The effect of silica sols on electrodeposited zinc coatings for sintered NdFeB

Abstract

We developed an in situ method to prepare silica particle-containing zinc coatings on NdFeB. Silica sols with silica nanoparticles of 30–150 nm in diameter were prepared through the hydrolysis and condensation of tetraethyl orthosilicate with an acid catalyst. Zinc coatings were prepared on NdFeB magnets using plating baths that contained the silica sols. The effect of the volume content of silica sols on the zinc coatings was studied via scanning electron microscopy (SEM), X-ray diffraction (XRD) and electrochemical analysis. The SEM results showed that the silica nanoparticles were embedded in the zinc particles. Compared with the zinc coatings that were electroplated in the other plating baths, the surface morphology of the zinc coating formed when the silica sol content in the plating bath was in a proper range exhibited better uniformity. The decrease in the corrosion current density value and the increase in the impedance value that were measured via electrochemical testing demonstrated that the embedded silica nanoparticles, which served as corrosion inhibitors, improved the corrosion resistance capability of the zinc coatings on the NdFeB magnets.

---

1. Introduction

Sintered NdFeB permanent magnets have been widely used for many applications in various fields, such as electronics, acoustics, communications, automation, magnetic resonance imaging, and biomedical applications, by virtue of their excellent magnetic properties.^1,2 However, their multiphase microstructure, which improves their magnetic properties, also makes NdFeB magnets vulnerable to corrosion. There are two types of protective coatings that can be applied to sintered NdFeB magnets: cathodic and anodic coatings.^3–8 A cathodic coating protects the sintered NdFeB magnet by segregating the substrate from the corrosive medium. Once the cathodic coating is ruptured and the sintered NdFeB is exposed to the corrosive medium, a large cathodic coupling forms with a small NdFeB anode, which can accelerate the corrosion of the sintered NdFeB magnet. An anodic coating insulates the substrate from the corrosive medium. Even when the anodic coating is damaged and the NdFeB is exposed to the corrosive medium, the anodic coating may act as a sacrificial anode and slow the corrosion of the metal substrate.⁹ The role of an anodic coating in corrosion protection is a combination of mechanical isolation and electrochemical protection. Zinc coatings have been widely used as anodic coatings for NdFeB magnets because of their low cost and ready adaptability to mass production.²

To improve the corrosion resistance of zinc coatings, various strategies focusing on combining zinc with other particles or corrosion inhibitors have been proposed over the past few years.^10–15 Silica and silicate materials have been shown to possess anti-corrosive properties.^16–20 A composite coating of zinc and silica is an effective means of improving the properties of zinc coatings.^21,22 Zinc and silica composite coatings can be prepared from silicate solutions or silica sols by dipping, spinning and electroplating silica films on top of zinc coatings.^2,23,24 It is also possible to obtain zinc and silica composite coatings from silica nanoparticles or a silicon-containing zinc solution using electroplating or sintering methods.^25–29 Electrodeposition is an effective method to obtain silica-containing zinc coatings with better protection performance on steel substrates.³⁰ The silica particles, which become embedded in the zinc coating, contribute to the dimensional stabilization of the protective zinc layer.¹³

Acid-catalyzed silica sols tend to form linear chains, which can be catalyzed to form silica nanoparticles by hydroxyl groups, and the cathodic electroplating process promotes the agglomeration and adsorption of these silica chains on the electrode surface.^31–33 Therefore, it is possible to control the migration and deposition of the silica particles on the surface of the cathode during the electrodeposition of the zinc coating.^34–39

In this work, we demonstrated a one-step in situ method for successfully preparing a uniform silica-containing zinc coating on the surface of an NdFeB magnet via electrodeposition by introducing silica sols into the plating bath. The content of silica sols in the plating bath was varied to investigate the effect of the silica sols on the electrodeposited zinc coatings. The sol–gel process of the acidic silica sols was promoted by the cathodic deposition reaction, leading to the generation of embedded silica nanoparticles. As observed from the morphologies of the zinc coatings formed from the silica sol-containing plating baths, the crystal size and distribution of the zinc particles were notably affected by the addition of silica sols. We also observed the distributions of silica throughout the zinc coatings and the different shapes of the silica nanoparticles between the zinc grain boundaries as the silica sol content in the plating baths increased. The silica nanoparticles embedded in the zinc coatings were found to improve the corrosion resistance of the zinc coatings. Additionally, it was demonstrated that silica sols are an effective additive during the electrodeposition reaction to enhance the uniformity of the zinc coating.

2. Experimental section

2.1 Materials

NdFeB magnets (Jingci Magnet Co, China) of grade 42H (Nd_10.1Tb_3.1Fe_77.4B_9.4) were used as substrates for the deposition of zinc coatings in silica sol-containing plating baths. Tetraethyl orthosilicate (TEOS, Kermel, highly pure, China), ethyl alcohol (EtOH, 99.7%), zinc chloride (ZnCl₂, 98%), potassium chloride (KCl, 99.5%) and HCl (11.9 mol L⁻¹) were used as received without any further purification.

