Surface Modiﬁcation of Mg0.8Ca Alloy via Wollastonite Micro-Arc Coatings: Signiﬁcant Improvement in Corrosion Resistance

: Biodegradable materials are currently attracting the attention of scientists as materials for implants in reconstructive medicine. At the same time, ceramics based on calcium silicates are promising materials for bone recovery, because Ca 2+ and Si 2+ ions are necessary for the mineralization process, and they take an active part in the formation of apatite. In the presented research, the protective silicate biocoatings on a Mg0.8Ca alloy were formed by means of the micro-arc oxidation method, and the study of their morphology, structure, phase composition, corrosion, and biological properties was carried out. Elongated crystals and pores were uniformly distributed over the surface of the coatings. The coated samples exhibited remarkable anti-corrosion properties in comparison with bare magnesium alloy because their corrosion current decreased 10 times, and their corrosion resistance increased almost 100 times. The coatings did not signiﬁcantly affect the viability of the cells, even without the additional dilution of the extract, and were non-toxic according to ISO 10993-5: 2009. In this case, there was a signiﬁcant difference in toxicity of the pure Mg0.8Ca alloy and the coated samples. Thus, the results demonstrated that the applied coatings signiﬁcantly reduced the toxicity of the alloy. chemical composition of wollastonite was as follows: wt.%: SiO 2 50.0–53.0, CaO 45.0–48.0, MgO 0.4–1.0, Al 2 O 3 0.1–0.3, and Fe 2 O 3 0.05–0.2. Wollastonite of this kind is characterized by high purity, which is important when creating biocoatings for implants.


Introduction
In the modern world, the number of diseases associated with damage to human bone tissue caused by injuries, fractures, and congenital genetic abnormalities has sharply increased [1]. Population aging and an increase in the amount of overweight people exacerbate this trend. Therefore, meeting the need for treatment and human bone replacement has become an urgent clinical, scientific, and socio-economic problem [1,2].
Micro-arc oxidation (MAO), also called plasma electrolytic or anode spark deposition, is a prospective method of modification of valve metals' surfaces via the formation of ceramic-like coatings [49,50]. MAO coatings can be used for various applications-for example, to form a hard and wear-resistant layer on aluminum, titanium, or magnesium alloys [50]; to create a chemically resistant anticorrosive layer [51][52][53]; or even for the coloring of metal surfaces [54]. By varying the parameters of the deposition process and the composition of the compounds included in the electrolyte, it is possible to form the coatings with a developed surface morphology and with high osteogenic properties [49,50]. Therefore, the aim of the present research was the synthesis of micro-arc coatings on the surface of a Mg0.8Ca alloy with the participation of natural wollastonite particles, as well as the investigation of how the structure, morphology, and phase composition of the coatings influence their corrosion resistance, behavior in biological fluid, and toxicity with respect to the 3T3 fibroblast cell line.

Sample Preparation
Mg-0.8 wt.% Ca (Mg0.8Ca) is the magnesium alloy, which was used as the material for the substrate. This alloy was produced at Helmholtz Zentrum (Geesthacht, Germany). We describe the process of the alloy's production elsewhere [27]. Samples from the Mg0.8Ca alloy were in the shape of a parallelepiped, with sizes of 10 × 10 × 1 mm 3 . Experimental samples were subjected to grinding with 1200 grit sandpaper containing silicon carbide. After this, the grinding samples were ultrasonically washed in distilled water for 10 min. Such treatment led to the roughness of the prepared samples being Ra = 0.3-0.6 µm.
The coatings were synthesized using the MAO method in the anodic potentiostatic regime for 5 min, under the applied voltages of 350-500 V, with the help of Micro-Arc 3.0 software, as reported in previous papers [27,31]. The electrolyte containing wollastonite (CaSiO 3 ), NaOH, Na 2 SiO 3 , and NaF was used for the coating's deposition. Wollastonite MIVOLL®05-96 (ZAO GEOKOM, Kaluga region, Russia) was used in the work. Endless chains of silicon-oxygen tetrahedra make up the structure of wollastonite. The Ca 2+ ions are located between the chains as if they were "stitching" them [55]. The recurrence period in wollastonite chains is three tetrahedrons (Figure 1a). Due to this structure, wollastonite crystals have an elongated shape, which is drawn-out along the chain (Figure 1b). In this study, wollastonite with the following particle sizes was used (L-length): average L av , maximum L max, and minimum L min equal to 35 µm, 130 µm, and 4 µm, respectively. The are located between the chains as if they were "stitching" them [55]. The recurrence pe in wollastonite chains is three tetrahedrons (Figure 1a). Due to this structure, wollasto crystals have an elongated shape, which is drawn-out along the chain (Figure 1b). In study, wollastonite with the following particle sizes was used (L-length): average maximum Lmax, and minimum Lmin equal to 35 μm, 130 μm, and 4 μm, respectively. chemical composition of wollastonite was as follows: wt.%: SiO2 50.0-53.0, CaO 45.0-MgO 0.4-1.0, Al2O3 0.1-0.3, and Fe2O3 0.05-0.2. Wollastonite of this kind is character by high purity, which is important when creating biocoatings for implants.

