Author's Accepted Manuscript

Enhancement of the mechanical properties of AZ31 magnesium alloy via nanostructured hydroxyapatite thin films fabricated via radio-frequency magnetron sputtering M.A. Surmeneva, A.I. Tyurin, T. Mukhametkaliyev, T.S. Pirozhkova, I.A. Shuvarin, M.S. Syrtanov, R.A. Surmenev www.elsevier.com/locate/jmbbm

PII: DOI: Reference:

S1751-6161(15)00068-5 http://dx.doi.org/10.1016/j.jmbbm.2015.02.025 JMBBM1402

To appear in: Journal of the Mechanical Behavior of Biomedical Materials

Received date:6 November 2014 Revised date: 19 February 2015 Accepted date: 23 February 2015 Cite this article as: M.A. Surmeneva, A.I. Tyurin, T. Mukhametkaliyev, T. S. Pirozhkova, I.A. Shuvarin, M.S. Syrtanov, R.A. Surmenev, Enhancement of the mechanical properties of AZ31 magnesium alloy via nanostructured hydroxyapatite thin films fabricated via radiofrequency magnetron sputtering, Journal of the Mechanical Behavior of Biomedical Materials, http://dx.doi.org/10.1016/j.jmbbm.2015.02.025 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Enhancement of the mechanical properties of AZ31 magnesium alloy via nanostructured hydroxyapatite thin films fabricated via radio-frequency magnetron sputtering

M.A. Surmeneva1, A.I. Tyurin2, T. Mukhametkaliyev1, T.S. Pirozhkova2, I.A. Shuvarin2, M.S. Syrtanov3, and R.A. Surmenev1,4

1

Department of Theoretical and Experimental Physics, National Research Tomsk Polytechnic University, 634050 Tomsk, Russia

2

NanoCenter “Nanotechnology and Nanomaterials”, G.R. Derzhavin Tambov State University, 392000 Tambov, Russia 3

Department of General Physics, National Research Tomsk Polytechnic University, 634050 Tomsk, Russia 4

Fraunhofer Institute for Interfacial Engineering and Biotechnology IGB, 70569 Stuttgart, Germany

e-mail: [email protected]

Abstract The structure, composition and morphology of a radio-frequency (RF) magnetron sputterdeposited dense nano-hydroxyapatite (HA) coating that was deposited on the surface of an AZ31 magnesium alloy were characterized using AFM, SEM, EDX and XRD. The results obtained from SEM and XRD experiments revealed that the bias applied during the deposition of the HA coating resulted in a decrease in the grain and crystallite size of the film having a crucial role in enhancing the mechanical properties of the fabricated biocomposites. A maximum hardness of 9.04 GPa was found for the HA coating, which was prepared using a bias of -50 V. The hardness of the HA film deposited on the grounded substrate (GS) was found to be 4.9 GPa. The elastic 1

strain to failure (H/E) and the plastic deformation resistance (H3/E2) for an indentation depth of 50 nm for the HA coating fabricated at a bias of -50 V was found to increase by ~30 and ~74 %, respectively, compared with the coating deposited at the GS holder. The nanoindentation tests demonstrated that all of the HA coatings increased the surface hardness on both the microscale and the nanoscale. Therefore, the results revealed that the films deposited on the surface of the AZ31 magnesium alloy at a negative substrate bias can significantly enhance the wear resistance of this resorbable alloy.

Key words: magnesium alloy, hydroxyapatite coating, RF magnetron sputtering, substrate bias, hardness, Young’s modulus, wear resistance

1.

Introduction

The successful application of different magnesium alloys as biodegradable biomaterials is critically dependent on the control of their degradation rate (Witte 2005). The fabrication of different coatings on the surface of different magnesium alloys (Mg, AZ31, AZ91, AZ91D, Mg4Y, Mg-Ca etc.) to prevent contact with their environment is the best way to avoid corrosion and has attracted the most interest (Phani 2005, Kannan Mathan 2013, Seyfoori 2013, Zhang 2014). Most studies are focused on investigating the influence of ceramic and organic films on the corrosion properties of magnesium alloys (Phani 2005, Gray-Munro 2009, Thomann 2010, Cui 2013, Kannan Mathan 2013). The most frequently used techniques to prepare protective CaP-based coatings on the surface of Mg alloys are biomimetic techniques (Yang , Lorenz 2009, Kannan 2011, Hornberger 2012), plasma electrolytic oxidation (Yao 2009, Bala Srinivasan 2010, Hornberger 2012), electrodeposition (Song 2008, Wen 2009), sol-gel (Roy 2011) etc. It should be mentioned that the prepared CaP coatings possess a thicknesses that is typically greater than several microns or even tens of microns (Hornberger 2012). However, the most important 2

limitation of these techniques is that the CaP coating adhesion strength is not sufficient for their practical applications. RF magnetron sputtering is an extremely useful technique to obtain wearresistant CaP coatings; moreover, a good biological response, both in vitro and in vivo, was found (Long 2002, Coelho 2005, Coelho 2009). To the best of our knowledge, there are no studies available that address dense nanostructured nanometre-thick RF magnetron sputterdeposited HA coatings on Mg alloys (Dorozhkin 2014, Surmenev 2014). Much of recent research includes studies on magnetron HA films deposited onto the surface of titanium, titanium alloys, NiTi, and silicon plates (Pichugin 2008, Surmenev 2011). Generally HA thin films deposited on titanium and its alloys using RF magnetron sputtering consist of a dense and nanocomposite structure, which provide a good biological response in vivo and in vitro and enhanced hardness of load-bearing surfaces (Surmenev 2012). HA-coated NiTi has previously been shown to significantly decrease Ni release in a 0.9% NaCl solution compared with the uncoated substrate. The thickness of the RF magnetron sputter deposited coating usually does not exceed 1 µm (Surmenev 2011, Surmeneva 2013a, Surmeneva 2013b). The average ion energy can be increased by applying a negative bias to the substrate to accelerate the ions from the plasma towards the substrate (Surmenev 2011). The kinetic energy of the ions is then converted to sputtering energy, thermal energy, implantation energy, and migration energy on the substrate surface for nucleation (Martin 1986, Surmeneva 2013a). Additionally, ions with very low energies may still influence coating nucleation and growth and enhance the chemical reactivity. In our previous studies, we have reported changes in the morphological properties and stoichiometry of HA coatings deposited on titanium as a function of the negative substrate bias (Surmenev 2011, Surmeneva 2013a). Therefore, additional substrate bias was used to vary the coating properties. The hardness of Mg alloys (Mg, AZ91D) is reported to be in the range of 0.5 to 0.75 GPa (Witte 2007, Zeng 2008). It is also reported that the biodegradation and mechanical properties of the magnesium alloy were an especially attractive combination for orthopaedic applications due 3