2.2 Synthesis of the silica sols

Silica sols were prepared via the acid-catalyzed reaction of tetraethyl orthosilicate. In the synthesis procedure, TEOS and ethanol were added to a mixture of deionized water, acid catalyst and ethanol, followed by stirring at 30 ± 1 °C for 24 h using a magnetic stirrer (RT5, IKA) at the speed of 240 rpm. The pH value of the mixture was adjusted to 4.6 through the addition of 1 mol L⁻¹ HCl solution, which served as the acid catalyst and stabilizer. The mixtures were aged at 30 ± 1 °C for another 24 h.

2.3 Electrodeposition

For the electrodeposition process, the silica sols were added to acidic zinc plating baths (ZnCl₂ 70 g L⁻¹, KCl 200 g L⁻¹). Silica sols were uniformly dispersed under stirring at 25 ± 1 °C for 10 min using RT5 at the speed of 300 rpm. The volume fraction of silica sols in each plating bath varied from 0 to 25%; specifically, contents of 0, 5, 15 and 25% were tested. The pH of all plating baths was adjusted to 4.5 before electrodeposition. The substrate material consisted of well-cleaned NdFeB magnets. The electrodeposition process was conducted at a current density of 1.2 A dm⁻² at room temperature.

2.4 Characterization

The particle distributions and surface roughness of the zinc coatings were investigated using field emission scanning electron microscopy (FESEM, JSM 7500). X-ray diffraction (XRD) analysis was performed using a Rigaku D/MAX-RB with Cu K radiation (K = 0.15406 nm), which was used to analyze the zinc grain growth. The zeta potential of the silica particles was calculated using an electrophoretic light scattering spectrometer (Zeta Sizer, Model 3000HS, Molvern instrument Ltd.). A CHI604A electrochemical workstation was used to perform electrochemical dynamic polarization and electrochemical impedance spectroscopy (EIS) in 3.5 wt% NaCl solution at room temperature. The tests were performed using a standard three-electrode system with platinum foil and a saturated calomel electrode as the counter and reference electrodes, respectively. The polarization measurements were carried out with a scanning rate of 1 mV s⁻¹. The electrochemical impedance spectroscopy data were obtained under potentiostatic control with a signal amplitude of 5 mV in the frequency range of 10⁵ Hz to 1 mHz. The samples were stabilized in 3.5 wt% NaCl solution for 30 minutes before EIS tests to ensure potential stability, and the open circuit potential was used as the initial potential. The electrochemical tests were performed at least three times to ensure the reproducibility of the results.

3. Results and discussion

Fig. 1 shows several representative low-magnification SEM images of the zinc coatings formed from the silica sol-containing plating baths (SSPBs) on the NdFeB magnet samples. It is evident that the silica sols significantly modified the size and distribution of the zinc particles via the adsorption effect. As shown in Fig. 1(A), the surface of the zinc coating produced from the silica sol-free plating bath was very heterogeneous, and had a broad size distribution of the zinc particles. When the zinc was deposited without silica sols, the quantity of nucleation points was lower, and the particle size was larger. The zinc particles aggregated on the substrate, and the grain boundaries of the zinc particles were remarkable. Fig. 1(B) shows the surface of the zinc coating obtained from the 5% SSPB, from which it is evident that the zinc particle size decreased considerably with the addition of silica sols. It was observed that the adsorption of silica sols limited the growth process of the zinc particles and generated more nucleation sites than were observed in the sample shown in (A). The inset in Fig. 1(B) shows that some silica particles aggregated and adsorbed on the surfaces of the zinc particles as well as at the grain boundaries. Fig. 1(C) shows the surface of the zinc coating obtained from the 15% SSPB. The zinc particle size was further reduced in comparison with that in (A) and (B), and the zinc coating evenly covered the NdFeB substrate. There was no zinc grain with an obvious geometric outline on the surface. Several silica sol particles were present on the surfaces of the zinc particles, and several embedded silica inclusions were distributed between the zinc grain boundaries. The inset shows that the adsorbed silica sols caused the formation of more nucleation sites and changed the crystallization direction of the zinc particles, which became extruded before they could grow to form typical hexagonal particles. In addition, the adsorbed silica nanoparticles could become wrapped with the deposited coating during the growth process of the zinc particles. When the silica sol content reached 25%, as shown in Fig. 1(D), the zinc particle size was further reduced. The silica particles that were adsorbed at the zinc grain boundaries were more visible, as seen from the inset in Fig. 1(D). Thus, as the silica sol content was increased from 0 to 25%, the silica nanoparticles became more uniformly distributed on the surface, whereas the agglomeration size of the silica decreased. The number of adsorbed silica nanoparticles that were embedded in the zinc grains also varied. At the same time, the size, number and crystallization direction of the zinc particles were also significantly affected by the increase in the silica sol content.

Fig. 1 Representative SEM images of (A) the zinc coating obtained from the 0% SSPB, (B) the zinc coating obtained from the 5% SSPB, (C) the zinc coating obtained from the 15% SSPB, and (D) the zinc coating obtained from the 25% SSPB. The insets are the corresponding high-resolution images.