Experimental Methods
The coating's microstructure, morphology, and elemental composition were stu using an LEO EVO 50 scanning electron microscope (SEM, Zeiss, Jena, Germ equipped with an energy dispersive X-ray spectroscope (EDX, INCA, Oxfrod In ments, Abingdon, UK), and a JEM-2100 transmission electron microscope (TEM, Jeol Musashino, Akishima, Tokyo, Japan) in the "Nanotech" Common Use Center of the I tute of Strength Physics and Materials Science Siberian Branch of the Russian Academ Sciences (Tomsk, Russia). The surface roughness was estimated as average roughness using a Hommel-Etamic T1000 profilometer (Jenoptic, Jena, Germany) at the Nationa search Tomsk Polytechnic University. The profile characteristics, such as traverse le and measuring rate, were 6 mm and 0.5 mm/s, respectively. The X-ray powder diffrac (XRD) method made it possible to determine the coating's phase composition (X DRON-7, Burevestnik, Russia, "Nanotech" center at ISPMS SB RAS) in the angular r 2θ = 5-90°, with a scan step of 0.02° with Co Kα radiation (λ = 0.17902 nm). The Data of the Joint Committee on Powder Diffraction Standards (JCPDS) was used for the p identification and interpretation of X-ray profiles. The obtained XRD profiles were to calculate the volume ratios of the crystalline and amorphous phases using Rietvel finement. Fourier-transform infrared spectroscopy (FTIRS) was performed using an pha IR-spectrometer (Bruker, Karlsruhe, Germany) in reflection mode, in the wave n ber range of 1500-500 cm −1 .

Experimental Methods
The coating's microstructure, morphology, and elemental composition were studied using an LEO EVO 50 scanning electron microscope (SEM, Zeiss, Jena, Germany) equipped with an energy dispersive X-ray spectroscope (EDX, INCA, Oxfrod Instruments, Abingdon, UK), and a JEM-2100 transmission electron microscope (TEM, Jeol Ltd., Musashino, Akishima, Tokyo, Japan) in the "Nanotech" Common Use Center of the Institute of Strength Physics and Materials Science Siberian Branch of the Russian Academy of Sciences (Tomsk, Russia). The surface roughness was estimated as average roughness (Ra) using a Hommel-Etamic T1000 profilometer (Jenoptic, Jena, Germany) at the National Research Tomsk Polytechnic University. The profile characteristics, such as traverse length and measuring rate, were 6 mm and 0.5 mm/s, respectively. The X-ray powder diffraction (XRD) method made it possible to determine the coating's phase composition (XRD, DRON-7, Burevestnik, Russia, "Nanotech" center at ISPMS SB RAS) in the angular range 2θ = 5-90 • , with a scan step of 0.02 • with Co Kα radiation (λ = 0.17902 nm). The Database of the Joint Committee on Powder Diffraction Standards (JCPDS) was used for the phase identification and interpretation of X-ray profiles. The obtained XRD profiles were used to calculate the volume ratios of the crystalline and amorphous phases using Rietveld refinement. Fourier-transform infrared spectroscopy (FTIRS) was performed using an Alpha IR-spectrometer (Bruker, Karlsruhe, Germany) in reflection mode, in the wave number range of 1500-500 cm −1 .

Electrochemical Studies
The electrochemical properties of the pure alloy and coated samples were studied using a Versa STAT MC system (Princeton Applied Research, Oak Ridge, TN, USA). The measurements were carried out in a three-electrode cell K0235, with a 0.9% NaCl aqueous solution. This methodology is described in previous works [27].
The linear polarization resistance experiment was performed, progressing from −30 mV to 30 mV vs. OCP, at a scan rate of 0.167 mV/s. The polarization resistance, R p , was calculated from the linear potential-current density plot as the R p = ∆E/∆j, as recommended in [56]. Potentiodynamic polarization curves were obtained at a scan rate of 1 mV/s, which is typical for Mg alloys, as opposed to 0.167 mV/s. The Levenberg-Marquardt (LEV) method was used to fit the experimental polarization curve (potential, E, vs. current density, j) with the following Equation (1): which gives the best fit values of corrosion potential, E C , and corrosion current density, j C [57]. The measurements of the electrochemical impedance spectroscopy (EIS) were carried out using a sinusoidal perturbation signal with an amplitude r.m.s. of 10 mV. Versa STUDIO (Princeton Applied Research, Oak Ridge, TN, USA), ZView, and CorrView software (Scribner Associates, AMETEK, Mahwah, NJ, USA) were used for the experiment's control and analysis. Each experiment was performed on the three different samples.

Biological Studies
We studied the biodegradability of samples in biological fluids (0.85 wt.% NaCl), according to ISO 10993-5. The samples were immersed in the solution at 37 • C for 16 days. The samples' weight loss was calculated by the formula (2): where m 0 is the mass before the dissolution and m i is the mass after the dissolution. The mouse 3T3 fibroblast line was used for evaluation of toxicity (State Research Center of Virology and Biotechnology "VECTOR" Novosibirsk, Russia). Cells were grown in a DMEM medium supplemented with 2 mM L-glutamine (HyClone, Logan, UT, USA), 10% fetal bovine serum FCS (HyClone, USA), and 1% penicillin/streptomycin (HyClone, USA). The samples were extracted for 3 h at a surface/volume ratio equal to 1 cm 2 /mL of cell medium. To assess the effect of particles on the viability of the cell lines, an MTT assay was performed. The incubation with MTT solution was carried out for 2 h at 37 • C and 5% CO 2 . The optical density was determined on a Thermo Scientific Multiskan FC microplate spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) at a wavelength of 570 nm. Parametric methods with a confidence level of p ≤ 0.05 were used for statistical data processing. The determination of the pH of the cell medium was performed during the preparation of the extract within a 24 h period.

