Mg-Phenolic Network Strategy for Enhancing Corrosion Resistance and Osteocompatibility of Degradable Magnesium Alloys.

Magnesium-based alloys are the most widely used materials for degradable metallic implants and have considerable potential for bone applications owing to their excellent stimulating effect on osteogenesis. However, their high corrosion rate limits their structural stability and causes oxygen deficiency and an excessive increase in the pH around the defect area during bone healing. Magnesium oxides, which are the main corrosion products of Mg, are nontoxic materials with useful effects on new bone formation and pH neutralization. Metal-phenolic networks were introduced recently as a cost-effective and efficient surface modifier and were fabricated by deposition of nanosized metal oxides on different types of substrates using the chemical reaction between phenolic groups and metallic ions. In this study, magnesium oxide films were formed successfully on a Mg-based substrate using Mg-phenolic networks. The effects of various coating parameters on the surface morphology, corrosion resistance, degradation behavior, wettability, and osteocompatibility of degradable metallic materials after surface modification with Mg-phenolic networks were thoroughly investigated for the first time. The results showed that the initial concentration of Mg ions was the main parameter affecting the corrosion resistance, which was almost as much as 3 times that of uncoated samples. Additionally, cytotoxicity and viability assessment and observation of the morphological changes in bonelike cells showed that the in vitro osteocompatibility was significantly enhanced by coatings with Mg concentrations of 2.4-3.6 mg mL-1. Finally, in vivo animal studies using the rat calvarial defect model confirmed that the proposed coating method mitigated the formation of gas cavities around the implantation area by reducing the corrosion rate of the Mg-based implant. The nanosized metal oxides produced by the Mg-phenolic network significantly improved the biodegradability and osteocompatibility of Mg alloys, suggesting a potential approach to advancing the clinical application of Mg alloys.

Introduction

Magnesium-based alloys are the most widely used degradable bone-related implant biomaterials for orthopedic applications.[1−3] In addition to the structural capability of these alloys to provide mechanical stability, Mg contributes to the metabolism of bone regeneration.[4,5] When Mg is implanted in bone defect areas, the contact between the Mg surface and blood plasma (the main component of blood that contains mostly water) results in a chemical reaction (see eq (6)). Consequently, Mg ions (Mg2+) are released into the surrounding blood. It has been proved that the Mg ions effectively enhance the proliferation and differentiation of osteogenic stem cells, which can support new bone formation.[7] However, the release dosage of Mg2+ ions should be controlled to prevent toxicity to the cells in the implantation area.[8] Most of the released Mg2+ ions react with hydroxide ions (OH–) and form a porous Mg(OH)2 film on the surface of the Mg substrate, which can change to MgO in the absence of water. Mg(OH)2 is relatively insoluble in water and has been used as an inorganic nontoxic medicine for pH neutralization of sour stomach.[9] Additionally, as the main degradation product of magnesium alloys in bone applications, Mg(OH)2 stimulates osteogenesis and reduces osteoporosis, resulting in a mature bone with a higher density.[10,11] Furthermore, it can be applied as an antibacterial agent with a high loading capability.[12]The low standard electrode potential of Mg makes it very reactive in the presence of electrolytes in blood plasma or other aqueous solutions.[13] Mg(OH)2 and/or MgO oxide films are not dense enough to protect Mg-based implants against the high corrosion rate in the physiological environment,[14] which causes an increase in crack initiation and propagation and ultimately decreases the mechanical stability in the defect zone.[1,15] According to eq , hydrogen gas (H2) is another corrosion product resulting from Mg degradation. The flow rate of blood in bone tissues is not naturally high enough to remove the insoluble H2 bubbles that are released. Therefore, H2 cavities are formed around Mg implants and prevent the arrival of sufficient oxygen and other necessary nutrients for adhesion, proliferation, and differentiation of bone cells at the implant surface, ultimately killing the cells before bone mineralization begins.[6,16,17] Hence, controlling the high corrosion rate of Mg will reduce the negative consequences of H2 cavity formation, especially in the days immediately following implantation, allowing the nutrient effect of the valuable corrosion products of Mg [i.e., Mg2+, Mg(OH)2] to stimulate bone ingrowth.

Apart from the design of new biocompatible Mg alloys by adding alloying elements such as Zn, Mn, and Ca[18,19] to reduce the corrosion rate of pure Mg, surface modification has been used as the most effective approach to passivate the surface of Mg-based alloys and enhance their corrosion resistance. Biomimetic degradable coatings such as calcium orthophosphate,[14,20] collagen,[21] and biocomposite coatings have been shown to be suitable approaches not only to delay the corrosion of Mg but also to generate a biocompatible interface for bone formation.[22] These strategies prolong the corrosion period by enhancing corrosion resistance of Mg alloy and providing barriers to corrosion electrolyte.[23] Recently, silk fibroin (SF) coating endowed Mg alloys with superior biocompatibility and significantly enhanced corrosion resistance as evidenced by a rabbit spine test for 180 days.[24] SF could also be composited with K3PO4 to form a self-healing coating on Mg alloys with pH sensitivity, preferable corrosion resistance, and osteogenic activity.[23] However, the main drawback of these surface passivation methods is that the positive role of Mg oxides in bone mineralization becomes negligible. Therefore, deposition of Mg(OH)2 and MgO coatings by various methods such as ion beam deposition,[25] physical vapor deposition,[26] anodization,[27] and pulse direct current (DC) magnetron sputtering[28] has been proposed to protect Mg alloys. However, most of these methods require special and expensive equipment to deposit Mg oxides on the Mg-based substrate. Furthermore, some of them can be applied only to substrates with simple geometries. Therefore, a cost-effective, simple, flexible, and efficient method to coat Mg-based substrates with Mg oxides could offer significant advantages over the methods mentioned above.

