Insight into the Influence of Cu on the Corrosion Mechanism of 1%Ni Weathering Steel in a Simulated Containing NaCl Atmospheric Environment

Abstract

It is difficult for traditional weathering steel (WS) to form a protective rust layer to withstand the chloride ions and high humidity. Hence, there is an urgent need to develop a new type of low-cost WS with excellent corrosion resistance in the containing NaCl environment. This study aims to determine the influence of Cu on the corrosion mechanism of 1%Ni WS in simulated containing NaCl atmospheric environments. By increasing the Cu content (0.15–0.55%), the corrosion resistance of WS is enhanced. The increasing Cu content promotes the formation of a stable and compact rust layer, significantly enriching the proportion of α-FeOOH to equip the rust layer with a physical barrier. The formation of CuO deposits in the holes and cracks make the rust layer more compact and uniform. The increased Cu content promotes the formation of CuFeO₂ and increases the content of NiFe₂O₄. The formation of CuFeO₂ and NiFe₂O₄ then equips the rust layer with a chemical barrier. Hence, the addition of Cu could enhance the resistance of 1%Ni WS to containing NaCl atmospheric environments.

1. Introduction

WS is a family of low-alloy structural steels, which has more corrosion resistance than typical carbon steel and is widely used in the construction of bridges, railways, carriages, containers, etc., and in rural, urban and industrial environments [1]. Recently, with the continuous development of marine engineering, the demand for weathering steel (WS) is increasingly gaining importance [1,2]. Traditional WS can be applied without paint in mild atmospheric environments. However, it is difficult for traditional WS to form a protective rust layer to withstand the chloride ions and high humidity [3,4]. Hence, advanced WS was developed to resist various marine atmospheres. Recently, it has been proven that 3%Ni WS can be applied paint-free in most marine atmospheres in Spain and Japan [5].

Ni-advanced WS forms a more compact and uniform rust layer than traditional WS in marine atmospheres [3,6]. Since the development of WS containing Ni, lots of studies have reached a consensus: the addition of Ni to WS is beneficial to the evolution of the compact rust layer, which forms a protective layer with physical and chemical barriers. It is believed that Ni has two effects on enhancing the corrosion resistance of WS [5,7,8]: (1) Ni can form NiFe₂O₄ in the rust layer, which can facilitate the nucleation of stable phases to form nanoscale α-FeOOH, causing the rust layer to form a more compact physical barrier; (2) Ni induces cation selectivity in the rust layer, making the resulting compact film an effective barrier to Cl⁻ ions, and making the rust layer act as a chemical barrier. However, increasing the content of Ni seems to be the most effective and fastest way to improve the corrosion resistance of WS. Cheng et al. [9] indicated that when the amount of Ni is higher than 3%, the corrosion resistance of WS does not significantly improve. The corrosion resistance of WS with a Ni content of more than 3.5% is similar to that of WS with a Ni content of 3%. In addition, due to the high cost of Ni and commercialization, researchers have focused on reducing an appropriate amount of Ni and other corrosion-resistant compounds added to improve the corrosion resistance of WS.

In addition, some refs. [9,10,11] discovered that adding about 1% Ni to WS also resulted in good corrosion resistance in coastal atmospheric environments. Xu et al. [11] found that the corrosion resistance of WS can be effectively improved by adding Al and Mo to 1%Ni WS. Hence, the addition of other corrosion-resistant elements could feasibly improve the corrosion resistance of 1%Ni WS. Some studies [12,13] have illustrated that a small amount of Cu added to WS can reinforce the effect of Ni in WS. Hao et al. [14] found that Cu can form CuFeO₂ in the rust layer to enhance the corrosion resistance of MnCuP WS. Kimura et al. [15] found that Cu, in the form of CuO, can jointly promote the formation of nanonetwork Fe(O,OH)₆ with NiFe₂O₄. Carius et al. [16] showed the effect of Cu(Ⅱ) ions on the formation of α/γ-FeOOH and that Cu can be transformed into nano metallic copper, which accumulated on the rust layer, reducing the corrosion rate. In conclusion, the addition of 1%Ni and Cu is helpful for enhancing the corrosion resistance of WS. Therefore, the synergistic effect of Cu and Ni can be explored to develop a new type of low-cost WS with excellent corrosion resistance in the containing NaCl environment.

