Rectangular blocks (40 mm × 20 mm × 5 mm) of Nd-Fe-B commercial magnets (N28) were studied, all of them with a nickel electrodeposit, Figure 1.

The magnet blocks were studied under magnetized (salt spray tests) and under demagnetized conditions (microstructure characterization, salt sprays tests and electrochemical tests). Two demagnetizing processes were used, heat treatment at 335˚C (HT) for one hour and under an opposite magnetic field (OMF).

The magnets microstructure characterization was made by a scanning electron microscopy-SEM (FEG-Field Emission Gun, FEI manufacturer, Quanta model 400F and X-EDS energy dispersion spectroscopy) before and after the salt spray and the electrochemical tests.

The magnets were studied by electrochemical tests. For this, demagnetized blocks were cut to obtain a cube with 1.0 cm of an uncoated side. This cube was embedded in Bakelite and the uncoated surface of 1.0 cm² was polished using diamond paste (1 μm) prior to the electrochemical tests. A conventional electrochemical system with three electrodes was used: the uncoated magnetic alloy as the working electrode, a saturated calomel electrode as a reference electrode and a rectangular platinum gauze as an auxiliary electrode. A solution containing 0.05 mol/L NaCl, pH = 6.7, was used as an electrolyte.

Open potential circuit measurements (OPC curves) for 10,600 s (3 h) and potentiodynamic polarization curves (PPC) were made. For PPC, a potentiostat/galvanostat PAR model 2273 was used, adopting a scanning rate of 0.166 mV/s. The anodic potentiodynamic polarization curves (APPC) started at the OCP up to −0.300 V/ECS and the cathodic potentiodynamic polarization curves (CPPC) swept from the OCP down to ‑1300 V/ECS. The tests were conducted at room temperature (23 ± 2)˚C.

Salt spray tests were done positioning three magnetized and three demagnetized samples (only triple-layer Ni coating) in a chamber in which a 5% NaCl solution was continuously sprayed at (35 ± 2)˚C for 72 h. The pH of the solution was between 6.5 and 7.2 and the relative humidity was 100%.

Figure 2 presents magnet SEM images of demagnetized samples by HT and by OMF, the gray areas are related to ferromagnetic phases (f) while the white areas are related to RE-rich phases.

In Table 1, we present the EDS obtained results in red highlighted areas in Figure 2(a) and Figure 2(b).

For both characterized samples, the RE content (Nd + Pr) was 29% and the iron content was 66% (Table 1). These values are close to the values of the alloys (Nd-Pr)-Fe-B observed in the literature [7] .

The HT demagnetizing should generate some cracks and holes in the alloy

microstructure (Figure 2(a)). In view of this, the demagnetizing by HT was abandoned. Therefore, salt spray and electrochemical tests were done with OMF treated samples.

Nickel coatings were analyzed by SEM-FEG and two types of coatings were observed: single layer (HT demagnetized) and triple layer (OMF demagnetized). SEM-FEG images are presented in Figure 3 in which the thicknesses of the layers were highlighted.

Figure 3(a) and Figure 3(b) images show intergranular corrosion inside the substrate near the alloy/coating interface for both kinds of nickel coatings. This corrosion probably occurred during the electrodeposition step and can affect the nickel coating adherence. Therefore, electrodeposition baths should be selected in such a way as to use formulations that do not damage the surface of the magnets to avoid the occurrence of intergranular corrosion, since Nd phase-rich is attacked.

The EDS analyses showed that the single layer coating was composed of only nickel (Figure 3(a)) and the triple layer coating composed of nickel (C), copper (B) and nickel (A) (Figure 3(b)). Table 2 shows the thickness values of the nickel coatings.

The triple layer coating was twice the thickness of the single layer nickel coating (Table 2). The triple layer coating frequently produces better protection than the single layer coating due to the increased thickness and also because of the overlapping layers. The triple layer coating will failure if an aligned damage occur in these three layers.

The electrochemical results are shown in Figure 4 and Figure 5. It is possible to see that the demagnetized sample (OMF) presented a typical active state behavior with the values of OCP falling during the first 1000 s of exposure to the solution. After this period, the sample surface stabilized and reached the stable potential of −0.78 V/SCE at the end of the 10,600 s (Figure 4).

The anodic potentiodynamic polarization curve (APPC) showed an increase of the current density indicating a sample active-state behavior during the scanning within the tested potential range (Figure 5).

The cathodic potentiodynamic polarization curve (CPPC) presented constant values of current densities between 10 − 5 A/cm² and 10 − 4.5 A/cm² within the potential range from −0.85 V/ECS to −1.10 V/ECS. This can be ascribed to the reduction of dissolved oxygen at the metal/electrolyte interface. From −1.10 V/ECS to −1.30 V/ECS, the current densities increased due to H3O⁺ (H⁺) ions reduction generating H₂ gas (Figure 5).

The samples were analyzed by SEM-FEG after both anodic and cathodic polatizations. The images are presented in Figure 6.

It is possible to see that the surface of the sample suffered strong intergranular corrosion after APPC. After the CPPC, the surface was roughened due to the period of exposure to the solution during the OCP monitoring which preceded the cathodic polarization.

The images of the samples before and after the salt spray tests are presented in Table 3 (magnetized samples) and Table 4 (OMF demagnetized samples).

The three magnetized samples suffered a slight superficial attack in the first 24 h of exposure. On the second day, there was one localized corrosion point on sample 2. After 72 h, two samples presented localized corrosion (samples 2 and 3) and sample 1 presented only spots on its surface, Table 3.

The three demagnetized samples suffered localized corrosion in the first 24 h of exposure. Although they were already corroded, the samples were kept in the chamber to verify whether other corrosion points might appear, but no more points occurred.

One of the magnetized samples was demagnetized (OMF) and cut transversely for SEM-FEG analyses of the cross section after salt spray tests (Figure 7).

Salt spray test made a damage on the substrate through discontinuities of the nickel coating triple layer. This corrosion occurred between the grains of the magnetic alloy in the rich RE phase, Figure 7.

Although the SEM images show intergranular corrosion, the ferromagnetic phase (iron-rich phase) must have also been corroded because it was possible to see red corrosion products on the samples after the immersion tests, Table 3 and Table 4.

Multilayer coatings aim at offering a great substrate protection because they present low probability of a substrate exposure. For the substrate corrosion to occur, the coating discontinuities of each coating layer must be aligned to expose the substrate [9] . During the salt spray tests, not all the samples suffered localized

corrosion, others were superficially stained. This fact is justified by the mentioned intrinsic characteristic of the multilayer nickel coating.