having elastic modulus of 40 to 50 GPa as compared to natural bone, which is in the range of 3 to 20 GPa (Staiger 2006). These mechanical and biodegradation properties are an important issue in evaluating the wear resistance of a surface for a long-term success of an implant. Moreover, enhancement of the H/E ratio (and thus the resistance of the material to plastic deformation H3/E2) of the implant surface may offer advantages, such as less potential for surface damage and increased durability. The term H/E can be considered to be a useful indicator of a good wear resistance of the coating (Leyland 2000). According to this study a coating with a high H/E ratio exhibited increased durability. The coatings with a high plastic resistance ratio H3/E2 are more likely to resist plastic deformation during low load contact events and exhibit higher yield strength (Musil 2000, Roy 2010, Surmeneva 2014). Few studies have been conducted that have focused on the influence of the surface substrate composition and the microstructure on the mechanical properties of HA thin films surfaces on titanium. The important parameters that are lacking are the hardness, the Young’s modulus and the resistance to plastic deformation (Dey 2013). The mechanical properties of the HA coating have been reported to be strongly affected by the film growth mechanism and the substrate microstructure. There is a lack of in-depth understanding of the mechanical properties as well as the wear resistance at the local nano- and micro-structural length scale of a HA coating obtained by RF magnetron sputtering on magnesium alloy. Therefore, the aim of this study is to evaluate the potential of a HA coating deposited by RF magnetron sputtering to enhance the hardness and resistance to plastic deformation of an AZ31 magnesium alloy. First, we will discuss the results of the influence of the applied substrate bias on the structure and morphology of the HA coating fabricated on the magnesium alloy. Subsequently, we will present results from nanoindentation tests and from the analysis of the correlation between the surface properties (roughness and microstructure) and obtained mechanical parameters (hardness, H, the Young’s modulus, E, elastic recovery, plastic resistance

4

parameter (H/E) and a resistance of the material to plastic deformation (H3/E2)), which provide information about the wear resistance of the HA-coated magnesium alloy.

2.

Materials and methods

AZ31 magnesium alloy substrates with the nominal mass composition of 96% Mg, 3% Al and 1% Zn were bought from GoodFellow (Germany). The size of the samples used in the experiments was set to 20 × 20 × 1 mm3 (width × length × thickness). The substrates were cleaned ultrasonically in a diluted acetone bath for 30 min at 100 °C. The substrates were ultrasonically cleaned a second time in ethanol followed by rinsing in distilled water. Pure HA target was prepared according to the procedures previously described (Surmenev 2011). A commercially available apparatus with an RF (13.56 MHz, COMDEL) magnetron source was used to deposit the HA coatings. Each coating was deposited at an RF-power level of 500 W in an argon atmosphere at a pressure of 0.1 Pa for 8 hours onto a substrate mounted in a grounded substrate (GS) holder or a substrate holder with a bias of -50 V. The application of the bias resulted in the coating thickness of 700±90 nm.

2.2

Coating characterisations

2.2.1. Coating thickness Optical ellipsometry (Ellipse 1891-S AG, Institute of Semiconductor Physics, Russian Academy of Sciences, Siberian Branch) was used to evaluate the thickness of the nanostructured HA coatings. Ellipsometric measurements were carried out at an incidence angle of 70° in the wavelength region of 250–1000 nm with a spectral resolution of 2 nm, which provided the accuracy of the measurements. The coating thickness was derived from the changes in the ellipsometric parameters between the bare and the coated substrates using a three-phase model 5

(substrate-layer-air) (Bu-Abbud 1986). Five measurements were done for each sample to reveal its thickness. For the purpose of thickness measurements three samples deposited under the same conditions were used. Then average values of thickness and standard deviation were calculated.

2.2.2. Phase analysis The phase composition and the structure of the CaP coatings were identified by X-ray diffraction (XRD-7000, Shimadzu, Japan) using CuKα radiation (λ=0.154 nm) in Bragg-Brentano mode and a 2Θ range from 10° to 60° with a scan speed of 2.0°/min, sampling pitch of 0.02°, preset time of 1.0 sec at 30 kV and at 30 mA. The average crystallites size, the degree of microstrain and the lattice parameters of the studied coatings were determined from the experimental XRD profiles taking into account the instrumental broadening through the use of the PowderCell-2.4 software. An instrumental broadening of 0.14 in 2θ was determined by the full width at half maximum for a silicon powder. As a reference for the pattern for the HA we used the ICDD database: #9-0432.

2.2.3. Microstructure study The surface morphology and composition of the deposited coatings were investigated using the MERLIN field emission scanning electron microscope (FE-SEM) equipped with energy dispersive X-ray spectroscopy (EDS, Carl Zeiss). Five different samples deposited under the same deposition conditions were studied. For each sample three measurements of EDS-spectra were done. Then an average value of the Ca/P ratio was calculated. The surface microstructure was additionally examined using atomic force microscopy (AFM) Solver HV (Russia), which was operating in tapping mode. Triangular golden silicon probes (NT-MTD) with a typical spring constant of 28 N⋅m-1 and a resonance frequency of 420 kHz were used. All of the images were collected in air at a typical frequency of 1.5 Hz and with 256 points per line. The roughness parameters of the surface were calculated three times at different spots for each specimen from 6

the AFM scans over the surface areas measure 5 × 5 µm2 using the Nova SPM software (NTMDT).

