Preparation and Corrosion Resistance of Microarc Oxidation-Coated Biomedical Mg-Zn-Ca Alloy in the Silicon-Phosphorus-Mixed Electrolyte.

Microarc oxidation (MAO) coating was prepared on the surface of the biomedical Mg-3Zn-0.2Ca alloy in a phosphate electrolyte with varying concentrations of Na2SiO3. The morphology, cross section, chemical composition, and corrosion resistance of the coatings were characterized by scanning electron microscopy (SEM), energy dispersive X-ray spectroscopy (EDS), X-ray diffraction (XRD), electrochemical polarization tests (EI), and in vitro immersion experiments. The addition of Na2SiO3 is performed to increase the thickness and compactness of the coating. When the Si/P atomic ratio is approximately equal to 1 (1.5 g/L Na2SiO3), the best corrosion resistance is achieved, while excessive addition may lead to coating defects such as voids and microcracks, resulting in decreased corrosion resistance. The competitive relationship between PO4 3- and SiO3 2- anions in the silicon-phosphorus microarc oxidation-mixed electrolyte is discussed. In this study, it was first proposed that, when Mg2SiO4 and Mg3 (PO4)2 phase contents were approximately the same, the synergistic improvement effect on coating corrosion resistance was the most effective.

Introduction

Due to the good biocompatibility of magnesium (Mg) and its alloys, many researchers have paid special attention to it.[1,52,53] The density[2,3] (1.74 g/cm3) and elastic modulus[4] (44 GPa) of Mg are close to that of a natural bone. There is no stress-shielding effect when the implant is implanted into the human body. In addition, Mg is an essential element of the human body, and about half of the Mg2+ ions in the human body are stored in bones.[7] The corrosion products produced in the body are also easily absorbed and metabolized by the human body.[5−7] Also, Mg2+ ions have been shown to promote the adsorption and growth of bone cells.[8] Zhang et al.[9] revealed the molecular mechanism of Mg in promoting bone repair. Their work examined the effective concentration of the Mg ion for activating PI3K phosphorylation via TRPM7. The authors found that Mg ions promote cell growth and survival, protecting cells against alkaline-stress-induced cytotoxicity caused by the degradation of Mg-based alloy implants.

Although many studies have shown that magnesium alloys are the most suitable metals for degradable implant materials,[1,10−12,48] their clinical application has been limited due to the rapid corrosion rate of magnesium alloys in humans.[13]

Results have shown that proper surface treatments can effectively reduce the corrosion rate of magnesium alloys.[14] At present, electrochemical deposition,[15] anodic oxidation,[16] biodegradable polymer coating,[17,46,54,57] chemical conversion coating,[18,55,56,59] microarc oxidation (MAO),[19,47,49,58] and other surface treatment technologies are used to improve magnesium alloy corrosion resistance.

MAO, also known as plasma electrolytic oxidation (PEO), is an emerging environmentally friendly technology for improving the corrosion resistance of magnesium and its alloys. In a high-energy electrolyte, the surface of the magnesium alloy is transformed into a ceramic oxide coating by a large number of partially discharge sparks.[20−22] In recent years, the most widely used electrolytes in microarc oxidation of magnesium alloys are phosphate[23−26] and silicate solutions.[27,28]

Zeng et al.[19] considered that the corrosion resistance of the coating is mainly related to the porosity of the MAO coating. Jia et al.[29] performed microarc oxidation treatment on a novel Mg–1Ca alloy in KF–silicate, KF–phosphate, and KF–silicate–phosphate electrolytes, and MTT assays analyzing in vitro toxicity against MG63 cells were subsequently carried out. It was found that the coatings prepared in the silicate and silicon-phosphorus-mixed electrolytes were thicker and more porous, but the microporous depth of the coating prepared in the silicon–phosphorus-mixed electrolyte was small. The coating prepared in the phosphate electrolyte was denser, and the cell survival rate was higher, thereby improving biosafety. Ghasemi et al.[31] found that SiO32– is more conducive than PO43– to the growth rate of MAO coatings by studying the effects of different ions on the microarc oxidation process and corrosion resistance of coatings in different electrolytes. In their study, the surface of the coating obtained from the silicate electrolyte had larger micropores than the phosphate electrolyte coating.

