Nanoscale modification of magnesium with highly textural lamellar nanosheets towards increasing the corrosion resistance and bioactivity

Surface & Coatings Technology 304 (2016) 425–437

Contents lists available at ScienceDirect

Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat

Nanoscale modification of magnesium with highly textural lamellar nanosheets towards increasing the corrosion resistance and bioactivity T.S.N. Sankara Narayanan, Min Ho Lee ⁎ Department of Dental Biomaterials and Institute of Biodegradable Materials, Institute of Oral Bioscience and School of Dentistry (Plus BK21 Program), Chonbuk National University, Jeonju 561-756, South Korea

a r t i c l e

i n f o

Article history: Received 15 March 2016 Revised 6 June 2016 Accepted in revised form 24 June 2016 Available online 27 June 2016 Keywords: Nanoscale modification Lamellar nanosheets Microarc oxidation Bioactivity Absorbable implants

a b s t r a c t The present study aims to modify the surface of microarc oxidation (MAO) coated Mg using an alkaline fluoride solution so as to impart a nanoscale surface feature, which would be beneficial to improve the corrosion resistance and to promote a better bioactivity. The MAO coated Mg is modified with the formation of a highly textural lamellar nanosheet-like morphology after immersion in 0.1 M NaF (pH: 8.40) at 25 °C for 120 min, which completely covered the porous structure with the formation of nanosheets along with some agglomerated crystals. Thin film X-ray diffraction measurement, chemical composition analysis and Fourier transform infrared spectroscopy studies confirmed that the nanosheets are primarily Mg(OH)2 − xFx while the agglomerates are NaMgF3. The complete coverage of the porous structure by the modified layer along with the formation of Mg(OH)2 − xFx and NaMgF3 has enabled a better corrosion resistance for MAO coated Mg modified by NaF. The higher surface area of the nanosheets favoured nucleation of monocalcium phosphate anhydrous and newberyite, both of which are biologically relevant. The unique morphological feature of the modified surface helped to achieve an improved cell adhesion and proliferation of MC3T3-E1 osteoblast-like cells. The relative growth rate of both uncoated and coated Mg are N75% on all the three days, fulfilled Grade 1 specification in terms of cytocompatibility as per ISO 10993-5 standard. The formation of nanoscale surface feature, improvement in corrosion resistance, better bioactivity and acceptable cytocompatibility point out that this methodology could be of immense help to modify the surface of Mg based absorbable implants. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Development of absorbable magnesium based implants is indeed challenging due to the rapid corrosion rate of Mg and its alloys that undermines their mechanical integrity [1,2]. Various surface modification methods are suggested to control the rate of corrosion of Mg and its alloys [3–5] and among these, microarc oxidation (MAO) is a viable approach. The mechanism of formation of MAO coatings and its characteristics are elaborated elsewhere [6]. The presence of micropores and cracks are inherent characteristics of the MAO coatings. A porous outer layer is considered to be beneficial to improve mechanical interlocking, to increase the bonding area, to distribute the stresses across the interface and, to achieve higher bond strength. Nevertheless, it increases the effective surface area and facilitates quicker infiltration of the corrosive medium through them down to the substrate, leading to a decrease in corrosion resistance. The strategies to improve the corrosion resistance of MAO coated Mg has been reviewed recently [7]. The

⁎ Corresponding author. E-mail addresses: [email protected] (S.N. T.S.N.), [email protected] (M.H. Lee).

http://dx.doi.org/10.1016/j.surfcoat.2016.06.070 0257-8972/© 2016 Elsevier B.V. All rights reserved.

pores can be sealed by various approaches to achieve an improved corrosion resistance. However, the benefits of realizing a better osseointegration could be lost due to the sealing of the pores. In recent years, numerous attempts are made to impart nanoscale surface features on Mg for biomedical applications [8,9]. One such approach is to treat Mg with NaOH to enlarge nanoscale surface roughness that enables an increase in osteoblast adhesion, cell proliferation, alkaline phosphatase activity, and deposition of calcium phosphates [8]. It has been demonstrated that treating Mg with NaOH helps to decrease the detrimental effects of the degradation products of Mg on osteoblast density [9]. Based on their findings, Weng and Webster [9] have recommended that creating nanoscale features on Mg could improve its use for many orthopedic applications. Zhu et al. [10] have also suggested that Mg alloy modified with the formation of a Mg(OH)2 film could serve as a potential material for orthopedic implants. Zhao et al. [11] have shown that an alkaline pretreatment in 5, 10 and 15 g/l NaOH at 80 °C for 30 to 90 min helps to realize a uniform, dense, crack-free 45S5 bioglass–ceramic coating on AZ31 Mg alloy with better bonding. The unevenness and presence of a number of microcracks on 45S5 coating deposited on Mg alloy surface without the alkaline pretreatment has previously been pointed out by Huang et al. [12]. Chen et al. [13] have

426

S.N. T.S.N., M.H. Lee / Surface & Coatings Technology 304 (2016) 425–437

advocated the benefits of developing a transitional layer of Mg:OH by alkaline pretreatment of Mg in 3 M NaOH at 60 °C for 12 or 24 h that helps to improve covalent bonding of phytic acid molecules, resulting in the formation of a dense and homogenous phytic acid coating. Waterman et al. [14] have reported that the formation of Mg(OH)2 layer on Mg by treating it in pure H2O at 100 °C for 15 or 30 min is beneficial for the nucleation of calcium phosphates in Hank's solution. Alkaline pretreatment by immersion in 0.05 to 4 M NaOH for 1 h has also been shown to promote apatite formation on non-woven poly(εcaprolactone) fabrics in simulated body fluid (SBF) [15]. According to Wei et al. [16], development of a hydroxyl functionalized surface over MAO coated Ti alloy, by chemical treatment in 5 M NaOH at 60 °C for 24 h, has favoured heterogeneous nucleation of the apatite. These studies clearly indicate that alkaline pretreatment of Mg and its alloys could offer considerable benefits. Nevertheless, it is imperative to manipulate the experimental conditions to generate nanoscale surface features to achieve the desired benefits. According to Waterman et al. [14], treatment of Mg in pure H2O at 100 °C for 15 min has resulted in the generation of Mg(OH)2 layer with a uniform flake-like structure while it becomes a dense layer with many defects and cracks after 30 min. The importance of developing a Mg(OH)2 layer with a specific surface structure is also recommended by Wei et al. [16]. The present study aims to modify the surface of MAO coated Mg using an alkaline fluoride solution, 0.1 M NaF (pH: 8.40), so as to impart a nanoscale surface feature. The choice of the electrolyte is made based on the ability of the alkaline solution to generate a layer of Mg(OH)2 with a flake-like structure. Alkaline treatment on Mg is performed mainly using NaOH [11,13,15] whereas the use of 0.1 M NaF has been suggested for the first time in this work. The structural and morphological characteristics of the modified surface, its corrosion behaviour in Hank's balanced salt solution (HBSS) and its ability to favour the deposition of calcium phosphates from a concentrated simulated body fluid (c-SBF) under biomimetic conditions are evaluated. 2. Materials and methods Commercially pure Mg (composition (in wt.%): Mg: 99.93; Al: 0.0032; Mn: 0.0128; Cu: 0.005; Fe: 0.0017; Si: 0.0228; Ni: 0.0003) having a dimension of 20 mm × 15 mm × 3 mm was used as substrate materials. They were surface ground using SiC coated abrasive (600 grit) paper, ultrasonically cleaned in ethanol and dried using a stream of compressed air. MAO of Mg was carried out using an alkaline silicate electrolyte that contained 5 g/l NaOH and 15 g/l Na2SiO3 under direct current mode at 250 V for 2 min. A large sheet of Pt (60 mm × 40 mm × 1 mm) was used as the counter electrode. During treatment, a water bath was used to control the temperature of the electrolyte b 40 °C. After deposition, the MAO coated Mg samples were rinsed thoroughly using deionized water and dried. The MAO coated Mg samples were modified by immersion in 0.1 M NaF (pH: 8.40) at 25 °C for various duration of time up to 180 min. Uncoated Mg samples were also treated under similar conditions to understand the mechanism of surface modification by NaF. After the stipulated time period, the surface modified Mg samples were rinsed using deionized water and dried. The morphological features of the MAO coating and those modified by NaF were evaluated by a scanning electron microscope (SEM) with field emission source (Hitachi – Analytical UHR Schottky Emission Scanning Electron Microscope SU-70). Before SEM analysis, all the samples were sputter coated with a thin layer of Pt to reduce charging as well as to increase the amount of secondary electrons that can be detected from the surface, which would help to enhance the signal to noise ratio. During SEM analysis, the images were acquired using an accelerating voltage of 5 and 10 kV while the working distance was varied between 4.9 mm and 16.3 mm, depending on the type of sample. The chemical composition of the coated surfaces was assessed by energy dispersive spectroscopy (EDS) attached with SEM. Background correction in EDS analysis was performed by applying background filtering

