was purchased from Guangdong Jiazhaoye New Materials Co., Ltd. (Jiangmen, China), with a flake size of 1–10 μm and a layer number of 1–3. Graphene oxide was purchased from Shenzhen Suiheng Technology Co., Ltd. (Shenzhen, China), with a flake size of 1–10 μm and a layer number of 1–3. AZ31 magnesium alloy square sheets (10 mm × 10 mm × 3 mm) were used. After grinding and polishing in sequence, the samples were ultrasonically cleaned in ethanol and dried for subsequent use. Epoxy resin was weighed and placed in a 50 °C water bath for 3 min. Ethanol and xylene were added to the preheated epoxy resin to prepare an epoxy solution.

Table 1. Contents of each component in coatings (wt.%).

	EP-G-1	EP-G-2	EP-G-3	EP-GO-1	EP-GO-2	EP-GO-3
epoxy resin	47.48	47.38	47.28	47.34	47.24	47.15
ethanol	Bal.	Bal.	Bal.	Bal.	Bal.	Bal.
xylene	14.28	14.28	14.28	14.24	14.24	14.24
KH550	_	_	_	0.2849	0.2849	0.2849
m-phenylenediamine	4.762	4.762	4.762	4.748	4.748	4.748
graphene	0.1428	0.2381	0.3333	_	_	_
GO	_	_	_	0.1424	0.2374	0.3324

shows the content of each component in the coating.

For the EP/G composite coating, an appropriate amount of epoxy solution was taken, and graphene powder and m-phenylenediamine were added. After magnetic stirring for 30 min, ultrasonic dispersion was performed for another 30 min. Finally, the mixture was uniformly coated onto the magnesium alloy surface using a KTQ-II film applicator(Dongguan, China), with the coating thickness controlled at 100 μm. The coated samples were cured at room temperature for 48 h in a dry environment to obtain magnesium alloy samples with EP/G composite coatings, labeled as EP-G-1, EP-G-2, and EP-G-3, respectively.

For epoxy/graphene oxide (EP/GO) composite coatings, an appropriate amount of epoxy solution was taken, and graphene oxide (GO) powder was added. Silane coupling agent KH550 (Macklin, Shanghai, China) was added dropwise, followed by m-phenylenediamine. The mixture was magnetically stirred for 30 min and then ultrasonically dispersed for 30 min. The same coating and curing procedures as described above were adopted, and the final samples were labeled as EP-GO-1, EP-GO-2 and EP-GO-3, respectively.

Four small holes were uniformly pierced in the coating using a fine needle to initiate the corrosion reaction. All experiments were conducted at room temperature using a 3.5 wt.% NaCl solution as the corrosive medium. We inverted the funnel filled with sodium chloride solution over the beaker containing the sample and sodium chloride solution, and observed the gas produced by the corrosion reaction. Electrochemical tests were performed on a CHI660e electrochemical workstation (Shanghai, China) with a standard three-electrode system: a saturated calcium chloride electrode (SCE) as the reference electrode, a platinum electrode as the counter electrode, and the coated magnesium alloy sample as the working electrode. Fourier transform infrared (FTIR) spectroscopy was conducted using a SPECTRUM TWO infrared spectrometer (PerkinElmer, Waltham, MA, USA). A Hitachi SU8081 scanning electron microscope (SEM, Tokyo, Japan) was employed for high-resolution observation of the coating surface morphologies (corroded and uncorroded).

3. Results

3.1. FT-IR

Figure 1 shows the FT-IR spectrum of the sample. The results indicate that the strong stretching peak of C=O near 1650 cm⁻¹ is more intense in EP-GO-x (x = 1, 2, 3), demonstrating that graphene oxide has been successfully grafted onto the epoxy resin. In addition, the intensity of the antisymmetric stretching peak of Si-O-C near 1050 cm⁻¹ is also enhanced, which further confirms that the modifier is bonded to the epoxy resin in EP-GO-x (x = 1, 2, 3).

3.2. Hydrogen Evolution Test

Figure 2 shows that hydrogen evolution volumes of EP-G-x (x = 1, 2, 3) decreased by 19.6%, 64.7%, and 46.7%, respectively, compared with the bare Mg substrate. This indicates that unmodified EP-G-x coatings with different graphene contents all enhance the corrosion resistance of magnesium alloys. The early and middle stages of the corrosion process were similar across EP-G-x samples, possibly because the magnesium alloy surface was extensively exposed after the coating lost its protective effect during these stages [15,16].