Fig. 2(A–D) show representative microstructural features of the zinc particles and grain boundaries in the zinc coatings formed from the plating baths with a silica sol content of 0, 5, 15 and 25%, respectively. As shown in Fig. 2(A), there were no silica nanoparticles at the zinc grain boundaries in the coating formed from the 0% SSPB. The zinc particles followed the crew dislocation growth mechanism, forming hexagonal grains. Fig. 2(B) shows the zinc coating grain boundaries in the coating formed from the 5% SSPB. Some silica agglomerates were adsorbed at the zinc grain boundaries. As shown in Fig. 2(C), in the coating formed from the 15% SSPB, the silica particles adsorbed on the zinc grains and filled the gaps between the zinc particles. There were also some visible silica nanoparticles embedded in the zinc grains. Fig. 2(D) shows that compared with the samples at a lower silica sol content, the agglomeration size of the adsorbed silica decreased in the coating formed from the 25% SSPB. This image also shows several silica nanoparticles with diameters below 100 nm, which were evenly adsorbed on the crystal plane of the zinc grains. Moreover, from Fig. 2, it is clear that the crystal orientation of the zinc grains gradually changed with the increase in the silica sol content.

Fig. 2 Representative SEM images of the zinc grain boundaries in (A) the zinc coating obtained from the 0% SSPB, (B) the zinc coating obtained from the 5% SSPB, (C) the zinc coating obtained from the 15% SSPB, and (D) the zinc coating obtained from the 25% SSPB.

As shown in Fig. 3, energy dispersive spectrometry (EDS) analyses were conducted to evaluate the distribution of the silica phase. Fig. 3(A) shows that in the coating produced from the 5% SSPB, some portions of the zinc grain boundaries were covered by aggregated silica nanoparticles. However, the distribution of the silica phase was not uniform, and several zinc grain boundaries lacked coverage by silica nanoparticles. As shown in Fig. 3(B), when the silica sol content of the SSPB reached 15%, more silica nanoparticles aggregated at the zinc grain boundaries, and the silica phase almost completely covered the zinc grains. Fig. 3(C) shows the zinc coating obtained from the 25% SSPB. The silica phase was uniformly adsorbed on this zinc coating. In the electrodeposition process, the electric charge tended to accumulate more on the zinc grain boundaries. For this reason, the strength of the electric field on the grain boundaries is stronger than that on the grain surfaces. The hydroxyl-rich silica particles are more likely to aggregate at the boundaries of the zinc grains. Because the outline of the silica phase distribution was more likely to form a hexagon, which is the typical geometrical shape of zinc grains, it was concluded that the silica nanoparticles were more likely to aggregate at the zinc grain boundaries in this case.

Fig. 3 EDS analyses of the zinc coatings obtained from (A) the 5% SSPB, (B) the 15% SSPB, and (C) the 25% SSPB.

Fig. 4(A–D) show the cross sections of the zinc coatings formed in the plating baths with the various silica sol content. To distinguish the substrates from the zinc coatings, the polished specimens were etched with ammonia and then ultrasonically cleaned in alcohol. The spectra of the liner EDS analyses, which scanned following the yellow arrows in the SEM images, show the distribution of different elements. Fig. 4(A) shows the zinc coating obtained from the 0% SSPB. The zinc grain boundaries could be clearly observed after etching. Fig. 4(B) shows that some aggregated silica nanoparticles, which could be distinguished from the zinc grains by their different contrast, were embedded in the zinc coating formed at a 5% silica sol content. However, the EDS analyses also show that the distribution of the silica phase was not uniform. As shown in Fig. 4(C), the agglomeration of the embedded silica nanoparticles decreased when the silica sol content was increased to 15%, and the silica phase was more uniformly distributed. Fig. 4(D) shows that when the silica sol content reached 25%, the embedded silica nanoparticles again aggregated more strongly near to the NdFeB substrate. The relative amount of silicon also increased in comparison with the other samples. In addition to the uniformly distributed silica nanoparticles, there existed several large-sized silica particles, which could exert a notable effect on the zinc coating’s performance. In summary, Fig. 4 demonstrates that aggregation of the silica nanoparticles embedded in the zinc coating occurred when the silica sol content in the SSPB was either too high or too low.

Fig. 4 Cross sectional views and EDS analyses of the zinc coatings obtained from (A) the 0% SSPB, (B) the 5% SSPB, (C) the 15% SSPB, and (D) the 25% SSPB.

To measure the effect of silica sols on the deposition process, the polarization curves of bare NdFeB cathodes were tested in plating baths with a different silica sol content. Fig. 5(A) shows the polarization curves recorded at the beginning of the electrodeposition process. From Fig. 5(A), it is apparent that the polarization curves exhibited a linear relationship at low current densities. At a fixed overpotential (η), the current density decreased with the increase in the silica sol content from 0 to 25%. The introduction of an increasing amount of silica sols led to the adsorption of more silica particles on the cathode surface. The results presented in Fig. 5(A) indicate that the adsorption of silica particles occurred at the beginning of the electrodeposition process. The adsorption of more silica particles caused a higher overpotential on the cathode surface. The zeta potentials of the silica particles from the 5% SSPB, 15% SSPB and 25% SSPB were −6.8 ± 0.4 mV, −4.4 ± 0.3 mV and −3.6 ± 0.3 mV. The silica particles may be covered by cations, like K⁺ and Zn²⁺, and moved towards the cathode under an electric field. Silica particles with a lower absolute value of the zeta potential more easily aggregated.⁴⁰ The adsorption process of the silica particles more easily occurred in the 25% SSPB. So a zinc coating obtained from a higher silica sol content plating bath has more adsorbed silica particles.