2.2.4. Nanoindentation study Nanoindentation tests were performed using a Nanotriboindenter TI-950 (Hysitron Inc., USA) equipped with a Berkovich indenter with an angle between the opposite faces of 142.3° and a tip radius of approximately 50 nm. Special software was used to calculate the nanohardness (H) and the reduced modulus (E) of the coatings according to the method of analysis given by Oliver and Pharr (Oliver 1992). Load–displacement curves with the load ranging from 10 µN tо 5 mN were obtained to determine the penetration depth (h), the elastic modulus (E) and the hardness (H) of the composites as a function of the applied load. A maximum force was achieved for 5 s and the dwell time was set to 2 s followed by 5 s of unloading. Repeated indentations were performed on different coated and uncoated AZ31 magnesium alloys, and the values of Н and Е were calculated as an average of 10 indentations.

3.

Results and discussion

Fig. 1 shows the AFM surface topography images of the HA coatings fabricated on AZ31 magnesium alloy. The roughness parameters of the HA coatings and of the uncoated substrates are summarized in Table 1. The Sq of the uncoated alloy was 37.6 and 79.5 nm for the scan areas of 5 × 5 and 35 × 35 µm2, respectively. The experimental data for the scan area of 35 × 35 µm2 estimated by the AFM revealed that the roughness of the substrate on the microscale level increased when the HA coating was deposited. According to Table 1, after HA film deposition onto the AZ31 substrate, the surface structure on the microscale level became smoother, which was indicated by the roughness parameter values for the of 5 × 5 µm2 scan areas. However, it was observed that the application of the bias resulted in a decrease in the surface roughness parameters for the HA coatings. Therefore, the Sq and Smax values of the film deposited at the GS 7

were determined to be 38.5.2±4.5 nm and 241.0±51.5 nm for a scan areas of 5 × 5 µm2. The HA coating obtained at the applied bias revealed slightly lower Sq and Smax values of 29.5±6.1 nm and 223.0±37.3 nm on the nanoscale level. The observed decrease in the surface roughness with an increase in the applied bias can be connected to an increase in the atomic movement and densification of the film material as a result of the increased flux and energy of the ions (Chang 2008, Kong 2011). The decrease in the surface roughness could also be connected to the surface re-sputtering effects, which was also observed in the study (Tsai 2012). The SEM images of the HA coating deposited onto the surface of the magnesium alloy at the GS revealed irregular grain shapes (Fig. 2). The average grain size of HA coating in case of an applied bias of -50 V was observed to be clearly reduced leading to an increased areal density on the surface. The effect of the applied bias was also observed in the case of other coatings (Kong 2011, Tsai 2012). Moreover, in the case of the substrate bias, grain boundaries are more clearly visible. No surface porosity is observed. The EDS results revealed that the Ca/P ratio for the coatings obtained at the GS and with the application of 50 V bias are 1.57±0.03 and 1.62±0.03, respectively. These values are close to 1.67, which corresponds to stoichiometric HA. The XRD patterns of the deposited coatings and uncoated substrates are presented in Fig. 3. The spectra were obtained in the grazing incidence mode. This is the reason for the increasing background intensity at decreasing low diffraction angles. For all of the coatings, the HA crystalline structure was revealed. According to JCPDS 9-432 standard the main peaks are from the planes of (200), (111), (002), (210), (211), (112), (300), and (310), of HA phase for 20 - 50 deg. 2θ. No diffraction peaks from other phases, such as calcium oxide, tricalcium phosphate or tetracalcium phosphate, were detected. The diffraction peak profiles are influenced by the crystalline size and the microstrain. Lattice parameters, crystalline size estimated for the coatings deposited via RF magnetron sputtering for different deposition conditions are presented in Table 2. Three samples were investigated deposited under the same deposition parameters. The values of crystalline size and microstrain were calculated for each sample. Then average values were 8

calculated. Based on the results obtained we can conclude that significant difference exists between the groups. The HA coating deposited at the GS holder consisted of crystallites with an average size of 30 nm (Table 2). The HA coatings deposited with a bias of -50 V demonstrated a nanocrystalline structure with the crystallites of an average size of 15 nm. The lower crystallite size in the latter case can be directly connected with an increase in the ion bombardment of the substrate, which results in the suppression of the crystallite growth and an increase in the microstrain. A substrate bias has been reported to promote enhanced nucleation and, as a result, a reduction in the grain size is observed (Eckert 2011). A decrease in the grain size was also observed in case of the negative substrate bias for CrN films (Lee 2006), multi-component (AlCrMoSiTi)N coatings (Chang 2008), (TiVCrZrHf)N coatings (Tsai 2012), and for TiC films (Wang 2006). The nanoindentation response of the thin film is known to be sensitive to such factors as the surface roughness and the presence of processing-induced surface residual stress as well as the texture, grain size, elastic anisotropy and the thickness of the film on the substrate (Gouldstone 2000). Moreover, the indentation in the case of a thin film has been known to be strongly affected by the nucleation of defects and by the plastic deformation characteristics of the films, which are functions of the film thickness and microstructure (Gouldstone 2000). The hardness demonstrated a maximum at the same grain sizes for the thin TiCxN1-x (0≤x≤1) films as the intrinsic stress was maximum (Karlsson 2000). The authors concluded that the increase in the intrinsic stress and the peak broadening in thin films indicated a difference in the density or the type of the residual defects. The dominating hardening mechanism was found to be hindering of the dislocation movement by the defects in the lattice. The typical load-displacement curves obtained for the uncoated substrate and for the CaP coating deposited on the AZ31 magnesium alloy are shown in Fig. 4. The load-displacement curves showed that the deformation behaviour of the coating and of the substrate was plastic with some elastic components, i.e., when the load was removed, a part of the deformation relaxed. 9

