2023 Volume 64 Issue 5 Pages 1011-1019
To improve the corrosion resistance of Mg alloys, composite coatings of layered double hydroxide intercalated with carbonate (LDH-CO3, hydrotalcite) particles and magnesium and aluminum hydroxide gel were formed on AZ31 Mg alloy by electrophoretic deposition (EPD) using the LDH-CO3 suspension with Mg(NO3)2 and Al(NO3)3 in total of 1.25–125 mmol/L. Wet-dry cyclic corrosion tests were performed for the coated specimens after ultrasonication. LDH-CO3 particle-rich layer and magnesium and aluminum hydroxide gel-rich layer almost alternately piled up to form a composite layer. The hydroxide gel adhered the particles together and to the substrate. At the bottom of the composite layer, flake-shape LDH-structured substance densely covered the substrate surface. The coverage and thickness of the composite layer increased from sub-micrometers to 30 µm with an increase of Mg and Al ion concentrations of the suspension. By ultrasonication, parts of the composite layer delaminated to expose the LDH-structured substance layer, while the coverage by the remaining composite layer increased with an increase of Mg and Al ion concentrations. The LDH-CO3 particles and hydroxide gel composite coatings prevented corrosion initiation even in the areas where the LDH-structured substance layer was exposed.
Magnesium is the lightest structural metal, and magnesium alloys show a high specific strength.1,2) Magnesium alloys are therefore a promising structural material which can reduce the weight of vehicles, aircrafts and mobile electronic devices for good energy efficiency and low CO2 gas emission. Practical application of Mg alloys has been slow due to their poor corrosion resistance. Surface coating is an effective means to improve the corrosion resistance of metallic materials. Therefore, various anti-corrosion coatings, especially non-chromate coatings like phosphate coatings, have been developed for Mg alloys by various coating method such as chemical conversion and micro-arc oxidation and so on.3)
As one of the non-chromate coatings, layered double hydroxide (LDH) coatings have attracted interest for Mg alloys, Al alloys and steel.4–14) LDH is a compound with a stacking structure consisting of basic layers of double hydroxides of divalent and trivalent metal ions, with various anions intercalated between the layers.15) A characteristic property of LDH is the anion-exchange ability.15) As the anion-exchange ability allows LDH to be intercalated various anions between the layers, LDH coatings are expected to serve as a reservoir of corrosion inhibitors.9,11,16) In addition, LDH structure is stable in Mg corrosion environments because corrosion products of Mg–3 mass% Al–1 mass% Zn (AZ31) exposed to the marine environment consisted mainly of hydrotalcite (LDH-CO3)17) which is a type of LDH having carbonate ions between Mg–Al double hydroxide layers.
LDH coatings are formed by various methods such as autoclaving and steam treatment,4,5) chemical conversion,7,11,12) sol-gel synthesis9) and electrochemical deposition.6) Introduction of a corrosion inhibitor between the layers has been conducted by the anion-exchange in aqueous solution after forming a LDH layer on the metal surface, or LDH with a corrosion inhibitor intercalated between the layers has been formed on the metal surface by co-precipitation.9,11,12,16) When performing anion-exchange in aqueous solution on the LDH formed on metal surfaces, there is a concern that the substrate metal can be corroded in the solution. In other words, the corrosion resistance of the substrate metal/alloys may limit the type of corrosion inhibitor which can be introduced between the LDH layers. If corrosion inhibitors are introduced into the LDH in advance and the inhibitor-loaded LDH is subsequently coated on the metal/alloy surface, a variety of corrosion inhibitors can be used for alloys with relatively low corrosion resistance. Therefore, we employed electrophoretic deposition (EPD) for the formation of LDH coating layers. EPD is an outstanding method of depositing a variety of ceramic particles with, sometimes, polymers and hydroxides on a variety of material surfaces.18–20) In our work, to glue the particles to substrate surfaces, hydroxide gel was intended to electrodeposit as an adhesive by adding metal ions to the suspension.21,22) Metal ions (Mn+) like Mg and Al ions in EPD suspensions precipitates as hydroxides (M(OH)n) because pH on the cathodically polarized substrate surface rises by water electrolysis according to the following reactions.
