In vivo Study on the Corrosion Behavior of Magnesium Alloy Surface Treated with Micro‐arc Oxidation and Hydrothermal Deposition

Abstract

Objective

To study the corrosion behavior of magnesium alloy surface treated with micro‐arc oxidation and hydrothermal deposition in living animals.

Methods

A magnesium oxide (MgO) layer was prepared on Mg alloy using micro‐arc oxidation technology, and then a composite coating composed of magnesium hydroxide, hydroxyapatite, and MgO was coated on the MgO layer using the hydrothermal deposition method for 2 h and 24 h. Male 3‐month‐old white New Zealand rabbits (n = 48) weighting 2200–2300 g, were divided into four groups randomly. The prepared Mg alloy samples with composite coatings were implanted into the femoral medullary cavity of rabbits. For the Mg group, bare Mg samples without any treatment were implanted; for the MgO group, bare Mg samples undergoing MAO treatment were implanted; for the HT2h group, samples of the MgO group undergoing hydrothermal treatment (HT) for 2 h were implanted; and for the HT24h group, samples of group MgO undergoing HT for 24 h were implanted. Then the in vivo corrosion behaviors of implants were evaluated by X‐ray observation, micro‐CT analysis and serum Mg ²⁺ examination.

Results

The X‐ray showed that samples implanted in animals were decreased as time went by. The micro‐CT showed that on the fourth week, the residual volume percentages (RVP) of samples of the Mg, MgO, HT2h, and HT24h groups were 72.81% ± 2.10%, 71.68% ± 1.49%, 81.14% ± 1.54%, and 82.04% ± 0.89%, respectively; on the eighth week, the RVP of four groups were 29.45% ± 1.06%, 41.82% ± 1.13%, 53.92% ± 0.37%, and 62.53% ± 2.06%, respectively; while on the 12th week, RVP were 8.45% ± 0.49%, 9.97% ± 0.75%, 37.09% ± 0.89%, 46.71% ± 1.87%. The RVP of the HT2h group and the HT24h group were higher than for the Mg group and the MgO group for all three time points (P < 0.05); the RVP for HT24h was higher than for HT2h at 8 and 12 weeks, and the differences were significant, indicating that the degradation of Mg alloy slowed down after composite coating. In addition, the composite‐coated Mg alloy by 24‐h hydrothermal treatment exhibited a slower degradation than that treated by 2 h. Serum Mg ²⁺ concentration results showed that on the second week, the Mg ²⁺ concentrations of the Mg, MgO, HT2h, and HT24h groups were 2.24 ± 0.10 mmol/L, 2.12 ± 0.07 mmol/L, 2.06 ± 0.11 mmol/L, and 2.15 ± 0.12 mmol/L, respectively. On the fourth week, these concentrations were 1.99 ± 0.33 mmol/L, 2.18 ± 0.06 mmol/L, 2.17 ± 0.09 mmol/L, and 2.13 ± 0.14 mmol/L, respectively. On the eighth week, the concentrations were 2.22 ± 0.09 mmol/L, 2.20 ± 0.17 mmol/L, 2.06 ± 0.11 mmol/L, and 2.14 ± 0.07 mmol/L, respectively. On the 12th week, the concentrations were 2.18 ± 0.04 mmol/L, 2.20 ± 0.08 mmol/L, 2.09 ± 0.02 mmol/L, and 2.16 ± 0.11 mmol/L.

Conclusion

The combination of micro‐arc oxidation and hydrothermal deposition can greatly improve the anti‐corrosion behavior of Mg alloy, and Mg alloy coated with this composite coating is a promising biomaterial with a satisfactory degradation rate.

Introduction

Metallic biomaterials widely used in orthopedic surgery can provide a stable fixation and stress conduction in the process of bone healing, but the inert shortcomings (including stress shielding effects, the need for a second removal procedure, and the generation toxic metallic ions) limit their wide applications in clinic1, 2. With satisfactory mechanical properties, unique biodegradability, and good biocompatibility, magnesium (Mg) and its alloy have been revealed as a promising candidate for orthopedic implants, and cardiovascular and ureteral stent applications3, 4, 5, 6, 7, 8. Some recent studies have evaluated the potential of Mg‐based alloys for orthopaedic applications9, 10, 11, 12. However, Mg alloy has a rapid degradation rate during its early implantation in vivo, and, therefore, does not meet the clinical needs of bone healing13; and it can produce hydrogen in the rapid degradation process, which may increase the risk of local infection after surgery3. Therefore, improving the corrosion resistance of Mg alloy in early implantation is challenging.