After 18 h of hydrogen evolution, the hydrogen evolution volumes of EP-GO-x (x = 1, 2, 3) decreased by 65.5%, 66.5%, and 85.6%, respectively, compared with the bare substrate. As shown in Figure 3, once corrosion initiated, EP-GO-x coatings exhibited a superior ability to enhance magnesium alloy corrosion resistance compared with EP-G-x coatings. However, by the final stage of corrosion, the coatings could barely provide any anti-corrosion protection.

3.3. Morphology of the Sample Surface

As observed in Figure 4a–c, the surfaces of EP-G-x are not smooth and exhibit numerous dot-like structures. In contrast, Figure 4d–f show that the number of dot-like or block-like structures on the surface of EP-GO-x coatings is significantly reduced; even under high-magnification observation, no pits or cracks are found on the EP-GO-x coating surfaces. The dot-like structures on unmodified EP-G-x coatings may result from the poor, uneven dispersion of graphene in the epoxy matrix. Unmodified graphene lacks active functional groups on its surface [17], and its sheets spontaneously aggregate via van der Waals forces, forming micron-sized agglomerates. After modification with KH550, the dot-like structures combine with EP through chemical bonds and are embedded in the epoxy matrix, resulting in stronger bonding between GO and EP.

3.4. Morphology of the Corroded Sample Surface

As shown in Figure 5a, the AZ31 bare substrate suffers the most severe corrosion. A rough surface with numerous post-corrosion cracks and a large accumulation of corrosion products are observed. For EP-G-1 (Figure 5b), the corroded surface lacks the extensive corrosion pits and skeleton-like corrosion structures seen on the bare substrate, but it exhibits significant height fluctuations and a large amount of uniformly dispersed corrosion products. For EP-G-2 (Figure 5c), the corroded surface also has uniformly dispersed corrosion products, but its flatness and integrity are significantly improved, with few areas showing large fluctuations or cracks. For EP-G-3 (Figure 5d), there is no excessive accumulation of dispersed corrosion products (as seen in EP-G-1 and EP-G-2), but the corroded surface begins to show uneven fluctuations and cracked areas similar to those of EP-G-1.

The difference in thermal expansion coefficients between graphene agglomerates and the epoxy matrix leads to interfacial stress concentration during curing, resulting in numerous defects in the coating. When local stress exceeds the yield strength of the resin, microcracks initiate in the coating; the extremely high stress intensity at crack tips accelerates crack propagation, making the sample more susceptible to corrosion. In contrast, carboxyl groups (-COOH) and epoxy groups (C-O-C) in the modified GO coating may undergo condensation reactions with amino groups (-NH₂) of KH550, forming amide bonds (-CONH-) and siloxane bonds (-Si-O-C-) [18]. This replaces van der Waals force-dominated agglomeration with chemical bonding, which is one reason for the slower corrosion rate of modified coatings.

Figure 5. Corroded microstructural morphology of samples under scanning electron microscope (a) matrix: (b) EP-G-1, (c) EP-G-2, (d) EP-G-3, (e) EP-GO-1, (f) EP-GO-2, and (g) EP-GO-3.

Figure 5. Corroded microstructural morphology of samples under scanning electron microscope (a) matrix: (b) EP-G-1, (c) EP-G-2, (d) EP-G-3, (e) EP-GO-1, (f) EP-GO-2, and (g) EP-GO-3.

The local accumulation of corrosion products on EP-GO-x (Figure 5e–g) can be attributed to the fact that only small areas of the coating peeled off first during corrosion. Since the total hydrogen release of each sample in the hydrogen evolution experiment was 10 mL, samples with better coating performance exhibited longer hydrogen release times. Compared with samples with poor corrosion resistance (where the coating peeled off rapidly over a large area), samples with good corrosion resistance showed continuous hydrogen release from early peeled areas, leading to significant accumulation of corrosion products in these regions, while other areas remained relatively flat. Even scattered, the incompletely peeled coating fragments observed on EP-GO-2 confirm this observation.