Fig. 5 Polarization curves of the NdFeB cathodes in plating baths with a different silica sol content: (A) in the low current density regime and (B) the overall curves.

As shown in Fig. 5(B), when the current density increased above 2 × 10⁻³ A cm⁻², the concentration polarization took on the dominant role. From the inset of Fig. 5(B), it is apparent that the silica sols exerted a clear effect on the concentration polarization. At a fixed current density, the cathode overpotential increased with the addition of more silica sols. The adsorbed silica particles caused the overpotential on the cathode to rise. According to the Erdey-Gruz and Valmer two-dimensional nucleation equation, the nucleation current density (i) and the overpotentiality (η_k) have the following relationship:

ln i = A − Bη_k⁻¹ (1)

where A and B are constants. There is a positive correlation between the number of crystal nuclei and the nucleation current density. In addition to the nucleation current density, the overpotential also affects the nucleation probability (W), as follows:

W = C exp(−c/η_k²) (2)

where C and c are constants.

According to the polarization curve data shown in Fig. 5(A), when the concentration polarization is not playing the dominant role, the tested current can be regarded as the nucleation current density. The model represented by eqn (1) satisfies the hypothesis test of linearity, and it matches the data well. The results of the linear regression for ln i and η_k⁻¹ are listed in Table 1. Low current density errors could be achieved for the fitted constants A and B for overpotentials below 200 mV, which indicated that the electrochemical reaction played the dominant role in the polarization effect. As shown in Fig. 5(B), when the overpotential increased above 250 mV, the role of the concentration polarization became dominant. The current density, which increased sharply with increasing overpotential, included both nucleation and concentration polarization currents. As shown in Fig. 5(B), the cathode overpotentials in different plating baths were read out for zinc coatings deposited at a current density of 1.2 A dm⁻². The fitted nucleation current densities calculated based on the linear regression results are listed in Table 1. Because the nucleation probability also affected the nucleation process, we assumed the nucleation probability of the zinc coating formed in the 25% SSPB to be 0.3. Then, according to eqn (2), the other nucleation probabilities corresponding to the 0, 5 and 15% SSPBs were 0.07, 0.11 and 0.18, respectively. Because the nucleation current densities in the different SSPBs were similar, the nucleation probability played the main role in determining the nucleation behavior of the zinc coatings.^41–43 The fitted results exhibited significant differences and were in good consistency with the images in Fig. 1. The adsorbed silica affected the crystallization of the zinc coatings during the deposition process. When the zinc was electrodeposited at the same current density, zinc coatings with higher overpotentialities tended to form more zinc grains.

Table 1 Fitted results for the electrodeposition of the zinc coatings

Silica sol content in plating bath (%)	Intercept (A)	Slope (B)	Adj. R-squared	Nucleation current density (A cm^−2)	Nucleation probability
0	−6.074	0.169	0.990	9.9 × 10^−4	0.07
5	−6.199	0.164	0.995	9.7 × 10^−4	0.11
15	−6.413	0.158	0.990	9.3 × 10^−4	0.18
25	−6.760	0.139	0.992	8.8 × 10^−4	0.3

---

The current efficiency for the NdFeB cathode in the varying SSPBs is shown in Fig. 6. The increased mass was measured and converted to a current efficiency (CE) from the relation:

where I_Zn was calculated from the actual increased mass measured and I_tot was the total current applied. A current density of 1.2 A dm⁻², which was in agreement with the previous paper, was used in the measurement. As shown in Fig. 6, the current efficiency grew higher with the increase of the silica sol content in the plating bath when the electrodeposition time was under 15 minutes. This may be caused by the adsorption of silica particles, which are visible in the cross section images. When the electrodeposition time was longer than 30 minutes, the current efficiency was slightly improved with the addition of silica sols. The results show that silica sols did not harm the electroplating efficiency.

Fig. 6 Current efficiency of the zinc coatings obtained from (A) the 0% SSPB, (B) the 5% SSPB, (C) the 15% SSPB, and (D) the 25% SSPB.

To evaluate the effects of the silica sol content on the zinc particles, the XRD patterns of the samples prepared in the 0, 5, 15 and 25% SSPBs were recorded; the results are shown in Fig. 7. All the XRD patterns have three pronounced diffraction lines in common, which agree with the (0 0 2), (1 0 1) and (1 0 2) diffraction lines of zinc (PDF65-3358). As indicated in the XRD diagrams, the diffraction lines varied with the increase of the silica sol content. For the coating made from the 5% SSPB (Fig. 7(b)), a stronger orientation along the (0 0 2) plane was observed in comparison with that recorded for the zinc coating obtained from the 0% SSPB (Fig. 7(a)).^44–47 The adsorbed silica nanoparticles at the grain boundaries may change the growth orientation of the zinc grains. When the silica sol content continued to increase from 15% to 25% (Fig. 7(c and d)), the relative intensities of the lattice parameters were slightly changed. The change in the X-ray diffraction intensity indicated that the silica sols modified the molecular orientation and crystallinity of the zinc particles in the coatings. The relative intensity of the (1 0 1) diffraction line grew higher with the increase of the silica sol content from 0 to 25% in the plating bath. The zinc coating with a more (1 0 1) crystal orientation showed better anti-corrosion ability.^48–50 Furthermore, the characteristic zinc diffraction line shifted to the right with the increase in the silica sol content from 0 to 15%. The diffraction line shifts of the XRD patterns may be caused by the adsorbed silica nanoparticles between the zinc grains, which caused a slight lattice distortion.^51,52

Fig. 7 XRD patterns of electrodeposited zinc coatings: (a) the zinc coating obtained from the 0% SSPB, (b) the zinc coating obtained from the 5% SSPB, (c) the zinc coating obtained from the 15% SSPB, and (d) the zinc coating obtained from the 25% SSPB.