Morphology, Structure, and Elemental Composition of the W-Coatings
An anodic potentiostatic mode was used for the W-coating deposition using the MAO method. The coatings were formed at fixed voltages equal to 350, 400, 450, and 500 V. The graphs of the current density at the time of coating deposition, recorded at different process voltages, are shown in Figure 2a. At the initial moment of the coating deposition, the current density reached its maximum and then decreased. The reason for this was an increase in the thickness of the ceramic-like coating. The decrease in the current density occurred at various rates for the different voltages. At a low voltage of 350 V, weak microarc discharges were realized in the system, and the current density dropped to a minimum value almost immediately, during the first 20 s of the process. At the voltages of 400-450 V, the current density decreased within 100 s according to the exponential law. At 500 V, the current density decreased evenly over the entire deposition period.
With the increase in the process voltage, the intensity of the micro-arc discharges increased and the temperature in the region of the micro-arc discharges rose. When the wollastonite was exposed to heat, at a temperature of 1125 • C, the polymorphic transition of β-wollastonite to α-wollastonite (pseudowollastonite) took place [30]. At a temperature of 1544 • C wollastonite melts [58]. In this case, ions passed into the system as charge carriers and then participated in the formation of micro-arc discharges. Mortazavi et al. [59] reported that the total applied current in this process consists of an electron current caused by sparking and an ion current caused by ion diffusion. Thus, a higher applied voltage led to a more intense melt formation in the system, and the MAO process continued for a With the increase in the process voltage, the intensity of the micro-arc discharges increased and the temperature in the region of the micro-arc discharges rose. When the wollastonite was exposed to heat, at a temperature of 1125 °C, the polymorphic transition of β-wollastonite to α-wollastonite (pseudowollastonite) took place [30]. At a temperature of 1544 °C wollastonite melts [58]. In this case, ions passed into the system as charge carriers and then participated in the formation of micro-arc discharges. Mortazavi et al. [59] reported that the total applied current in this process consists of an electron current caused by sparking and an ion current caused by ion diffusion. Thus, a higher applied voltage led to a more intense melt formation in the system, and the MAO process continued for a longer period. The increase in the voltage of the MAO process from 350 to 500 V led to the increase in the thickness and roughness of the coatings from 40 to 150 μm, and from 6.5 to 10.5 μm, respectively ( Figure 2b). The surface morphology of the W-coatings' surfaces, and their cross-section image, are shown in Figure 3.
Elongated crystals can be seen on the coating's surface (Figure 3a,b,d,e). This form of crystal is common for natural mineral wollastonite. Moreover, pores with sizes of 2-7 μm are observed in the coatings. Spherical formations (bubbles) can be seen in the coating deposited at 500 V, in the field of the intense micro-arc discharge output. This "boiling" of the coating substance [60,61] is due to the melting of electrolyte components and the formation of the melt. SEM images of the cross-sections of the coatings show that the coatings have a loose porous structure (Figure 3c

O Kα Si Kα
Mg Elongated crystals can be seen on the coating's surface (Figure 3a,b,d,e). This form of crystal is common for natural mineral wollastonite. Moreover, pores with sizes of 2-7 µm are observed in the coatings. Spherical formations (bubbles) can be seen in the coating deposited at 500 V, in the field of the intense micro-arc discharge output. This "boiling" of the coating substance [60,61] is due to the melting of electrolyte components and the formation of the melt. SEM images of the cross-sections of the coatings show that the coatings have a loose porous structure (Figure 3c The element distribution tracks along the cross-section of the coating are shown in Figure 5. Mg and O prevail in the boundary between the coating and the magnesium substrate, which is explained by the oxide layer MgO formation. Closer to the surface layer, the amount of Mg increases again, as well as the amounts of Si and O.  The element distribution tracks along the cross-section of the coating are shown in Figure 5. Mg and O prevail in the boundary between the coating and the magnesium substrate, which is explained by the oxide layer MgO formation. Closer to the surface layer, the amount of Mg increases again, as well as the amounts of Si and O.

O Kα Si Kα
The energy spectrum ( Figure 5b) shows that Mg, Si, and O are the predominant elements in the cross-section of the coating, while the amount of Ca is lower inside the coating than on the coating's surface. Fluorine is also present, mostly in the coating's crosssection, because the MgF2 was formed together with the oxide in the protective layer between the substrate and the coating. The presence of sodium was revealed in the coatings. It is possible that it diffused from the electrolyte and embedded into the amorphous phase of the coatings. Increasing the applied voltage led to an increase in the amount of Ca, while the amount of Mg decreased both in the surface and in the cross-section of the coatings (Table  1). This indicates that with the increase in voltage a greater amount of the wollastonite was involved in the coating synthesis process.   The energy spectrum (Figure 5b) shows that Mg, Si, and O are the predominant elements in the cross-section of the coating, while the amount of Ca is lower inside the coating than on the coating's surface. Fluorine is also present, mostly in the coating's cross-section, because the MgF 2 was formed together with the oxide in the protective layer between the substrate and the coating. The presence of sodium was revealed in the coatings. It is possible that it diffused from the electrolyte and embedded into the amorphous phase of the coatings.
Increasing the applied voltage led to an increase in the amount of Ca, while the amount of Mg decreased both in the surface and in the cross-section of the coatings (Table 1). This indicates that with the increase in voltage a greater amount of the wollastonite was involved in the coating synthesis process.