technique. The structural characteristics as well as the nature of functional groups were determined by thin film X-ray diffraction (TF-XRD) (PANalytical X'pert MRD) measurement and Fourier transform infrared (FT-IR) spectroscopy (Perkin–Elmer, Spectrum GS) in attenuated total reflectance (ATR) mode. The 3-dimensional surface topography of the MAO coated Mg and those modified by NaF was assessed using atomic force microscopy (AFM) (Bruker NanoScope V multimode 8 scanning probe microscopy). The AFM imaging was performed in tapping mode over a scanning area of 5 μm × 5 μm. The corrosion resistance of the uncoated Mg, MAO coated Mg and MAO coated Mg modified by NaF in HBSS at 25 °C was evaluated by potentiodynamic polarization and electrochemical impedance (EIS) studies. The details of the characterization studies and evaluation of corrosion behaviour were elaborated in our earlier paper [17]. The uncoated, MAO coated Mg and MAO coated Mg modified by NaF were also subjected to immersion test in HBSS at 37 °C for 168 h. After completion of the immersion test for 168 h, the samples were taken out, rinsed with de-ionized water and dried. The corrosion products as well as the remnant coating present on the samples were removed by immersion in a mixture of 200 g/l CrO3 with 10 g/l AgNO3 for 30 min, washed thoroughly with de-ionized water and dried. The surface morphology of the corroded region was assessed by SEM with field emission source to compare the extent of corrosion attack and to understand the corrosion mechanism. In order to evaluate the ability of the MAO coated Mg and those modified by NaF to promote deposition of calcium phosphates, they were subjected to immersion in c-SBF under biomimetic conditions at 37 °C for 24 h. The c-SBF solution contains 1.65 g/l CaCl2; 0.30 g/l KH2PO4; 7.0 g/l NaCl; 0.40 g/l Na2HPO4·4H2O; and 0.35 g/l NaHCO3. The constituents of the c-SBF essentially comprised of ions that are commonly present in physiological solutions. However, the concentrations of these ions are relatively higher in c-SBF so as to facilitate an early deposition of calcium phosphates. A similar composition was used previously by Waterman et al. [14]. After deposition, the coated samples were heat-treated at 300 °C for 1 h to promote crystallization of the calcium phosphate. A similar procedure was followed in a previous study by Yang et al. [18]. The cytocompatibility of uncoated Mg, MAO coated Mg and those modified by NaF was evaluated by an indirect assay method as per ISO 10993-5 standard using MC3T3-E1 osteoblast-like cell line. Minimum Essential Medium Eagle (α-MEM) (Sigma, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco Co., USA) and 1% antibiotic served as the cell culture medium. The uncoated Mg, MAO coated Mg and those modified by NaF were incubated in cell culture medium for 72 h under physiological conditions (37 °C; 5% CO2; 95% relative humidity) as per ISO 10993-12 standard. The ratio of surface area of the sample to the cell culture medium was maintained at 1.25 cm2/ml. After incubation for 72 h, the samples were removed from the cell culture medium, washed with deionized water and dried. The extracts were centrifuged at 14,000 rpm for 15 min and supernatant solution was separated for further use (100% extract). It is diluted with α-MEM to prepare 10% extracts. MC3T3-E1 cells were seeded with a cell density of 1 × 104 cells/ ml of the medium in 24-well plates and incubated for 24 h under physiological conditions. Subsequently, the medium was replaced with 100 μl of 10% extracts and 1 ml of negative control (cell culture medium alone) and the MC3T3-E1 cell lines were incubated again under physiological conditions for 1, 2 and 3 days. 10 μl of 2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2.4-disulfophenyl)-2H-tetrazolium monosodium salt (WST-8) solution was added into each well at the end of each time interval and they were again incubated for 2 h. The absorbance of the solution mixture in each well was measured at 490 nm using an enzyme-linked immunosorbent assay (ELISA) microplate reader. The relative growth rate (RGR) of the cells was calculated using the following equation: RGRð%Þ ¼ ðODt =ODn Þ  100%

S.N. T.S.N., M.H. Lee / Surface & Coatings Technology 304 (2016) 425–437

where ODt and ODn correspond to the optical density of the samples and the negative control, respectively. The cell morphology and extent of cell growth after 1, 2 and 3 days of cell culture were assessed by crystal violet staining method using an inverted phase contrast microscope. 3. Results The surface morphology of the MAO coating formed on Mg using alkaline silicate electrolyte at 250 V for 2 min and, the evolution of its morphological features after immersion in 0.1 M NaF as a function of time, are shown in Fig. 1. The MAO coated Mg reveals a porous structure

427

(Fig. 1(a)), which by virtue of its formation mechanism is a common trait of such type of coatings. The pores are distributed all over the surface with their diameters ranging from 0.30 to 1.50 μm. The average thickness of the MAO coating formed on Mg (determined at the cross section by SEM) is ~ 8 to 10 μm. After immersion in 0.1 M NaF for 30 min, the porous structure of the MAO coating is still evident. However, there appears to be a slight reduction in the pore diameter towards 0.25 to 0.80 μm. In addition, many nanowhiskers (white spots distributed all over the surface) are formed all over the coated surface (Fig. 1(b)). With a further increase in immersion time from 30 to 60 min, majority of the pores are covered by the nanowhiskers/nanosheets, which are

Fig. 1. (a) Surface morphology of the MAO coating formed on Mg; and (b–e) evolution of its morphological features after immersion in 0.1 M NaF as a function of time; (b) 30 min; (c) 60 min; (d) 120 min; (e) magnified view of matrix area shown in ‘d’; (f) micrograph acquired at the cross section of the modified layer formed over MAO coated Mg after immersion in 0.1 M NaF for 120 min; (g, h) surface morphology of uncoated Mg after immersion in 0.1 M NaF for 120 min.