As shown in Fig. 8, when the TEOS was hydrolyzed under acid-catalyzed conditions, part of the Si–OC₂H₅ transformed into Si–OH. Moreover, under acidic conditions, the activity of the Si–OH groups is sufficiently high to undergo condensation under stirring. During the aging process, the –OH groups that formed aggregated under further hydrolysis, and the silica chains formed silica nanoparticles that were covered with –OH groups. The hydroxyl-wrapped silica nanoparticles moved to the surface of the cathode during the electroplating process. The latter usually involves application of cathodic potentials to the substrates, yielding a cathodic reduction of:

O₂ + 2H₂O + 4e⁻ → 4OH⁻ (4)

2H₂O + 2e⁻ → 2OH⁻ + H₂ (5)

Fig. 8 Growth model for silica particles under acid catalysis.

These cathodic reactions raise the pH near the conducting substrate by generating OH⁻ ions. The electrochemically induced OH⁻ ions on the cathode catalyze the condensation of the silica sol to a gel on the substrate which thus leads to particle formation. The intent was for the nanoparticles to become embedded in the zinc particles to improve the physical and chemical properties of the coating.

The silica sols were added to the acidic zinc plating baths, the pH of which was close to that of the silica sol system, allowing the silica particles to remain stable in the mixture. When the electrodeposition process was performed in such a mixed plating bath, the deposition of zinc occurred. Simultaneously, the hydroxyl-rich silica sols became attached to the cathode surface due to the effect of the external electric field.^14,23 The condensation of the silica sols on the cathode surface was promoted by the hydroxyl groups that were generated through the side reaction of the zinc electrodeposition, leading to the formation of silica nanoparticles. These silica nanoparticles could then be adsorbed onto the surfaces of the electroplated zinc particles. During the continuous zinc ion crystallization process, some of these adsorbed silica nanoparticles were covered and became embedded into the zinc particles. These embedded silica nanoparticles modified the crystal characteristics and size distribution of the zinc particles.^38,39

In general, the uniformity of their distribution and their adsorption properties are both important to the corrosion resistance performance of zinc coatings. The standard solution with a content of 3.5 wt% NaCl was chosen as the corrosion medium for the determination and comparison of the corrosion behavior in the present work. The polarization curves for the zinc-coated NdFeB substrates exposed to 3.5 wt% NaCl solutions are shown in Fig. 9(A). Under high overpotential conditions, the electrochemical reaction rate under kinetic control could be represented simply by the Tafel relationship. Therefore, by fitting the data via linear extrapolation at a high overpotential (over 100 mV), we were able to obtain the corrosion current density (j_corr) from the intersection with the ordinate and the corrosion potential (E_corr) from the intersection with the abscissa. The E_corr values and j_corr values of the coatings formed at the different silica sol content are listed in Table 2. Compared with the zinc coating obtained from the 0% SSPB, it is notable that the E_corr of the zinc coating from the 15% SSPB and 25% SSPB was shifted to more positive potentials: approximately 36 mV and 40 mV, respectively. Given that E_corr is the measure of corrosion susceptibility and is independent on the geometry of the sample, the shift in E_corr to a more positive direction suggests an increase in the corrosion resistance due to the presence of silica particles in the zinc coating. The j_corr value of the zinc coating obtained from the 0% SSPB was distinctly higher than those of the coatings formed in the other baths. In the aqueous NaCl solution, the Nd-rich phase dissolved first in the anode reaction (eqn (6) and (7)), which most likely produced a passivation film between the substrate and the coating, thereby decreasing the corrosion rate. Simultaneously, the zinc could dissolve to form another passivation film on the surface of the coating (eqn (8)).⁸ Nevertheless, the aggressive Cl⁻ could migrate to the interface to destroy the passivation film thus formed, thereby accelerating the dissolution of the NdFeB. However, the anodic current density of the zinc coating obtained from the 15% SSPB was more than one order of magnitude lower than that of the zinc coating obtained from the 0% SSPB. This suggests that the zinc coatings that were deposited onto NdFeB from the SSPBs remarkably decreased the dissolution of the NdFeB. The decrease in the current density was likely caused by the embedding with silica nanoparticles, which filled the gaps between the zinc particles. The j_corr values of the coatings decreased as the silica sol content was increased from 0 to 15%. However, when the silica sol content reached 25%, the corrosion rate of the zinc coating again increased because of the aggregation of the silica particles, leading to a decline in the filling effect.