Irregular pits are observed on the surface of some magnesium alloy samples in Figure 4: these pits have varying depths, no internal corrosion products, and flocculent corrosion products around them, which is consistent with the pitting corrosion mechanism of magnesium alloys. In addition, some figures show relatively uniform corrosion products with homogeneous morphologies, corresponding to the uniform corrosion mechanism of magnesium alloys. Discontinuous cracks are also observed, with corrosion products accumulated around them—this is attributed to the intergranular corrosion mechanism of magnesium alloys. All these mechanisms are potential corrosion behaviors of magnesium alloys in 3.5 wt.% NaCl, and the addition of coatings does not introduce new corrosion mechanisms.

3.5. Electrochemical Test

The polarization curve data presented in this paper were all obtained by the Tafel extrapolation method.

Table 2. Corrosion potential and free corrosion current density of the substrate and specimens.

	Matrix	EP-G-1	EP-G-2	EP-G-3	EP-GO-1	EP-GO-2	EP-GO-3
Ecorr (V)	−1.39179	−0.99377	−1.05023	−1.02406	−0.66471	−0.72417	−0.77591
icorr (A/cm2)	7.863 × 10−5	7.912 × 10−5	2.591 × 10−5	5.223 × 10−5	4.088 × 10−6	1.294 × 10−6	8.295 × 10−7

presents the corrosion-related data for the samples, and Figure 6 shows the polarization curves of the samples. As shown in Table 2, the self-corrosion current densities of EP-G-2 and EP-G-3 are lower than that of the bare AZ31 substrate, indicating that the applied coatings significantly improve the corrosion resistance of magnesium alloys. The self-corrosion current density of EP-G-1 is similar to that of the bare substrate, but its self-corrosion potential is significantly higher—this confirms that the corrosion resistance of EP-G-x is significantly stronger than that of the bare substrate. Among the EP/G coatings, EP-G-2 exhibits the best corrosion resistance, followed by EP-G-3, both of which outperform EP-G-1.

The self-corrosion current densities of EP-GO-x are significantly lower than those of EP-G-x, and their self-corrosion potentials are significantly higher. As the GO content increases, the self-corrosion current densities of EP-GO-x increase accordingly, but no significant agglomeration (as observed in EP-G-3) occurs. This indicates that the added KH550 silane coupling agent effectively improves the dispersibility of GO and the bonding strength between GO and the epoxy matrix [18].

Figure 6. Polarization curves of AZ31 matrix and samples.

Figure 6. Polarization curves of AZ31 matrix and samples.

The impedance arc sizes of different samples can be observed in Figure 7. The capacitive arc diameter of EP-G-1 is relatively small, which may be due to low graphene content leading to insufficient coating compactness and high porosity, making it difficult to achieve effective anti-corrosion performance [12]. The capacitive arc diameter of EP-G-2 increases significantly, with a steeper slope in the high-frequency region—this is because increased graphene content improves coating compactness, and graphene effectively blocks the penetration of corrosive media. The EIS curve of EP-G-3 shows a slightly smaller capacitive arc diameter than EP-G-2, which can be attributed to agglomeration caused by excessive graphene; this generates microcracks inside the coating and accelerates local corrosion [19,20].

The capacitive arc diameters of EP-GO-x are significantly larger than those of EP-G-x and increase with increasing GO content. This result is consistent with the polarization curve analysis, indicating that the KH550 coupling agent enhances GO dispersibility via chemical bonding. It not only fills the pore defects in EP-G-1 (low G content) but also enables coatings with high G content to form a multi-layer physical barrier structure through the interfacial bridging effect of KH550. This effectively eliminates the agglomeration phenomenon in EP-G-3 and significantly improves corrosion resistance [21,22,23].

The equivalent circuit shown in Figure 8 is quite appropriate. Except for the substrate, the impedance values are consistent with those shown in Figure 7. During the curing process of epoxy resin, tiny defects can be generated, which significantly affect the characteristic diffusion time and diffusion impedance. Compared with the uncoated sample, the epoxy resin/graphene composite exhibits increased surface roughness, which in turn leads to a longer characteristic diffusion time and higher diffusion impedance. At this point, the diffusion process is closer to the ideal state, indicating that the epoxy resin/graphene coating shows an excellent physical barrier effect [24]. For graphene oxide, chemical bonding occupies corresponding sites during the curing of epoxy resin, altering the cured structure, thereby changing the surface roughness and affecting the variation in Ws-P. The data presented in

Table 3. Equivalent circuit fitting data of samples.