In this work, a wet–dry cycling corrosion test was carried out to simulate containing NaCl atmospheric environments to study the function of different contents of Cu-alloying elements in 1 wt% Ni WS. A scanning electron microscope (SEM) was used to observe the cross-section of the rust layer. X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS) were used to study the rust layer’s phase composition. The electrochemical behavior of the WS after different exposure times in the simulated containing NaCl atmospheric environment was investigated by EIS and potentiodynamic polarization. In conclusion, the influence of Cu on the corrosion mechanism of 1%Ni WS in simulated containing NaCl atmospheric environments was determined.

2. Materials and Methods

2.1. Sample Preparation

This study used a new type of 1%Ni weathering steel, which was melted using a 100 kg vacuum induction furnace. The chemical compositions of three 1%Ni steels with different Cu contents are shown in

Table 1. Chemical compositions of the test steel (wt.%).

Samples	C	Si	Mn	P	S	Ni	Mo	Cu	Fe
Q235	0.150	0.22	0.64	0.003	0.003	-	-	-	Balance
0.15Cu	0.045	0.21	0.91	0.006	0.008	1.21	0.28	0.14	Balance
0.35Cu	0.046	0.21	0.89	0.003	0.003	1.23	0.30	0.35	Balance
0.55Cu	0.050	0.22	0.90	0.003	0.003	1.20	0.31	0.58	Balance

. Q235 steel was used as a contract test steel.

2.2. Wet–Dry Cyclic Corrosion Test

According to TB/T 2375-1993, “Test Method for Periodic Infiltration Corrosion of Weathering Resistant Steel for Railway” [17], the test samples needed to be cut to the dimensions of 60 × 40 × 4 mm [18]. For the wet–dry cyclic corrosion test, each group of periodic processing required 4 parallel samples, 3 samples for corrosion kinetics analysis, and 1 sample for corrosion resistance analysis of the corrosion product layer. All samples had to be polished step by step with #150~1500 sandpaper until the surface of the sample was smooth and flat. After polishing, ultrasonic washing was carried out with acetone, dried and labeled with deionized water and alcohol, and stored in a drying box for 24 h until testing. The parameters of wet–dry cyclic corrosion test was shown in

Table 2. Parameters of wet–dry cyclic corrosion test.

-	Experimental Parameters
Solution	35 g/L ± 1 g/L (3.5%NaCl solution)
pH	7.5–8.2
Temperature	45 ± 2 °C
Relative humidity	70 ± 5%RH
Cycle time	60 ± 3 min, infiltration time 12 ± 1.5 min
Test period	24 h, 96 h, 168 h, 216 h, 360 h
Reference standard	TB/T 2375-1993

.

The corrosion kinetics were obtained by measuring the weight loss. The test samples were measured by a precision balance to obtain the original weight, W₀. Then, they were immersed in a solution (500 mL HCl + 500 mL deionized water + 3.5 g hexamethylenetetramine) in an ultrasonic cleaner for 600 s, with vigorous stirring at 25 °C to remove the surface corrosion products and obtain the final weights, W_t. This step was needed to ensure accuracy every time. Then, the corrosion rate (R, mg/cm²·h) of the WS was calculated as follows:

$$R={\displaystyle \frac{w_0-w_t}{St}}$$

(1)

where (W₀ − W_t) represents the weight loss of the WS, mg. S represents the exposed area of the WS, cm². t represents the processing time, h.