| \begin{equation} 2H_{2}O + 2e^{-} \to H_{2} \nearrow + 2\textit{OH}^{-} \end{equation} | (1) |
| \begin{equation} M^{n+} + \textit{nOH}^{-} \to M(\textit{OH})_{n} \searrow \end{equation} | (2) |
The purpose of this study was to apply EPD method to form composite coating layers of LDH particles and magnesium and aluminum hydroxide gel on Mg–3Al–1Zn (AZ31). Considering the crystal structure of LDH, the charge of LDH particles in the suspension depends on the ratio between divalent and trivalent metal ions of the base layers, so that the charge of LDH is expected not to be so dependent on the anion species intercalated between base layers. Thus, the EPD coating method developed in this study can be applicable to LDH particles intercalated with various anions. The corrosion resistance of the LDH-CO3 particles and hydroxide gel composite-coated AZ31 was examined by a wet-dry cyclic corrosion test. Before the corrosion tests, weakly glued particles were removed by ultrasonication because weakly glued particles are not expected to contribute the corrosion resistance. The wet-dry cyclic corrosion test was conducted to examine the corrosion resistance for atmospheric corrosion.
AZ31 (Al 2.9 mass%, Zn 1.05 mass%, Mn 0.41 mass%, Si 0.01 mass%, Cu <0.01 mass%, Ni <0.001 mass%, Fe 0.001 mass%, Mg Bal.) magnesium alloy plates of 40 mm × 30 mm or 50 mm × 30 mm and 2 mm thick were used as substrates. The surface was ground with SiC papers (PSI, USA) up to #1200 and ultrasonically rinsed in acetone. The entire back surface and ground surface of the plates was covered with Teflon™ tape, leaving a test surface of 20 mm × 30 mm. Immediately before the EPD treatment, the substrate was immersed in 0.25 mol/L C10H14N2Na2O8·2H2O (EDTA-Na, FUJIFILM Wako) solution for 10–20 sec to remove the air formed oxide films.
2.2 EPD with different electrolyte concentrationsEthanol and ultrapure water were mixed with a volume ratio of 4:1 to make a total volume of 500 mL. Mg(NO3)2·6H2O (FUJIFILM Wako) and Al(NO3)3·9H2O (FUJIFILM Wako) were dissolved at concentrations of 0–100 mmol/L and 0–25 mmol/L, respectively, in a concentration ratio of 4:1 for Mg to Al ions. Subsequently, LDH-CO3 powder with a particle diameter of about 1 µm (hydrotalcite, Sigma Aldrich) was suspended in the solution at a concentration of 30 g/L. The suspensions were stirred for several minutes and then ultrasonicated for 60 min. Composition of the suspensions is summarized in Table 1.
AZ31 plate specimen and 304 stainless steel plate were used as the deposition (working) and counter electrodes, respectively, and mounted in the EPD cell with a distance of 10 mm between the electrodes. Immediately after immersing the electrodes in the suspension, a square-wave pulse voltage with a 10 V-to-0 V voltage was applied between the electrodes for 10 min. Frequency and duty of the pulse voltage were 0.1 Hz and 98%, respectively. The suspension was kept being stirred during the EPD. The pulse voltage was applied because hydrogen gas bubbles adhering to the specimen surface were expected to be detached by the solution flow during the 0 V for 0.2 s. The EPD-treated specimens were dried at 100°C for 1 h and observed using a digital camera. The composite layers were characterized using scanning electron microscope (SEM, Miniscope TM3030Plus, Hitachi) in backscattered electron mode, X-ray diffraction (XRD, D2 Phaser, Bruker) and laser Raman spectroscopy (Renishaw, inVia Reflex) with a laser wavelength of 532 nm. To remove weakly glued LDH-CO3 particles that are not expected to contribute to the corrosion resistance, the EPD-treated specimens were ultrasonicated in ethanol for 1 min. Specimens that were just dried after the EPD treatment were denoted EPD-only specimens, while specimens that were subsequently ultrasonicated were denoted EPD-US specimens. Specimen names and the EPD conditions are shown in Table 1.