Mori et al.[30] conducted microarc oxidation treatment on the AZ31B alloy in the mixed electrolyte with different concentration ratios of phosphate and silicate. Corrosion resistance analysis found the coating with the best corrosion resistance in the silicon–phosphorus-mixed electrolyte had a phosphate ratio of about 20% (P/Si = 20:80). Gao et al.[34] performed MAO treatment on the AZ31B alloy in different silicate electrolytes with a Na2SiO3 concentration. The bond strength of the coating was shown to gradually decrease with the increase in the Na2SiO3 concentration in the electrolyte. The optimal concentration of Na2SiO3 in the phosphorous–silicon-mixed electrolyte requires a balance ratio of phosphate and silicate and may be different for magnesium alloys with different components. Chen et al.[38] argued that the presence of Si-containing phases could improve the bioactivity of the MAO coating.

Although various studies have been carried out on the influence of different electrolyte compositions of magnesium alloy microarc oxidation,[31−33] most applications are not in the field of biomaterials. Therefore, no reports on the use of Na2SiO3 as a trace additive in phosphate electrolytes have been published. In phosphate solution, it is easier to obtain a more dense coating with higher a cell survival rate and better biosafety,[29] but the wear resistance is poor.[26] This study systematically describes the role of trace amounts of Na2SiO3 in electrolytes, with a primary focus on the effect of the Si/P atomic ratio on corrosion resistance in coatings. Concentrate on the synergistic effect of phosphate and silicates on the formation of MAO coating. Other studies have not focused on the effect of the Si/P atomic ratio on corrosion resistance.

In this study, the biomedical Mg–3Zn–0.2Ca alloy was treated with MAO in the phosphate electrolyte with different concentrations of Na2SiO3. Na2SiO3 is added to improve the growth rate and[31] coating thickness and to increase the hardness and wear resistance of the coating. Also, the phase composition, microstructure, and corrosion resistance of the coating in the simulated body fluid were studied.

Materials and Experimental Method

Material and Sample Pretreatment

In a vacuum melting furnace, the Mg–Zn–Ca alloy was smelted at 700 °C under argon. The alloy composition was 3.0 wt % Zn, 0.2 wt % Ca, and the balance was Mg. After the alloy ingot was homogenized at 350 °C for 24 h, extruded into a Φ 8 rod at 300 °C by a YQ 32-315 extruder, and the cutting for a Φ 8 × 3 mm sample. Before the microarc oxidation (MAO) treatment, all the samples were polished with 320#, 800#, and 1500# SiC sandpapers, and then the sample was ultrasonically cleaned with distilled water and acetone to remove grease and placed in a dry box.

Preparation of MAO Coating

Microarc oxidation treatment was performed using a 50 kW pulse power source (Chengdu Tongchuang Electric Equipment Co., Ltd.). Taking the Mg–3Zn–0.2Ca alloy sample as an anode and the stainless steel plate as a cathode inside an electrolyte containing Na3PO4·12H2O, NaOH, and Na2SiO3. Electrolyte compositions are shown in Table . The sample was subjected to microarc oxidation treatment under conditions of a positive voltage of 400 V, a negative voltage of 20 V, a frequency of 600 Hz, and a duty ratio of 30% for 15 min. Then, the sample was taken out, ultrasonically cleaned with distilled water and ethanol, and put into a dry box for drying.

Table 1

Electrolyte Composition

sample	Na3PO4·12H2O (g/L)	Na2SiO3 (g/L)	NaOH (g/L)	temperature (°C)
M1	12	0	6	30
M2	12	0.5	6	30
M3	12	1	6	30
M4	12	1.5	6	30
M5	12	2	6	30

Microstructure Characterization

A field emission scanning electron microscope (FE-SEM, JOEL 6700F, Japan) and an energy-dispersive X-ray energy spectrometer (EDS) were used to characterize the microstructures, thicknesses, and elemental distributions of the samples. The phase composition of the coated sample was analyzed by an X-ray diffractometer (XRD, Rigaku D/max/2500 pc, Japan) using Cu Kα radiation (λ = 0.154 nm) with a small angle mode, scanning from 10 to 80° at a speed of 5°/min.