The approach of surface coating on Mg alloy to improve its corrosion resistance has been accepted by the majority of researchers14. The application of micro‐arc oxidation (MAO) to form a magnesium oxide (MgO) layer is considered a feasible way to form a protective coating on Mg alloy, but the protective effect is limited15. Hydroxyapatite (HA) is a Ca‐P biocaramic with poor biodegradation, so HA coating is a feasible way to protect basic material from being rapidly degraded16. Wen et al. coated HA on AZ31 alloy using the electro‐deposition technique, and the results in vitro and in vivo demonstrated that the corrosion resistance improved after HA coating17, 18. Chen et al. applied the MAO and electro‐deposition technique to form a composite coating on Mg–Ca–Zn alloy with MgO as the inner layer and HA as the outer layer; the in vivo tests showed that the composite coating could improve the anti‐corrosion property of Mg alloy19. However, the HA coating is too brittle to bond with the basal Mg‐alloy tightly, and tends to exfoliate from the substrate20, 21, 22, 23. In our previous study, we applied the MAO technique to produce an evenly distributed and tightly‐bonded MgO layer on the surface of the Mg alloy as the base coating24, 25; and then we used the hydrothermal deposition method to form a composite coating composed of HA, MgO, and magnesium hydroxide (Mg(OH)₂), which was in situ growing in the holes of MgO. The composite coating was proven to increase the bonding strength between the composite coating and the substrate, as the holes of the MgO coating acted as the binder.

Research on degradation is very important to evaluate the protective effect of coating on basal material. In the present study, the Mg alloys with different composite coatings were implanted into the femoral medullary cavity of New Zealand rabbits, and the in vivo corrosion behavior was evaluated by X‐ray examination, micro‐CT analysis, and serum Mg ions examination, to study the corrosion behavior of magnesium alloy surface treated with micro‐arc oxidation and hydrothermal deposition in living animals.

Materials and Methods

Sample Preparation

Commercial magnesium (Mg, Luoyang City Xinyou Magnesium Alloys Scientific Technology, LuoYang, China) was processed into columns (Φ3.2 mm × 12 mm) and divided into four groups. Each group was treated as follows.

• Mg Group: Bare Mg without any treatments, as the control group, aliased as Mg.

• MgO Group: Bare Mg undergoing MAO treatment, aliased as MgO. For the MAO treatment, a pulsed direct current power supply (400 A) was employed, and an Mg sample was used as the anode, while a stainless steel plate was taken as the cathode in the electrolytic cell. The applied voltage, pulse frequency, duty ratio, and duration time were fixed at 450 V, 100 Hz, 26%, and 10 min, respectively.

• HT2h Group: Samples of MgO group undergoing hydrothermal treatment (HT) for 2 h, aliased as HT2h. In this process, 7.8 mL of aqueous solution containing 0.1 mol/L C₁₀H₁₂CaN₂Na₂O₈ (Ca‐EDTA) and 0.5 mol/L NaOH was added into the autoclave, and the MgO were immersed in the solution to receive HT at 90°C.

• HT24h Group: Samples of MgO group undergoing HT for 24 h, aliased as HT24h.

Animal Surgery

Ethical approval for undertaking this animal experiment was obtained from the institutional ethics board of the Second Affiliated Hospital of Xi’an Jiaotong University. Male 3‐month‐old white New Zealand rabbits (n = 48) weighting 2200–2300 g were divided into four groups randomly (Mg, MgO, HT2h, and HT24h, with 12 in each group). After being anesthetized with 3% pentobarbital sodium (1 mL/kg), the animals were in a supine position, the hind legs were sterilized with Betadine, and the body was covered with a sterile sheet. Then a 1‐cm long longitudinal surgical incision was made around the greater trochanter of the femur. After the greater trochanter was exposed, a hole (Φ = 3.2 cm) was drilled on the depression of the greater trochanter of the femur, and then a tunnel parallel to the femur was drilled before the samples were implanted. After implanting the samples, the incision was sutured layer by layer. The femur on the opposite side was treated similarly.