The non-coated magnesium alloy substrates exhibited yielding at the maximum load and allowed for considerable plastic deformation. The HA coating deposited at an impulse substrate bias of 50 V was significantly more resistant to penetration by the indenter tip. The maximum displacements at the maximum force for the bias-deposited samples are less than those for the HA coating fabricated at the grounded substrate. At a load of 1 mN, the penetration depth of the indenter tip into the HA coating was ~47 nm when the bias was applied, while for the HA coating deposited at the grounded substrate the depth was ~67 nm. In the case of uncoated substrate, displacement excursions (pop-ins) of different lengths were observed. These pop-ins likely occurred due to incipient plasticity as the crystal deformation transitioned from being elastic to plastic (Wang 2004). The improved resistance to plastic deformation afforded by the uncoated and HA-coated magnesium alloy is clearly visible. As the maximum load is increased to 1 mN, the HA films remain almost purely elastic whereas the substrate has a strong increased permanent displacement (Fig. 4). The calculated nanohardness values, H, and the Young’s modulus, E, with their dependence on the indentation depth are presented in the Fig. 5. The hardness, H and the Young’s modulus, E, for all of the HA films were found to decrease with an increasing penetration depth due to an enhanced influence of the substrate. The effect of substrate and penetration depth on the mechanical properties, such as the hardness and Young’s modulus, of the thin films has been studied elsewhere (Saha 2002). Two types of substrates and films were defined, namely hard and soft substrates and hard and soft films. In the case of hard films on a soft substrate, a reduced H was observed with an increasing indentation depth. Moreover, the indentation hardness H in the case of a hard film on a soft substrate could never approach a constant value even when the indentation depth reached the film thickness (Zhang 2007). This result is different from the indentation of bulk metals and is clearly due to the effect of the soft substrate (Zhang 2007). Here, the magnesium alloy is used as a substrate, which is much softer than the HA film. The average values of E and H obtained from loading–unloading 10

measurements for the penetration depths of 50 and 100 nm are presented in Table 3. Here, an indentation depth of 50 and 100 nm corresponded to approximately 6 and 12% of the HA film thickness, respectively. The values of the elastic strain to failure ratio (H/E) for the HA films deposited at the different biases and for the uncoated substrates are also summarized in Table 3. The magnesium alloy substrate revealed a nanohardness of 1.19± 0.32 GPa and a Young’s modulus of 44.24 ± 3.40 GPa for a penetration depth of 100 nm. The HA coated alloy substrate showed an increased Young’s modulus and nanohardness compared with the uncoated substrate. The HA coating, for the case of a -50 V bias, showed the greatest nanohardness of 6.89 ± 1.23 GPa. The HA coating deposited on the GS had an average hardness of 3.08 GPa and an elastic modulus of 79.03 GPa. The observed differences are likely attributed to the different coating structure prepared at the GS holder and the substrate bias of -50 V. The effect of surface roughness can also be taken into consideration. A decrease in the surface roughness with an increase in the substrate bias was also observed in the study (Lai). In this study, smoother surface topographies were observed in the case of the coatings deposited at a bias of 50 V compared with the grounded case (Table 1). The nanoindentation hardness, the Young's modulus, and the toughness of the films increase as the substrate bias increases (Wang 2006). A similar trend was observed in the case of multi-element (AlCrTaTiZr)N coatings in a previous study (Lai). An increase in the hardness by 10% between 40 and 70 V was observed in case of a TiAlN PVD coating (Ahlgren 2005). As the substrate bias increases to approximately -150 V, the internal stress, the hardness and the Young's modulus of ZrN films reach a maximum (Niu E.W. 2007). A maximum hardness was obtained for a TiN–MoSx composite coating deposited at a substrate bias voltage of -40 V (Gangopadhyay 2010). The TiN-deposited Co–Cr with a negative substrate bias demonstrated a very high hardness of 44.7±1.7 GPa, which was much greater than the hardness of the bare Co–Cr (4.2±0.3 GPa) and TiN-deposited Co–Cr without a negative substrate bias (23.6±2.8 GPa) (Pham 2011).

11

Very little information is available in the literature devoted to the mechanical properties of different Mg alloys coated with CaP or other protective coatings. However, the H and E values obtained in this study are in good agreement with the results reported other authors when other coating deposition techniques were used. However, one should mention that the primary advantage of the CaP coating is that it resembles the composition bone tissue. The hardness and the Young’s modulus of micro-arc oxidized (MAO) AZ31B alloy were ~3 GPa and ~90 GPa, respectively (Dey 2013). The Vickers hardness of MAO-deposited coatings on pure Mg was found to be 5.2 GPa (530 HV), which was 10 times greater than the uncoated substrate. The microhardness of the Y1 magnesium alloy surface reached 4.02 GPa (410 HV), which was much greater than the original Y1 magnesium alloy without MAO. Depth-sensing nanoindentation tests were performed on an AZ91D alloy with a 2 µm thick ZrO2-CeO2 and resulted in a hardness and elastic modulus of 4.5 GPa and 98 GPa, respectively (Phani 2005). The wear resistance of a solid can be adjusted by tailoring its elastoplastic properties (Leyland 2000). To predict the wear resistance of the HA coating, the ratio H/E was determined from the data obtained for different penetration depths. From the H/E values, shown in Table 3, it is observed that the applied bias tends to increase the wear resistance of the HA coatings. H3/E2 increased from 4.68 x 10-3 to 48.4 x 10-3 GPa at a penetration depth of 100 nm when a bias of -50 V was applied. The nanohardness, H, the elastic strain to failure (H/E) and the plastic deformation resistance (H3/E2) for an indentation depth of 100 nm in case of the HA coating fabricated with a bias of 50 V was found to increase by ~55, ~54 and ~90 %, respectively, compared with the coating deposited at the GS. According to the authors of the study in (Johnson), the H3/E2 parameter was associated with the film resistance to plastic deformation. The HA film in the case of bias of a -50 V showed a slightly greater value of Е for a penetration depth of 50 nm but significantly greater values of H/E and H3/E2 than the parameters obtained for the samples with the HA coating fabricated at the GS. From the nanoindentation results discussed here, it is possible to infer that all of the HA coatings increased the nanohardness and 12