Electrochemical Test

Electrochemical testing was performed using a Zennium electrochemical workstation (Germany) in the simulated body fluid (SBF) (2.2683 g·L–1 NaHCO3, 0.3676 g·L–1 CaCl2·2H2O, 0.2681 g·L–1 0.3728 g·L–1 KCl, Na2HPO4, 6.5453 g·L–1 NaCl, 0.026 g·L–1 NaH2PO4, 0.1 g·L–1 Na2SO4, 6.057 g·L–1 (CH2OH)3CNH2). Electrochemical testing was performed in a three-electrode system. The MAO-coated samples (working area: 0.503 cm2) were the working electrodes, the counter electrode was graphite, and the reference was a saturated calomel electrode (SCE). The samples were tested immediately after immersion for 30 min in SBF. The potential polarization test was performed at a scan rate of 1 mV/s. By using the Tafel extrapolation method[39,40] (Zview software 3.1) to fit the Tafel curve, corrosion potential (Ecorr), and corrosion current density (Icorr) were obtained. The results are derived from the average of three samples per group.

The corrosion rate (Pi) is proportional to the corrosion current density (Icorr) and is calculated according to the following equation:[41,42]

The polarization resistance (Rp) is inversely proportional to the corrosion current density (Icorr) and is calculated using the formula:[43]

Matthews et al.[19,44] estimated the conductive porosity (F) of the coating based on the empirical equation and complementary equation. The total porosity (F) of the coating can be determined according to eq :where F is the total porosity of the coating, Rpm is the polarization resistance of the substrate, Rp is the polarization resistance of the coating, Ecorr is the corrosion potential difference between the coating and the substrate, βa is the anode Tafel slope of the matrix alloy, and βc is the cathode Tafel slope of the matrix alloy.

Immersion Test

To test the degradation of different coated samples in vitro, the immersion experiment was conducted in SBF of 37 ± 0.5 °C and pH = 7.4, and the ratio of immersion was 35.5 mL/cm2. After soaking for 1, 3, 7, and 14 days in the SBF solution, the sample was taken out of the SBF solution, rinsed with distilled water, and dried in a dry box. After the immersion, the corrosion products on the surface of the sample were removed with chromic acid solution (5 g K2Cr2O7 + 10 mL H2O + 90 mL H2SO4), then rinsed with distilled water and absolute ethyl alcohol, and dried in a dry box. A stereomicroscope (LCM, OLS 4000, Olympus, Japan) was used to observe the sample morphology immersed in SBF solution for different days. The corrosion rate was calculated from the weight loss using the following equation:where CRavg is the average corrosion rate, A is the surface area exposed to SBF, ρ is the density of alloy (≈1.74 g/cm3), t is the immersion time, and W0–W1 is the weight loss.

In addition, the pH of the SBF solution was measured with a digital pH meter (STARTER 3100, OHOUS), and all tests were performed on three parallel samples to determine their repeatability.

Results

Characterization

Figure shows the XRD patterns of M1, M2, M3, M4, and M5 samples after microarc oxidation treatment. The XRD patterns of M1 coating prepared from a single phosphate electrolyte shows multiple diffraction peaks of MgO and Mg3(PO4)2 and a broad dispersion peak at 2θ = 15–25°, indicating Mg3(PO4)2 formed in the M1 coating did not completely crystallize in the microarc oxidation process and most Mg3(PO4)2 exists in the amorphous form. This is consistent with the results of Mori et al.[30] Their work concluded that amorphous Mg3(PO4)2 exists in MAO coating obtained from the phosphate electrolyte, which plays an important role in improving the corrosion resistance of the coating. The broad dispersion peak disappeared in the XRD patterns of the M2–M5 coatings, which were formed in mixed electrolytes of phosphate and silicate. The M2, M3, M4, and M5 coatings contained crystalline Mg3(PO4)2, MgO, Mg2SiO4, and possibly amorphous Mg3(PO4)2. The XRD patterns of M4- and M5-coated samples showed diffraction peaks of Mg2SiO4 at 2θ = 34.36° and 48.09°. With the increase with Na2SiO3 in the electrolyte, the number of diffraction peaks of Mg3(PO4)2 in the coating gradually decreased, the relative content of Mg2SiO4 increased, and the content of amorphous phase Mg3(PO4)2 may decrease. Therefore, the dispersion broad peak caused by the amorphous phase Mg3(PO4)2 is not shown in the XRD pattern of M4 and M5 coating samples.