Phase Composition of the W-Coatings
The XRD results demonstrated that wollastonite was the main crystalline phase ( Figure 6). Reflexes of wollastonite in the XRD patterns of the coatings were almost exactly repeated from the reflexes observed in the spectrum of the initial wollastonite ( Figure 6b). In addition, magnesium oxide (periclase) and magnesium silicate (forsterite) were identified in the coatings. The formation of the oxide layer took place at the initial moment of the micro-arc oxidation process, and was located at the boundary between the substrate and the coating. This fact is confirmed by the results of the elemental analysis ( Figure 5). Forsterite was formed by the interaction of silicate ions with a magnesium substrate, according to the reaction (3) [59].
With an increase in the applied voltage, the reflexes of periclase practically disa peared, and the intensity of the magnesium reflexes became lower. A larger amount of th amorphous phase appeared in the coatings, as evidenced by an increase in the halo. 10   Studies of the coating's structure were carried out using the Rietveld method [62,63 The volume ratio of the crystalline and amorphous phases was calculated (Figure 7). fundamental change in the W-coating's structure from crystalline to amorphous-crysta line occurred when the voltage increased from 350 V to 400 V. As Figure 2a demonstrate  With an increase in the applied voltage, the reflexes of periclase practically disappeared, and the intensity of the magnesium reflexes became lower. A larger amount of the amorphous phase appeared in the coatings, as evidenced by an increase in the halo.
Studies of the coating's structure were carried out using the Rietveld method [62,63]. The volume ratio of the crystalline and amorphous phases was calculated (Figure 7). A fundamental change in the W-coating's structure from crystalline to amorphous-crystalline occurred when the voltage increased from 350 V to 400 V. As Figure 2a demonstrates, this may be due to the fact that in this case the initial current density has almost doubled.
Studies of the coating's structure were carried out using The volume ratio of the crystalline and amorphous phases fundamental change in the W-coating's structure from cryst line occurred when the voltage increased from 350 V to 400 V this may be due to the fact that in this case the initial current Accordingly, the intensity of the micro-arc discharges melting and rapid solidification of the coating substance in crystalline layer [64]. The XRD results consistent with the dat method ( Figure 3).
Through the useof transmission electron microscopy (TE (SAD) patterns, light-field, and dark-field TEM images of t deposited at the different voltages were obtained (Figure 8).  Accordingly, the intensity of the micro-arc discharges increased, which led to the melting and rapid solidification of the coating substance in the form of the amorphouscrystalline layer [64]. The XRD results consistent with the data were obtained via the SEM method ( Figure 3).
Through the useof transmission electron microscopy (TEM), selected area diffraction (SAD) patterns, light-field, and dark-field TEM images of the fragments of W-coatings deposited at the different voltages were obtained ( Figure 8).
The  (Figure 8d) were 200-300 µm. Akermanite crystallites are located mainly on the surface of wollastonite particles. Therefore, it is assumed that the formation of akermanite occurred at the interface when wollastonite interacted with magnesium, according to the reaction (4).
On the phase diagram of the three-component system CaO-MgO-SiO 2 [58], the crystallization fields of wollastonite and akermanite, as well as those of akermanite and forsterite, border one another, confirming the possibility of the formation and simultaneous presence of these compounds in the coatings. Figure 9 shows the IR spectra of the wollastonite (a) and W-coatings formed at voltages of 350 and 500 V (b).
Metals 2021, 11, x FOR PEER REVIEW 10 of 21 300 μm. Akermanite crystallites are located mainly on the surface of wollastonite particles. Therefore, it is assumed that the formation of akermanite occurred at the interface when wollastonite interacted with magnesium, according to the reaction (4). 2CaSiO3 + Mg 2+ + 2OH -→ Ca2MgSi2O7 + H2O On the phase diagram of the three-component system CaO-MgO-SiO2 [58], the crystallization fields of wollastonite and akermanite, as well as those of akermanite and forsterite, border one another, confirming the possibility of the formation and simultaneous presence of these compounds in the coatings. In the FTIR spectra of natural wollastonite, the IR absorption peaks were observed in the region of 1056-1060 cm −1 , corresponding to the asymmetric stretching mode of Si-O-Si. The IR absorption peaks located at 960 and 896 cm −1 can be connected with the non-bridging silicon-oxygen bond of Si-O. The absorption bands located at 469 cm −1 correspond to the bending vibrational mode of Si-O-Si [32]. In addition, the bands of Si-O bonds' vibrations are superimposed on the non-bridging bonds of Ca-O in the range of 450-500 cm -1 [65]. In the FTIR spectra of metasilicates, the region of 750-550 cm −1 is of the greatest interest. The number of bands in this region makes it possible to judge the number of silicon-oxygen tetrahedrons in the chain recurrence period [66]. The recurrence period in the wollastonite chain is three tetrahedrons (Figure 1a), and three bands are observed in the IR spectrum: they are 681, 642, and 564 cm −1 , respectively. Figure 9b demonstrates a decrease in intensity and broadening of the main bands in the IR spectra of the W-coatings. Moreover, if the bands corresponding to the vibrations of bonds in the wollastonite chain are still preserved for the coating formed at 350 V, these bands are no longer present in the spectrum of the coating deposited at 500 V. This indicates a distortion of the crystalline structure and its partial amorphization. 2021, 11, x FOR PEER REVIEW 11 Figure 9 shows the IR spectra of the wollastonite (a) and W-coatings formed at ages of 350 and 500 V (b). In the FTIR spectra of natural wollastonite, the IR absorption peaks were observ the region of 1056-1060 cm −1 , corresponding to the asymmetric stretching mode of S Si. The IR absorption peaks located at 960 and 896 cm −1 can be connected with the bridging silicon-oxygen bond of Si-O. The absorption bands located at 469 cm −1 c spond to the bending vibrational mode of Si-O-Si [32]. In addition, the bands of bonds' vibrations are superimposed on the non-bridging bonds of Ca-O in the ran 450-500 cm -1 [65]. In the FTIR spectra of metasilicates, the region of 750-550 cm −1 is greatest interest. The number of bands in this region makes it possible to judge the nu of silicon-oxygen tetrahedrons in the chain recurrence period [66]. The recurrence p in the wollastonite chain is three tetrahedrons (Figure 1a), and three bands are obs in the IR spectrum: they are 681, 642, and 564 cm −1 , respectively. Figure 9b demons a decrease in intensity and broadening of the main bands in the IR spectra of the W ings. Moreover, if the bands corresponding to the vibrations of bonds in the wollas chain are still preserved for the coating formed at 350 V, these bands are no longer pr in the spectrum of the coating deposited at 500 V. This indicates a distortion of the talline structure and its partial amorphization.