elasticity. This association improved the elasticity index and the resistance to plastic deformation, which suggests a better tribological performance of the surface for medical applications. In studies (Viswanath 2007, Saber-Samandari 2009 , Zamiri 2011) the indentation tests were performed to investigate the mechanical properties of HA single crystals. These studies showed interesting scale-dependent behavior. At the microscale, HA crystals were very brittle and crack during microindentation (Viswanath 2007). However, nanoindentation reveals plastic response due to dislocation activity (Viswanath 2007). As shown in our previous work the HA-based coating fabricated vie RF-magnetron sputtering exhibited the nanocomposite structure. Some generalizations may be made regarding the deformation mechanisms of the nanocomposite HA films. Nanocomposite structure affects the plastic deformation processes of dislocation emission/ propagation and grain boundary sliding. The initial purely elastic deformation also can be due to additional atomic movements caused by the indenter load on the film surface. Thus, in the study (Ievlev 2014) the structure of nanocrystalline HA and amorphous calcium phosphate coatings fabricated via RF-magnetron sputtering on the titanium and silicon substrates in the indentation zone has been investigated using high-resolution transmission electron microscopy. The plastic deformation is interpreted in terms of the cluster (Ca9(PO4)6, Ca3(PO4)2, and PO43-) representation of the HA structure and amorphous calcium phosphates of the same elemental composition and cluster-boundary sliding during the deformation. To evaluate the ability of the HA coating to resist contact damage, indentations on the surface of the film deposited at the GS, which were obtained at a load of 5 mN using a Berkovich indenter, were studied (Fig. 6). The radial cracks were absent at the corners of indentations formed when the Berkovich indenter penetrated into the HA-coatings deposited on the surface of the AZ31 alloy. No apparent pile-up and sink-in could be observed at the indent edges of HA-coated magnesium alloy, Fig. 6 (b). The surface of the HA-coated remained unchanged, as it can be seen in Fig. 6 a. The HA-coating surface revealed densely packed grains (Fig. 2), as mentioned 13

before. Therefore, it could be concluded, that the values of nanohardness obtained for the HA films on the AZ31 substrate are correct. However, it is important to note that changes in the film density are very difficult to detect using AFM, because most of the voids can be located under the surface. This effect was observed in the study (Tsui 1999), where investigations of the plastic deformation mechanism around plane strain indentations made in hard thin films deposited on soft substrates were reported. The results indicated that the plastic deformation around the plane strain indentations was dominated by the soft substrate used in the study. There was a small change in the film thickness around the indentation. Thus, we can conclude that the surface mechanical properties reported in this study improved after HA film deposition. In further, investigations of the adhesion strength and deformation behavior of the deposited on the surface of magnesium alloy HA coatings will be done. The biological effect of the fabricated nanostructured films on the surface of bioresorbable AZ31 magnesium alloy will also be studied in vitro.

Conclusions The nanoscale hardness and the Young’s modulus of the nanostructured RF magnetron sputterdeposited HA coatings fabricated via RF magnetron sputtering method on an AZ31 alloy were measured using a low load nanoindentation technique from which the elastic strain to failure (H/E) and the plastic resistance ratio H3/E2, and the elastic recovery of the displacement upon unloading %R was calculated. Moreover, the effect of the bias on the structure, roughness and morphology of the HA coatings was investigated using XRD, SEM and AFM. Several trends could be extracted from the nanoindentation experiments. The nanoindentation test demonstrated that all of the HA coatings increased the hardness of the magnesium alloy on both the micro- and nanoscale. The nanohardness of the HA coatings increased when a bias was applied during deposition, and the nanohardness reached a value of 9.04 GPa primarily due to the strengthening 14

effect offered through the decrease of the grain and crystallites size. The elastic strain to failure (H/E) and the plastic deformation resistance (H3/E2) for an indentation depth of 50 nm for the HA coating fabricated at a bias of -50 V was found to increase by ~30 and ~74%, respectively, compared with the coating deposited at the GS holder.

Acknowledgements The authors acknowledge the support of the Russian Science Foundation (project number 14-1300274). The authors are thankful to Mrs. A.A. Ivanova for help with the measurements of the coating thickness.

References: Ahlgren, M., Blomqvist, H., 2005. Influence of bias variation on residual stress and texture in TiAlN PVD coatings. Surf. Coat. Technol. 200, 157-160. Bala Srinivasan, P., Liang, J., Blawert, C., Stoermer, M., Dietzel, W, 2010. Characterization of calcium containing plasma electrolytic oxidation coatings on AM50 magnesium alloy. Appl. Surf. Sci. 256, 4017-4022. Bu-Abbud, G.H., Bashara, N.M., Woollam, J.A., 1986. Thin Solid Films 138, 27-41. Chang, H.-W., Huang, P.-K., Yeha, J.-W., Davison, A., Tsau, C.-H., Yang, C.-C., 2008. Influence of substrate bias, deposition temperature and post-deposition annealing on the structure and properties of multi-principal-component (AlCrMoSiTi)N coatings. Surf. Coat. Technol. 202, 3360-3366. Coelho, P.G., Cardaropoli, G., Suzuki, M., Lemons, J.E., 2009. Early Healing of Nanothickness Bioceramic Coatings on Dental Implants. An Experimental Study in Dogs. J. Biomed. Mater. Res. Part B: Appl Biomater. 88, 387-393.