428

S.N. T.S.N., M.H. Lee / Surface & Coatings Technology 304 (2016) 425–437

distributed uniformly throughout the surface, resulting in further reduction in the pore diameter (from 0.20 to 0.50 μm) (Fig. 1(c)). Moreover, agglomeration of crystals could be observed at some locations (marked within the ‘◯’ in Fig. 1(c)). The amount of nanowhiskers/ nanosheets is increased with an increase in immersion time to 60 min and this trend is further continued beyond 60 min. After 120 min of immersion, the MAO coated Mg reveals the formation of a highly textural lamellar nanosheet-like morphology (Fig. 1(d)). The nanosheets are interlaced with each other and lay perpendicular to the substrate, resulting in a wormhole-like structure. They are distributed homogeneously over the entire surface and completely cover the porous structure of MAO coated Mg. Every individual nanosheet resembles a typical rose petal. The nanosheets have an average width of 450–550 nm while their thickness ranges from 40 to 60 nm. The size as well as the population density of the agglomerates is increased with an increase in immersion time from 60 to 120 min (Fig. 1(d)). A closer look at the agglomerated crystal (Fig. 1(d)) reveals that its appearance is totally different from the nanosheets. The magnified view of the area covered by the nanosheets in Fig. 1(d) shows the formation of spherical-like structure underneath the nanosheets (marked within the ‘◯’ in Fig. 1(e)). The size as well as the population density of agglomerates is increased further when the treatment time is increased from 120 to 180 min (not shown in Fig. 1). Since this would affect the uniformity of the modified surface, the treatment time is limited to 120 min. The average thickness of the surface modified layer, which is a combination of the MAO coating and the layer with the lamellar nanosheet-like morphology formed after immersion in 0.1 M NaF for 120 min (determined at the cross section by SEM) is ~12 to 14 μm (Fig. 1(f)). There is no distinct interface between the MAO coating and the modified layer obtained after immersion on 0.1 M NaF for 120 min. The morphological evolution of the MAO coated Mg after immersion in 0.1 M NaF as a function of time clearly indicates a decrease in the pore diameter and an increase in surface coverage by the nanosheets. For a better understanding of the formation of nanosheets and agglomerates, uncoated Mg is also subjected to immersion in 0.1 M NaF for 120 min. Despite the similarity in the formation of highly textural lamellar structure, the size of the nanosheets formed on uncoated Mg (Fig. 1(g)) is much smaller than those formed over MAO coated Mg (Fig. 1(d)). Agglomeration of the crystals is also noticed at some locations (marked within the ‘◯’ in Fig. 1(g) and (h)) of the modified surface on uncoated Mg. However, they appear to be a combination of a few nanosheets and drastically different from those formed over MAO coated Mg (Fig. 1(d)). The variation in the size of the nanosheets formed over MAO coated Mg (Fig. 1(d)) and uncoated Mg (Fig. 1(g)) suggests a significant change in their reactivity with NaF. To understand the difference in reactivity between them, Bode plots are recorded in sequence during the first 15 min (one scan per min) of their immersion in 0.1 M NaF (Fig. 2). Uncoated Mg shows the formation of a coating during the first few scans, as evidenced the shift in |Z| from 1000 to 5000 Ω·cm2. However, with a further increase in the number of scans, the variation in | Z | between 3000 and 5000 Ω·cm2 suggests a continuous dissolution-reprecipitation of the coating (Fig. 2(a)). In contrast, for MAO coated Mg, the first few scans led to a decrease in | Z| from 60,000 to 30,000 Ω·cm2, which suggests a possible dissolution of a portion of the coated layer in 0.1 M NaF. However, the continuous oscillation of |Z| between 30,000 and 40,000 Ω·cm2 points out the dissolution-reprecipitation of the coated layer during the subsequent scans and an increase in surface coverage with time (Fig. 2(b)). The EDS analysis performed at different regions over the surface of the MAO coated Mg and those modified after immersion in 0.1 M NaF for 120 min along with their chemical composition are shown in Fig. 3. The MAO coated Mg shows the presence of O, Mg and Si as the major elements (Fig. 3(c)). MAO coated Mg modified by NaF points out the presence of F besides O, Mg and Si (Fig. 3(d) and (e)). Elemental analysis performed over the agglomerated crystals (Fig. 3(d)) indicates a presence of Na as well as F besides O, Mg and Si. The amount of F is

Fig. 2. Bode impedance plots recorded in sequence during the first 15 min of immersion of (a) uncoated Mg; and (b) MAO coated Mg in 0.1 M NaF.

relatively higher over the agglomerated crystals (Fig. 3(d)) when compared to those observed over the area of nanosheets (Fig. 3(e)). The Xray elemental mapping of the MAO coated Mg and those modified by NaF are shown in Fig. 4. The elemental maps of O, Mg and Si depict uniform distribution of these elements over a large area of the surface, thus confirming the uniformity of the MAO coating. A comparison of the elemental maps of O, Mg and Si of MAO coated Mg modified by NaF acquired over a large area, nanosheets and agglomerated crystals further confirm that the chemical nature of the nanosheets and the agglomerated crystal are completely different. The traces of FT-IR spectra of the MAO coated Mg and those modified by NaF are shown in Fig. 5. The MAO coated Mg shows the presence of four IR bands at 3419 (broad), 1419, 919 and 547 cm− 1, which could be attributed to the O–H stretching of H2O/surface adsorbed –OH, asymmetrical stretching of (CO2– 3 ) carbonates adsorbed on the surface, Si–O stretching and Mg–O stretching vibrations, respectively. The IR bands pertaining to these functional groups are also observed for MAO coated Mg modified by NaF, which appear at 3378 cm−1 (broad), 1427 cm−1, 975 cm−1 and 483 cm−1, respectively. Besides these four bands, some additional IR bands at 3647 cm−1 (sharp), 1738 cm−1 (shoulder), 1650 cm−1 and 703 cm− 1 are observed for MAO coated Mg modified by NaF. The sharp band at 3647 cm−1 can be correlated to the O–H stretching of Mg(OH)2. The shoulder at 1738 cm−1 and the IR band at 1650 cm−1 correspond to H–O–H bending and rotation of H2O, while the IR band at 703 cm−1 can be attributed to the (Si–O–Si) symmetric stretching in the crystalline silicate or metal-oxygen lattice vibrations. The 3-dimensional surface topographic images of MAO coated Mg and those modified using NaF along with the parameters of surface roughness are shown in Fig. 6. It is evident that modification of the MAO coated Mg by NaF has led to an increase in the root mean square roughness

S.N. T.S.N., M.H. Lee / Surface & Coatings Technology 304 (2016) 425–437

429

Fig. 3. EDS analysis performed at different regions of the MAO coated Mg and those modified using 0.1 M NaF for 120 min along with their chemical composition: (a, c) secondary electron micrograph and EDS spectrum of MAO coated Mg; and (b, d, e) secondary electron micrograph and EDS spectra of MAO coated Mg modified using NaF.