2.3 Wet-dry cyclic corrosion testWet-dry cyclic corrosion tests were carried out for the EPD-US and uncoated specimens. 200 µL of 0.3 mass% NaCl-75% ethanol solution was dropped onto the specimen surface to place 1 g/m2 NaCl on the surface. Then, the specimens were placed in a chamber, and the relative humidity in the chamber was controlled at room temperature as follows. Four wet-dry cycles were repeated, with one cycle of 8 hours at 30%RH, 8 hours at 95%RH and 8 hours at 30%RH again. After the wet-dry corrosion tests, the specimen surface was observed using a digital camera and SEM and characterized using XRD. Corrosion products and the composite layers were chemically removed from the corroded specimens using chromate–AgNO3 solution and the surface appearance and surface topography was observed using digital camera and one-shot 3D microscope (VR-3100, Keyence). The surface morphology was also observed using SEM in secondary electron and back-scattered electron modes, and the composition was analyzed using energy dispersive X-ray spectroscopy (EDS: X-stream-2, Oxford) equipped to SEM.
Figure 1 shows the optical images of the EPD-only specimens prepared in the suspensions with Mg and Al ions in total of 0–125 mmol/L. In the absence of Mg and Al ions, the substrate surface was uniformly and thinly covered with LDH-CO3 particles. In the presence of Mg and Al ions, the substrate surface was more obviously covered with LDH-CO3 particles than in the absence of Mg and Al ions. The 1.25 mM- and 12.5 mM-EPD only specimens showed smooth surfaces except at the periphery of the specimens, while the coating layer of the 125 mM-EPD-only specimen was coarse.
Optical images of EPD-only specimens treated in 3% LDH-CO3 suspension with Mg(NO3)2 and Al(NO3)3 in total of (a) 0 mmol/L, (b) 1.25 mmol/L, (c) 12.5 mmol/L and (d) 125 mmol/L.
Figure 2 shows the surface and cross-section SEM images of the EPD-only specimens prepared with 1.25–125 mmol/L Mg and Al ions. The coverage by LDH-CO3 particles on the 1.25 mM-EPD-only specimen was low, while the substrate surface was densely covered with fine flake-shape substance (Fig. 2(d)). On the 12.5 mM- and 125 mM-EPD-only specimens, dark contrasting gel was observed between LDH-CO3 particles, indicating that the gel glued particles with each other (a typical location is indicated by an arrowhead). The amount of gel deposited with LDH-CO3 particles apparently increased with an increase of Mg and Al ions. The current value during 10 V application a few minutes after the start of EPD was about 5, 20 and 120 mA for 1.25, 12.5 and 125 mmol/L Mg and Al ions, respectively. The deposition of hydroxide gel depends on the amount of OH− ions generated in the water reduction. The apparent increase of gel deposition coincides with the increase in current corresponding to the generation rate of OH− ions.
(a)–(d) Surface and (e)–(i) cross-section SEM images of EPD-only specimens prepared in 3% LDH-CO3 suspension with Mg(NO3)2 and Al(NO3)3 in total of (a), (d) and (g) 1.25 mmol/L, (b), (e) and (h) 12.5 mmol/L and (c), (f) and (i) 125 mmol/L. (d) is a magnified image of the area squared in (a). (h) and (i) are magnified image of (e) and (f), respectively. Typical deposited gel and gel-rich layers are indicated by white arrowheads and particles are indicated by white arrows.
According to the Smoluchowski equation (eq. (3)), the electrophoretic mobility of particles is proportional to the zeta potential.
| \begin{equation} \zeta = \frac{4\pi \eta}{\varepsilon} \cdot \frac{v}{E} \end{equation} | (3) |
The cross-section SEM images (Fig. 2) showed that gel-rich layers (indicated by white arrow heads) and particle-rich layers (indicated by white arrows) almost alternately piled up in the composite layers which thickened with an increase of Mg and Al ions from sub-micrometers to about 30 µm. The thickness of particle-rich layer increased towards the outer side (Fig. 2(h)–(i)). The volume fraction of gel then appeared to be higher on the inner side than on the outer side (Fig. 2(h) and (i)). This morphology indicates that immediately after the start of EPD, the precipitation of Mg and Al ions as hydroxides occurs preferentially over the deposition of LDH-CO3 particles. The 125 mM-EPD-only specimen showed delamination between the layers which was due to the shrinkage of the gel during drying. Additionally, the cross-section images exhibited that submicron to micron pores were formed preferentially at the boundary between the composite layer and the substrate due to H2 gas generated in the water reduction reaction during EPD. Number and size of the micro-pores increased with an increase of Mg and Al ions.