The vein bloods of each rabbit were extracted before surgery and after surgery at 2, 4, 8, and 12 weeks to measure the concentrations of serum magnesium ions. At each time point post‐operation, three rabbits in each group were killed, and the femurs on both sides were collected for micro‐CT scan.

X‐ray Observation

X‐ray photos were taken at 2, 4, 8, and 12 weeks post‐implantation to observe the morphology, position, and degradation condition of implants. The voltage, current, distance, and time were fixed at 50 kV, 50 mA, 500 mm, and 0.2 s.

Micro‐CT Analysis

Bilateral femurs of each rabbit were collected and scanned by micro‐CT system (eXplore Locus SP, GE company, Fairfield, CT, US) in 2‐D planes and 3‐D reconstructions 2, 4, 8, and 12 weeks after surgery. The resolution, voltage, and current were 14 μm, 80 kV, and 80 μA, respectively. The 3‐D pictures of implanted samples were obtained by reconstruction using reconstruction software (ABA, version 2.1.2, GE Company, Fairfield, CT, USA). Then the residual volumes of the implanted samples at different time points were quantitatively analyzed using the micro‐CT system. The residual volume percentage (RVP) and the degradation rate (DR) were calculated according to the following equations:

$$\mathrm{RVP}=\left(\mathrm{V}1/\mathrm{V}\right)\times 100\%$$ (1)

$$\mathrm{DR}=\Delta V/\Delta t,$$ (2)

where V (3.2046 × 3.2046 × 11.6057 mm) was the standard volume pre‐implantation, V1 was the residual volume of the implanted samples at different time points, ΔV is the degraded volume between different time points, and Δt is time intervals (week).

Examination of Serum Mg²⁺ Concentration

The Mg²⁺ concentrations in the serum of rabbits in each group were measured by inductively coupled plasma atomic emission spectroscopy 2, 4, 8, and 12 weeks after the procedure (ICP‐AES; Varian, Palo Alto, California, USA).

Statistical Analysis

Data are expressed as means ± standard deviation, and the normality of the data distribution was assessed with the Shapiro–Wilk test. A two‐way analysis of variance followed by Fisher’s test was conducted to assess differences among groups using SPSS 16.0 (SPSS, Chicago, IL, USA). Statistical significance was set at a P‐value less than or equal to 0.05.

Results

X‐ray Observations

Two weeks after surgery, pneumatosis (white arrow) could be found in soft tissues around the femur containing Mg (Fig. 1A1), indicating that Mg alloy had a rapid early degradation; however, no such pneumatosis was found in MgO, HT2h, and HT24h groups. From 2 to 4 weeks after surgery, the volume of Mg became smaller (Fig. 1B1), indicating that Mg was degraded continuously; while in MgO and HT2h groups, no apparent volume change but just some coarse edges of implants could be found (Fig. 1B2, B3), indicating that the materials began to be degraded. No coarse edges of implants could be found in the HT24 group at 4 weeks (Fig. 1B4). Eight weeks post‐implantation, obvious degeneration of materials could be found in Mg and MgO groups (Fig. 1C1, C2); meanwhile, clear general shapes of implants in HT2h and HT24h groups could still be observed (Fig. 1C3, C4), only with some coarse edges. Twelve weeks after surgery, the implants of Mg and MgO could barely be found (Fig. 1D1, D2); in comparison, the implants of HT2h and HT24h could be found, although they seemed much smaller than before (Fig. 1D3, D4).

Figure 1.

The X‐ray observation of the four groups. The results of Mg group at four time points are showed in A1, B1, C1, and D1, those of MgO group are showed in A2, B2, C2, and D2, those of HT2h are showed in A3, B3, C3, and D3, and those of HT24h are showed in A4, B4, C4, and D4. All samples were around the greater trouchanter of the femur. It could be roughly found that Mg samples were smallest at all time points, while HT24h samples were largest.