Nd → Nd³⁺ + 3e⁻ (6)

Nd³⁺ + 3OH⁻ → Nd(OH)₃ (7)

Zn → Zn²⁺ + 2e⁻ (8)

Fig. 9 (A) Tafel plots, (B) Nyquist plots, (C) equivalent electric circuit and (D) Bode plots for (a) the zinc coating obtained from the 0% SSPB, (b) the zinc coating obtained from the 5% SSPB, (c) the zinc coating obtained from the 15% SSPB, and (d) the zinc coating obtained from the 25% SSPB on the NdFeB magnets.

Table 2 Electrochemical parameters derived from the polarization curves of the NdFeB-based zinc coatings in 3.5 wt% NaCl solution

Silica sol content in plating bath (%)	Ecorr (mV, vs. SCE)	jcorr (A cm^−2)
0	−1148	6.3 × 10^−6
5	−1158	3.2 × 10^−6
15	−1112	4.0 × 10^−7
25	−1108	7.9 × 10^−7

---

Comparisons of the EIS data acquired for the zinc coatings formed in the different SSPBs are shown in Fig. 9(B and D). These tests were also performed in neutral 3.5 wt% NaCl solution. In Fig. 9(B and D), the experimental data are marked as points and the fitted impedance data obtained using the proposed electrical equivalent circuit (EEC) are marked as continuous lines. The arc curves indicate that the electron transfer process was inhibited on the electrode surface. Furthermore, a larger radius of the arc indicates a higher anti-corrosion resistance and a greater difficulty of electron transfer. The shape of the Nyquist plots (Fig. 9(B)) for the samples with a different silica sol content indicates that the anti-corrosive property of the coatings was notably improved due to the tightly embedded silica nanoparticles when the silica sol content in the plating bath was 15%. A more detailed interpretation of the EIS measurement was performed by fitting the experimental plots using the EEC depicted in Fig. 9(C). In Fig. 9(C), R_s, R_cor, R_ct are the solution resistance, the resistance of the corrosion product and the charge transfer resistance to the ionic current through the defects in the zinc film. CPE₁ and CPE₂ are the constant phase elements that represent the corrosion product capacitance and double-layer capacitance due to the roughness and inhomogeneity of the electrode surface, respectively. The CPE impedance is given by the following equation:

Z_CPE(ω) = Y₀⁻¹(jω)⁻ⁿ (9)

where Y₀ is a constant that is independent of frequency, ω is the angular frequency, , and n is the exponential index that represents the dispersion of relaxation. W is the Warburg impedance that represents the diffusion-controlled process at low frequencies and is attributed to mass transfer limitations through the pits on the zinc coating.^2,3,11 The values of the main EEC parameters for the different samples are presented in Table 3. All of the standard deviation values (SDV) were less than 1% in value, which means the equivalent circuit used for fitting the AC impedance data was appropriate. As shown in Table 3, the zinc coating obtained from the 15% SSPB had an R_ct value of 8850 Ω cm², which was significantly higher than the others, implying superior anti-corrosive properties.

Table 3 EEC parameters for the different samples

Silica sol content in plating bath (%)	Rs (Ω cm^2)	CPE1 (Ω^−1 cm^−2 s^−n)	n	Rcor (Ω cm^2)	CPE2 (Ω^−1 cm^−2 s^−n)	n	Rct (Ω cm^2)	W (Ω^−1 cm^−2 s^−0.5)	SDV
0	1.2	9.6 × 10^−6	1	38.4	2.6 × 10^−3	0.89	238	1.9 × 10^−2	9.6 × 10^−3
5	1.2	6.0 × 10^−6	1	47.3	2.1 × 10^−4	0.86	616	3.6 × 10^−2	6.2 × 10^−3
15	1.2	5.1 × 10^−6	1	211.6	9.1 × 10^−5	0.92	8850	5.2 × 10^−2	6.9 × 10^−3
25	1.3	9.9 × 10^−6	1	310.8	7.8 × 10^−5	0.91	5688	7.3 × 10^−2	8.3 × 10^−3

---

Fig. 9(D) shows the Bode magnitude plots of the coatings formed in the different SSPBs. At low frequency regions, the impedance values of the zinc-coated NdFeB samples prepared in the plating baths with a non-zero silica sol content were increased by approximately one order of magnitude compared with that of the zinc-coated NdFeB sample prepared in the 0% SSPB. At intermediate frequencies, |Z| displayed the effects of the double-layer capacitance of the interface. The |Z| behavior of the zinc coatings formed in the presence of silica sols resembled that of the zinc coating obtained from the 0% SSPB because both surfaces favored the collection of charges in a similar manner. Additionally, Fig. 9(B and D) show that the anti-corrosive properties of the coatings initially increased and then decreased as the silica sol content was increased from 0 to 25%, with the best performance achieved at a 15% silica sol content by volume. This finding is also consistent with the results from the polarization curves. Moreover, at high frequencies, the |Z| behavior demonstrated the resistance of the protective coatings; specifically, the zinc coatings formed in the presence of silica sols had higher |Z| values compared with that of the zinc coating formed on the sample coated in the 0% SSPB. This change in resistance can be attributed to the electrical surface resistance of the silica particles. Because a larger |Z| value implies a lower susceptibility to corrosion, the EIS behaviors indicated that the zinc coatings formed in the presence of silica sols were most resistant to corrosion when the silica particles were uniformly distributed in the coatings. The EIS results further illustrated that the silica sols were acting primarily as inhibitors of the anodic reaction (ionic barriers) for NdFeB. The dominant feature of the decrease in the corrosion rate was the reduction in the kinetics of metallic ionization.