Figure 3 shows the XRD patterns of the EPD-only specimens prepared in the suspensions with 1.25–125 mmol/L Mg and Al ions. With an increase of Mg and Al ions, the intensity of diffraction peaks from LDH-CO3 particles increased and a broad peak at around 18–19 degrees corresponding to Mg(OH)2 (JCPDS card No. 00-44-1482) and Al(OH)3 (JDPDS card No. 03-0145) appeared. The XRD pattens of the 12.5 mM- and 125 mM-EPD-only specimens showed a very broad peak ranging from 32 degrees to 40 degrees which corresponds to amorphous Mg and Al hydroxides. Thus, the dark contrasting gel observed in the SEM images is attributed to amorphous Mg and Al hydroxides.
XRD patterns of EPD-only specimens treated in 3% LDH-CO3 suspension with Mg(NO3)2 and Al(NO3)3 in total of 1.25–125 mmol/L. (a) Wide range and (b) narrow range for (003)LDH plane.
Figure 3(b) shows magnified diffraction peaks from (003) plane of LDH structure. The (003)LDH peaks were asymmetrical owing to a broad shoulder peak on the lower angle side. The presence of the shoulder peak indicates that a compound taking LDH structure is formed. The interlayer spacing (d003) of Mg–Al–LDH-NO3− synthesized by co-precipitation was 0.78 nm which is slightly larger than 0.76 nm of Mg–Al–LDH-CO3−.23) From thermodynamic calculations, NO3− ion has its molecular plane tilted to the basic layers of LDH while CO32− ion keeps the orientation parallel to the basic layers of LDH, indicating the interlayer spacing of LDH-NO3− can be larger than that of LDH-CO3−.24) The shoulder peak at the lower angle is thus attributed to the formation of LDH-NO3.
To confirm the presence of compound taking LDH structure other than LDH-CO3 particles, the flake-shape substance exposed in the area without LDH-CO3 particles on the 1.25 mM-EPD-only specimen was analyzed using laser Raman spectroscopy as shown in Fig. 4. As a comparison, the LDH-CO3 particles adhered to the substrate surface and the untreated surface were analyzed. LDH-CO3 particles showed Raman peaks at ca. 1050 cm−1 and 540 cm−1, which correspond to carbonate ions bonded to trivalent metal and stretching of OH–O between carbonate ions and water molecules, respectively.25) The area without LDH-CO3 particles showed small broad peaks at ca. 1070 cm−1 and 520 cm−1 which indicate that the flake-shape substance has a LDH structure. It was reported that Mg and Al ions in the suspension precipitate as Mg and Al hydroxide gel during EPD.21,22) Water reduction reaction continues to occur on the substrate surface during EPD, indicating that the initially precipitated Mg and Al hydroxides continues to be exposed to a high pH environment in the presence of NO3− ions. It is thus highly expected that the precipitated Mg and Al hydroxides transformed to Mg–Al–LDH having nitrate ions (LDH-NO3). On the other hand, CO2 in the air could be dissolved in the suspension and incorporated into LDH as CO32−. Therefore, further investigation is necessary to determine the composition of the flake-shape substance.
(a) and (b) Raman spectra of LDH-CO3 particles (point 1 on optical image (c)) and an area in absence of LDH-CO3 particles (point 2 on optical image (c)) on 1.25 mM-EPD-only specimen. (c) Optical image showing analysis points. Raman spectra of uncoated substrate AZ31 is shown as a reference.