2.3. Characterization and Analysis of the Corrosion Products

In this experiment, the digital camera was carried out to obtain surface morphologies of four test samples in the simulated containing NaCl atmospheric environment. Rigaku D/max-2500/PC X-ray diffractometer was used to detect the rust layer of four kinds of test steel by XRD (Rigaku Tokyo, Japan). The test angle under the Cu target was 10–70°, and the method of step scanning was adopted. The test voltage was 40 KV, and the current was 200 mA. The Hitachi S-3400N scanning electron microscope was used to observe the surface morphology of the rust layer and the cross-sections of the four test steels at different corrosion periods (HITACHI, Tokyo, Japan). The voltage and current during the test were 15 KV and 40 mA, respectively. XPS (EscaLab 250xi, Thermo Scientific, Waltham, MA, USA) was employed to investigate the elemental states using an Al Kα X-ray source. The binding energy was corrected by referencing the C1s peak (284.8 eV).

2.4. Electrochemical Measurements

Electrochemical analysis was carried out in a three-electrode system at room temperature (25 °C). A sample size of 10 × 10 × 4 mm was cut from the test steel with wire cutting. Five surfaces, other than the rusty surface, were polished with sandpaper, and the wire was welded on the back of the steel with an electric soldering iron. Those same five sides were sealed in a PVC pipe with epoxy resin as the working electrode. A Pt sheet was used as the counter-electrode, and a saturated calomel electrode (SCE) was used as the reference electrode. The solution used in the electrochemical test was a 3.5% sodium chloride solution. Before the test, in order to ensure the stability of the experimental environment and the surface state of the sample during the test, the sample was soaked in 3.5% NaCl solution for 1 h before the electrochemical open-circuit potential was measured for 600 s. After the open-circuit voltage stabilized, the sample was tested using the electrochemical impedance spectrum. EIS measurements were performed at the OCP, with the voltage perturbation amplitude of 10 mV in the frequency range from 10⁵ Hz to 10⁻² Hz. Potentiodynamic polarization measurements were carried out with a scan rate of 0.001667 V s⁻¹ starting from −1.0 V up to −0.1 V vs. SCE.

3. Results

3.1. Corrosion Kinetics

Figure 1a shows the weight loss of the four test samples in the simulated containing NaCl atmospheric environment. The weight loss of the four test samples increased with each corrosion cycle. The weight loss of Q235 was the highest out of the four test samples. Meanwhile, the weight loss decreased with the increase in the Cu content in the same cycles. In the five cycles (360 h), the weight loss of Q235 was about 200 mg/cm², and the weight loss of the 0.55Cu WS was about 105 mg/cm², which is half as much as that of Q235. This indicates that the addition of Ni and Cu effectively improved the corrosion resistance of the WS. As shown in Figure 1b, the corrosion rates of the four test samples first rose in the initial corrosion cycles (<96 h) and then decreased in the later corrosion cycles (>96 h). Meanwhile, the corrosion rate decreased with the increase in the Cu content in the same cycles, which indicates that the addition of Cu was beneficial for enhancing the corrosion resistance. These preliminary results suggest that increasing the Cu content within a certain range benefits the corrosion resistance of 1%Ni WS in a simulated containing NaCl atmospheric environment.

3.2. Morphological Studies of Corrosion Products

Figure 2 presents the surface morphologies of the four test samples in the simulated containing NaCl atmospheric environment after different exposure times. Wang et al. [19] determined that the surface morphology of WS could quantify the evolution of the protective properties of the rust layer during the corrosion process. Hence, the corrosion resistance of the rust layer could be preliminarily judged by observing the surface morphology. In the early stage of the corrosion process, the surface morphologies of the four test samples were covered in an orange rust layer, which was identified as no protective ability. As the corrosion process continued, the surface morphologies of Q235 and 0.15Cu WS showed that the orange rust layer transformed into a black rust layer with caking, which indicates that the protective ability was poor. However, the surface morphology of 0.55Cu presented a dark brown rust layer with a small amount of caking, which means it was equipped with some protective ability. This phenomenon indicates that increasing the Cu ⁱcontent within a certain range benefits the corrosion resistance of 1%Ni WS, which is very consistent with the corrosion kinetics of the four test samples.