Figure 1

Grazing incidence XRD patterns of M1-, M2-, M3-, M4-, and M5-coated samples.

Morphologies of the M1–M5-coated samples are shown in Figure . All sample surfaces are porous. The surface of the M1 coating prepared in the phosphate electrolyte is distributed with pores of different sizes, with a maximum diameter of 4–5 μm. With the addition of Na2SiO3 in the electrolyte (M2–M5), the distribution of micropores became more uniform and the pore diameter gradually decreased. When the concentration of Na2SiO3 reached 2 g/L (M5), the surface coating showed cracking and the diameter of micropores increased. The cracks were generated by the thermal stress between molten oxide and cold electrolytes.[30]

Figure 2

(A) Morphology of M1, (B) M2, (C) M3, (D) M4, (E) M5 MAO-coated samples and (F) Si/P atomic ratio in the coatings and electrolytes.

The phase of the coating depends on the composition of the electrolyte. As can be seen from Table , the M1 coating was composed of Mg, O, and P elements, and the M2–M5 coatings were composed of Mg, O, P, and Si elements. The Si/P atomic ratios of the M1–M5 coatings were 0:3.88, 1.38:3.65, 2.62:3.45, 3.88:3.22, and 5.16:3.02, respectively. With the increase with Na2SiO3 in the electrolyte, the content of silicon in the coating gradually increased, the content of phosphorus gradually decreased, and the Si/P atomic ratio continuously increased. As shown in Figure F, the increase in the Si/P atomic ratio in the coating was significantly greater than that in the electrolyte. It indicates that there is a competition between PO43– and SiO32– in the electrolyte during the formation of the coating and the reaction rate of Mg2+ and SiO32– is faster than that of PO43–.

Table 2

EDS Elemental Analyses of the M1–M5 MAO-Coated Samples

	atomic %
	element
sample	O	Mg	P	Si
M1 (area 1)	52.47	43.65	3.88
M2 (area 2)	52.47	42.50	3.65	1.38
M3 (area 3)	52.85	41.09	3.45	2.62
M4 (area 4)	53.95	38.96	3.22	3.88
M5 (area 5)	54.66	37.15	3.02	5.16

No break point between the coating and the substrate can be seen from the cross-sectional morphology indicating that the coating has good adhesion (Figure ). As the concentration of Na2SiO3 in the electrolyte increased, the thickness of the coating gradually increased, becoming more dense and uniform. Similarly, the coating growth rate accelerated as the concentration of Na2SiO3 in the electrolyte increased (Figure ). When the concentration of Na2SiO3 in the electrolyte was 2 g/L (M5), the thickness and growth rate of the coating showed a significant decline. Moreover, the micropore size increased, and the continuity decreased. Different growth rates may be caused by the difference in the reactivity of Mg2+ with PO43– and SiO32– ions.[28] This indicates that adding an appropriate amount of Na2SiO3 to the phosphate electrolyte has the effect of promoting coating growth up to a point. The EDS elemental analysis (Figure ) shows the M1 coating was mainly composed of a homogeneous distribution of Mg, O, and P, indicating the MgO and Mg3(PO4)2 phases in the coating may grow simultaneously. The M2–M5 coatings were composed of a homogeneous distribution of Mg, O, P, and Si, indicating the SiO32– ion participates in the formation of the coating, and the Mg2SiO4, MgO, and Mg3(PO4)2 may be growing at the same time.

Figure 3

Cross-sectional morphology and EDS element analysis of (A, B) M1-, (C, D) M2-, (E, F) M3-, (G, H) M4-, and (I; J) M5-coated samples.

Figure 4

Thickness and growth rate of M1–M5 MAO-coated samples.

Electrochemical Polarization Testing (EI)

The dynamic potential polarization curves of the M1–M5-coated samples are shown in Figure . The relevant electrochemical analysis results obtained from the electrodynamic polarization curves are shown in Table . The Mg alloy had a low corrosion potential (−1749 mV) and high corrosion current density (84.9 μA/cm2). After microarc oxidation treatment, the corrosion potential shifted to approximately 86 mV and the corrosion current density decreased to 7.28 μA/cm2. The high corrosion potential and low corrosion current density indicate that the corrosion resistance of the substrate improved after the microarc oxidation treatment. Increasing the Na2SiO3 concentration in the electrolyte increased the corrosion potential and decreased the corrosion current density. At 1.5 g/L Na2SiO3, the coating had the highest corrosion potential (−1550 mV), the minimum corrosion current density (2.38 μA/cm2), and the maximum polarization resistance (12.93 kΩ cm2). The porosity and corrosion rate of this coating were minimized and are nearly 3 times smaller than that of the coating without Na2SiO3 (M1).