Bioresorption of the Pure Mg0.8Ca Alloy and W-Coatings
The samples of the pure Mg0.8Ca and of the coated alloy were immersed in the NaCl solution at 37 °C for 16 days. It was revealed that the dissolution rate of the c samples was significantly lower than that of the pure magnesium alloy (Figure 10a) Figure 10b-g shows the optical images of the Mg0.8Ca and of the coated samp fore the dissolution, after 7-day dissolution, and after 16-day dissolution in 0.9% N The pure magnesium alloy dissolved explosively. The deposition of dissolution pro was observed on the surface of the magnesium alloy as a white bloom, despite the i

Bioresorption of the Pure Mg0.8Ca Alloy and W-Coatings
The samples of the pure Mg0.8Ca and of the coated alloy were immersed in the 0.9% NaCl solution at 37 • C for 16 days. It was revealed that the dissolution rate of the coated samples was significantly lower than that of the pure magnesium alloy (Figure 10a).   the dissolution, after 7-day dissolution, and after 16-day dissolution in 0.9% NaCl. The pure magnesium alloy dissolved explosively. The deposition of dissolution products was observed on the surface of the magnesium alloy as a white bloom, despite the intensive bioresorption of the samples (Figure 10e-g). The coating retained its integrity for 16 days (Figure 10b-d). No large cracks or deep corrosion pits were observed in the optical images of the coating surface. SEM images of the alloy surface and of the W-coating after 16-day dissolution are shown in Figure 11. Lamellar crystals forming into rosettes are observed at high magnification on the surface of the magnesium alloy. In the XRD patterns ( Figure 12a) it can be seen that the intensity of magnesium reflexes decreased significantly after bioresorption, and some new reflexes appeared. These reflexes are related to the crystalline phases of Mg(OH)2 and MgCO3. TEM results confirmed the phase composition of the precipitate deposited on the magnesium alloy ( Figure  12b). Light-field TEM images show that the diameter of the lamellar crystals is 100 nm and their thickness is 20 nm.
The dissolution of the coating occurred more intensively in the regions free from the wollastonite particles (Figure 11b). In the SEM image, a yellow line marks the area of the coating's dissolution. The pores and cracks have become deeper; however, the crystals of the wollastonite retained their characteristic elongated shape.
The comparative analysis of the XRD patterns of the coating before and after dissolution (Figure 12c) shows that the intensity of the wollastonite reflexes decreases while the intensity of those of the magnesium substrate increases. This means that the wollastonite crystals also participated in the dissolution process.
The energy EDX spectra of the W-coating (Figure 12d) demonstrate that after dissolution the amount of silicon and calcium decreased, while the amount of magnesium increased. In this case, as a result of the dissolution, the W-coating became thinner and the magnesium substrate shone through more intensely. In the XRD patterns (Figure 12a) it can be seen that the intensity of magnesium reflexes decreased significantly after bioresorption, and some new reflexes appeared. These reflexes are related to the crystalline phases of Mg(OH) 2 and MgCO 3 . TEM results confirmed the phase composition of the precipitate deposited on the magnesium alloy (Figure 12b). Light-field TEM images show that the diameter of the lamellar crystals is 100 nm and their thickness is 20 nm.
The dissolution of the coating occurred more intensively in the regions free from the wollastonite particles (Figure 11b). In the SEM image, a yellow line marks the area of the coating's dissolution. The pores and cracks have become deeper; however, the crystals of the wollastonite retained their characteristic elongated shape.
The comparative analysis of the XRD patterns of the coating before and after dissolution (Figure 12c) shows that the intensity of the wollastonite reflexes decreases while the intensity of those of the magnesium substrate increases. This means that the wollastonite crystals also participated in the dissolution process.
The energy EDX spectra of the W-coating (Figure 12d) demonstrate that after dissolution the amount of silicon and calcium decreased, while the amount of magnesium increased. In this case, as a result of the dissolution, the W-coating became thinner and the magnesium substrate shone through more intensely.