Metal–phenolic networks (MPNs) were recently introduced as a flexible and cost-effective surface modification method for deposition of multiple types of nanoscale biointerfaces on various substrates.[29] MPNs are generally formed by dynamic chemical reactions between metal ions and phenolic molecules using tannic acid (TA) or phytic acid, which are the two main natural polyphenol products.[30,31] Chen et al.[32] proposed a Mg-ion-integrated phytic acid coating to form three-dimensional (3D) networks of Mg oxides. They found that a bonelike structure formed on the Mg substrate, which can effectively improve the corrosion resistance and osteocompatibility, simultaneously stimulating the bioactivity of bonelike apatite precipitation and osteoblast cell adhesion and proliferation in vitro. Although some studies have investigated MPNs as surface coatings on Mg alloys,[33,34] the effects of experimental parameters such as the type and concentration of the Mg ion source and phenolic molecules, the number of coating cycles, and the characteristics of the intermediate layer on the morphology, density, homogeneity, and stability of the deposited film are not fully understood. Furthermore, the effects of those parameters on the degradation behavior, release of Mg ions, corrosion resistance, cytotoxicity, and cell differentiation need to be elucidated to obtain an osteocompatible surface.

In this research, Mg–phenolic networks were deposited on a Mg–Zn alloy and the effects of coating variables such as the concentration of Mg ions and number of coating cycles on the surface morphology and characteristics, corrosion resistance, biodegradability (i.e., ion release), and cytotoxicity were studied to identify the optimal coating parameters for surface passivation of Mg-based implants. Finally, samples were fabricated using the optimal coating parameters and in vivo animal studies were conducted to compare their degradation behavior with that of uncoated (raw) samples.

Results and Discussion

Characterization of Mg–Phenolic Networks

Surface Morphology

The surface morphologies of the coated samples after the coating processes are presented in Figure a. The initial step of the coating process using the NaOH solution produced a uniform nanosized porous film with tiny scales on the surface. The effect of Mg concentration on the morphology after three coating cycles is obvious. The tiny scales on the intermediate layer of step 1 were successfully thickened at the lowest Mg concentration (0.6%). The tiny scales formed in step 1 acted as an intermediate passivation layer that reduced Mg substrate corrosion by allowing Mg ions to react with TA to form the 3D phenolic structure. Additionally, the gaps between the tiny scales were almost filled in 0.6%Mg@TA. When the initial concentration of Mg was increased from 0.6 to 1.2%, the gaps in the intermediate layer were completely filled by Mg–phenolic networks. For the 2.4% Mg group, a cauliflower-like structure was formed and ultimately became connected to form a uniform secondary layer as the concentration increased to 3.6%. Moreover, the image to 0.6%Mg@TA sample shows the formation of cracks on the surfaces of the coated samples after three cycles of coating at this concentration. The micro- and nanocracks indicated release of Mg ions from the substrate and subsequent corrosion. This result indicated that Mg ions at the initial concentration of 0.6% did not react with the free oxygen ions of TA. The remaining oxygen ions would react with the substrate, resulting in a lack of Mg ions and fewer Mg–phenolic networks. Figure b compares the thickness of the coating layer in different samples using backscattered electron (BSE) imaging. There was no significant difference between the NaOH- and 0.6%Mg@TA-coated samples. Moreover, the thickness of the coating layer was increased from 6 to 7.5, 9, and 10 μm by increasing the initial concentration of Mg during step 2 of the coating process from 0.6 to 1.2, 2.4, and 3.6%, respectively.

Figure 1

Scanning electron macrographs of (a) surface morphology of the coated samples in different process parameters using a secondary electron detector (SED). (b) Cross section of the coated samples using backscattered electron scanning.

Figure S1a,b illustrates the effect of cycle number on the coating morphology of the 3.6% Mg sample. The thin scales resulting from step 1 were filled after the first cycle (Figure a); they were further thickened by Mg phenols and became denser at the end of cycle two. Figure S1b shows that the cauliflower-like networks and then began to appear. Thus, step 2 of the coating process led to filling and thickening of the intermediate layer, initiation of the formation of the 3D Mg–phenolic structure, and the further formation of a uniform, dense coating.

Chemical Composition Analysis of the Mg–Phenolic Film

Figure a shows the chemical composition of the NaOH- and 3.6%Mg@TA-coated samples, as representative of the results of steps 1 and 2 of the coating procedure (Figure ). C, O, Mg, and Zn were present in both coatings. Figure b shows the high-resolution Mg 2p spectra of the coating after steps 1 and 2. The Mg 2p spectra of the NaOH sample showed that step 1 introduced a mixture of MgO and Mg(OH)2. In contrast, MgO was the main component of the 3.6%Mg@TA sample owing to the chemical reaction between free O2– ions from TA and Mg2+ ions, which led to the formation of cauliflower-like Mg–phenolic networks.

Figure 2

Surface characterization of different samples. X-ray photoelectron spectroscopy (XPS) results: (a) wide scan spectrum, (b) corresponding Mg 2p high-resolution spectrum; (c) average surface roughness (Sa); (d) water contact angle; significance is shown as *** and +++ p < 0.001 compared to raw and NaOH samples, respectively.

Figure 10

Schematic of the assembly of Mg–phenolic networks on Mg-based alloys.