15

Coelho, P.G., Suzuki, M., 2005. Evaluation of an IBAD thin-film process as an alternative method for surface incorporation of bioceramics on dental implants. A study in dogs. J. Appl. Oral Sci. 13, 87-92. Cui, W., Beniash, E., Gawalt, E., Xu, Z., Sfeir, C., 2013. Biomimetic coating of magnesium alloy for enhanced corrosion resistance an d calcium phosphate deposition. Acta Biomater. 9, 8650–8659. Dey, A., Uma Rani, R., Thota, H.K., Sharma, A.K., Bandyopadhyay, P., Mukhopadhyay, A.K., 2013. Microstructural, corrosion and nanomechanical behaviour of ceramic coatings developed on magnesium AZ31 alloy by micro arc oxidation. Ceramics International 39, 3313-3320. Dorozhkin, S., 2014. Calcium orthophosphate coatings on magnesium and its biodegradable alloys. Acta Biomater. 10, 2919-2934. Eckert, M., Mortet, V., Zhang, L., Neyts, E., Verbeeck, J., Haenen, K., Bogaerts, A., 2011. Theoretical Investigation of Grain Size Tuning during Prolonged Bias-Enhanced Nucleation. Chem. Mater. 23, 1414-1423. Gangopadhyay, S., Acharya, R., Chattopadhyay, A.K., Paul, S., 2010. Effect of substrate bias voltage on structural and mechanical properties of pulsed DC magnetron sputtered TiNMoSx composite coatings. Vacuum 84, 843-850. Gouldstone, A., Koh, H.-J., Zeng, K.-Y., Giannakopoulos, A.E., Suresh, S., 2000. Discrete and continuous deformation during nanoindentation of thin films. Acta Mater. 48, 2277-2295. Gray-Munro, J.E., Seguin, C., Strong, M., 2009. Influence of surface modification on the in vitro corrosion rate of magnesium alloy AZ31. J. Biomed. Mater. Res. 91A, 221-230. Hornberger, H., Virtanen, S., Boccaccini, A.R., 2012. Biomedical coatings on magnesium alloys – A review. Acta Biomater. 8, 2442-2455.

16

Ievlev, V.M., Kostyuchenko, A.V., Darinskii, B.M., Barinov, S.M., 2014. Hardness and microplasticity of nanocrystalline and amorphous calcium phosphate coatings. Physics of the Solid State 56, 321-329. Johnson, K.L. Contact Mechanics, Cambridge University press, London, UK, ISBN: 0-52134796-3, 1985, 464 p. Kannan Mathan, B., Liyanaarachchi, S., 2013. Hybrid coating on a magnesium alloy for minimizing the localized degradation for load-bearing biodegradable mini-implant applications. Mater. Chem. Phys. 142, 350-354. Kannan, M.B., Orr, L., 2011. In vitro mechanical integrity of hydroxyapatite coated magnesium alloy. Biomed. Mater. 6, 045003. Karlsson, L., Hultman, L., Johansson, M.P., Sundgren, J.-E., Ljungcrantz, H., 2000. Growth, microstructure, and mechanical properties of arc evaporated TiCxN1-x (0≤x≤1) films. Surface and Coatings Technology 126, 1-14. Kong, Q., Ji, L., Li, H., Liu, X., Wang, Y., Chen, J., Zhou, H., 2011. Influence of substrate bias voltage on the microstructure and residual stress of CrN films deposited by medium frequency magnetron sputtering. Materials Science and Engineering B 176, 850-854. Lai, C.-H., Lin, S.-J., Yeh, J.-W., Davison, A., 2006. Effect of substrate bias on the structure and properties of multi-element (AlCrTaTiZr)N coatings. J. Phys. D: Appl. Phys. 39, 46284633. Lee, J.-W., Tien, S.-K., Kuo, Y.-C., 2006. The effects of pulse frequency and substrate bias to the mechanical properties of CrN coatings deposited by pulsed DC magnetron sputtering. Thin Solid Films 494, 161-167. Leyland, A., Matthews, A., 2000. On the significance of the H/E ratio in wear control: a nanocomposite coating approach to optimised tribological behavior. Wear 246, 1-11.

17

Long, J.D., Xu, S., Cai, J.W., Jiang, N., Lu, J.H., Ostrikov, K.N., Diong, C.H., 2002. Structure, bonding state and in-vitro study of Ca-P-Ti film deposited on Ti6Al4V by RF magnetron sputtering. Mater. Sci. Eng. C 20, 175-180. Lorenz, C., Brunner J.G., Kollmannsberger P., Jaafar L., Fabry B., Virtanen S., 2009. Effect of surface pre-treatments on biocompatibility of magnesium." Acta Biomater. 5, 2783-2789. Martin, P.J., 1986. Review ion based methods for optical thin film deposition. J. Mater. Sci. 21, 1-25. Musil, J., 2000. Hard and superhard nanocomposite coatings. Surf. Coat. Technol. 125, 322-330. Niu E.W., Li, L., Lv., G.H., Chen, H., Feng, W.R., Fan, S.H., Yang, S.Z., Yang, X.Z., 2007. Influence of substrate bias on the structure and properties of ZrN films deposited by cathodic vacuum arc. Mater. Sci. Eng. A 460-461, 135-139. Oliver, W.C., Pharr, G.M., 1992. An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. J. Mater. Res. 7, 1564-1583. Pham, V.-H., Yook, S.-W., Li, Y., Jeon, G., Lee, J.-J., Kim, H.-E., Koh, Y.-H., 2011. Improving hardness of biomedical Co–Cr by deposition of dense and uniform TiN films using negative substrate bias during reactive sputtering. Mater. Lett. 65, 1707-1709. Phani, A.R., Gammel, F.J., Hack, T., Haefke, H., 2005. Enhanced corrosioon resistance by solgel-based ZrO2-CeO2 coatings on magnesium alloys. Materials and Corrosion 56, 77-82. Pichugin, V.F., Surmenev, R.A., Shesterikov, E.V., Ryabtseva, M.A., Eshenko, E.V., Tverdokhlebov, S.I., Prymak, O., Epple, M., 2008. The preparation of calcium phosphate coatings on titanium and nickel-titanium by rf-magnetron sputtered deposition: composition, structure and micromechanical properties. Surf. Coat. Technol. 202, 39133920. Roy, A., Singh, S., Lee, B., Ohodnicki, J., Kumta P.N., 2011. Novel sol-gel derived calcium phosphate coatings on Mg4Y alloy. Mater. Sci. Eng. B 176, 1679-1689. 18