	EP-G-1	EP-G-2	EP-G-3	EP-GO-1	EP-GO-2	EP-GO-3	Matrix
Rs (Ω·cm2)	18.44	13.01	17.93	18.38	17.29	13.26	21.47
R1 (Ω·cm2)	8.261	5.678	3.112	13.2	15.62	6.346	15.86
CPE1-T (Ω·cm·sP)	7.4362 × 10−4	2.8245 × 10−8	3.2095 × 10−5	5.3565 × 10−7	7.893 × 10−7	6.652 × 10−9	1.7877 × 10−6
CPE1-P	1.948	1.214	1.185	1.421	1.407	1.3	1.01
WS-R (Ω·cm2)	45.11	19.65	52	27.19	58.13	157.8	-
WS-T (s)	6.8562 × 10−4	1.4316 × 10−4	7.3256 × 10−4	2.6625 × 10−4	3.8034 × 10−4	3.8651 × 10−4	-
WS-P	0.48244	0.5052	0.48328	0.48691	0.49646	0.5763	-

can verify this viewpoint. It must be pointed out that the value range of CPE-P is typically 0 to 1. The occurrence of CPE-P values greater than one in this study is attributed to parameter compensation by the fitting software, and such values have no practical physical meaning. The underlying reason is that the coating surface can no longer be described by the ideal interfacial electric double-layer model. The introduction of the epoxy resin composite coating directly causes changes in the dielectric constant and plate distance of the system, which in turn leads to the emergence of this fitting phenomenon.

Two mainstream viewpoints exist regarding the anti-corrosion mechanisms of graphene and GO. Physical blocking effect: Dispersing G/GO in EP prevents direct contact between magnesium alloys and the external environment, achieving corrosion protection. Influence of π bonds [25]: Both graphene and GO contain π bonds, and some electrons in these bonds can participate in the corrosion process of magnesium alloys. During corrosion, after magnesium atoms lose electrons, π bonds (with a large number of electrons) can provide electrons to them; meanwhile, highly chemically active sites on G/GO can also participate in water-related reactions.

To better describe the physical process of the protective effect exerted by the coating, we established the physical model illustrated in Figure 9. Based on the impedance results and relevant studies, the EP-G/GO-x layer provides a favorable covering effect and does not exhibit impedance spectrum characteristics corresponding to ion penetration into the interior of the EP-G/GO-x layer. In the low-frequency region, graphene undergoes a process closer to ideal diffusion without ion exchange occurring. In contrast, graphene oxide deviates from ideal diffusion in the solution, which may be attributed to the occurrence of mass transfer processes (caused by chemical reactions) or a rougher surface compared with graphene.

4. Conclusions

In this study, EP/G and EP/GO composite coatings were prepared, and the relationship between coating composition and corrosion resistance was systematically investigated. The main conclusions are as follows:

(1)

In the 18 h hydrogen evolution experiment, the hydrogen evolution volumes of unmodified EP-G-x (x = 1, 2, 3) are reduced by 19.6%, 64.7%, and 46.7%, respectively, compared with the bare substrate. When the graphene content reaches 0.7 wt.%, graphene agglomeration tends to occur, which increases the coating’s porosity and results in the degradation of corrosion resistance. For EP-GO-x (x = 1, 2, 3), the hydrogen evolution volumes are reduced by 65.5%, 66.5%, and 85.6%, respectively, compared with the bare substrate. Meanwhile, the self-corrosion current densities of EP-GO-x (x = 1, 2, 3) are significantly lower than those of EP-G-x, and the self-corrosion potentials are obviously positively shifted.

(2)

Graphene is dispersed in the epoxy resin system and exhibits a dot-like distribution after curing. In contrast, graphene oxide can bond with epoxy resin molecules through chemical bonds and be embedded in the cured epoxy matrix. The corrosion types of the coated magnesium alloy mainly include uniform corrosion, pitting corrosion and intergranular corrosion, and the coating itself does not change the corrosion mechanism of the magnesium alloy. Stress concentration at the tip of microcracks is the main cause of coating failure.

(3)

Graphene composite coatings primarily protect magnesium alloys through physical barriers, while graphene oxide protects magnesium alloys through both physical barriers and potential chemical reactions.