Figures 5 and 6 show optical and surface SEM images of the EPD-US specimens prepared with different concentrations of Mg and Al ions. After ultrasonication, almost all the LDH-CO3 particles detached from the specimen prepared in the absence of Mg and Al ions. On the 1.25 mM-EPD-US specimen, only minute amounts of particles remained (Fig. 5(b) and 6(a)), while the flake-shape substance layer, most likely Mg–Al–LDH-NO3 layer, remained on the entire surface of the substrate as shown in Fig. 6(g). On the 12.5 mM-EPD-US specimen, the outer part of the composite layer delaminated, and the inner part remained (Fig. 6(b)). There were the round shape areas with a diameter of ca. 200 µm where most of the composite layer delaminated leaving the bottom layer consisting of the LDH-structured substance and a small amount of particles (Fig. 6(h)). Also, on the 125 mM-EPD-US specimen, the outer part delaminated, and the inner part remained in most areas (Fig. 6(c)). The round areas, where most of the composite layer delaminated leaving the LDH-structured substance layer, were fewer than those on the 12.5 mM-specimen (Fig. 6(b), (c), (h) and (i)). It is suggested that LDH-structured substance formed at the bottom of the composite layer has good adhesiveness to the substrate surface.
Optical images of (a) 0 mM-EPD-US specimen, (b) 1.25 mM-EPD-US specimen, (c) 12.5 mM-EPD-US specimen and (d) 125 mM-EPD-US specimen EPD-treated and subsequently ultrasonicated.
SEM images of (a), (d) and (g) 1.25 mM-EPD-US specimen, (b), (e) and (h) 12.5 mM-EPD-US specimen and (c), (f) and (i) 125 mM-EPD-US specimen EPD-treated and ultrasonicated. (d)–(f) are magnified images of areas where LDH-CO3 particles-gel composite layer remained. (g)–(i) are magnified images of areas where most of the composite layer delaminated leaving the bottom layer.
Figure 7 shows the surface appearance of the EPD-US and uncoated specimens after the wet-dry cyclic corrosion tests. The uncoated specimen clearly showed several corrosion pits with a millimeter size and brown spots due to corrosion (Fig. 7(a)). The 1.25 mM-EPD-US specimen showed a few corrosion pits with a sub-millimeter size and a lot of brown spots due to corrosion (Fig. 7(b)). The 12.5 mM- and 125 mM-EPD-US specimens did not show apparent corrosion pits but showed brown spots due to corrosion (Fig. 7(c) and (d)).
Optical images after wet-dry cyclic corrosion tests for (a) uncoated AZ31 specimen, (b) 1.25 mM-EPD-US specimen, (c) 12.5 mM-EPD-US specimen and (d) 125 mM-EPD-US specimen.
Figure 8 shows the XRD patterns of the EPD-US and uncoated specimens after the wet-dry corrosion tests. The XRD patterns after the corrosion tests did not show significant changes in the type of detected compounds compared to the EPD-only specimens (Fig. 3), except for NaCl peaks at 31.8 degrees and 45.5 degrees (JCPDS card No. 5-628). This indicates that the amount of corrosion products was too small to be detected by the XRD measurement. The magnified (003)LDH peak of the uncoated specimen (Fig. 8(b)) showed a small diffraction peak at 10.9 degrees, indicating that corrosion products become LDH during the corrosion test.
XRD patterns after wet-dry cyclic corrosion tests of uncoated and EPD-US specimens. (a) Wide range and (b) narrow range for (003)LDH plane. * XRD peak intensity of uncoated specimen in (b) is shown doubled to clarify the presence of a diffraction peak.
To observe the corrosion morphology of the substrate AZ31, the corrosion products and the composite layers were chemically removed, and the surface appearance and topography were observed as shown in Fig. 9. Profiles of the black dashed lines i)–vii) in Fig. 9(a)–(d) are shown in Fig. 10(a) and (b). Secondary electron and back scattered electron SEM images of the areas *1–*7 in Fig. 9(a)–(d) are shown in Fig. 11. Dark contrast substance in corrosion pits and on substrate were observed in the back scattered electron images (Fig. 11(e)–(g)), which were oxide according to EDS analysis. Since both corrosion products and composite layers mainly consist of Mg, Al, and O, it was not possible to separate them by composition. From their morphology, dark contrast substance in Fig. 11(e)–(g) was corrosion products and that in Fig. 11(h) is parts of the composite layer remained. Those are observed as red dots on the topography images.
(a)–(d) Optical and (e)–(h) topography images of (a) and (e) uncoated specimen, (b) and (f) 1.25 mM-EPD-US specimen, (c) and (g) 12.5 mM-EPD-US specimen and (d) and (h) 125 mM-EPD-US specimen after wet-dry cyclic corrosion tests and removing corrosion products and composite layers.