Roy, M.E., Whiteside, L.A., Xu, J., Katerberg, B.J., 2010. Diamond-like carbon coatings enhance the hardness and resilience of bearing surfaces for use in joint arthroplasty. Acta Biomater. 6, 1619-1624. Saber-Samandari, S., Gross, K.A., 2009. Nanoindentation reveals mechanical properties within thermally sprayed hydroxyapatite coatings. Surf. Coat. Technol. 203, 1660-1664. Saha, R., Nix, W.D., 2002. Effects of the substrate on the determination of thin film mechanical properties by nanoindentation. Acta Mater. 50, 23-38. Seyfoori, A., Mirdamadi, Sh., Seyedraoufi, Z.S., Khavandi, A., Aliofkhazraei, M., 2013. Synthesis of biphasic calcium phosphate containing nanostructured films by micro arc oxidation on magnesium alloy. Mater. Chem. Phys. 142, 87-94. Song, Y.W., Shan, D.Y., Han, E.H., 2008. Electrodeposition of hydroxyapatite coating on AZ91D magnesium alloy for biomedical application. Mater. Lett. 62, 3276-3279. Staiger, M.P., Pietak, A.M., Huadmai, J., Dias, G., 2006. Magnesium and its alloys as orthopedic biomaterials: a review. Biomaterials 27, 1728-1734. Surmenev, R.A,. 2012. A review of plasma-assisted methods for calcium phosphate-based coatings fabrication. Surf. Coat. Technol. 206, 2035-2056. Surmenev, R.A., Surmeneva, M.A., Ivanova, A.A., 2014. Significance of calcium phosphate coatings for the enhancement of new bone osteogenesis – a review. Acta Biomater. 10, 557-579. Surmenev, R.A., Surmeneva, M.A., Evdokimov, K.E., Pichugin, V.F., Peitsch, T., Epple, M., 2011. The influence of the deposition parameters on the properties of an rf-magnetrondeposited nanostructured calcium phosphate coating and a possible growth mechanism. Surf. Coat. Technol. 205, 3600-3606 Surmeneva, M.A., Kovtun, A., Peetsch, A., Goroja, S.N., Sharonova, A., Pichugin, V.F., Teresov, A.D., Koval, N.N., Buck, V., Wittmar, A., Ulbricht, M., Prymak, O., Epple, M., Surmenev, R.A., 2013a. Preparation of a silicate-containing hydroxyapatitebased coating 19

by magnetron sputtering : structure and osteoblast-like MG63 cells in vitro study. RSC Adv. 3, 11240-11246. Surmeneva, M.A., Chaikina, M.V., Zaikovskiy, V.I., Pichugin, V.F., Prymak, O., Epple, M., Surmenev, R.A., 2013b. The structure of an RF-magnetron sputter-deposited silicatecontaining hydroxyapatite-based coating investigated by high-resolution techniques. Surf. Coat. Technol. 218, 39–46. Surmeneva, M.A., Surmenev, R.A., Tyurin, A.I., Mukhametkaliyev, T.M., Teresov, A.D., Koval, N.N., Pirozhkova, T.S., Shuvarin, I.A., Oehr, C., 2014. Comparative study of the radiofrequency magnetron sputter deposited CaP films fabricated onto acid-etched or pulsed electron beam-treated titanium. Thin Solid Films 571, 218-224. Thomann, M., Krause, C., Angrisani, N., Bormann, D., Hassel, T., Windhagen, H., MeyerLindenberg, A., 2010. Influence of a magnesium-fluoride coating of magnesium-based implants (MgCa0.8) on degradation in a rabbit model. J. Biomed. Mater. Res. 93A, 16091619. Tsai, D.-C., Liang, S.-C., Chang, Z.-C., Lin, T.-N., Shiao, M.-H., Shieu, F.-S., 2012. Effects of substrate bias on structure and mechanical properties of (TiVCrZrHf)N coatings. Surf. Coat. Technol. 207, 293-299. Tsui, T.Y., Vlassak, J., Nix, W.D., 1999. Indentation plastic displacement field: Part II. The case of hard films on soft substrates. J. Mater. Res. 14, 2204-2209. Viswanath, B., Raghavan, R., Ramamurty, U., Ravishankar, N., 2007. Mechanical properties and anisotropy in hydroxyapatite single crystals. Scr. Mater. 57, 361-364. Wang, H., Zhang, S., Li, Y., Sun, D., 2006. Bias effect on microstructure and mechanical properties of magnetron sputtered nanocrystalline titanium carbide thin films. Thin Solid Films 516, 5419-5423. Wang, M.G., Ngan, A.H.W., 2004. Indentation strain burst phenomenon induced by grain boundaries in niobium. J. Mater. Res. 19, 2478-2486. 20

