*2.1. Sample Preparation*

In order to eliminate the influence of other alloying elements, ultra-low carbon steel was used as a raw material in this study. Copper-bearing steels were prepared in a highfrequency vacuum induction furnace.

Samples with a side length of about 50 mm and a thickness of around 10 mm were cut from each ingot. The surface of the samples were polished smooth, and multipoint (≥3) composition tests were carried out by direct reading spectrometer (ARL 3460, Thermo Fisher, New York, America), and the average value of the test values was taken as the alloy composition. The chemical compositions (in wt%) of the test samples are listed in Table 1, where sample No.0 is a raw steel. At present, the high content of copper in commercial copper steel is about 3%. In this experiment, two samples with higher content of copper were specially added for comparative study.


**Table 1.** Chemical composition of test steel (in wt%).

Rectangular specimens with the approximate dimensions of 25 mm × 25 mm × 3 mm (with a small hole to hang the sample during the corrosion test) were cut from the ingots, and then mechanically ground using SiC papers in succession up to 800 grit. Afterwards, they were treated by descaling, cleaning, and rinsing in acetone, and then drying.

#### *2.2. Wet-Dry Cyclic Accelerated Test*

Dry-wet cycle accelerated test is a conventional method to test the corrosion resistance of steels at 25 ◦C [12]. This method is used to test the corrosion resistance of test steels with different copper content, and to corrode iron on the surfaces of steels to obtain enriched copper.

Before and after each corrosion test, the specimens were weighed using an analysis balance with an accuracy of 1 mg. The samples were subjected to a periodic immersion wet/dry cyclic corrosion test, which was an accelerated corrosion test under an artificially simulated atmospheric environment. Each cycle of this wet/dry cyclic corrosion test lasted 24 h and included three stages: (1) Put the dried samples one by one on the filter paper in the high-precision balance, and handle the samples gently to prevent the rust layer from falling off, (2) immersing the specimens into a 5% sodium chloride solution with a pH value of 6.5–7.2 (as measured using a digital pH tester) at 25 ◦C for 3 s, and (3) drying the specimens immediately by hanging them in a test chamber with a constant humidity of RH60% (±5%) for 24 h. The total corrosion test lasted 288 h. No direct connection between the different metals occurred during the entire testing process.

#### *2.3. Descaling and Observation*

For all the corroded samples, the rust was removed with citric acid (30 g) and sodium dihydrogen phosphate anhydrous (50 g), which were dissolved in 1 L of deionized water [13]. The pH value of the descaling solution was approximately 1.6–1.7, and the descaling process lasted 24 h at 25 ◦C.

The samples were mounted with a slope, because pure copper is softer than steel and is easy to wear off. The microstructures of the mounted samples were observed using an optical microscope (OM, Axioskop 2 MAT, Carl Zeiss AG, Jena, Germany). A scanning electron microscope (SEM, JSM-6500, JEOL, Tokyo, Japan) equipped with an energy dispersive X-ray spectroscopy (EDS) feature (Quanta FEG250 FEI, Hillsboro, OR, USA) was used to obtain the elemental distribution at different areas of the coating microstructure. X-ray diffraction (XRD D/max-2500) was used to determine the surface phase compositions of the descaled samples. The working voltage is 40 KV, the working current is 40 mA, the radiation target is CuKa, and the scanning speed is 6◦/min. Scanning ranges from 10◦ to 90◦ (2θ). JCPDS (Joint Committee on Powder Diffraction Standards) cards are used as reference to identify crystal phases in XRD patterns. The JCPDS card numbers used in this paper are: Iron 01-089-7194, Copper 01-070-3038, Fe3O4 00-003-0863, Fe2O3 00-004-0755, and FeOOH 01-081-0463.

#### *2.4. Electrochemical Corrosion Test*

The steel samples (with and without Cu coating) were rectangular in shape (10 × 8 × 3 mm3). All samples were suitably embedded in epoxy resin, and the exposed surface was 0.8 cm2. In this study, polarization curves were measured by Autolab electrochemical station (ECO CHEMIE BV BST7276). The samples with an electric contact were used as working electrodes in a standard three-electrode cell, i.e., platinum counter electrode, silver/silver chloride (Ag/AgCl/sat. KCl) reference electrode, and a working electrode. The latter was placed in a capillary in such a way that it remained outside the cell at 25 ◦C, while the capillary tip was immersed in the testing solution close to the working electrode surface.

Before experiment, the working electrode surface was degreased with acetone followed by rinsing with deionized water, and a 3.5% sodium chloride solution with pH value of 6.5–7.0 was used at 25 ◦C. Potentiodynamic polarization were conducted from a potential of −1.0 V to −0.1 V versus Ag/AgCl with a scan rate of 0.05 V/min after achieving a stationary value for the open circuit potential (potential variation not higher than 0.1 mV/s).