4. Conclusions

Zinc coatings were obtained via electroplating from plating baths containing a different silica sol content. The morphological characteristics and corrosion resistance of the zinc coatings were influenced by the variation in the silica sol content. Moreover, the size and distribution of the silica nanoparticles embedded in the zinc particles also varied considerably. As observed from SEM images, the distributions of the embedded silica particles were most uniform when the content of the silica sols in the plating bath was 15% by volume. The XRD patterns of the different coatings showed that the crystallization characteristics of the zinc coatings were also affected by the addition of silica sols. The results of electrochemical tests revealed that in comparison with the zinc coating obtained from a silica sol-free plating bath, the zinc coatings formed in the presence of silica sols were able to provide a more effective protective barrier for the sintered NdFeB magnets in corrosive environments. The improvement in corrosion resistance of the zinc coatings arises from two effects of the silica particles: (i) as inert and insulating materials, the silica particles protect the zinc matrix against corrosion attack, essentially forming a localized barrier film and (ii) the incorporated silica particles could also support corrosion products and slow down subsequent corrosion. The controlled crystal orientation, better uniformity and doped silica particles, which were caused by the addition of silica sols, contributed to the improvement of the protection ability together.

These results demonstrate that zinc coatings that contain silica nanoparticles are a promising new means of providing corrosion protection for sintered NdFeB. Additionally, the described method is an effective way to prepare other types of silica sol-containing metal layers formed in situ. The deposition of highly uniform and compact zinc coatings from sol-containing plating baths may also provide an effective means of protecting various other metallic surfaces.

Acknowledgements

The authors acknowledge the financial support of the National Natural Science Foundation of China (51371020).