Micro‐CT Analysis

Residual Volume Percentage

Four weeks post‐implantation (Figs 2A,3A,4), the samples of Mg and MgO were in an incomplete main shape, and 72.81% ± 2.10% was left in Mg and 71.68% ± 1.49% was left in MgO. Samples of the HT2h group remained in a complete shape, and the RVP (81.14% ± 1.54%) was obviously higher than those of Mg and MgO by 11.44% and 13.20%, respectively (F = 37.40, P < 0.05); HT24h has the most complete main shape among the four groups, and the RVP (82.04% ± 0.89%) was obviously higher than those of Mg and MgO by 12.68% and 14.45%, respectively (F = 37.40, P < 0.05).

Figure 2.

The gap between samples and bone from micro‐CT cross‐section observation of four groups on the fourth, eighth, and twelfth week. The results of Mg group at three time points are showed in A1, B1, and C1, those of MgO group are showed in A2, B2, and C2, those of HT2h are showed in A3, B3, and C3, and those of HT24h are showed in A4, B4, and C4.

Figure 3.

Reconstruction of samples of four groups. The results of Mg group at three time points are showed in A1, B1, and C1, those of MgO group are showed in A2, B2, and C2, those of HT2h are showed in A3, B3, and C3, and those of HT24h are showed in A4, B4, and C4. It is found that Mg degraded faster than the other three groups and HT2h and HT24h degraded more slowly. Sample shapes of HT2h and HT24h were more complete than those of Mg and MgO at all three time points.

Figure 4.

The relationship of residual volume percentages (RVP) and time. The results showed that HT2h and HT24h groups were higher than that of Mg and MgO groups at all three time points (P < 0.05), and that of the HT24h group was higher than that of the HT2h group (P < 0.05).

Eight weeks post‐implantation (Figs 2B,3B,4), the materials in the Mg group had been degraded in large quantities, and the RVP (29.45% ± 1.06%) was significantly lower than those of the MgO, HT2h, and HT24h groups, by 29.58%, 45.38%, and 52.90%, respectively (F = 367.28, P < 0.05); the MgO materials had lost their main shape, the length became shorter and the diameter became smaller (only 41.82% ± 1.13% was left); the corrosive pits in HT2h and HT24h become deeper and the main shape was also lost, and the RVP of HT2h (53.92% ± 0.37%) and HT24h (62.53% ± 2.06%) were clearly higher than those of Mg and MgO; in addition, the RVP of HT24h was significantly higher than that of HT2h by 15.97% (F = 367.28, P < 0.05).

Twelve weeks post‐implantation (Figs 2C,3C,4), Mg (RVP, 8.45% ± 0.49%) and MgO (RVP, 9.97% ± 0.75%) were almost completely degraded, and only a small amount of degrading debris was left; HT2h and HT24h were seriously degraded, with shorter length and smaller diameter; the RVP of HT2h was 37.09% ± 0.89%, which was higher than those of Mg and MgO by 338.93% and 272.02%, respectively (F = 880.24, P < 0.05); the RVP of HT24h was 46.71% ± 1.87%, which was higher than those of Mg, MgO, and HT2h by 452.78%, 368.51%, and 25.94%, respectively (F = 880.24, P < 0.05).

Degradation Rate

Degradation rates (DR) of different groups calculated from micro‐CT analysis are shown in Fig. 5. At 4 weeks, DR of Mg, MgO, HT2h, and HT24h groups were 6.80% ± 0.53%, 7.16% ± 0.37%, 4.72% ± 0.39%, and 4.49% ± 0.22%, respectively; at 8 weeks, 10.84% ± 0.79%, 7.38% ± 0.11%, 6.80% ± 0.29%, and 4.88% ± 0.58%; and at 12 weeks, 5.25% ± 0.17%, 7.96% ± 0.37%, 4.21% ± 0.19%, and 3.96% ± 0.19%, respectively. Mg showed a faster degradation rate than the other groups. For MgO, HT2h, and HT24h groups, the degradation rates increased within 4 weeks and became relatively stable after 4 weeks. Comparison of degradation rates between different groups showed a tendency of Mg > MgO > HT2h > HT24 at 8 weeks, and MgO > HT2h > HT24h at 4 and 12 weeks.