(Rq), average surface roughness (Ra) and the maximum vertical distance between the highest and lowest data points in the image (Rmax) (Fig. 6). The TF-XRD patterns of the MAO coated Mg and those modified by NaF are shown in Fig. 7. The MAO coated Mg indicates the formation of MgO, Mg2SiO4 and MgSiO3 (curve (a) in Fig. 7). The peaks from Mg (base metal) are very intense since the MAO coating is relatively thin and X-rays could easily penetrate through them up to the base metal. MAO coated Mg modified by NaF points out the formation of MgO, Mg2SiO4, MgSiO3, Mg(OH)2 − xFx and NaMgF3 (curve (b) in Fig. 7). The presence of F besides O, Mg and Si over the area covered by the nanosheets as well as the predominance of Na, F and Mg over the agglomerated crystals inferred from the EDS analysis as well as the formation of Mg(OH)2 − xFx and NaMgF3 phases in the XRD pattern corroborates well with each other. The considerable decrease in the intensity of the peaks pertaining to Mg for coatings modified by NaF

(curve (b) in Fig. 7) further confirms a better surface coverage, substantiating the inferences made from the surface morphology (Fig. 1). Peaks pertaining to Mg(OH)2 − xFx shows broadening, suggesting that they could be nanocrystalline. The potentiodynamic polarization curves of uncoated Mg, MAO coated Mg and MAO coated Mg modified by 0.1 M NaF, in HBSS, are shown in Fig. 8. The electrochemical parameters, namely, Ecorr and icorr, calculated from the polarization curves by Tafel extrapolation method are compiled in Table 1. The Nyquist and Bode plots of uncoated Mg, MAO coated Mg and MAO coated Mg modified using 0.1 M NaF, in HBSS, recorded at their respective OCP's are shown in Fig. 9(a) and (b), respectively. Different equivalent electrical circuit models are built to analyze the EIS data with the use of ZSimpWin 3.21 software to identify the most suitable one based on the lowest χ2 value and % error. An equivalent electrical circuit model that involves two time constants is proposed to explain the corrosion behaviour of uncoated Mg (Fig.

430

S.N. T.S.N., M.H. Lee / Surface & Coatings Technology 304 (2016) 425–437

Fig. 4. X-ray elemental mapping acquired on MAO coating formed on Mg and over different regions of MAO coated Mg modified using 0.1 M NaF for 120 min.

9(c)) while another model that involves three time constants is suggested for MAO coated Mg as well as MAO coated Mg modified using 0.1 M NaF for 120 min (Fig. 9(d)). The proposed equivalent electrical circuit models seem to be valid as they exhibit the most satisfactory nonlinear least-squares fitting for the respective experimental data (Fig. 9(a) and (b)). In these circuits, Rs is the solution resistance between the Ag/AgCl/KCl (satd.) electrode and uncoated/coated Mg electrode. CPEf, CPEdl and CPEdiff represent capacitance of the film, electrical double layer, and that arise due to the diffusion process, respectively. Rf, Rct and Rdiff refer to the film resistance, charge transfer resistance, and diffusion resistance, respectively. In these circuits, CPE (constant phase

element) is used in place of pure C (capacitance) to account for the surface roughness and surface heterogeneity. The impedance of CPE is represented by ZCPE = 1/Q(jω)n, Where ‘n’ is the CPE exponent. When n = 1, the capacitance element Q (CPE) will be pure capacitance whereas it will be pure resistance when n = 0. Q is called as CPE when the value of ‘n’ lies between 0.5 and 1. The values of the individual electrical components after fitting the EIS data are compiled in Table 2. The surface morphology of uncoated Mg, MAO coated Mg and MAO coated Mg modified by NaF after immersion in HBSS at 37 °C for 168 h and after removal of corrosion products is shown in Fig. 10. The formation of corrosion products along with the precipitation of calcium

S.N. T.S.N., M.H. Lee / Surface & Coatings Technology 304 (2016) 425–437

Fig. 5. FT-IR spectra of (a) MAO coating formed on Mg; and (b) MAO coated Mg modified using 0.1 M NaF for 120 min.

phosphate after immersion in HBSS for 168 h could be observed in all the samples (Fig. 10(a), (c) and (e)). Nevertheless, the tendency for the precipitation of calcium phosphates is much higher on MAO coated Mg modified by NaF (Fig. 10(c)) than on MAO coated Mg (Fig. 10(b)). A comparison of the extent of corrosion attack on uncoated Mg, MAO coated Mg and MAO coated Mg modified by NaF after the removal of corrosion products reveals severe corrosion attack on uncoated Mg, as evidenced by the formation of deep pits (Fig. 10(b)). In contrast, the degree of damage is relatively less on MAO coated Mg and those modified by NaF, which is substantiated by the shallow nature of the pits formed on them (Fig. 10(d) and (f)). The mechanism of corrosion is similar in all the cases, confirming that it is governed by localized corrosion attack by the high concentration of Cl− ions present in HBSS.

Fig. 6. Surface topographic images and the parameters of surface roughness of (a) MAO coated Mg; and (b) MAO coated Mg modified using 0.1 M NaF for 120 min (images are acquired at 5 × 5 μm scale).

431

Fig. 7. TF-XRD patterns of (a) MAO coating formed on Mg; and (b) MAO coated Mg modified using 0.1 M NaF for 120 min.

The surface morphology and chemical composition of MAO coated Mg as well as those modified by NaF after immersion in c-SBF under biomimetic conditions at 37 °C for 24 h followed by heat-treatment at 300 ° C for 1 h is shown in Fig. 11. The deposition of calcium phosphate is evident on both types of coatings (Fig. 11(a)-(c)). Nevertheless, the amount of deposition of calcium phosphate is relatively higher on MAO coating modified by NaF. For MAO coated Mg, the higher concentration of O and Mg with smaller amounts of Ca and P suggest that its surface is rich in Mg(OH)2 with a lower amount of calcium phosphate. For MAO coated Mg modified by NaF, the higher concentration of O, Ca and P with a lower amount of Mg suggest that its surface is rich in calcium phosphate with a lower amount of Mg(OH)2. The Ca/P ratio of 0.44 (for MAO coated Mg) and 0.37 (for MAO coated Mg modified by NaF) indicate that monocalcium phosphate anhydrous (Ca/P ratio: 0.50) could have formed. Since the Ca/P is b0.50, the formation of magnesium phosphate is also possible. It is apparent that the nanosheets served as an effective base for the deposition of calcium phosphate (Fig. 11(b) and (c)). A high magnification image of the modified surface clearly reveals the precipitation of calcium phosphates on the top surface of the nanosheets (Fig. 11(c)). The X-ray elemental maps acquired over the calcium phosphate coatings deposited on MAO coated Mg modified by NaF shows uniform distribution of Ca (Fig. 11(d)) and P (Fig. 11(e)). The presence of F on the surface of MAO coated Mg modified by NaF after immersion in c-SBF could also be due to the formation of CaF2.