Figure 3 shows the microscopic morphologies of four test samples in the simulated containing NaCl atmospheric environment with different corrosion cycles. In the early corrosion process, lots of cracks and holes existed in the rust layer, which resulted in an easy entrance for corrosion particles to increase the corrosion rate. Meanwhile, feather-shape-like (γ-FeOOH) and rosette-like (β-FeOOH) structures were observed, which are considered open structures that allow for corrosion particles to easily pass through the rust layer, leading to an increased corrosion rate [20,21]. Open structures and cracks cause a high corrosion rate in the early corrosion process. However, compared to the microscopic morphology of Q235, those of 0.35Cu and 0.55Cu WS contained fewer cracks and holes, which led to a lower corrosion rate than that of Q235, as shown in Figure 1. As the corrosion process continued, needle-shaped and cotton-ball-shape structures (α-FeOOH) appeared on the surfaces of the 0.35Cu and 0.55Cu WS. These structures can be dense, impeding corrosion particles by building a physical barrier to decrease the corrosion rate. Hence, in the later corrosion process, Q235 still contained lots of cracks and holes, while the 0.35Cu and 0.55Cu WS formed a physical barrier to decrease the corrosion rate. Meanwhile, it was observed that the holes contained in the 0.35Cu and 0.55Cu WS gradually decreased as the corrosion process continued, which may indicate that the Cu could repair the holes. Hence, this phenomenon indicates that the addition of Cu is beneficial for the formation of α-FeOOH and makes the rust layer more compact, resulting in better corrosion resistance.

Figure 4 shows the cross-sectional microscopic morphologies of the four test samples in the simulated containing NaCl atmospheric environment. As shown in Figure 4, there were many cracks and holes in the rust layer, which shows that it was not equipped with a protective ability in the early corrosion process. As the corrosion process continued, the cross-sectional microscopic morphology of Q235 was still similar to that of the early rust layer. Meanwhile, compared to Q235, the rust layer for 0.35Cu and 0.55Cu WS was more uniform and compact, which formed a physical barrier to decrease the corrosion rate. This is the main reason for the decreased corrosion rate in the later corrosion cycles.

3.3. Composition of the Rust Layer

Figure 5 shows the XRD spectra of the four test samples in the simulated containing NaCl atmospheric environment. As shown in Figure 5, all rust layers were composed of γ-FeOOH, β-FeOOH, α-FeOOH, and Fe₃O₄. Although the composition of the rust layer was the same, the contents of each component were different. The reference intensity ratio method and the Rietveld method were carried out to determine the relative amount of each component [22]. In the early corrosion stage, the rust layer was mainly composed of lots of γ-FeOOH and β-FeOOH and a small number of α-FeOOH and Fe₃O₄. As the corrosion process continued, the contents of α-FeOOH and Fe₃O₄ increased, and the γ-FeOOH and β-FeOOH decreased. In addition, the content of α-FeOOH increased with the increase in Cu content. This indicates that the Cu element could promote the generation of α-FeOOH, which is very consistent with previous studies [23,24].

Kamimura [22] indicated that the ratio of α (α-FeOOH) to γ* (the total content of γ-FeOOH, β-FeOOH, and Fe₃O₄) could be calculated to evaluate the protective ability of the rust layer. A higher α/γ* means that the rust layer has better corrosion resistance. In addition, when the α/γ* is greater than 1, it means that the rust layer has a protective ability. As shown in Figure 6, the α/γ* values of Q235, 0.15Cu, 0.35Cu, and 0.55Cu were 0.15, 0.21, 0.28, and 0.33, respectively. These ratios were all lower than 1, which shows that the rust layer was not equipped with a protective ability. In the later corrosion process, the α/γ* values of Q235, 0.15Cu, 0.35Cu, and 0.55Cu were 0.29, 0.79, 0.91, and 1.41, respectively. This indicates that the addition of Cu and Ni effectively improved the corrosion resistance. In addition, the α/γ* increased with the increasing Cu content, indicating that the protective ability of the rust layer also improved with the increasing Cu content. These results are consistent with the corrosion kinetics.