Figure 5.

The relationship of degradation rates. For groups of MgO, HT2h, and HT24h, the degradation rates increased before 4 weeks and became relatively stable after 4 weeks. Comparison of degradation rates between different groups showed a tendency of Mg > MgO > HT2h > HT24h at 8 weeks, and MgO > HT2h > HT24h at 4 and 12 weeks (P  < 0.05).

Concentration of Serum Mg Ions

The results of serum Mg²⁺ concentration are presented in Fig. 6. On the second week, the Mg²⁺ concentrations of Mg, MgO, HT2h, and HT24h groups were 2.24 ± 0.10 mmol/L, 2.12 ± 0.07 mmol/L, 2.06 ± 0.11 mmol/L and 2.15 ± 0.12 mmol/L, respectively. On the fourth week, these concentrations were 1.99 ± 0.33 mmol/L, 2.18 ± 0.06 mmol/L, 2.17 ± 0.09 mmol/L and 2.13 ± 0.14 mmol/L, respectively. On the eighth week, these concentrations were 2.22 ± 0.09 mmol/L, 2.20 ± 0.17 mmol/L, 2.06 ± 0.11 mmol/L and 2.14 ± 0.07 mmol/L, respectively. On the 12th week, those were 2.18 ± 0.04 mmol/L, 2.20 ± 0.08 mmol/L, 2.09 ± 0.02 mmol/L and 2.16 ± 0.11 mmol/L. The concentration of Mg²⁺ in all groups appeared to have an increasing tendency from 0 to 4 weeks and from 8 to 12 weeks. From 2 to 8 weeks, the Mg²⁺ concentration in MgO, HT2h, and HT24h groups became stable. Except that the Mg²⁺ of the MgO group was higher than that of the HT24h group by 1.85% at 12 weeks (F = 1.622, P < 0.05), no significant differences in the Mg²⁺ concentration between different groups were found (P > 0.05).

Figure 6.

The results of Mg²⁺ concentration. Except for MgO and HT24h at 12 weeks (P < 0.05), no significant differences in the Mg²⁺ concentration between different groups were found (P > 0.05).

Discussion

Magnesium alloy is a class of Mg‐based degradable metallic material with appropriate mechanical properties and good biocompatibility. However, the degradation process of Mg alloy is too rapid and can generate hydrogen during early implantation in vivo, which does not meet the clinical needs on bone healing3. It has been reported that the mechanical integrity of magnesium alloy is only maintained for 6–8 weeks, with the release of hydrogen during the corrosion process15. At present, the hot spots and difficulties are focused on how to improve the corrosion resistance of the Mg alloy. The surface coating technique has become a popular method to modify Mg alloy18, 26, 27, 28.

Micro‐arc oxidation is a useful anodic oxidation technique for depositing a ceramic coating on the surface of valve metals, such as Al, Ti and Zr, and their alloys29. In our previous study, micro‐arc oxidation treatment generated an MgO layer with a two‐layer structure, a dense layer and a porous layer24. With a low porosity, the formed MgO layer bonded tightly with the Mg alloy substrate and was considered to be an effective basal protective layer for Mg alloy materials. In this study, the degradation of MgO was found to be slower than that of Mg by X‐ray and micro‐CT analyses, the residual volume percentage of MgO was significantly higher than that of Mg, and the degradation rate of MgO at 2 and 8 weeks was lower than that of Mg. All these results proved that the MgO layer could effectively protect Mg alloy from being rapidly degraded. When implantation time was prolonged, the MgO layer was disintegrated and stripped, and could not protect Mg from contacting with the body fluid, so material in the MgO group was also degraded rapidly after 4 weeks and only a small number of degrading debris was left (Figs 1, 2, 3)4, 30. MAO technology has been used in Mg alloy, and in vivo and in vitro studies show that the MAO layer could significantly increase the corrosion resistance, but increasing effect is limited20, 31, 32, 33.