Fig. 8. Potentiodynamic polarization curves of uncoated Mg, MAO coated Mg and MAO coated Mg modified using 0.1 M NaF for 120 min, in HBSS.

432

S.N. T.S.N., M.H. Lee / Surface & Coatings Technology 304 (2016) 425–437

Table 1 Electrochemical parameters derived from potentiodynamic polarization curves of uncoated Mg, MAO coated Mg and MAO coated Mg modified using 0.1 M NaF for 120 min, in HBSS.

Type of sample Uncoated Mg MAO coated Mg MAO coated Mg modified using 0.1 M NaF for 120 min

Corrosion potential Ecorr V vs. Ag/AgCl/KCl (sat.)

Corrosion current density icorr (A/cm2)

−1.54 −1.47 −1.53

1.40 × 10−5 1.70 × 10−6 1.10 × 10−7

The TF-XRD pattern of the MAO coated Mg modified by NaF after immersion in c-SBF under biomimetic conditions at 37 °C for 24 h followed by heat-treatment at 300 °C for 1 h (Fig. 12) indicates the presence of MgO, Mg(OH) 2, Mg2SiO 4, MgSiO3, Mg(OH)2 − x Fx and NaMgF 3, MgHPO 4·3H2 O and Ca(H2 PO 4) 2 . The presence of MgO, Mg2SiO4, MgSiO3, Mg(OH)2 − xFx and NaMgF3 are identified in the MAO coated Mg modified after immersion in 0.1 M NaF for 120 min (curve (b) in Fig. 7). Hence, it is clear that Mg(OH)2, MgHPO4·3H2O and Ca(H2PO4)2 are formed during the reaction of the coated surface with the c-SBF. The formation of CaF2 could not be detected in the XRD patterns. The RGR of MC3T3-E1 cells after cell culture using 10% extract of uncoated Mg, MAO coated Mg and MAO coated Mg modified by 0.1 M NaF, for 1, 2 and 3 days is shown in Fig. 13. Both uncoated and coated Mg exhibit N75% of RGR on all the three days. Hence, their cytotoxicity can be classified as Grade 1 as per ISO 10993-5 standard. The cell morphology and the extent of cell growth after 2 days of cell culture using the 10% extract are shown in Fig. 14. The extent of cell growth is much pronounced over MAO coated Mg and MAO coated Mg modified by 0.1 M NaF (Fig. 14(b) and (c)) than those observed over uncoated Mg (Fig. 14(a)). 4. Discussion The pores in the MAO coating on Mg (Fig. 1(a)) are generated due to the plasma discharges during the process while their size and

distribution (0.30 to 1.50 μm) depend on the type and intensity of the discharges as well as the number of points at which the dielectric oxide film is broken down. The morphological evolution of the MAO coated Mg after immersion in 0.1 M NaF as a function of time involves the following stages: (i) formation of nanowhiskers all over the surface, which is subsequently developed as nanosheets (Fig. 1(b)); (ii) crossing of the nanosheets with each other, thus generating a uniform porous network (Fig. 1(c)); and (iii) development of a highly textural lamellar nanosheet-like morphology, completely covering the porous structure of the MAO coating (Fig. 1(d)). The evolution of morphology as a function of time suggests a heterogeneous nucleation of the nanowhiskers along with a simultaneous occurrence of their self-assembly. The lamellar nanosheet-like morphology has been observed previously for Mg(OH)2 prepared by ED [19–23]. Sufficient surface nucleation density, slow nucleation and/or availability of sufficient time for self assembly of the nanosheets are likely to favour the formation of the lamellar-like Mg(OH)2 [19,24,25]. It is interesting to note that the TF-XRD patterns of the MAO coated Mg after immersion in 0.1 M NaF for 120 min fails to show the characteristic peaks of Mg(OH)2 that was aimed for, rather it indicates the presence of Mg(OH)2 − xFx (Fig. 7). The formation of Mg(OH)2 − xFx could be due to the instability of Mg(OH)2 at pH b 10.50 [26] and the strong affinity of F− ions with the Mg2+ ions [27]. This leads to the question that if Mg(OH)2 − xFx is formed during the surface modification by NaF, then how it results with a lamellar morphology, which is typical of Mg(OH)2? It is believed that upon immersion of the MAO coated Mg in 0.1 M NaF, dissolution of a portion of the oxide coating generates Mg2+ ions along with a raise in interfacial pH that results in the formation of Mg(OH)2 in the form of nanowhiskers, which subsequently develops as nanosheets. The Mg(OH)2 reacts with the F− ions and enables the formation of Mg(OH)2 − xFx while a small fraction of Mg(OH)2 helps to retain the morphology [28]. The presence of peaks pertaining to Mg(OH)2 could not be distinctly identified by TF-XRD probably due to the overlapping with the peaks of Mg(OH)2 − xFx (Fig. 7). However, the sharp peak at 3647 cm−1 in the FT-IR spectra (Fig. 5), attributed to the O–H stretching of Mg(OH)2, strongly supports the possibility of the presence of a small fraction of Mg(OH)2, which helped to retain the lamellar morphology.

Fig. 9. (a) Nyquist; and (b) Bode plots of uncoated Mg, MAO coated Mg and MAO coated Mg modified using 0.1 M NaF for 120 min, in HBSS, recorded at their respective open circuit potentials along with the equivalent electrical circuit models: (c) for uncoated Mg; and (d) for MAO coated Mg and MAO coated Mg modified using 0.1 M NaF for 120 min.

S.N. T.S.N., M.H. Lee / Surface & Coatings Technology 304 (2016) 425–437

433

Table 2 Electrochemical parameters derived from the Nyquist plots of uncoated Mg, MAO coated Mg and MAO coated Mg modified using 0.1 M NaF for 120 min, in HBSS, recorded at their respective open circuit potentials.

Type of sample Uncoated Mg MAO coated Mg MAO coated Mg modified using 0.1 M NaF

Rs (Ω·cm2) 93.6 94.6 280.2

CPEf (S·sn·cm−2) −5

2.67 × 10 1.23 × 10−5 2.83 × 10−6

nf 0.84 0.79 0.71

Rf (Ω·cm2) 774 3465 8434

Agglomeration of the crystals observed at some locations (marked within the ‘◯’ in Fig. 1(c) and within the ‘□’ in Fig. 1(d)) and the increase in their size with time from 60 to 120 min suggest the occurrence of yet another reaction other than the formation of Mg(OH)2 − xFx. Based on the chemical composition analysis by EDS (Fig. 3), the agglomerated crystals could be NaMgF3, which is further supported by the enrichment of F and Na at these locations in the X-ray elemental maps (Fig. 4). The size of the lamellar formed over uncoated Mg (Fig. 1(g)) is much smaller than those formed over MAO coated Mg (Fig. 1(d)). The passivation ability of Mg in fluoride containing solutions is well documented and it forms the basis of fluoride conversion coatings on Mg and its alloys [17,29–32]. The increase in | Z | of uncoated Mg from