To further analyze the role of Cu in the evolution of the rust layer during the corrosion process, XPS analysis was conducted to study the Cu 2p and Ni 2p in the corrosion products for 0.55Cu contents WS after 360 corrosion cycles, as shown in Figure 7. As shown in Figure 7, Cu 2p spectra of the 0.55Cu contents WS after 360 corrosion cycles exhibited two fitting peaks at 932.6 ± 0.1 and 933.6 ± 0.1 eV, corresponding to CuFeO₂ and CuO, respectively. The Ni 2p spectrum displayed two fitting peaks at 855.5 ± 0.1 and 858.7 ± 0.1 eV, corresponding to NiFe₂O₄ and Ni(OH)₂, respectively.

3.4. Electrochemical Properties of the Rust Layer

The general behavior of the four test samples (rusted steels) in 3.5% NaCl after different exposure times in the simulated containing NaCl atmospheric environment was investigated by recording the potentiodynamic polarization curves over a wide potential range (Figure 8). Figure 8 shows that the addition of Cu and Ni influences the anodic curves. The anodic and cathodic reactions were mainly controlled by the metal dissolution and the reduction reaction of the corrosion products and oxygen, respectively. The relationship between the current density i and the polarization potential E under Tafel should follow the Butler–Volmer equation, as follows:

$$i=i_{corr}[exp\left(2.303{\displaystyle \frac{△E}{b_a}}\right)-exp(2.303{\displaystyle \frac{△E}{b_c}}\left)\right]$$

(2)

$$△E=E-E_{corr}$$

(3)

where i_corr is the corrosion current density, b_a is the Tafel constant for an anode and b_c is the Tafel constant for a cathode (i_corr, b_a and b_c). These reflect the corrosion mechanism and kinetics and are considered electrochemical corrosion parameters. Hence, this equation can be carried out to obtain the electrochemical parameters, as shown in Figure 8d and

Table 3. Tafel curve parameters with the four test samples.

Samples	Ecorr/V(vs. SCE)	Icorr/μA·cm−2	bc/mV·dec−1	ba/mV·dec−1
24 h
Q235	−0.710 ± 0.008	154.33 ± 4.95	−144.3 ± 0.2	257.0 ± 1.0
0.15Cu	−0.674 ± 0.005	104.35 ± 6.12	−201.6 ± 0.6	147.7 ± 0.6
0.35Cu	−0.659 ± 0.009	87.46 ± 5.79	−236.9 ± 0.2	133.8 ± 0.5
0.55Cu	−0.646 ± 0.007	65.31 ± 3.24	−259.0 ± 0.5	127.2 ± 0.5
96 h
Q235	−0.732 ± 0.005	103.86 ± 7.34	−118.3 ± 0.3	130.0 ± 0.6
0.15Cu	−0.655 ± 0.008	89.38 ± 7.22	−125.1 ± 0.7	122.2 ± 0.7
0.35Cu	−0.623 ± 0.011	65.68 ± 7.13	−178.5 ± 0.4	132.6 ± 0.5
0.55Cu	−0.623 ± 0.009	51.23 ± 4.59	−121.6 ± 0.2	194.1 ± 0.2
360 h
Q235	−0.667 ± 0.004	78.32 ± 8.01	−136.7 ± 0.5	132.8 ± 0.4
0.15Cu	−0.555 ± 0.011	67.84 ± 6.06	−125.3 ± 0.2	254.2 ± 0.4
0.35Cu	−0.541 ± 0.005	58.63 ± 8.34	−148.5 ± 0.1	217.3 ± 0.5
0.55Cu	−0.541 ± 0.011	32.87 ± 9.22	−172.7 ± 0.1	194.9 ± 0.3