Surface Roughness

Figure c summarizes the average roughness (Sa) of various samples observed by atomic force microscopy (AFM) in a scanning range of 20 × 20 μm2. The average Sa value was approximately 10 nm for the polished raw samples and increased to 23 nm after step 1 of coating. The roughness of the samples completely coated with Mg–phenolic networks increased to 88, 90, 107, and 116 nm for the 3.6%Mg@TA, 2.4%Mg@TA, 1.2%Mg@TA, and 0.6%Mg@TA samples, respectively. The roughness values of the 2.4%Mg@TA and 3.6%Mg@TA samples were quite similar. Moreover, the surface roughness decreased as the initial concentration of Mg increased owing to smoothing at higher concentrations.

Wettability

The results of wettability assessment are shown in Figure d. The average water contact angle of the raw samples was ∼88°. The NaOH coating decreased the water contact angle by 41% compared to that of the raw samples (p < 0.001). It is also clear that regardless of the initial Mg concentration, the addition of Mg–phenolic networks enhanced the hydrophilicity. Hydrophilic surfaces with water contact angles of 20–40° can reportedly provide the highest levels of cell attachment.[35−38] As shown in Figure d, all of the coated samples exhibited moderate water contact angles of 21–27°, which were significantly different from that of the raw sample (p < 0.001). Furthermore, the deposited Mg–phenol networks also significantly increased the hydrophilicity compared to that of the NaOH coating (p < 0.001).

Corrosion Resistance Analysis

Figure a,b shows the electrochemical impedance spectroscopy (EIS) and potentiodynamic polarization (PDP) curves. The Nyquist impedance spectra clearly show the kinetics of electrochemical corrosion of the samples. Figure c shows the corresponding equivalent circuit for the coated samples,[39−41] where Rs, Qdl, Rct, Qp, and Rp represent the solution resistance, double-layer capacitance, reaction resistance associated with the interfacial charge-transfer reaction, coating capacitance, and coating resistance, respectively. For the raw sample, Rct, Qp, and Rp were ignored. Using the equivalent circuit and obtained EIS curves, curve fitting analysis was performed using EC-Lab software; the corresponding data for each element of the circuit were extracted and are summarized in Table . Rct is known to have the largest effect on the impedance of the corrosion reaction.[32] As shown in Table , the Rct value of the 0.6%Mg@TA sample was even lower than that after the first step of coating (i.e., coating with NaOH), possibly owing to cracking during the coating process (Figure ). The increase in the Mg concentration caused an increase in Rct. Therefore, it can be concluded that the porous coating of Mg–phenolic networks with the optimal Mg concentration can significantly enhance the corrosion resistance of Mg-based alloys. The obtained curves from PDP tests (Figure b) and the corresponding values for the Ecorr and Icorr extracted from the polarization curves using the Tafel method (Table ) had almost the same trend as that of EIS curves (Figure a) and the Rct value among the considered samples (Table ). The 2.4%Mg@TA and 3.6%Mg@TA samples delivered the lowest value for the corrosion current density (Icorr) among all of the samples (Table ). The value of the Icorr for these samples is almost 5 times lower than that of the raw sample. As far as Icorr is directly proportional to the corrosion activities,[42] we can claim that the Mg–phenolic networks with a proper coating procedure deliver proper electrochemical stability and promote anticorrosion performance of the Mg-based substrate kinetically. Moreover, the free corrosion potential (Ecorr) of the coated samples shifted toward the more positive potentials, indicating the thermodynamic reaction activity of the coated samples in phosphate-buffered saline (PBS) was more suppressed compared to that of the uncoated sample.[42]

Figure 3

Comparison between electrochemical behavior of different samples: (a) Nyquist EIS spectra and (b) PDP curves. (c) Equivalent circuit model used to interpretation of the corrosion behavior of the coated samples.

Table 1

Corresponding Parameters Determined from Curve Fitting of the EIS and PDP Results

	Nyquist EIS spectra					polarization curves
samples	Rs (Ω cm2)	Qdl (μF cm–2 sn–1)	Rct (kΩ cm2)	Qp (μF cm–2 sn–1)	Rp (kΩ cm2)	Ecorr (mVSCE)	Icorr (μA cm–2)
raw	22.0	71.1	1.619			–1896	5.616
NaOH	32.5	19.9	2.042	629	1.235	–1790	4.046
0.6%Mg@TA	19.8	18.3	1.860	1371	0.537	–1876	5.186
1.2%Mg@TA	16.4	13.9	2.893	961	1.056	–1589	2.190
2.4%Mg@TA	23.7	13.7	4.543	1074	0.903	–1598	1.205
3.6%Mg@TA	30.3	14.5	4.549	1039	0.942	–1561	1.191

Taken together, all of these results can be considered as convincing evidence to prove that performing Mg–phenolic network strategy with a proper initial Mg concentration during the coating procedure can serve a proper corrosion protection for the Mg-based substrate. Furthermore, the Rct and Icorr values of the 2.4%Mg@TA and 3.6%Mg@TA samples were all almost similar (Table ), indicating that initial Mg concentrations exceeding 2.4% would not further improve corrosion resistance.

Degradation Behavior

Figure a,b shows the Mg concentration and the volume of H2 gas evolution within 3 weeks and at five time points during immersion of the samples in PBS solution at 37 ± 0.5 °C, respectively. The raw sample clearly released the highest quantity of Mg ions and hydrogen evolution, indicating that it had the highest degradation rate. With increasing initial Mg concentration in the coating procedure, the degradation rate decreased that led to less release of Mg ions and hydrogen evolution. Additionally, for all of the samples, the Mg concentration in the extracted solution increased dramatically from day 1 to day 3 and then gradually increased from day 3 to day 7. The reason may be the formation of insoluble salty clusters on the sample surfaces, which decreased after the surface area was exposed to the solution.[43] It is worth mentioning that the trend of changing in the quantity of the measured hydrogen evolution (Figure b) are in agreement with that of the Mg concentration (Figure a) in all of the considered time points.