Electrochemical Properties
To determine the electrochemical properties of the bare magnesium alloy and the coated samples, potentiodynamic curves were obtained and electrochemical parameters were calculated. The polarization curve of the alloy Mg0.8Ca is typical for this type of material (Figure 13a) [27,67]. For the samples with W-coatings, polarization curves were in the field of lower currents compared to the curve for bare Mg alloy. In addition, significant suppression of anodic and cathodic reactions was found.
Electrochemical parameters are presented in Table 2. The corrosion current values for the coated samples were almost ten times lower than that for the Mg0.8Ca. At the same time, the corrosion resistance values of the coatings were by an order of magnitude-or even two orders of magnitude-higher than that of the magnesium alloy. The W-coating deposited at the voltage of 500 V was the most resistant, because it had the lowest corrosion current value. In this case, the polarization resistance value and impedance modulus were the highest.

Electrochemical Properties
To determine the electrochemical properties of the bare magnesium alloy and the coated samples, potentiodynamic curves were obtained and electrochemical parameters were calculated. The polarization curve of the alloy Mg0.8Ca is typical for this type of material (Figure 13a) [27,67]. For the samples with W-coatings, polarization curves were in the field of lower currents compared to the curve for bare Mg alloy. In addition, significant suppression of anodic and cathodic reactions was found.
Electrochemical parameters are presented in Table 2. The corrosion current values for the coated samples were almost ten times lower than that for the Mg0.8Ca. At the same time, the corrosion resistance values of the coatings were by an order of magnitude-or It should be mentioned that after the formation of the coating at 350 V the corrosion potential value shifts to lesser values in comparison with the bare alloy, along with an increase in polarization resistance. This can be explained as follows: The coating had complicated phase composition and developed surface morphology. Some of the soluble compounds in the coating composition can interact with the NaCl solution, changing the potential-determining reaction. At the same time, this coating had a thickness of 40 µm, and demonstrated high corrosion resistance, shielding the metal surface from contact with the aggressive medium. The coatings formed at 400 and 500 V are characterized by slightly nobler values of corrosion potential, due to the transition from crystalline to crystalline-amorphous structure, and more pronounced barrier properties explained by their greater thickness.  It should be mentioned that after the formation of the coating at 350 V the corrosion potential value shifts to lesser values in comparison with the bare alloy, along with an increase in polarization resistance. This can be explained as follows: The coating had complicated phase composition and developed surface morphology. Some of the soluble compounds in the coating composition can interact with the NaCl solution, changing the potential-determining reaction. At the same time, this coating had a thickness of 40 μm, and demonstrated high corrosion resistance, shielding the metal surface from contact with the aggressive medium. The coatings formed at 400 and 500 V are characterized by slightly nobler values of corrosion potential, due to the transition from crystalline to crystallineamorphous structure, and more pronounced barrier properties explained by their greater thickness.
The Bode plots (Figure 13b,c) show the results of impedance measurements in the form of dependences of the impedance modulus |Z| and phase angle theta versus the frequency.
The presence of a thin film of oxide or hydroxide on the surface of the magnesium alloy sample has an effect on the dependence of the phase angle on the frequency in the middle and low ranges. The value of the impedance modulus |Z|f→0 Hz of Mg0.8Ca in the low-frequency region was equal to 2 × 10 2 Ω cm 2 (Table 2). Thus, the bare magnesium alloy is highly reactive, and the protective coatings are necessary in order to increase its corrosion resistance.
For modelling the impedance spectra, the equivalent electrical circuits (EEC) for the Mg0.8Ca alloy and W-coatings were created (Figure 14). To demonstrate the samples' electrochemical behavior, the CPE were used in the EEC: where ϖ is an angular frequency (ϖ = 2πf), j is an imaginary unit, and n and Q are the exponential coefficient and the frequency independent constant, respectively. In this case,  The Bode plots (Figure 13b,c) show the results of impedance measurements in the form of dependences of the impedance modulus |Z| and phase angle theta versus the frequency.
The presence of a thin film of oxide or hydroxide on the surface of the magnesium alloy sample has an effect on the dependence of the phase angle on the frequency in the middle and low ranges. The value of the impedance modulus |Z| f→0 Hz of Mg0.8Ca in the low-frequency region was equal to 2 × 10 2 Ω cm 2 (Table 2). Thus, the bare magnesium alloy is highly reactive, and the protective coatings are necessary in order to increase its corrosion resistance.
For modelling the impedance spectra, the equivalent electrical circuits (EEC) for the Mg0.8Ca alloy and W-coatings were created ( Figure 14). To demonstrate the samples' electrochemical behavior, the CPE were used in the EEC: where is an angular frequency ( = 2πf ), j is an imaginary unit, and n and Q are the exponential coefficient and the frequency independent constant, respectively. In this case, n shows the deviation from ideal capacitive behavior. When n = 0 and 1 the physical meaning of Q became classic elements of conductivity (1/R) and capacitance (C), respectively. The presence of capacitive in the middle and inductive in the low frequency range loops was revealed for the bare Mg0.8Ca and for the coated samples. The R 1 -CPE 1 capacitive loop for Mg0.8Ca was connected with the charge transfer resistance and the capacitance of the electrical double layer (Figure 14a). For the samples with W-coatings, the R1-CPE 1 capacitive loop corresponded to the electrolyte resistance in the pores and the geometrical capacitance of the oxide layer, whereas R 2 -CPE 2 exhibited parameters of the barrier layer of the W-coating (Figure 14b). The parameters of the EECs elements are presented in Table 3. The presence of capacitive in the middle and inductive in the low frequency range loops was revealed for the bare Mg0.8Ca and for the coated samples. The R1-CPE1 capacitive loop for Mg0.8Ca was connected with the charge transfer resistance and the capacitance of the electrical double layer (Figure 14a). For the samples with W-coatings, the R1-CPE1 capacitive loop corresponded to the electrolyte resistance in the pores and the geometrical capacitance of the oxide layer, whereas R2-CPE2 exhibited parameters of the barrier layer of the W-coating (Figure 14b). The parameters of the EECs elements are presented in Table 3. The RL-L chain was connected with the relaxation processes that took place during the dissolution of magnesium and deposition of corrosion products on the phase boundary of the "alloy/electrolyte" interface for untreated samples, and of the "alloy/coating/electrolyte" interface for the coated samples [68].
Thus, the W-coatings demonstrated the ability to reduce the corrosion rate of the magnesium alloy and showed excellent corrosion resistance, especially those formed at 500 V.