[25,26]. Meanwhile, the cathodic process is controlled by the reduction of rust and dissolved O₂, and the anodic process is controlled by the metal dissolution. It was found that with the increase in the Cu content and the continuation of the corrosion cycle, the corrosion potential of the rust layer gradually became positive, and the corrosion current density gradually decreased. This indicates that the corrosion resistance of the rust layers was gradually enhanced. The results of the XRD analysis show that the composition of each phase in the rust layer was different; the higher the Cu content, the higher the α-FeOOH content. Hedenstedt et al. [27] investigated the kinetics of the hydrogen evolution reaction for α-FeOOH and γ-FeOOH. The results show that the corrosion potential of α-FeOOH is more positive than that of γ-FeOOH. Thus, it can be considered that the corrosion potential shifted positively with the increase in the Cu content. This suggests that the corrosion resistance of the rust is enhanced as the Cu content of the 1%Ni WS increases. It can be seen from the polarization curve that increasing the Cu content in the 1%Ni WS is beneficial to improving the corrosion resistance of corrosion products.

In order to further evaluate the electrochemical mechanism of the rust layer, the EIS was carried out, as shown in Figure 9. The figure shows that the Nyquist diagram consists of two incomplete semi-arcs, namely the capacious reactance arc in the high-frequency region and the diffusion mass transfer arc in the low-frequency region. With the continuation of the corrosion process, the arc radii of the four test samples increased gradually, and in the different corrosion stages, the arc radii of the test steels with more Cu content were greater than those of the test steels with less Cu content. The above results show that with the extension of corrosion time, the charge transfer resistance of the rust layer of the WS increased continuously, and the corrosion resistance of the product layer increased continuously. The test steel containing Cu, especially the 0.55Cu WS, has more obvious improvement in the densification degree and corrosion resistance of the corrosion product layer. Figure 9(a2–c2,a3–c3) show the Bode diagrams of the four test samples with different corrosion periods, which are mainly composed of impedance mode diagrams and phase angle diagrams. Due to the porosity, mass transfer, and relaxation effects in the corrosion products, the maximum phase angle measured was less than 90°, which is not an ideal capacitor [21,28].

Based on the EIS test results, a fitting equivalent circuit diagram of the electrochemical impedance spectrum was proposed, as shown in Figure 10. Qrust and Rrust represent the constant phase components and resistors of the rust layer, respectively. Rs, Qdl, Rct, and W₁ represent the solution resistance, constant phase element of the double-layer capacitor, charge transfer resistance, and the Warburg impedance, respectively [29]. The Warburg impedance is always recognized as the diffusion of dissolved O₂ for the process of mass transport which meaning to equipment the phase compositions of the rust layer. This phenomenon has been extensively researched on the generation of rust layer for WS, where the mass transport of the rust layer is restricted by the diffusion of oxidation products. Hence, it is reasonable for a straight line to appear with a slope approaching 1 in the low-frequency area due to limitations in the diffusion of oxidation products [21]. The smaller value of W1 indicates the rust layer is equipped with an excellent phase composition.

The corrosion parameters obtained from the EIS for the rust layers in different corrosion periods are shown in

Table 4. The fitting results of the EIS curves of the four test samples after different corrosion cycles.