Figure 4

Variation of four different parameters during 3 weeks degradation of different samples in PBS solution at 37 ± 0.5 °C: (a) concentration of released Mg element, (b) hydrogen gas evolution, (c) pH value, (d) surface morphology; (e) energy-dispersive X-ray (EDX) spectrum and weight% elemental quantitative data representative of the insoluble cluster, the hydroxide layer, and substrate fragments.

Figure c shows the pH of the collected PBS solution after immersion of the samples. The differences in the degradation rates of the samples resulted in differences in the hydroxide ion (OH–) release and pH.[4] The pH for the raw and 0.6%Mg@TA samples reached 8.75 owing to the absence of a suitable passivation layer for the raw sample and cracking that occurred during coating of the 0.6%Mg@TA sample. The NaOH-treated and 1.2%Mg@TA samples produced moderate increases in pH to 8.70 and 8.74, whereas the 2.4%Mg and 3.6%Mg samples caused the smallest pH changes, to 8.34 and 8.40, respectively, on day 1. As the immersion time increased to 3 days, the pH of all of the samples increased. When the immersion time increased to 7 days, a temporary hydroxide passivation layer that was thermodynamically stable at pH values above 8.3 may have formed (Figure d).[43] This layer caused the degradation rate to increase more slowly and protected against significant changes in the pH, especially for the raw sample. Figure d shows the surface morphology of the samples after 7 days of immersion in PBS. Although a more slatelike structure formed on the surface of the raw sample, it still had a larger contact area than the coated samples owing to severe cracking. Figure e shows the EDX profiles of the insoluble cluster shown in Figure d. Despite the formation of cracks with submicron widths, the Mg–phenolic films with different initial Mg concentrations were quite stable and decreased the corrosion rate, especially for the 1.2%Mg@TA, 2.4%Mg@TA, and 3.6%Mg@TA samples. Even the NaOH-treated sample maintained its nanoscale porous coating on day 7 (Figure d). After day 7, the samples exhibited different degradation rates. The Mg ion release of the raw sample continued to increase, but the amount of release tended to be steady at days 15 and 21. For the NaOH-treated sample, the release of Mg ions increased more slowly, indicating better corrosion resistance compared to the raw sample. For the samples coated with the Mg–phenolic networks, the degradation rate decreased from day 7 to day 21. Figure e shows the EDX profiles of these hydroxide films. However, the presence of dissolved chloride ions (Cl–) at high concentrations in the PBS solution, which mimicked that of blood plasma,[44] would accelerate the corrosion rate of Mg and convert some of the temporary hydroxide coating into soluble MgCl2, further increasing the exposed area.[14] At high pH values, formation and dissolution of the hydroxide layer are typically competing processes. Free Cl– ions played a prominent role in the increase in pH. Furthermore, the cracks on the substrate released more Mg ions and thus increased the pH to 9.23, 9.22, 9.24, 9.11, 9.06, and 9.0.3 for the raw, NaOH, 0.6%Mg@TA 1.2%Mg@TA, 2.4%Mg@TA, and 3.6%Mg@TA samples, respectively. The pH of the raw samples reached 9.36 at day 15, whereas the pH of the modified samples decreased compared to that on day 7. Insoluble salty clusters formed on the surface of the raw sample. Despite the cracking, the coating layer was present and connected to the substrate of the 2.4%Mg@TA and 3.6%Mg@TA samples. In contrast, the Mg–phenolic networks were separated from the intermediate layer of the 1.2%Mg@TA samples. A uniform porous oxidized layer was formed on the surface of the 0.6%Mg@TA and NaOH-treated samples. However, the 0.6%Mg@TA sample was more severely corroded than the NaOH sample owing to its highly porous surface. This result indicates that the Mg–phenolic coating was not stable enough to passivate the substrate at day 15.

The cracks on the surfaces of all of the samples were wider on day 21 (Figure d), and more phosphate-based cluster structures had formed. The 2.4%Mg@TA and 3.6%Mg@TA samples still had stable Mg–phenolic networks. In the 1.2%Mg@TA samples, the Mg–phenolic networks had begun to peel off from the intermediate layer. The tiny and large clusters of the 0.6%Mg@TA samples decreased the corrosion activity. The NaOH-treated samples were divided into coated and uncoated areas as the intermediate coating layer was removed from the Mg substrate. As on days 7 and 15, the morphology of the raw sample consisted of the oxidized substrate and cluster structure. The pH of all of the samples on day 21 was lower than that on day 15 owing to the precipitation of more insoluble salty clusters containing phosphate on the surface of the samples resulting from the high concentration of phosphate in the PBS solution.[43] Furthermore, part of the Mg–phenolic passivation film was still present on the 2.4%Mg@TA and 3.6%Mg@TA samples.

Overall, the changes in Mg ion release, H2 gas evolution, and pH were clearly consistent. In light of the surface morphology, the intermediate and Mg–phenolic layers as well as the formation of insoluble phosphate-based clusters were the main factors affecting the degradation behavior.