References

1. E. Isotahdon, E. Huttunen-Saarivirta, V. T. Kuokkala, M. Paju and L. Frisk, J. Alloys Compd., 2014, 585, 203–213.
2. F. Liu, Q. Li, X. K. Yang, Y. Dai, F. Luo, S. Y. Wang and H. X. Zhang, Mater. Corros., 2011, 62, 1141–1148.
3. S. Mao, H. Yang, Z. Song, J. Li, H. Ying and K. Sun, Corros. Sci., 2011, 53, 1887–1894.
4. X.-k. Yang, Q. Li, J.-y. Hu, X.-k. Zhong and S.-y. Zhang, J. Appl. Electrochem., 2009, 40, 39–47.
5. K. Majima, S. Sunada, H. Ito and Y. Kaneko, J. Alloys Compd., 2006, 408–412, 1426–1428.
6. X. K. Yang, Q. Li, S. Y. Zhang, X. K. Zhong, Y. Dai and F. Luo, Mater. Corros., 2009, 61, 618–625.
7. H. Zhang, Y. W. Song and Z. L. Song, Mater. Corros., 2008, 59, 324–328.
8. W. He, L. Zhu, H. Chen, H. Nan, W. Li, H. Liu and Y. Wang, Appl. Surf. Sci., 2013, 279, 416–423.
9. J. Zhang, C. X. Sun, Z. H. Yu, J. B. Cheng, W. Z. Li and J. Z. Duan, Int. J. Electrochem. Sci., 2014, 9, 5712–5721.
10. M. Musiani, Electrochim. Acta, 2000, 45, 3397–3402.
11. S. M. Tamborim Takeuchi, D. S. Azambuja, A. M. Saliba-Silva and I. Costa, Surf. Coat. Technol., 2006, 200, 6826–6831.
12. S. Zheng and J. Li, J. Sol-Gel Sci. Technol., 2010, 54, 174–187.
13. Z. F. Lin, D. Zhang, Y. Wang and P. Wang, Mater. Corros., 2012, 63, 416–420.
14. A. Hovestad, R. J. C. H. L. Heesen and L. J. J. Janssen, J. Appl. Electrochem., 1999, 29, 331–338.
15. J. Chen, Y. Liu, B. Chi, J. Pu and J. Li, J. Power Sources, 2014, 256, 312–318.
16. K. Kotti and C. Kiparissides, Macromol. React. Eng., 2010, 4, 347–357.
17. L.-K. Wu, X.-F. Zhang and J.-M. Hu, Corros. Sci., 2014, 85, 482–487.
18. M. L. Zheludkevich, I. M. Salvado and M. G. S. Ferreira, J. Mater. Chem., 2005, 15, 5099.
19. S. Dalbin, G. Maurin, R. P. Nogueira, J. Persello and N. Pommier, Surf. Coat. Technol., 2005, 194, 363–371.
20. X.-X. Zhang, B.-B. Xia, H.-P. Ye, Y.-L. Zhang, B. Xiao, L.-H. Yan, H.-B. Lv and B. Jiang, J. Mater. Chem., 2012, 22, 13132.
21. A. E. Raevskaya, Y. V. Panasiuk, O. L. Stroyuk, S. Y. Kuchmiy, V. M. Dzhagan, A. G. Milekhin, N. A. Yeryukov, L. A. Sveshnikova, E. E. Rodyakina, V. F. Plyusnin and D. R. T. Zahn, RSC Adv., 2014, 4, 63393–63401.
22. Y. Y. Fan, Y. J. Zhang and P. Dong, in Multi-Functional Materials and Structures Ii, Pts 1 and 2, ed. Y. S. Yin and X. Wang, Trans Tech Publications Ltd, Stafa-Zurich, 2009, vol. 79–82, pp. 1903–1906.
23. H. Yuan, J. Wang and J. Chen, Corros. Eng., Sci. Technol., 2014, 49, 197–203.
24. E. Vanea, C. Moraru, A. Vulpoi, S. Cavalu and V. Simon, J. Biomater. Appl., 2014, 28, 1190–1199.
25. L.-K. Wu, J.-T. Zhang, J.-M. Hu and J.-Q. Zhang, Corros. Sci., 2012, 56, 58–66.
26. R. Anedda, C. Cannas, A. Musinu, G. Pinna, G. Piccaluga and M. Casu, J. Nanopart. Res., 2008, 10, 107–120.
27. C. Zhu and M. J. Panzer, ACS Appl. Mater. Interfaces, 2015, 7, 5624–5628.
28. W. Widiyastuti, I. Maula, T. Nurtono, F. Taufany, S. Machmudah, S. Winardi and C. Panatarani, Chem. Eng. J., 2014, 254, 252–258.
29. N. C. López Zeballos, M. C. García Vior, J. Awruch and L. E. Dicelio, Dyes Pigm., 2015, 113, 145–150.
30. M. Azizi, W. Schneider and W. Plieth, J. Solid State Electrochem., 2005, 9, 429–437.
31. C.-X. Zhao and A. P. J. Middelberg, RSC Adv., 2013, 3, 21227.
32. D. D. Lovingood, J. R. Owens, M. Seeber, K. G. Kornev and I. Luzinov, ACS Appl. Mater. Interfaces, 2012, 4, 6875–6883.
33. S. Cai, Y. Zhang, H. Zhang, H. Yan, H. Lv and B. Jiang, ACS Appl. Mater. Interfaces, 2014, 6, 11470–11475.
34. A. Tarat, R. Majithia, R. A. Brown, M. W. Penny, K. E. Meissner and T. G. G. Maffeis, Surf. Sci., 2012, 606, 715–721.
35. J. Du, R. Zhao, S. Chen, H. Wang, J. Li and Z. Zhu, ACS Appl. Mater. Interfaces, 2015, 7, 5870–5878.
36. O. Kammona, K. Kotti, C. Kiparissides, J. P. Celis and J. Fransaer, Electrochim. Acta, 2009, 54, 2450–2457.
37. A. Pepe, M. Aparicio, S. Ceré and A. Durán, Mater. Lett., 2005, 59, 3937–3940.
38. M. Tamilselvi, P. Kamaraj, M. Arthanareeswari and S. Devikala, Appl. Surf. Sci., 2015, 327, 218–225.
39. J. D. Brassard, D. K. Sarkar, J. Perron, A. Audibert-Hayet and D. Melot, J. Colloid Interface Sci., 2015, 447, 240–247.
40. M. Ramezani, M. R. Vaezi and A. Kazemzadeh, Appl. Surf. Sci., 2014, 317, 147–153.
41. M. Romero, M. A. del Valle, R. del Rio, F. R. Diaz, F. Armijo and E. A. Dalchiele, J. Electrochem. Soc., 2013, 160, G125–G134.
42. O. Gladysz and P. Los, Electrochim. Acta, 2008, 54, 801–807.
43. M. Rezaei, S. H. Tabaian and D. F. Haghshenas, J. Mater. Chem. A, 2014, 2, 4588.
44. G. Piga, A. Santos-Cubedo, S. Moya Solà, A. Brunetti, A. Malgosa and S. Enzo, J. Archaeol. Sci., 2009, 36, 1857–1868.
45. A. Nakata, H. Murayama, K. Fukuda, T. Yamane, H. Arai, T. Hirai, Y. Uchimoto, J.-i. Yamaki and Z. Ogumi, Electrochim. Acta, 2015, 166, 82–87.
46. V. S. K. Valiveti, F. Singh, S. Ojha and D. Kanjilal, Appl. Surf. Sci., 2014, 317, 1075–1079.
47. M. Ashuri, F. Moztarzadeh, N. Nezafati, A. Ansari Hamedani and M. Tahriri, Mater. Sci. Eng. C, 2012, 32, 2330–2339.
48. L. E. Morón, A. Méndez, F. Castañeda, J. G. Flores, L. Ortiz-Frade, Y. Meas and G. Trejo, Surf. Coat. Technol., 2011, 205, 4985–4992.
49. C. Cachet, F. Ganne, S. Joiret, G. Maurin, J. Petitjean, V. Vivier and R. Wiart, Electrochim. Acta, 2002, 47, 3409–3422.
50. H. Jianping, L. Xinyi and L. Shunlin, Chin. Rare Earths, 2003, 24, 24–27.
51. T. Sawabe, M. Akiyoshi, K. Yoshida and T. Yano, J. Nucl. Mater., 2011, 417, 430–434.
52. K. Meena, K. Muthu, V. Meenatchi, M. Rajasekar, A. Aditya Prasad and S. Meenakshisundaram, Spectrochim. Acta, Part A, 2014, 125, 308–312.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra19517a

---