Wen, C., Guan, S., Peng, L., Ren, C., Wang, X., Hu, Z., 2009. Characterization and degradation behavior of AZ31 alloy surface modified by bone-like hydroxyapatite for implant applications. Appl. Surf. Sci. 255, 6433-6438. Witte, F., Feyerabend, F., Maier, P., Fischer, J., Stoermer, M., Blawert, C., Dietzel, W., Hort, N., 2007. Biodegradable magnesium–hydroxyapatite metal matrix composites. Biomaterials 28, 2163-2174. Witte, F., Kaese, V., Haferkamp, H., Switzer, E., Meyer-Lindenberg, A., Wirth, C.J., Windhagen, H., 2005. In vivo corrosion of four magnesium alloys and the associated bone response. Biomaterials 26, 3557-3563. Yang, J.X., Cui, F.Z., Lee, I.-S., Zhang, Y., Yin, Q.S., Xia, H., Yang, S.X., 2011. In vivo biocompatibility and degradation behavior of Mg alloy coated by calcium phosphate in a rabbit model. J. Biomater. Appl. http://dx.doi.org/10.1177/0885328211398161. Yao, Z., Li, L., Jiang, Z., 2009. Adjustment of the ratio of Ca/P in the ceramic coating on Mg alloy by plasma electrolytic oxidation. Appl. Surf. Sci. 255, 6724-6728. Zamiri, A., De, S., 2011. Mechanical properties of hydroxyapatite single crystals from nanoindentation data. J. Mechan. Behav. Biomed. Mater. 4, 146-152. Zeng, R., Dietzel, W., Witte, F., Hort, N., Blawert, C., 2008. Progress and challenge for magnesium alloys as biomaterials. Adv. Eng. Mater. 10, B3-B14. Zhang, F., Saha, R., Huang, Y., Nix, W.D., Hwang, K.C., Qu, S., Li, M., 2007. Indentation of a hard film on a soft substrate: Strain gradient hardening effects. Int. Journal of Plasticity 23. 25-43. Zhang, X., Li, X.-W., Li, J.-G., Sun, X.-D., 2014. Preparation and Characterizations of Bioglass Ceramic Cement/Ca−P Coating on Pure Magnesium for Biomedical Applications. Appl. Mater. Interf. 6, 513−525.

21

Figures and figures’ captions

а) initial uncoated substrate

\ b) substrate coated with CaP at the GS

22

c) substrate coated with CaP at the bias -50 V Figure 1. The typical 3D and 2D views of 5 × 5 µm2 morphologies for (a) uncoated AZ31 magnesium alloy and HA coatings deposited on the (b) GS and (c) at a bias of -50 V. Images obtained using AFM.

23

Figure 2. SEM micrographs of HA coatings on the surface of the AZ31 substrate deposited by RF magnetron sputtering at the GS a) and at a substrate bias of -50 V b)

24

Fig. 3. The typical XRD patterns of the samples with an incident beam angle of 1.0°: а) uncoated AZ31 magnesium alloy, b) HA deposition at the GS for 480 min, c) HA deposition at a bias of 50 V for 480 min.

25

c) b)

1000

a)

Load / µN

800 600 400 200 0

0

30

60 90 120 150 Penetration depth / nm

180

Fig. 4. The typical load-unloading curve patterns obtained for the uncoated substrate a) the HA coating deposited at the GS holder b) and with a negative substrate bias of 50 V c) at a load of 1 mN.

26

a)

b)

Fig. 5. The hardness, H, and the Young’s modulus, E, values plotted as a function of the indentation depth for the HA coatings deposited on the GS a) and with a bias of -50 V b).

27

а)

b)

Fig. 6. A typical image of an indentation obtained to study the damage resistance of the HA coating: (a) an indentation obtained via a Berkovich indenter at a load of 5 mN, imaged by AFM, (а) a typical surface profile (b).

28

List of tables and tables captions

Table 1. The surface roughness parameters of uncoated AZ31 magnesium alloy and HA coatings deposited at the bias - 50 V and uncoated substrate. Scan area 5 × 5 µm

2

35 × 35 µm2

Roughness

Uncoated

HA coating

HA coating,

parameters

substrate

GS

-50 V

Sa, nm

28.8±0.9

30.7±1.5

22.3±2.1

Sq, nm

37.6±3.9

38.5±4.5

29.5±6.1

Sz, nm

304.3±34.6

217.1±41.2

208.2±25.2

Smax, nm

375.2±41.1

241.0±51.5

223.0±37.3

Sa, nm

62.8±3.9

98.5±3.5

94.5±4.3

Sq, nm

79.5±7.8

123.1±8.5

120.2±7.1

Sz, nm

650.3±38.2

847.1±47.7

828.2±33.6

Smax, nm

738.0±48.1

992.2±56.5

906.0±22.1

29

Table 2. Lattice parameters, crystalline size estimated for the coatings deposited via RF magnetron sputtering for different deposition conditions. Mode of deposition Lattice parameters, Å Crystallite size, nm Microstrain, ×10-3 GS

a = 9.4396

30±6

3±1

15±4

5±1

c = 6.8532 bias -50 V

a = 9.4390 c = 6.8924

30

Table 3. The values of the penetration depth, hmax, nanohardness, H, and the Young’s modulus, E, the H/E ratio, and parameter H3/E2 for the СaP coatings and the uncoated AZ31 magnesium alloy. Sample AZ31 magnesium alloy

E, GPa 44.24 ± 3.40

H/E 0.027

H3/E2, GPa 0.9×10-3

0.90 ± 0.80 3.08 ± 2.00 4.90 ± 2.50

37.01 ± 22.30 79.03 ± 11.00 78.04 ± 16.00

0.024 0.039 0.063

0.5×10-3 4.68×10-3 19.2×10-3

6.89 ± 1.23 9.04 ±1.59

82.22 ± 7.53 100.40 ± 12.13

0.084 0.090

48.4×10-3 73.29×10-3

hc, nm H, GPa 100 1.19 ± 0.32

50 AZ31 magnesium 100 alloy+coating deposited 50 at GS AZ31 magnesium 100 alloy+coating deposited 50 at bias -50 V

31

• The hydroxyapatite films increased the hardness on the microscale and nanoscale. • The substrate bias resulted in a decrease in the crystallite size of the coating. • The coating significantly enhanced the wear resistance of the AZ31 alloy. • Maximum hardness was found for the HA coating prepared using a negative bias.

32