In Vitro Osteocompatibility Assessment

Short-Term Cytotoxicity Assessment

Figure a summarizes the results of short-term cytotoxicity assessment of the MC3T3-E1 cells using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay after 6, 12, and 24 h. Among all of the groups, the 2.4%Mg@TA and 3.6%Mg@TA samples showed an acceptable level of viability during the first 24 h of incubation of the bonelike cells, indicating that they provide a suitable physiological ambient for stimulating cell adhesion and proliferation. Figure b shows the pH values of the culture media containing MC3T3-E1 cells added to the samples during 24 h. The initial pH value of the control group (i.e., the clear medium with cells only) was 7.77 and varied between 7.78 and 7.92 during that time period. The 2.4%Mg@TA and 3.6%Mg@TA samples clearly caused the least fluctuation in the pH of the culture medium, and the corrosion rate was stable. Culture media with alkaline pH values of 8.40–8.50 support the highest osteoblast viability in vitro.[45−47] The average pH values of the 2.4%Mg@TA and 3.6%Mg@TA samples during the first day of cell culture were 8.42 and 8.39, respectively.

Figure 5

(a) Average absorbance of MC3T3-E1 cells seeded on different samples after 6, 12, and 24 h.***, ●●●, ■■■, +++, and ###p < 0.001, differences from the control group, 3.6%Mg@TA, 2.4%Mg@TA, 1.2%Mg@TA, and NaOH at each time point, respectively. (b) pH value changes of the culture media during 24 h incubation of MC3T3-E1 cells with different samples.

Morphology Assessment of MC3TC-E1 Cells

The morphological changes in the seeded MC3T3-E1 cells on the samples were investigated using scanning electron microscopy (SEM) imaging to compare the effects of the Mg–phenolic coating on the adhesion and proliferation of the bonelike cells.

Figure presents SEM micrographs showing the morphology of MC3T3-E1 cells cultured on the samples at the end days 1 and 3. On day 1, the 3.6%Mg@TA, 2.4%Mg@TA, and 1.2%Mg@TA samples showed high osteocompatibility, with cell attachment and suitable spreading owing to the stable moderately alkaline pH and suitable surface wettability of the protective coating.

Figure 6

SEM images of the morphology of MC3TC-E1 cells cultured on different samples at day 1 and day 3 of incubation.

On day 3, the morphology of cells cultured on the samples indicated cell proliferation. A few cells became attached to the raw and 0.6%Mg@TA samples, indicating that they could not provide a suitable biocompatible physiological environment for cell proliferation. In contrast, the 3.6%Mg@TA, 2.4%Mg@TA, 1.2%Mg@TA, and NaOH-treated samples exhibited improved biocompatibility, with more cells and acceptable levels of proliferation. Specifically, the 3.6%Mg@TA and 2.4%Mg@TA samples showed the best osteocompatibility among all of the groups. The above results were consistent with the short-term cytotoxicity assessment (Section ), wettability results (Section ), and corrosion resistance experiment (Section ). That is, the 3.6%Mg@TA and 2.4%Mg@TA samples with the proposed porous osteocompatible Mg–phenolic networks exhibited acceptable corrosion resistance by mitigating the negative effect of excessive release of H2 bubbles and exhibited suitable wettability for adhesion and proliferation of bonelike cells.

Indirect Viability Assessment

Surface properties such as roughness and wettability, the pH of the culture medium, the concentration of Mg ions, and oxygen deficiency due to the release of H2 gas bubbles can be considered to be the four main parameters directly affecting the in vitro viability, in particular, the medium-term cytocompatibility, of the Mg alloys. Indirect viability assessment is a common method of evaluating the in vitro medium-term osteoblastic cytocompatibility of such alloys and their capacity to mitigate the effects of these surface properties, increase of pH, and H2 gas evolution. That is, indirect viability assessment can provide better information about the medium-term influence of the degradation of Mg-based alloys on the surrounding physiological environment in vitro. Therefore, to focus on the effect of the release of Mg2+ ions on the viability of osteoblast-like cells, indirect viability assessment of extracted media mixed with 1× Dulbecco’s modified Eagle’s medium (DMEM) at a 1:1 ratio was performed. The average pH values of the media collected from the 3.6%Mg@TA, 2.4%Mg@TA, 1.2%Mg@TA, 0.6%Mg@TA, NaOH-treated, and raw samples were 8.12, 8.18, 8.25, 8.39, 8.30, and 8.41, respectively. Figure shows the results of the MTT assay after 1 and 3 days of cell culture with the prepared culture medium for each group. The absorbance of all of the samples obviously differed significantly from that of the control group (p < 0.001) on day 1. This finding confirms the stimulating effect of Mg corrosion products (especially Mg2+ ions) on the proliferation of bonelike cells reported by Wu et al.[7] Furthermore, Galow et al.[46] have reported that culture media with alkaline pH enhance the viability of osteoblast-like cells in vitro. They claimed that higher pH values of up to 8.4 cause the higher viability.

Figure 7

Average absorbance of MC3T3-E1 cells after 1 and 3 days incubation with different culture media. ***, ●●●, ■■■, +++, and ###p < 0.001, differences from the control group, 3.6%Mg@TA, 2.4%Mg@TA, 1.2%Mg@TA, and NaOH at each time point, respectively.

Figure shows confocal images of the morphology of the cultured MC3T3-E1 cells in the media extracted on day 1. Among all of the groups, the highest numbers of adhered and proliferated cells were observed on the raw and 0.6%Mg@TA samples, which is consistent with the MTT results (Figure ). On day 3, the same trend as day 1 was observed for MTT results. There was still a significant difference between the control group and the other groups (p < 0.001). The increase in the absorbance of all of the groups from day 1 to day 3 indicates enhancement of the proliferation mechanisms of the cells. The confocal images of the cells on day 3 are shown in Figure . The cells clearly proliferated well. The differences in cell proliferation among the groups can be explained in terms of the difference in the medium pH.

Figure 8

Confocal images of the cytoskeletal arrangement of MC3TC-E1 cells incubated with different culture media at days 1 and 3.