CPEdl (S·sn·cm−2) −4

8.79 × 10 8.57 × 10−6 1.55 × 10−5

nct 0.83 0.76 0.77

Rct (Ω·cm2) 503 1890 3717

CPEdiff (S·sn·cm−2) – 2.93 × 10−4 8.19 × 10−5

ndiff – 0.90 0.79

Rdiff (Ω·cm2) – 1356 4062

χ2

% error −4

8.51 × 10 1.03 × 10−3 2.08 × 10−4

b2.91 b3.03 b1.44

1000 to 5000 Ω·cm2 (Fig. 2(a)) during the first few impedance scans in 0.1 M NaF validates such an occurrence. The lack of sufficient concentrations of Mg2 + ions could have been the reason for the absence of NaMgF3 on uncoated Mg within 120 min of treatment (Fig. 1(g) and (h)). In contrast, for MAO coated Mg, the presence of a thin barrier layer beneath the porous layer could have prevented passivation of Mg. Instead, the fluoride ions could have attacked the MAO coating, which is supported by the decrease in | Z| of MAO coated Mg from 60,000 to 30,000 Ω·cm2 (Fig. 2(b)) during the first few impedance scans in 0.1 M NaF. The availability of sufficient concentration of Mg2+ ions following dissolution of the MAO coated layer could facilitates a faster nucleation of Mg(OH)2 nanowhiskers with a relatively higher size. Based on the

Fig. 10. Surface morphology of (a, b) uncoated Mg, (c, d) MAO coated Mg and (e, f) MAO coated Mg modified using 0.1 M NaF for 120 min: (a, c, e) after immersion in HBSS at 37 °C for 168 h; and (b, d, f) after removal of corrosion products.

434

S.N. T.S.N., M.H. Lee / Surface & Coatings Technology 304 (2016) 425–437

Fig. 11. (a–c) Surface morphology; and (d, e) X-ray elemental mapping along with the chemical composition of (a) MAO coated Mg; and (b–e) MAO coated Mg modified by NaF after immersion in the c-SBF solution at 37 °C for 24 h followed by heat-treatment at 300 °C for 1 h; (c) magnified view of a selected region in ‘b’; (d) X-ray mapping of calcium; (e) X-ray mapping of phosphorus.

inferences made in the present study, a pictorial model is proposed to explain the various stages of surface modification of MAO coated Mg after immersion in 0.1 M NaF, as a function of time (Fig. 15).

The formation of nanosheets and agglomerates of crystals on the MAO coated Mg is considered responsible for the increase in surface roughness (Fig. 6). Since the nanosheets exhibit uniformity over a

Fig. 12. TF-XRD pattern of MAO coated Mg modified by NaF after immersion in the c-SBF at 37 °C for 24 h followed by heat-treatment at 300 °C for 1 h.

Fig. 13. Relative growth rate (RGR) of MC3T3-E1 cells on (a) uncoated Mg; (b) MAO coated Mg; and (c) MAO coated Mg treated using 0.01 M NaF for 120 min, after cell culture using 10% extracts for 1, 2 and 3 days.

S.N. T.S.N., M.H. Lee / Surface & Coatings Technology 304 (2016) 425–437

435

Fig. 14. Cell morphology and the extent of cell growth on (a) uncoated Mg; (b) MAO coated Mg; and (c) MAO coated Mg treated using 0.01 M NaF for 120 min, after cell culture for 2 days using the 10% extract.

large area, the extent of increase in Rq (from 85 nm to 113 nm) and Ra (from 64 nm to 87 nm) are rather limited. However, the large increase in Rmax (from 709 nm to 1189 nm) is due to the presence of agglomerates at various regions of the modified surface. An increase in Rq and Ra would be beneficial to achieve an improved cell growth and proliferation. However, unevenness in the surface (large Rmax) could be deleterious. Considering a steady increase in the size as well as the population density of agglomerates with an increase in treatment time from 60 to 180 min, a treatment time of 120 min appears to be optimal for the modification of MAO coated Mg in 0.1 M NaF.

The decrease in icorr from 1.40 × 10−5 to 1.70 × 10−6 A/cm2 (Table 1) confirm the ability of MAO coated Mg to offer a better corrosion resistance than the uncoated one. Modification of the surface of MAO coated Mg with 0.1 M NaF at 25 °C for 120 min has led to a further decrease in icorr from 1.70 × 10−6 to 1.10 × 10−7 A/cm2 (Table 1), confirming the ability of the modified surface to offer an improved corrosion resistance. The higher Rf and Rct values along with a corresponding lower CPEf and CPEdl values obtained for MAO coated Mg as well as MAO coated Mg modified using 0.1 M NaF than those obtained for uncoated Mg (Table 2) confirms their ability to resist the dissolution of Mg to a greater

Fig. 15. Pictorial model depicting the modification of the surface of MAO coated Mg after immersion in 0.1 M NaF (pH: 8.40) at 25 °C as a function of time.

436

S.N. T.S.N., M.H. Lee / Surface & Coatings Technology 304 (2016) 425–437

extent. The inferences made from polarization and EIS studies are well supported by the lower extent of corrosion damage for MAO coated Mg and MAO coated Mg modified by NaF. The extent of corrosion protection offered by coatings is a function of their thickness, porosity, and chemical composition. Formation of MAO coating has led to an increase in corrosion resistance of Mg. Nevertheless, the limited thickness (~ 8 to 10 μm), permeation of HBSS thorough the pores, and lower chemical stability of MgO limit the extent of corrosion protection offered by it. In contrast, the relatively higher thickness (~ 12–14 μm), complete coverage of the porous structure by the modified layer and the formation of Mg(OH)2 − xFx and NaMgF3 has enabled a better corrosion resistance for MAO coated Mg modified by NaF. The morphological features, chemical composition and TF-XRD patterns confirm the deposition of monocalcium phosphate anhydrous and newberyite on the surface of the MAO coated Mg modified by NaF after immersion in c-SBF at 37 °C for 24 h, followed by heat-treatment at 300 °C for 1 h (Figs. 10 and 11). The higher surface area of the nanosheets favoured the nucleation of the monocalcium phosphate anhydrous, which is evidenced by their formation on the top of the walls. The ability of nanoplates of Mg(OH)2 to offer a high surface area has been reported previously [27]. Both monocalcium phosphate anhydrous and newberyite are biologically relevant compounds. The bioactivity of calcium phosphates is well established [33–36]. Newberyite has also been shown to exhibit a low cytotoxicity and positive results towards biocompatibility for osteoblastic cells [37]. Hence, the ability of MAO coated Mg modified by NaF to exhibit a better bioactivity is confirmed. The heat-treatment procedure at 300 °C for 1 h is adopted in this study to substantiate the ability of the MAO coated Mg modified by NaF to promote nucleation and growth of calcium phosphates. However, it is cautioned that adopting the heat-treatment step could cause a deleterious influence on the mechanical properties of Mg. The RGR, which is a measure of the cell viability is relatively lower for uncoated Mg than those observed over MAO coated Mg and MAO coated Mg modified by 0.1 M NaF (Fig. 13). The higher corrosion rate of uncoated Mg in the cell culture medium than the coated ones is considered responsible for this behaviour. The increase in concentration of Mg2+ ions, following the higher corrosion rate of uncoated Mg could change the osmolality of the cell culture medium leading to rupture of cell membranes and cell death [38]. The existence of a good correlation between higher corrosion rate of pure Mg as well as uncoated WE43 Mg alloy with their poor cytocompatibility has already been established by Jung et al. [39]. Cell proliferation and growth is evident in all the samples and the cell morphology appears to be similar (Fig. 14). However, the extent of cell growth is much pronounced over MAO coated Mg and MAO coated Mg modified by 0.1 M NaF (Fig. 14(b) and (c)) than those observed over uncoated Mg (Fig. 14(a)). This is due to the ability of the coated ones to resist the extent of corrosion attack, which is likely to create an environment amenable for cell proliferation and growth. In addition, it is also clear that the chemical constituents of MAO coating on Mg, namely, MgO, Mg2SiO4, MgSiO3 as well as Mg(OH)2 − xFx and NaMgF3 present in MAO coating modified by 0.1 M NaF did not cause any deleterious influence on the cytocompatibility in cell culture medium. The present study highlights modification of the surface of MAO coated Mg with the formation of a highly textural lamellar nanosheetlike morphology, homogenous over the surface without any cracks, by a simple and cost-effective chemical treatment at room temperature without the need for any template or additives. The modified surface endowed several beneficial attributes that include complete coverage of the porous structure of MAO coating, a significant improvement in corrosion resistance and a better bioactivity due to the higher surface area of the nanosheets. Moreover, the unique morphological feature of the modified surface is more appropriate to achieve better cell adhesion and proliferation. Since the top surface of the modified surface resembles that of a scaffold, it could be used for drug loading. The ability of the modified surface to improve the corrosion resistance as well as the