Figure 5

Dynamic potential polarization curves of bare Mg alloy and M1–M5-coated samples.

Table 3

Analysis Results Obtained from the Electrodynamic Polarization Curves

sample	Ecorr (mV) vs SCE	Icorr (μA/cm2)	–βc (mV/decade) vs SCE	βa (mV/decade) vs SCE	Rp (kΩ cm2)	Pi (mm/year)	F
Mg alloy	–1749	84.9	210	141	0.432	1.94
M1	–1663	7.28	138	138	4.12	0.167	10.47
M2	–1651	4.24	139	148	7.35	0.097	5.86
M3	–1617	3.07	141	148	10.226	0.071	4.21
M4	–1550	2.38	134	150	12.93	0.055	3.33
M5	–1564	5.57	155	166	6.257	0.126	6.89

The MAO coating formed on the surface of the substrate impedes the penetration of the corrosive medium into the substrate, and the corrosion resistance is improved. The cross-sectional morphology of the coating (Figure ) shows a uniform and dense coating. This uniform and dense structure can effectively reduce diffusion of the corrosive medium into the substrate, thereby reducing the corrosion rate of the alloy. In general, the corrosion of the substrate by corrosive ions occurs via defects, micropores, and cracks in the coating, and the coating porosity can be used for qualitative evaluation of the corrosion resistance of the coating. The 1.5 g/L Na2SiO3 electrolyte sample (M4) had the lowest coating porosity, in agreement with SEM results (Figure ). When the Na2SiO3 concentration reached 2.0 g/L (M5), the polarization resistance, compactness, and thickness of the coating decreased and the corrosive medium is more likely to contact the magnesium matrix.

Degradation of Coatings In Vitro

Figure shows the corrosion morphology of the bare Mg alloy and MAO-treated samples. The bare Mg alloy (Figure A) shows the most severe corrosion damage, the surface is covered with etch pits with different depths and the edges are detached. The size and number of corrosion pits on the surface of M1–M5 MAO-coated samples were significantly reduced. Among them, the surface of M1 (Figure B) is relatively rough, and a large area of corrosion occurs. However, the degree of corrosion shedding at the edge of the sample is significantly reduced compared with that of the bare Mg alloy. The surface distribution of the uniform area of corrosion pits of the sample decreased by adding Na2SiO3 into the electrolyte. After immersion in SBF, the area and number of corrosion pits were gradually reduced and the edge of the sample was more complete. The M4 and M5 coatings (Figure E,F) remained relatively intact, and fewer corrosion pits were observed. The corrosion occurs preferentially near the micropores on the surface of the sample, and the corrosive medium contacts the substrate along the extending direction of the micropores to form a pitting morphology. The M4 coating had more uniform corrosion and a less corrosion pit distribution corrosion surface. Cracks appeared on the surface of the M5-coated sample allowing the corrosive medium to pass through the coating to the substrate.

Figure 6

(A) Morphology of the bare Mg alloy and (B–F) M1–M5-coated samples immersed in SBF for 168 h after removing the corrosion product.

Figure shows the pH curve of SBF with samples immersed up to 14 days. Studies have shown that pH changes are related to the corrosion resistance of magnesium alloys.[45] Bare Mg alloy reacts violently with SBF, and the alkalinity of the solution increased significantly. The pH value reached 8.77 after 4 days of immersion indicating a large amount of Mg(OH)2 alkaline corrosion product was generated on the surface of the sample. This produced a protective effect on the Mg alloy allowing the pH value to increase slowly reaching 9.47 after 14 days.

Figure 7

pH value of bare Mg alloy and M1–M5-coated samples immersed in SBF at 37 °C for different times (mean ± SD, n = 3).