Figure 9

(a) Schematic of the conducted rat cranial bone defect surgery. Two 5 mm diameter critical-sized defects were made on the skull; (b) comparison between uncoated and coated Mg implants in terms of the formation of subcutaneous hydrogen gas cavity in the implantation area; (c) front and back views of the reconstructed microcomputed tomography (μCT) images obtained at 4 weeks after implantation surgery. The blue circle indicates the initial size of the defects.

In summary, the indirect viability test showed that the dose of Mg ions from uncoated Mg-based alloys (i.e., the raw sample) resulted in an alkaline pH in vitro, enhancing proliferation. Furthermore, the 3.6%Mg@TA and 2.4%Mg@TA samples have not only an effective coating of Mg–phenolic networks and the most osteocompatible surface among all the groups, which results in reduced cytotoxicity, but also can still release corrosion products, improving the viability of the biological fluid in the zone around the implantation area.

In Vivo Implantation of Mg Disks in Rat

After the implantation surgery according to the procedure describes in Section (Figure a) and during the 4 week period before the final surgery for harvesting the skull bone blocks, all of the rats were in good health and showed wound healing without any severe side effects.

Observation of Subcutaneous Hydrogen Gas Cavity

For Mg-based implants with a higher degradation rate, subcutaneous hydrogen gas cavities can be easily observed in the region around the implantation area.[6,16,17] Regardless of the size of the Mg-based implant, the formation of such cavities can be considered convincing evidence for the high degradation rate and undesirable corrosion resistance, which can limit the biocompatibility of such biomaterials. In previous studies, for example, in Lin et al.,[48] photographs of the shaved skin covering the defects in the calvarial region have been used to compare the degradation behavior of different Mg-based implants. Figure b compares the formation of subcutaneous hydrogen gas pockets in the uncoated and coated Mg implants 4 weeks after implantation. An obvious gas cavity formed in the implantation area of the uncoated Mg disk. In contrast, no obvious gas pockets formed near the Mg disks coated with Mg–phenolic networks, indicating that the passivation coating decreased the degradation rate of the Mg implants, allowing sufficient time for blood flow to remove the created hydrogen gas bubbles in the implantation area and prevent accumulation of these bubbles. Preventing or at least mitigating the formation of H2 gas cavities can increase the oxygen supply to the defect area and improve the biocompatibility of Mg-based implants.

Micro-CT Findings

Figure c compares focused front and back views of the reconstructed micro-CT images of defect area with uncoated and coated Mg implants as well as the sham control group obtained 4 weeks after implantation. Regardless of the volume of bone formed around the Mg implant, the difference in the degradation of the uncoated and coated samples is obvious. In the uncoated Mg implant, some localized corrosion pits and/or holes were formed, indicating a higher corrosion rate than that of the implants coated with Mg–phenolic networks, on which there were no obvious voids or holes. The accumulation of hydrogen gas near the uncoated sample can be attributed to these corrosion features (Figure b). As the degradation rate increased, the formation of subcutaneous gas pockets became more likely.

Conclusions

In this research, nanostructured Mg–phenolic networks were formed on a Mg-based alloy by dip-coating in a solution containing Mg and phenol groups. The two main goals were to enhance the corrosion resistance and osteocompatibility of Mg-based degradable implant materials. The following results were obtained:

The concentration of Mg ions in the coating solution was found to be the main factor affecting the performance of the coating. The corrosion resistance increased with increasing concentration of Mg ions in the coating solution. The corrosion resistance of the 3.6%Mg@TA and 2.4%Mg@TA samples increased by 180% compared to that of the raw sample.

Observations of the degradation behavior showed that the Mg–phenolic network coating can be quite stable in the first few weeks of immersion in PBS solution.

Direct short-term cytotoxicity assessment indicated that the 3.6%Mg@TA and 2.4%Mg@TA coatings significantly decreased the cytotoxicity of the Mg-based substrate in the first day after implantation.

SEM images of the morphology of osteoblast-like cells showed that the Mg-based substrate with a suitable coating of Mg–phenolic networks had an osteocompatible surface that significantly enhanced cell adhesion and proliferation in the first few days after cell seeding.

Indirect medium-term viability assessment showed that the media extracted from even the samples with the highest Mg concentration in the passivation coating (i.e., 3.6%Mg@TA and 2.4%Mg@TA) enhanced the viability of the bonelike cells.

In vivo animal experiments revealed that coating Mg-based implants with appropriate passivation layers containing Mg–phenolic networks can control the degradation rate of the implants and mitigate the formation of subcutaneous hydrogen gas cavities in the implantation area.

Overall, passivation of Mg-based alloys with 3.6%Mg@TA and 2.4%Mg@TA coating solutions produced a surface with excellent osteocompatibility that protected the alloys from severe degradation and prevented the formation of excessive H2 gas bubbles and the resulting oxygen deficiency. Thus, our coating procedure with Mg–phenolic networks shows promise for widening the application of Mg-based degradable implants for bone regeneration.