Biological Research
The study of the effect of the sample extracts on 3T3 fibroblast cell viability showed that when cells contacted the extracts of the W-coating and of the bare magnesium alloy, the number of surviving cells was equal to 85% and 28% respectively (Figure 15a).  The R L -L chain was connected with the relaxation processes that took place during the dissolution of magnesium and deposition of corrosion products on the phase boundary of the "alloy/electrolyte" interface for untreated samples, and of the "alloy/coating/electrolyte" interface for the coated samples [68].
Thus, the W-coatings demonstrated the ability to reduce the corrosion rate of the magnesium alloy and showed excellent corrosion resistance, especially those formed at 500 V.

Biological Research
The study of the effect of the sample extracts on 3T3 fibroblast cell viability showed that when cells contacted the extracts of the W-coating and of the bare magnesium alloy, the number of surviving cells was equal to 85% and 28% respectively (Figure 15a).
When the coating extracts were diluted 10 and 100 times, the number of surviving cells increased to 97%, and approached the control. In this case, dilution of extracts from the pure magnesium samples resulted in a threefold increase in surviving cells, but their number remained low in comparison with coated samples (Figure 15a). The sample extracts were diluted to simulate the flow conditions of biological fluids in the body.
The performed studies showed that the coated samples did not greatly affect the viability of the cell line, even without additional dilution of the extracts, and were non-toxic according to ISO 10993-5: 2009 (the decrease in viability does not exceed 20% relative to the negative control). Thus, it was found that the W-coatings significantly reduce toxicity of the magnesium alloy.
An important condition for maintaining cell viability is the maintenance of the medium's pH in the range of neutral values. When the test samples were immersed in a cell culture medium, a change in pH was observed (Figure 15b). Hydrogen evolution during the dissolution of the magnesium alloy led to an increase in the pH of the cell medium to 9.8 within 25 h of exposure. The coating helped to reduce the pH level, which prevented the medium from alkalizing in the implantation site (Figure 15b). The obtained data correlated well with the toxicity data. When the coating extracts were diluted 10 and 100 times, the number of surviv cells increased to 97%, and approached the control. In this case, dilution of extracts fr the pure magnesium samples resulted in a threefold increase in surviving cells, but t number remained low in comparison with coated samples (Figure 15a). The sample tracts were diluted to simulate the flow conditions of biological fluids in the body.
The performed studies showed that the coated samples did not greatly affect the ability of the cell line, even without additional dilution of the extracts, and were non-t according to ISO 10993-5: 2009 (the decrease in viability does not exceed 20% relativ the negative control). Thus, it was found that the W-coatings significantly reduce toxi of the magnesium alloy.
An important condition for maintaining cell viability is the maintenance of the dium's pH in the range of neutral values. When the test samples were immersed in a culture medium, a change in pH was observed (Figure 15b). Hydrogen evolution dur the dissolution of the magnesium alloy led to an increase in the pH of the cell medium 9.8 within 25 h of exposure. The coating helped to reduce the pH level, which preven the medium from alkalizing in the implantation site (Figure 15b). The obtained data related well with the toxicity data.
Optical microscopy enables us to assess the changes in the cell morphology after t Optical microscopy enables us to assess the changes in the cell morphology after their contact with the W-coating extracts. As shown in Figure 15c,d, the cells cultured in plates in the presence of the sample extract did not differ morphologically from the cells in the control group cultured only in nutrient medium. Thus, the extracts of the samples did not cause changes to the morphology of the 3T3 cell line.