The porous structure of the MAO coating makes the corrosive medium easy to infiltrate; therefore, a more dense coating produces better corrosion resistance. With the increase with Na2SiO3 in the electrolyte, the increase in the pH value in SBF at the initial stage of immersion was significantly reduced, which further proved that the density of the coating from M1 to M4 was indeed increased, which had a strong protective effect on the substrate. As the porosity of the M5 coating increased, the protective effect on the substrate decreased. After immersion for 1 day, the pH increase of all coated samples slowed. After 14 days of immersion, the pH of the M1 coating (0 g/L Na2SiO3) reached 8.96 and the pH of the M4 coating (1.5 g/L Na2SiO3) was 8.13. The M4 coating had the best corrosion resistance base on the pH curve, consistent with electrochemical data.

The weight loss and the average corrosion rate of the bare Mg alloy and M1–M5 MAO-coated samples during in vitro immersion are shown in Figure . The weight loss of all the samples increased rapidly within 5 days and then slowed due to the formation of surface corrosion products. After immersion for 14 days, the weight loss (0.095 ± 0.003 g) and corrosion rate (8.04 ± 0. 24 mm/year) of the bare alloy were higher than that of MAO-coated samples. MAO treatment slowed the weight loss and corrosion rate, producing a protective effect on the coating. With the increase with the Na2SiO3 concentration in the electrolyte, the weight loss and corrosion rate gradually decreased. At 1.5 g/L Na2SiO3 (M4 coating), the minimum weight loss (0.017 ± 0.001 g) and corrosion rate (1.44 ± 0.13 mm/year) were nearly 6 times smaller than the bare Mg alloy.

Figure 8

(A) Weight loss and (B) average corrosion rate of bare Mg alloy and M1–M5-coated samples after immersion in SBF (mean ± SD, n = 3).

Discussion

Effect of Na2SiO3 Concentration on MAO Coating Phase

The XRD results (Figure ) show SiO32– and PO43– ions in the electrolyte participate in the formation of MAO coating and form Mg2SiO4 and Mg3(PO4)2 phases in the coating. The formation of phases in the coating is based on the following reactions.[30,34,35,63,64]

Under high voltage conditions, the magnesium matrix and oxygen ions in the electrolyte are converted to Mg2+ and O2– ions, respectively. Due to the presence of an electric field,[36,60,63] Mg2+ migrates outward from the magnesium matrix to the discharge channel and O2– migrates inward from the electrolyte to the discharge channel to form MgO, as shown in eqs and 6. Mg3(PO4)2 is formed at high temperatures[25] from Mg2+ migrating outward from the substrate to the discharge channel and PO43– migrating inward from the electrolyte to the discharge channel, as shown in the eq . Mg2SiO4 is formed by the plasma chemical oxidation reaction between the magnesium substrate and electrolyte[37,38,61] in the discharge channel, as shown in eqs , 9, and 10.

Coating Growth Process in the Silicon–Phosphorus-Mixed Electrolyte Solution

The Mg2SiO4, MgO, and Mg3(PO4)2 phases are the main components of the coating obtained from the mixed electrolyte of silicon and phosphorus, and these phases may be grown simultaneously. The reaction mechanism of MgO and Mg3(PO4)2 is shown in eqs and 7. Mg2SiO4 is formed by the plasma chemical oxidation reaction between the substrate and electrolyte[45,60−63] in the discharge channel, as shown in eqs , 9, and 10. As shown in Figure , the addition of an appropriate amount of Na2SiO3 to the phosphate electrolyte promotes coating growth and increases the compactness of the coating. With the increase in Na2SiO3 concentration in the electrolyte, the Si element content in the coating increased gradually, while the content of the P element decreased gradually, and the Si/P atomic ratio increased continuously. This indicates that there is a competitive relationship between PO43– and SiO32– anions in the electrolyte during the formation of the coating. As shown in Figure B, both the PO43– and SiO32– ions react with Mg2+ obtained by substrate ionization in the silicon–phosphorus-mixed electrolyte. Under the same electrical parameters, the amount of substrate Mg2+ ionization is constant. Therefore, the concentration of PO43– and SiO32– in the mixed electrolyte becomes the main factor in components of the coating.

Figure 9

(A) Schematic diagram of MAO coating growth of magnesium alloy in phosphate electrolyte and (B) silicon-phosphorus-mixed electrolyte.