Materials and Methods

Substrate Preparation

Casted ingots of Mg–2.8%Zn alloy were extruded to billets with a diameter of 10 mm. Then, the samples with a diameter of 6 mm and a thickness of 1 mm were machined from the extruded billets. All of the mentioned fabrication processes were carried out in Shaanxi Key Laboratory of Biomedical Metal Materials of Northwest Institute for Nonferrous Metal Research (Xi’an, China). Afterward, the disks were polished with SiC papers down to 4000 grit. Then, all of the samples were ultrasonically cleaned in 100% acetone (10 min) and 100% ethanol (10 min) and finally rinsed in deionized (DI) water to remove any surface contaminations.[49]

Mg–Phenolic Network Deposition

The coating procedure is summarized schematically in Figure . The Mg–phenolic networks were coated on disk samples in two steps. In the first step, the prepared Mg–2.8%Zn disks were immersed in NaOH with a pH of 10 for 90 min. A uniform nanosized film of Mg oxides [both Mg(OH)2 and MgO] can form on Mg substrates immersed in NaOH solution.[32,50] Then, the samples were dip-coated in a prepared solution containing 1 mg mL–1 TA (analytical grade, purity = 99%, Sigma-Aldrich) combined with MgCl2 (analytical grade, purity = 99%, Thermo Fisher Scientific, Australia) at four concentrations (0.6, 1.2, 2.4, and 3.6 mg mL–1) in NaOH solution (pH = 10). After 30 min, the samples were removed and washed 3 times with the NaOH solution to remove any salty liquid from their surfaces. In the second step, this procedure was repeated using a fresh solution to perform the second and third cycles and to form a uniform Mg–polyphenolic network surface. To reduce undesirable oxidation, the coated samples were kept in a vacuum desiccator. The NaOH + 1% TA + 3.6% MgCl2, NaOH + 1% TA + 2.4% MgCl2, NaOH + 1% TA + 1.2% MgCl2, NaOH + 1% TA + 0.6% MgCl2, and NaOH coating solutions and the samples coated with them are denoted 3.6%Mg@TA, 2.4%Mg@TA, 1.2%Mg@TA, 0.6%Mg@TA, and NaOH, respectively.

Surface Morphology Assessment of the Mg–Phenolic Film

To observe the surface morphology of the coated samples after the coating procedure mentioned in Section , the samples were coated with a thin layer (∼10 nm) of gold using a sputter (EM SC005 Gold Coater, Leica). Then, the surface morphology of the different samples was observed using scanning electron microscopy (SEM, JEOL 7001F, Japan). A secondary electron detector (SED) was used to capture the detailed shape and surface information. Parallel samples were prepared to observe the cross-sectional view of the coated samples. To that aim, the samples were mounted inside the epoxy resin and after the cutting and polishing procedures the cross section of the samples was coated with a very thin layer (∼4 nm) of platinum using a sputter (EM ACE600 Platinum Coater, Leica). Finally, the cross-sectional view of the coating layer of the different samples was captured using SEM (SEM, JEOL 7001F, Japan). Backscattered electron (BSE) imaging was applied to extract chemical contrast information.

Surface Characterization of the Mg–Phenolic Film

The chemical composition of the sample surfaces was examined by X-ray photoelectron spectroscopy (XPS) using the XPSpeak 4.1 package to analyze the high-resolution spectra and fit the peaks to bonding states. The wettability of the surfaces was assessed by conducting a simple water contact angle test at room temperature and 50% relative humidity using the sessile-drop method on an inclined plate. The roughness of the samples was quantified by atomic force microscopy (AFM) in a scanning range of 20 × 20 μm2. Each experiment was repeated at least 3 times for statistical analysis.

Corrosion Resistance Assessment

The effect of the Mg–phenolic coating on the corrosion resistance was examined by electrochemical impedance spectroscopy (EIS) with an electrochemical unit (VSP, Bio-Logic Science Instruments, France) sample as the working electrode, a saturated calomel electrode as the reference electrode, and a platinum foil as the counter electrode. Phosphate-buffered saline (PBS) solution was used as the electrolyte. The side and bottom surfaces of the disk samples were covered by silicon glue, and only the top surface (28.3 mm2) was exposed to the electrolyte. All of the tests were performed at 37 ± 0.5 °C. The EIS data were recorded from 200 kHz to 100 mHz with a 10 mV sinusoidal perturbing signal.[33]

Accordingly, the potentiodynamic polarization (PDP) tests were conducted at a scanning rate of 1 mV s–1. Then, the Tafel method with linear extrapolation of the active polarization zone in the cathodic polarization section of PDP curves was applied to extract the current density (Icorr) and corrosion potentials (Ecorr).[32] The EC-Lab software package (V11, Bio-Logic Science Instruments, France) was applied to analyze the EIS and PDP results.

The degradation behavior was examined by immersing the samples in PBS solution (pH = 7.4) for 3 weeks at 37 °C. The surface/volume ratio was 56 mm2 mL–1. The time checkpoints were set as day 1, day 3, week 1, week 2, and week 3. The PBS solutions were refreshed every 2 days with 500 μL of a fresh solution using a method similar to that mentioned in refs (51) and (52). At the end of each interval, the concentration of Mg2+ ions released into the extracted PBS solution was determined using inductively coupled plasma optical emission spectroscopy (ICP-OES 700, Agilent) after measuring the pH value of the extracted using a pH meter (Thermo Fisher Scientific) in each time checkpoints. Parallely, similar samples were prepared to measure the evolution of H2 gas during the degradation of the samples in PBS with a test area of 0.283 cm2 using an eudiometer and based on the methodology presented in ref (32). For surface morphology assessment, the samples were ultrasonically washed with distilled water for 5 min to remove the salty crystals deposited on the surface from the PBS solution. Scanning electron microscopy (SEM) imaging and energy-dispersive X-ray (EDX) spectroscopy were used to assess the surface morphology and chemical composition of the samples, respectively. Three samples were tested for each interval.

In Vitro Biological Tests

Osteoblast Cell Morphological Analysis

To study the in vitro osteocompatibility of the coated samples, the cell morphology, adhesion, and proliferation of the MC3T3-E1 osteoblast cell line (Sigma-Aldrich) on the coated samples were monitored. A suspension of 8 × 103 cells in 1× Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum as well as 1% pericyte growth supplement was seeded on the surfaces of the samples, which were held under 5% CO2 at 37 °C. After the samples were cultured for 3 days, the culture medium was extracted, and all of the samples were washed with PBS, fixed with 3% glutaraldehyde for 30 min, and then washed again 3 times with PBS. Next, the samples were dehydrated according to previous protocols.[53] Finally, the morphology of the cells was observed using SEM (JEOL 7001F, Japan).