Hydroxyapatite is the naturally occurring mineral form of calcium apatite and is the major mineral component of bones and teeth34. HA and octacalcium phosphate (OCP) coatings were formed on AZ31 magnesium alloy using a single‐step chemical solution deposition method by http://www.sciencedirect.com/science/article/pii/S1742706114004097, and the HA and OCP coating were found to suppress the corrosion of AZ31 in vitro and in vivo 35. HA are generally stiff and brittle, potentially leading to the early failure of the bone–implant interface36. Chen et al. made use of MAO technology and electrochemical precipitation to form MgO as the inner layer of the composite coating and HA as the outer layer on the surface of Mg–Ca–Zn alloy; the in vivo study revealed that the composite coating can greatly improve the corrosion resistance of Mg alloys19. Chen and his co‐workers applied MAO and electrochemical deposition (ED) to form a composite coating with HA and OCP in ED layers and MgO, Mg₃(PO₄)₂ in MAO layers on Mg–Zn–Ca alloy to improve the corrosion resistance, and animal results showed that the composite coatings could reduce the degradation rate of the substrate19. However, the calcium phosphate coating prepared using this electrochemical method was brittle and showed low adhesion strength with the substrate23, 34, 37.

In our previous study, a homogenous MgO layer was formed on the surface of Mg alloy using the MAO technique, and then the HA coating grew in situ in the porous structure of the MgO layer by treatment in the electrolyte solution rich in Ca‐P24, 25. At various treatment times, sheaf‐like or lamellar HA was formed, as well as Mg(OH)₂ in the hydrothermal process. Eventually, a composite coating containing MgO, Mg(OH)₂ and HA was formed. This technology was beneficial to improve the adhesive strength between the coating and the substrate, and could overcome the disadvantages of the brittleness of pure HA coating and the low adhesive strength with the substrate. In our present study, the in vivo degradation test revealed that the composite coating could delay the degradation of Mg alloy within 12 weeks. After 12 weeks, there were 37.09% HT2h and 46.71% Ht24h of the original material left, but most material of the Mg and MgO groups were degraded. From both qualitative and quantitative results, it could be found that the HT24 group had a lower degradation rate than that of the HT2h group, which indicated that a longer period of hydrothermal deposition brought a stronger anti‐corrosion effect. Hiromoto synthesized HA on a Mg surface without pre‐treatment by immersion in a solution of Ca‐EDTA, KH₂PO₄, and NaOH; the immersion times varied between 6 and 24 h to control the coating thickness, and the authors showed that the corrosion resistance increased with growing coating thickness38. All these results demonstrated that the composite coating containing HA, MgO, and Mg(OH)₂ could improve the corrosion resistance greatly, compared to uncoating and MgO coating. With a very low degradability, the composite coating containing HA on the surface of Mg alloy protected the substrate from contacting the body fluid, and then delayed the degradation process of Mg alloy. In addition, the composite coating grew in the holes of the MgO layer, which made the coating more stable and provided long‐term protection.

We inferred that the Mg(OH)₂ in the composite coating had low anti‐corrosion properties. The existence of Mg(OH)₂ in the holes of the MgO layer may affect the in situ formation of HA as the closure effect, and Mg(OH)₂ could produce MgCl₂ in the environment of Cl⁻, but MgCl₂ has no corrosion‐protective effect39. Therefore, it is necessary to reduce the amount of Mg(OH)₂ and increase the amount of HA in future studies.

In the present study, we did not evaluate the biomechanical properties of samples. However, biomechanical properties such as compression strength and elasticity modulus are important properties for orthopedic implants. The ideal orthopedic absorbable implants should be able to facilitate an excellent stress replacement process after implantation, which means that during the process of the absorption of the implants and the filling of bone tissue, the biomechanical properties of the bone‐implants compound system should be able to remain stable and as close to natural bone as possible. We will evaluate the biomechanical properties of samples in our next work.

Conclusions

Mg alloy was treated with micro‐arc oxidation first to form an MgO layer and then treated by hydrothermal deposition to coat a composite coating composed of Mg(OH)₂, HA, and MgO in the holes of the MgO surface. The in vivo anti‐corrosion behavior revealed that the degradation of Mg alloy with composite coating was slower than for those with no coating or only MgO coating, and Mg alloy with longer periods of hydrothermal deposition exhibited stronger anti‐corrosion properties. With its strong corrosion resistance, the Mg alloy with a composite coating is considered to be a promising orthopedic implant material.