bioactivity suggest that it could be of immense help to modify the surface of Mg based absorbable implants. 5. Conclusions The findings of this study reveal that a highly textural lamellar nanosheet-like morphology can be formed over the surface of MAO coated Mg by immersion in 0.1 M NaF at 25 °C for 120 min, without the need for any template or additives. TF-XRD patterns of the modified surface confirms the formation of Mg(OH)2 − xFx and NaMgF3, besides the MgO, Mg2SiO4 and MgSiO3, which are originated from the MAO coating. In spite of the increase in surface roughness, the complete coverage of the porous structure of the MAO coating by the modified layer along with the formation of Mg(OH)2 − xFx and NaMgF3 has enabled a better corrosion resistance in HBSS. The higher surface area of the nanosheets favoured the nucleation of the monocalcium phosphate anhydrous along newberyite after immersion in c-SBF. The RGR of both uncoated and coated Mg are N 75% on all the three days, meeting Grade 1 specification in terms of cytocompatibility as per ISO 10993-5 standard. The extent of cell growth is much pronounced over MAO coated Mg and MAO coated Mg modified by 0.1 M NaF than those observed over uncoated Mg. The unique morphological feature of the modified surface helps to achieve an improved cell adhesion and proliferation. Hence, it can be concluded that surface modification of MAO coated Mg by immersion in 0.1 M NaF at 25 °C for 120 min could be of immense help to improve the performance of Mg based absorbable implants. Acknowledgements This work was financially supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (2011-0028709, 2013R1A1A2012322 and 2014R1A4A1005309) and Regional Strategic Industry project (2013-R0002274). This paper was also supported by the research funds of Chonbuk National University. The help rendered by Ms. Bae Soo Kyung, Research Facility Center, Engineering in SEM analysis and by Mr. Kim Dong-Hyun and Mr. Han YeonSoo, Physical Lab, Center for University-wide Research Facilities in recording the FT-IR and XRD spectra, are gratefully acknowledged. References [1] F. Witte, N. Hort, C. Vogt, S. Cohen, K.U. Kainer, R. Willumeit, F. Feyerabend, Degradable biomaterials based on magnesium corrosion, Curr. Opin. Solid State Mater. Sci. 12 (2008) 63–72. [2] R. Zeng, W. Dietzel, F. Witte, N. Hort, C. Blawert, Progress and challenge for magnesium alloys as biomaterials, Adv. Eng. Mater. 10 (2008) B3–B14. [3] H. Hornberger, S. Virtanen, A.R. Boccaccini, Biomedical coatings on magnesium alloys – a review, Acta Biomater. 8 (2012) 2442–2455. [4] J. Wang, J. Tang, P. Zhang, Y. Li, J. Wang, Y. Lai, L. Qin, Surface modification of magnesium alloys developed for bioabsorbable orthopedic implants: a general review, J. Biomed. Mater. Res. B 100 (2012) 1691–1701. [5] T.S.N. Sankara Narayanan, I.S. Park, M.H. Lee, in: T.S.N. Sankara Narayanan, I.S. Park, M.H. Lee (Eds.), Surface modification of magnesium and its alloys for biomedical applications: opportunities and challenges, Surface Modification of Magnesium and its Alloys for Biomedical Applications, vol. 1, Woodhead Publishing, UK (Elsevier Science Inc.) 2015, pp. 29–87 (Chapter 2). [6] C. Blawert, S.P. Sah, N. Scharnagl, M.B. Kannan, in: T.S.N. Sankara Narayanan, I.S. Park, M.H. Lee (Eds.), Lasma electrolytic oxidation/micro-arc oxidation of magnesium and its alloys, Surface Modification of Magnesium and its Alloys for Biomedical Applications, vol. 2, Woodhead Publishing, UK (Elsevier Science Inc.) 2015, pp. 193–234 (Chapter 8). [7] T.S.N. Sankara Narayanan, I.S. Park, M.H. Lee, Strategies to improve the corrosion resistance of microarc oxidation (MAO) coated magnesium alloys for degradable implants: prospects and challenges, Prog. Mater. Sci. 60 (2014) 1–71. [8] L. Weng, T.J. Webster, Increased osteoblast functions on nanostructured magnesium, Nanotechnology 23 (2012) 485105. [9] L. Weng, T.J. Webster, Nanostructured magnesium has fewer detrimental effects on osteoblast function, Int. J. Nanomedicine 8 (2013) 1773–1781. [10] Y. Zhu, G. Wu, Y.-H. Zhang, Q. Zhao, Growth and characterization of Mg(OH)2 film on magnesium alloy AZ31, Appl. Surf. Sci. 257 (2011) 6129–6137. [11] H. Zhao, S. Cai, S. Niu, R. Zhang, X. Wu, G. Xu, Z. Ding, The influence of alkali pretreatments of AZ31 magnesium alloys on bonding of bioglass–ceramic coatings and corrosion resistance for biomedical applications, Ceram. Int. 41 (2015) 4590–4600.