Discussion
The correlation of the analysis of the elemental-and structural-phase composition of the coatings, investigated using the EDX, XRD, TEM, and IR methods, suggests the following: Owing to the peculiarities of the micro-arc oxidation process in the unipolar potentiostatic mode in an electrolyte containing both solutes and a dispersed phase, the coatings were formed with a specific structure. At the initial moment of the MAO process (Figure 2a), the current density reached its highest value. The most powerful microarc discharges occurred at the interface between the metal substrate (oxide film) and the electrolyte. This led to intense melting of the electrolyte substance in the micro-arc discharge channel, and the formation of a porous amorphous-crystalline layer. As Khan et al. [64] reported, because of the action of short-lived micro-discharges, non-stationary processes take place, such as a very fast heating/cooling, thermochemical interaction between the substrate and the electrolyte, instant melting, and the subsequent solidification of the coating.
Within 50 s of the coatings' formation, the intensity of micro-arc discharges decreased (Figure 2a). At the final stage, when the current density reached its minimal value and the micro-arc discharges were extinguished, wollastonite particles were deposited on the coating's surface ( Figure 16). Previous studies have also shown that when the coatings were formed via the MAO method, using a potentiostatic mode in an electrolyte containing dispersed particles, these particles were transferred from the electrolyte to the coating's surface [27,67]. Thus, it is assumed that the main crystalline phase of wollastonite was presented on the coating surface. In addition, small quantities of akermanite nanocrystallites were formed on the wollastonite particles, but they were detected only by TEM.
was presented on the coating surface. In addition, small quantities of akermani crystallites were formed on the wollastonite particles, but they were detected TEM.
Magnesium silicate, forsterite, and magnesium oxide were mainly formed in coating, since Mg, Si, and O prevailed in the elemental composition of the coating section. The authors of [69,70] wrote about the formation of forsterite and pericla interaction of the electrolyte containing sodium silicate with a magnesium substr thermore, as was revealed using the SEM EDX method, in addition to the oxide, M formed in the boundary protective layer between the coating and the substrate halo was observed in the XRD pattern of the coatings, one can conclude that the contained an amorphous phase. Moreover, with an increase in the voltage up t the amount of the amorphous phase in the coatings' phase composition increased ically. The dissolution of the coating in the solution of 0.9% NaCl occurred mo sively in the regions between the crystals, where the amorphous phase was pred ( Figure 16). On the other hand, with an increase in the process voltage, the amount of ca the coatings rises, which indicates an increase in the involvement of wollastoni MAO process. The corrosion resistance of the coatings increased in this case, b was found that coatings applied at a process voltage of 500 V had the highest c resistance. Sainz et al. [34] described the dissolution process of wollastonite in S reaction (6) took place in this case: Magnesium silicate, forsterite, and magnesium oxide were mainly formed inside the coating, since Mg, Si, and O prevailed in the elemental composition of the coating's cross-section. The authors of [69,70] wrote about the formation of forsterite and periclase in the interaction of the electrolyte containing sodium silicate with a magnesium substrate. Furthermore, as was revealed using the SEM EDX method, in addition to the oxide, MgF 2 also formed in the boundary protective layer between the coating and the substrate. Since a halo was observed in the XRD pattern of the coatings, one can conclude that the coatings contained an amorphous phase. Moreover, with an increase in the voltage up to 500 V, the amount of the amorphous phase in the coatings' phase composition increased dramatically. The dissolution of the coating in the solution of 0.9% NaCl occurred most intensively in the regions between the crystals, where the amorphous phase was predominant ( Figure 16).
On the other hand, with an increase in the process voltage, the amount of calcium in the coatings rises, which indicates an increase in the involvement of wollastonite in the MAO process. The corrosion resistance of the coatings increased in this case, because it was found that coatings applied at a process voltage of 500 V had the highest corrosion resistance. Sainz et al. [34] described the dissolution process of wollastonite in SBF. The reaction (6) took place in this case: When wollastonite dissolved, calcium ions were released from the surface. In this case, many silanol groups (Si-OH) were formed. These groups were the centers of the formation and growth of hydroxyapatite [28,29]. The released calcium ions, in turn, intensified apatite nucleation. Similar processes occurred during the dissolution of Ca-and Mg-silicates, such as akermanite [39] and forsterite [69].
Zakaria et al. [30] showed that the wollastonite coating had good biological activity, both in the in vitro tests with a model of biological fluid, and in the in vivo tests on real implants for dogs. The mechanism of HA growth on the surface of a wollastonite coating was presented. Zhai et al. [37] reported that bioceramics in the system Ca-Mg-Si have a stimulating effect on osteogenesis and angiogenesis. Dissolved Ca ions can induce osteoblast proliferation, while Mg stimulates the mineralization of calcified tissues and affects mineral metabolism [71].
In the presented work, the wollastonite contained in the coatings contributed to the improvement of the biocompatibility of the magnesium samples, since there was a significant difference in the number of viable cells after interacting with the bare magnesium alloy and with the coated samples.

Conclusions
Porous coatings containing crystalline phases such as wollastonite, akermanite, forsterite, magnesium oxide, and amorphous phase were synthesized on the surface of the Mg0.8Ca alloy by means of the MAO method.
Elongated crystals of wollastonite were uniformly distributed over the surface of the coatings. It is assumed that akermanite was formed as a result of the interaction of wollastonite with the magnesium substrate during the micro-arc discharges, while forsterite was formed when the magnesium substrate interacted with sodium silicate contained in the electrolyte.
The thickness and roughness of the coatings increased from 40 to 150 µm, and from 6.5 to 10.5 µm, respectively, with an increase in the process voltage in the range of 350-500 V. The rate of dissolution of the coated samples was significantly lower than that of the pure magnesium alloy. The corrosion current of the coated Mg0.8Ca decreased ten times, and its corrosion resistance increased almost a hundred times.
Thus, it was shown that due to the special structure of the micro-arc coatings, their complex amorphous-crystalline structure, and their phase composition, the W-coatings exhibited significantly low bioresorption rates and remarkable corrosion resistance. In addition, the W-coatings were not cytotoxic to the 3T3 fibroblast cell line, and improved the biocompatibility of the magnesium alloy.
The developed biodegradable biocomposite based on the Mg0.8Ca alloy and the W-coating is a promising material for use in implants in reconstructive medicine.