Samples	Cycles (h)	Rs (Ω·cm2)	Qrust (Y0) × 10−3 (Ω−1·cm−2·sn)	nrust	Rrust (Ω·cm2)	Qdl (Y0) × 10−3 (Ω−1·cm−2·sn)	ndl	Rct (Ω·cm2)	W1 (Ω−1·cm2·s0.5)	χ2 (×10−5)
Q235	24	44.72	14.37	0.33	35.75	17.33	0.29	30.11	0.79	3.24
	96	49.11	25.67	0.54	44.79	39.47	0.47	60.00	0.65	5.31
	360	57.24	40.47	0.53	59.47	50.11	0.53	71.22	0.51	3.14
0.15Cu	24	98.47	17.53	0.52	51.47	12.79	0.44	70.39	0.76	5.42
	96	134.85	29.47	0.53	55.79	23.57	0.45	88.47	0.64	1.31
	360	182.14	58.71	0.67	72.31	39.27	0.60	85.32	0.48	1.10
0.35Cu	24	84.21	22.49	0.54	49.72	11.11	0.45	79.49	0.77	2.41
	96	79.14	28.79	0.67	68.14	19.89	0.55	90.10	0.59	2.11
	360	90.08	53.48	0.67	88.79	21.37	0.66	90.31	0.44	3.14
0.55Cu	24	62.71	23.47	0.45	62.77	13.85	0.43	88.71	0.75	2.79
	96	75.58	54.57	0.71	73.81	20.11	0.60	130.47	0.55	9.57
	360	93.42	69.97	0.76	92.31	24.37	0.72	139.49	0.32	5.41

, and the fitting errors were less than 10⁻⁴. R_S refers to the physical impedance in addition to the reaction impedance in the entire test system, including the impedance of the electrolytic cell system, the contact impedance generated by the connection with the electrochemical workstation, the impedance of the wire, etc. Hence, the changes of R_s may be caused by the change in the number of ions in the solution. Because the surface of the rust layer is uneven and rough and porous, the double electric layer capacitor cannot become an ideal capacitor element. Therefore, the constant phase components Qrust and Qdl are used to represent the double electric layer capacitor, which usually represents the surface roughness of the corrosion system, mainly due to the uneven surface caused by the corrosion products generated by the complex electrochemical corrosion reaction. The N value, which represents the phase shift, characterizes different surface phenomena, such as surface heterogeneity resulting from surface roughness, dissolution of the metal, impurities, distribution of the active sites, inhibitor adsorption or porous layer formation. As shown in Table 4, the n values are all less than 1, which demonstrates that the rust layer is very heterogeneous. It is clear that both R_rust and Rct increased with the longer corrosion time, revealing that the resistances of the corrosion product films (barrier properties) to the charge transfer were enhanced, and the resistance of the rust layer was increased. This indicates that the corrosion resistance of the rust layer was improved, which hinders the invasion of the corrosion ions [7]. The values of Rct and Rrust of the Cu-containing WS were higher than that of Q235. The results show that Cu can improve the resistance of the corrosion product film and the charge transfer resistance of 1%Ni WS.

4. Discussion

Refs. [4,5] reported that Ni can effectively improve corrosion resistance in the marine atmospheric environment when the content of Ni is greater than 3%. Ni reacts with H₂O and O₂ in the air to form Ni(OH)₂. A part of Ni(OH)₂ is hydrolyzed to form NiO, and another part of Ni replaces Fe²⁺ in Fe₃O₄ to form NiFe₂O₄, as shown in Figure 11. Nickel ferrate is a substance that promotes the spontaneous conversion of γ-FeOOH to a stable phase of α-FeOOH at the nanoscale. The chemical reactions involved are as follows [30]:

$$2\mathrm{Ni}+{\mathrm{O}}_2+{2\mathrm{H}}_2\mathrm{O} \text{→}{2\mathrm{Ni}\left(\mathrm{OH}\right)}_2$$

(4)

$${\mathrm{Ni}\left(\mathrm{OH}\right)}_2\text{→}\mathrm{NiO}+{\mathrm{H}}_2\mathrm{O}$$

(5)

$${\mathrm{Ni}\left(\mathrm{OH}\right)}_2+2{\mathrm{Fe}\left(\mathrm{OH}\right)}_2\text{→}\mathrm{Ni}{\mathrm{Fe}}_2{\mathrm{O}}_4+{3\mathrm{H}}_2\mathrm{O}$$

(6)

Some researchers have shown that Ni²⁺ can replace Fe²⁺ in the Fe₃O₄ octahedral structure to form a NiFe₂O₄ phase [31]. It provides many sites for the nucleation of γ-FeOOH and α-FeOOH, promotes the spontaneous conversion of γ-FeOOH to the nanoscale α-FeOOH stable phase, and can form an electronegative rust layer inside the corrosion product layer to prevent the intrusion of Cl⁻, thus improving the corrosion resistance of WS. Therefore, the addition of Ni to WS could have stronger corrosion resistance than ordinary weathering steel in the marine atmosphere environment. Considering the cost factor, this method could appropriately reduce the content of Ni, but the corrosion resistance of the WS does not satisfy the requirements in the marine atmospheric environment.