Direct Cytotoxicity Assessment

Three timelines (6, 12, and 24 h) were used, and 1.6 × 104 MC3T3-E1 cells were directly seeded on the samples.[32] The 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, Sigma-Aldrich, Australia) assay was conducted to assess the cytotoxicity.[54] Briefly, after cell seeding, the samples were kept in the incubator at 37 °C and 5% CO2. Then, 15 μL of the MTT solution with a concentration of 5 mg mL–1 was added to each well and incubated at 37 °C and 5% CO2 for 4 h. Next, 150 μL of dimethyl sulfoxide (DMSO, Fisher Scientific, U.K.) was added to each well and incubated at room temperature for 10 min. Finally, a microplate reader (Benchmark Plus, Bio-Rad) at 570 nm was applied to measure the absorbance. Simultaneously with the cell culture experiments, parallel samples were prepared to measure the pH value of the culture medium using a pH meter (Thermo Fisher Scientific) to investigate the change in pH with time.

The MTT assay similar to that described in Section was conducted to assess the cell viability. Each sample was first soaked in 150 μL of the prepared culture medium for 24 h at 37 °C and 5% CO2. Then, the extracted media were diluted with 1× DMEM to a 1:1 ratio and used for subsequent experiments, including fixed-cell confocal imaging and viability assessment. For confocal imaging, 1.6 × 104 MC3T3-E1 cells were seeded on a round coverslip slide. At the same time, 8 × 103 MC3T3-E1 cells were seeded on a 96-well plate. Three wells were used per group for each experiment. Fresh media were used as the control groups in both experiments. The cells were cultured for 1 and 3 days at 37 °C and 5% CO2. At the end of each timeline, the viability of each group of samples was assessed, and the absorbance was measured using the procedure described in Section .

To investigate the cytoskeletal arrangement of the cells using multiphoton confocal microscopy, similar samples in different groups were prepared using the fixation protocol presented by Guo et al.[55] and Marchesano et al.[56] Briefly, at the end of each time point, the culture medium was extracted and the MC3T3-E1 cells were washed with 1× PBS and fixed with 4% paraformaldehyde. Next, 0.1% Triton-X was added to permeabilize the cells (1 h). Then, the cells were kept in tetramethylrhodamine (TRITC)-conjugated phalloidin for 1 h. Finally, the cell nuclei were stained by incubation in 4′,6-diamidino-2-phenylindole for 10 min at room temperature.

In Vivo Animal Tests

A total of 10 albino rats of the Wistar strain with initial weights of 350–375 g were used in the present work. Each rat was kept and housed in a separate cage and fed with a standard diet and water. All animal procedures including animal selection, preparation, and surgical protocol were conducted based on protocols approved by the local ethical committee and laboratory animal administration rules of China.

Study Design

On the basis of the results of the material characterization and in vitro experiments, an in vivo animal study was designed to investigate the effect of the Mg–phenolic network coatings on the corrosion resistance and biocompatibility of Mg-based implants using the rat calvarial model. Two critical-sized defects each 5 mm in diameter were created in the parietal bone of each rat calvarium. One defect was used as the sham group, and a Mg disk was implanted on the other. Two types of Mg disks were used: uncoated (raw) and coated with Mg–phenolic networks at a Mg concentration of 3.6 mg mL–1. Five rats were allocated to each of the two Mg implant groups for 4 weeks of observation after surgery.

Surgical Procedure

The surgical procedure was based on the steps mentioned in refs (48) and (57). The surgical procedure was based on the steps mentioned in refs (40) and (41). Briefly, the rats were sedated in a chamber with 4% isoflurane in 100% O2 and then anesthetized by intraperitoneal injection of 15 mg kg–1 Zoletile and 10 mg kg–1 Rompun. Under local anesthesia with 2% lidocaine hydrochloride containing 1:100 000 epinephrine, after disinfection with povidone iodine, a middle skin incision was made on a shaved skin area on the skull and a full-thickness flap was reflected. Under copious saline irrigation, two standardized round defects each 5 mm in diameter were created on the left and right lateral parietal bones of the rat skull using a trephine bur. After the two critical-size calvarial bone defects, a Mg disk with a diameter of 5 mm and thickness of 0.8 mm was implanted in the right defect and the left defect was used as the sham group. Then, the periosteum and shaved skin were repositioned and sutured properly. After a healing period of 4 weeks, five rats in each group were euthanized in a CO2 chamber. Subsequently, block sections of the rat skull will were collected and fixed in a 4% paraformaldehyde solution.

Photographs of hydrogen gas accumulation in the implantation area of the Mg disk and any subcutaneous hydrogen gas cavities that formed within 4 weeks after implantation were obtained before the animals were harvested.

Micro-CT Analysis

A micro-computed tomography (μCT) scanner was chosen (μCT 40, Scanco Medical, Brüttisellen, Switzerland) to assess the harvested and fixed rats’ cranial bone blocks at an isotropic voxel size of 18 μm. The X-ray source voltage and current were controlled at 90 kV and 110 μA, respectively. Finally, the images were reconstructed and processed using CTVox volume rendering software (Bruker).

Statistical Analysis

Statistical analyses were performed via one-way analysis of variance (ANOVA) using SigmaPlot ver13.0 (Systat Software, Inc.) software package.