Table shows the Si/P atomic ratios of M1–M5 coatings were 0:3.88, 1.38:3.65, 2.62:3.45, 3.88:3.22, and 5.16:3.02, respectively. When the coating has a Si/P atomic ratio of 3.88:3.22 (approximately equal to 1), the M4-coated sample has the fastest growth rate. The content of P in the M1–M3 coating is higher than that of Si. In the formation of the coating, it is possible to promote the formation of the coating by using Na3PO4. As the concentration of Na2SiO3 increases, more SiO32– ions participate in the coating formation process. At the same electrical parameters, SiO32– and PO43– ions will compete to form compounds. In the M4 and M5 MAO coatings, the Si content is higher than that of P. Although only a small amount of Na2SiO3 was added to the electrolyte, the content of the Si element in the M4 and M5 coatings was higher than that of the P element. It may be that the ability of the SiO32– ion to form a compound with Mg2+ is stronger than that of the PO43– ion. During the formation of the M5 coating, Na2SiO3 became the main coating formation promotion agent, which forms a competitive relationship with Na3PO4, resulting in a decrease in the coating density and thickness (Figure ). Therefore, the best concentration addition of Na2SiO3 plays a role in improving the density and thickness of the coating.

The micropores of the coating prepared in the silicate electrolyte are a single structure, while the micropores of the coating in the phosphate electrolyte are interconnected structures. With the increase in the Na2SiO3 concentration in the electrolyte, the SiO32– ion gradually plays a role in the coating forming process, which changes the microporous structure of the coating. As shown in Figure , the coating gradually changes from an interconnected microporous structure to a single structure, which reduces the porosity of the coating and improves the compactness and corrosion resistance of the coating.

Corrosion Resistance Mechanism of Coatings

Mori et al.[30] concluded that the corrosion resistance of coating generated in the phosphate electrolyte is the result of the amorphous Mg3(PO4)2 phase spontaneously transforming into Mg3(PO4)2·22H2O in solution with a self-repair mechanism. Liang et al.[26] compared the corrosion resistance of the two coatings. The first consisted of MgO and Mg2SiO4, and the second was MgO. Their results showed the coating containing Mg2SiO4 had good corrosion resistance. In this study, the corrosion resistant phases in the coating prepared by the silicon–phosphorus-mixed solution may be amorphous Mg3(PO4)2 and Mg2SiO4 phases. Other researchers also believe that Mg2SiO4[47,50] and Mg3(PO4)2[51] play a beneficial role in improving the corrosion resistance of MAO coating. Adding an appropriate amount of Na2SiO3 can increase the thickness and compactness of the coating, while excessive addition may lead to coating defects such as voids and microcracks.

According to electrodynamic polarization and in vitro immersion tests, the formation of Mg2SiO4 and Mg3(PO4)2 in MAO coating improves the corrosion resistance of the coating, the corrosion resistance increased and then decreased with increasing concentrations of Na2SiO3. When the Na2SiO3 concentration was 1.5 g/L (M4), the coating was the most thick and dense, and the corrosion resistance was maximized. The Si/P atomic ratio, the ratio of Mg3(PO4)2 to Mg2SiO4 of the M4 coating, was about 1. According to different corrosion resistance mechanisms, Mg3(PO4)2 to Mg2SiO4 synergistically enhance the corrosion resistance of the substrate. If the concentration of Na2SiO3 continues to increase, the Si/P atomic ratio increases, the coating thickness decreases, and microcracks were observed. As the amount of Mg2SiO4 in the coating increased, the amorphous Mg3(PO4)2 phase decreased, and the corrosion resistance may decrease. Therefore, when Si/P atomic ratio in the coating is approximately equal to 1, it has the best corrosion resistance.

Conclusions

MAO coatings on the Mg alloy were prepared in the phosphorous electrolyte with various concentrations of Na2SiO3. The coatings were composed of Mg3(PO4)2, MgO, and Mg2SiO4. Increasing the concentration of Na2SiO3 in the electrolyte increased the content of silicon and decreased the content of phosphorus. The atomic ratio of Si/P increased with increasing Na2SiO3 concentrations, indicating a competitive relationship between PO43– and SiO32– anions in the electrolyte during the formation of the MAO coating. In 1.5 g/L Na2SiO3 electrolyte, the Si/P atomic ratio in the coating was approximately equal to 1, and the coating was uniform and compact. This sample also demonstrated the fastest growth rate and the best corrosion resistance, which contributes to the clinical application of magnesium alloys.