(a)–(d) Secondary electron and (e)–(h) back scattered electron SEM images of the areas *1, *3, *5 and *7 in Fig. 9(a)–(d). (i)–(l) Secondary electron SEM images of the area *2, *4 and *6 in Fig. 9(a)–(c). (a), (e) and (f) uncoated specimen, (b), (f) and (j) 1.25 mM-EPD-US specimen, (c), (g) and (k) 12.5 mM-EPD-US specimen and (d), (h) and (l) 125 mM-EPD-US specimen.
The uncoated specimen showed dark and light brown areas (Fig. 9(a)). The profiles of lines i) and ii) showed a pit of about 20 µm depth in the respective dark brown area. The dark brown area *1 showed a pit with remaining corrosion product (Fig. 11(a) and (e)). The light brown area *2 showed very shallow filiform corrosion (Fig. 11(i)).
Almost the entire surface of the 1.25 mM-EPD-US specimen showed light brown spots. The profile of line iii) showed similar roughness to that of line i) on the uncoated specimen. The area *4 showed very shallow filiform corrosion (Fig. 11(j)). The profile of line iv) on the light brown area shows a pit of about 5 µm depth and the area *3 around the line iv) shows shallow filiform corrosion (Fig. 11(b)). These results indicate that slight corrosion occurred in the light brown areas.
On the 12.5 mM-specimen, the area *6 covered by a relatively thick composite layer showed a metallic luster and no corrosion (Fig. 11(k)). The area *5 covered by a thinner layer showed filiform corrosion (Fig. 11(c)) while the profile of the corresponding line vi) show the presence of pits (Fig. 10(b)), indicating that the filiform corrosion was shallower than 1 µm.
The color of brown spots on the 125 mM-EPD-US specimen (Fig. 9(d)) was lighter than on the other specimens. Brown dots with a size of 100–300 µm corresponded to the remaining composite layer as shown in Fig. 11(d) and (h). The profile of the line vii) was somewhat smoother than those of the other specimens and no obvious corrosion was observed in the area *7 as shown in Fig. 11(l).
Total area of pits deeper than polishing scars and the maximum corrosion depth were obtained from the topography images and are summarized in Fig. 10(c) and (d), respectively. The maximum corrosion depth was determined by excluding the periphery of the specimen plates and the boundary area with the Teflon tape. The total area of pits, corresponding to corroded area, significantly decreased by the composite coatings formed at concentrations above 12.5 mmol/L by less than 1/5. The maximum corrosion depth decreased by about 1/5 by the composite coatings. The maximum corrosion depth apparently did not depend on the Mg and Al ion concentration of the EPD suspension.
The above results revealed that the LDH-CO3 particle and hydroxide gel composite layers formed by EPD can show corrosion protection ability and that the corrosion resistance of the composite layers increases with an increase of Mg and Al ion concentration of the EPD suspension. As shown in Fig. 6, the relatively thick composite layers had micro-pores and there were the areas of a few hundred micrometers where the LDH-structured substance layer was exposed. This microstructure of the composite layers presumably allowed NaCl solution to permeate through the composite layer during the wet-dry corrosion tests. The decrease in corrosion depth on the 1.25 mM-EPD-US specimen demonstrated that the thin dense LDH-structured substance layer, most likely LDH-NO3 layer, can show corrosion resistance. Therefore, the high corrosion protection ability of the composite layers is considered to highly depend on the bottom LDH-structured substance layer.
The LDH-CO3 particles and magnesium and aluminum hydroxide gel composite layers were formed on AZ31 by EPD using the 3% LDH-CO3 particles and ethanol-water suspension containing Mg(NO3)2 and Al(NO3)3 in total of 1.25, 12.5 and 125 mmol/L. After ultrasonication of the EPD-treated specimens, the wet-dry cyclic corrosion tests were performed. Followings findings were derived.
It was eventually demonstrated that the composite layers of LDH particles and hydroxide gel can be formed on the Mg alloy by EPD and that the composite layers show corrosion protection ability. Further investigation is necessary to improve the uniformity and adhesiveness of the composite layer to the substrate surface.
The authors are grateful to Ms. N. Nishikawa for her sincere support during the experiments.