S.N. T.S.N., M.H. Lee / Surface & Coatings Technology 304 (2016) 425–437 [12] K. Huang, S. Cai, G.H. Xu, X.Y. Ye, Y. Dou, M.G. Ren, X.X. Wang, Preparation and characterization of mesoporous 45S5 bioactive glass–ceramic coatings on magnesium alloy for corrosion protection, J. Alloys Compd. 580 (2013) 290–297. [13] Y. Chen, G. Wan, J. Wang, S. Zhao, Y. Zhao, N. Huang, Covalent immobilization of phytic acid on Mg by alkaline pre-treatment: corrosion and degradation behavior in phosphate buffered saline, Corros. Sci. 75 (2013) 280–286. [14] J. Waterman, A. Pietak, N. Birbilis, T. Woodfield, G. Dias, M.P. Staiger, Corrosion resistance of biomimetic calcium phosphate coatings on magnesium due to varying pretreatment time, Mater. Sci. Eng. B 176 (2011) 1756–1760. [15] H.-S. Kim, S.-H. Um, S.-H. Rhee, The evaluation of hydroxyl ions as a nucleating agent for apatite on electrospun non-woven poly(ϵ-caprolactone) fabric, J. Biomat. Sci. 23 (2012) 1325–1338. [16] D. Wei, Y. Zhou, D. Jia, Y. Wang, Chemical treatment of TiO2-based coatings formed by plasma electrolytic oxidation in electrolyte containing nano-HA, calcium salts and phosphates for biomedical applications, Appl. Surf. Sci. 254 (2008) 1775–1782. [17] T.S.N. Sankara Narayanan, I.S. Park, M.H. Lee, Tailoring the composition of fluoride conversion coatings to achieve better corrosion protection of magnesium for biomedical applications, J. Mater. Chem. B 2 (2014) 3365–3382. [18] J.X. Yang, F.Z. Cui, Q.S. Yin, Y. Zhang, T. Zhang, X.M. Wang, Characterization and degradation study of calcium phosphate coating on magnesium alloy bone implant in vitro, IEEE T. Plasma Sci. 37 (2009) 1161–1168. [19] G. Yuan, J. Zheng, C. Lin, X. Chang, H. Jiang, Electrosynthesis and catalytic properties of magnesium oxide nanocrystals with porous structures, Mater. Chem. Phys. 130 (2011) 387–391. [20] G. Zou, R. Liu, W. Chen, Highly textural lamellar mesostructured magnesium hydroxide via a cathodic electrodeposition process, Mater. Lett. 61 (2007) 1990–1993. [21] G. Zou, W. Chen, R. Liu, Z. Xu, Morphology-tunable synthesis and characterizations of Mg(OH)2 films via a cathodic electrochemical process, Mater. Chem. Phys. 107 (2008) 85–90. [22] C.F. Li, M.J. Wang, W.H. Ho, H.N. Li, S.K. Yen, Effects of electrolytic MgO coating parameters on corrosion resistance of AZ91D magnesium alloy, J. Electrochem. Soc. 158 (2011) C11–C16. [23] Y. Lv, Z. Zhang, Y. Lai, Y. Liu, Electrodeposition of porous Mg(OH)2 thin films composed of single-crystal nanosheets, J. Electrochem. Soc. 159 (2012) D187–D189. [24] Y. Gao, H. Wang, Y. Su, Q. Shen, D. Wang, Influence of magnesium source on the crystallization behaviors of magnesium hydroxide, J. Cryst. Growth 310 (2008) 3771–3778. [25] G. Zou, R. Liu, W. Chen, Z. Xu, Preparation and characterization of lamellar-like Mg(OH)2 nanostructures via natural oxidation of Mg metal in formamide/water mixture, Mater. Res. Bull. 42 (2007) 1153–1158. [26] M.C. Zhao, M. Liu, G.L. Song, A. Atrens, Influence of pH and chloride ion concentration on the corrosion of Mg alloy ZE41, Corros. Sci. 50 (2008) 3168–3178.

437

[27] K.Y. Chiu, M.H. Wong, F.T. Cheng, H.C. Man, Characterization and corrosion studies of fluoride conversion coating on degradable Mg implants, Surf. Coat. Technol. 202 (2007) 590–598. [28] L.C. Wu, Y.J. Chen, M.L. Mao, Q.H. Li, M. Zhang, Facile synthesis of spike-piece-structured Ni(OH)2 interlayer nanoplates on nickel foam as advanced pseudocapacitive material for energy storage, ACS Appl. Mater. Interfaces 6 (2014) 5168–5174. [29] J. Zhang, N. Kong, J. Niu, Y. Shi, H. Li, Y. Zhou, G. Yuan, Influence of fluoride treatment on surface properties, biodegradation and cytocompatibility of Mg–Nd–Zn–Zr alloy, J. Mater. Sci. Mater. Med. 25 (2014) 791–799. [30] L. Mao, L. Shen, J. Chen, Y. Wu, M. Kwak, Y. Lu, Q. Xue, J. Pei, L. Zhang, G. Yuan, R. Fan, J. Ge, W. Ding, Enhanced bioactivity of Mg−Nd−Zn−Zr alloy achieved with nanoscale MgF2 surface for vascular stent application, ACS Appl. Mater. Interfaces 7 (2015) 5320–5330. [31] L. Wolters, S. Besdo, N. Angrisani, P. Wriggers, B. Hering, J.M. Seitz, J. Reifenrath, Degradation behaviour of LAE442-based plate–screw-systems in an in vitro bone model, Mater. Sci. Eng. C 49 (2015) 305–315. [32] T.F. da Conceiçao, N. Scharnagl, in: T.S.N. Sankara Narayanan, I.S. Park, M.H. Lee (Eds.), Fluoride conversion coatings for magnesium and its alloys for the biological environment, Surface Modification of Magnesium and its Alloys for Biomedical Applications, vol. 2, Woodhead Publishing, UK (Elsevier Science Inc.) 2015, pp. 3–21 (Chapter 1). [33] S. Shadanbaz, G.J. Dias, Calcium phosphate coatings on magnesium alloys for biomedical applications: a review, Acta Biomater. 8 (2012) 20–30. [34] R.A. Surmenev, M.A. Surmeneva, A.A. Ivanova, Significance of calcium phosphate coatings for the enhancement of new bone osteogenesis – a review, Acta Biomater. 10 (2014) 557–579. [35] S.V. Dorozhkin, Calcium orthophosphate coatings on magnesium and its biodegradable alloys, Acta Biomater. 10 (2014) 2919–2934. [36] S.V. Dorozhkin, Calcium orthophosphate deposits: preparation, properties and biomedical applications, Mater. Sci. Eng. C 55 (2015) 272–326. [37] F. Tamimi, D.L. Nihouannen, D.C. Bassett, S. Ibasco, U. Gbureck, J. Knowles, A. Wright, A. Flynn, S.V. Komarova, J.E. Barralet, Biocompatibility of magnesium phosphate minerals and their stability under physiological conditions, Acta Biomater. 7 (2011) 2678–2685. [38] J. Fischer, D. Profrock, N. Hort, R. Willumeit, F. Feyerabend, Improved cytotoxicity testing of magnesium materials, Mater. Sci. Eng. B 176 (2011) 830–834. [39] O. Jung, R. Smeets, D. Porchetta, A. Kopp, C. Ptock, U. Müller, M. Heiland, M. Schwade, B. Behr, N. Kroeger, L. Kluwe, H. Hanken, P. Hartjen, Optimized in vitro procedure for assessing the cytocompatibility of magnesium-based biomaterials, Acta Biomater. 23 (2015) 354–363.