The above experimental results show that it is feasible to add Cu to 1%Ni WS to enhance its corrosion resistance in the containing NaCl atmospheric environment. The morphology results of the corrosion products confirm that the addition of Cu effectively reduced the diameters of the holes and cracks, indicating that Cu could repair the rust layer, as shown in Figure 3. As refs. [24,32] reported, the Cu reacts with H₂O and O₂ in the air to form Cu(OH)₂. A part of Cu(OH)₂ is hydrolyzed to form CuO. The generation of CuO plays a key role in improving the corrosion resistance of WS. The Schematic diagram of corrosion mechanism of 1%Ni WS containing Cu is shown in Fig.11. Firstly, it is deposited in the holes and cracks, making the rust layer more compact and uniform, which could impel the entrance of the corrosion particles to form a physical barrier. Additionally, the CuO provides nucleation sites for amorphous oxyhydroxide to promote nano-sized oxyhydroxide network, which not only traps the corrosion particles, but also facilitates the formation of α-FeOOH, leading to the rust layer forming a physical barrier with a protective ability.

In addition, as shown in Figure 1, the higher the Cu content is, the lower the corrosion weight loss. In the later corrosion stages, parts of the initial corrosion products (FeOOH) transform into Fe₃O₄. Meanwhile, the content of Fe₃O₄ is relatively reduced under the conditions of high Cu content, which indicates that Cu reacts with Fe₃O₄. As ref. [32] indicated, the higher Cu content could be attributed to the formation of CuFeO₂, which is generated by one Cu⁺ replacing one Fe³⁺ in the spinel oxide Fe₃O₄. The formation of CuFeO₂ plays a similar role as NiFe₂O₄ in the rust layer of WS. The CuFeO₂ gives the rust layer an electronegative layer, and then impels the Cl⁻ to close to the WS. Moreover, the Gibbs free energy of CuFeO₂ (−479,900 J/mol) is higher than that of Fe₃O₄ (−101,545 J/mol), which means that CuFeO₂ is more thermodynamically stable than Fe₃O₄ [33,34]. Hence, a Cu addition could promote the formation of CuFeO₂, causing the rust layer to form a chemical barrier and impelling the Cl⁻ to close the WS, thus enhancing the corrosion resistance of the WS. Overall, adding a small amount of Cu can make up for a lack of corrosion resistance due to the reduction in Ni content. A rust layer could form with physical and chemical barriers to enhance the corrosion resistance in the marine atmospheric environment.

5. Conclusions

The addition of Cu to 1%Ni WS effectively improves the densification and stability of the rust layer, and the corrosion rate decreases with an increase in Cu content. The 0.55Cu WS showed better corrosion resistance in a simulated containing NaCl atmosphere. The conclusions are as follows:

(1)

An increasing Cu content promotes the formation of a stable and compact rust layer, significantly enriching the proportion of α-FeOOH to equip the rust layer with a physical barrier.

(2)

The formation of CuO deposits in the holes and cracks make the rust layer more compact and uniform. This provides nucleation sites for amorphous oxyhydroxide to promote a nano-sized oxyhydroxide network, which not only traps the corrosion particles but also facilitates the formation of α-FeOOH, leading to the rust layer forming a physical barrier with a protective ability.

(3)

Cu promotes the formation of CuFeO₂, and the content of NiFe₂O₄ increases with an increase in Cu content. The formation of CuFeO₂ and NiFe₂O₄ equips the rust layer with a chemical barrier.