Materials Science and Engineering C 35 (2014) 314–321
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Effect of surface mechanical attrition treatment on biodegradable Mg–1Ca alloy N. Li a,f, Y.D. Li b, Y.X. Li c, Y.H. Wu d, Y.F. Zheng a,c,d,f,⁎, Y. Han e,⁎⁎ a
State Key Laboratory for Turbulence and Complex System, College of Engineering, Peking University, Beijing 100871, China DongGuan EONTEC Co., Ltd, Yin Quan Industrial District, Qing Xi, DongGuan 523662, China Center for Biomedical Materials and Engineering, Harbin Engineering University, Harbin 150001, China d Center for Biomedical Materials and Tissue Engineering, Academy for Advanced Interdisciplinary Studies, Peking University, Beijing 100871, China e State-key Laboratory for Mechanical Behavior of Materials, Xi'an Jiaotong University, Xi'an 710049, China f Department of Materials Science and Engineering, College of Engineering, Peking University, Beijing 100871, China b c
a r t i c l e
i n f o
Article history: Received 2 September 2013 Received in revised form 24 October 2013 Accepted 2 November 2013 Available online 15 November 2013 Keywords: Biomedical Mg alloys Mg–Ca SMAT Grain refinement Corrosion resistance
a b s t r a c t Surface mechanical attrition treatment (SMAT) is considered to be an effective approach to obtain a nanostructured layer in the treated surface of metals. In this study, we evaluated the effect of SMAT on the microstructure, mechanical properties and corrosion properties of biodegradable Mg–1Ca alloy, with pure Mg as control. Grain refinement layers with grain size at the nanometer scale in the topmost surface were successfully prepared on Mg–1Ca alloy using SMAT technique, similar to pure Mg. The SMAT not only refined the surface layer of Mg–1Ca alloy, but also promoted the re-dissolution of the Mg2Ca phase into the matrix. As a result, the microhardness of the SMATed samples in the near-surface region was considerably enhanced, and the surface roughness and wettability of the SMATed samples were increased. However, the SMAT led to high density of crystalline defects such as grain boundaries (subgrain boundaries) and dislocations, which severely weakened the corrosion resistance of Mg–1Ca alloy, same as pure Mg. © 2013 Elsevier B.V. All rights reserved.
1. Introduction In the past decade, severe plastic deformation (SPD) approaches, such as equal channel angular pressing (ECAP) [1–7], high-pressure torsion (HPT) [8] and groove pressing (GP) [9], have been applied to produce ultra-fine grained (UFG) magnesium and its alloys. The resulting UFG Mg materials usually exhibit enhanced mechanical strength due to the grain refinement, however different corrosion resistances were found with different SPD methods: HPT treated Mg–Zn–Ca alloy [8] and groove pressed AZ31 alloy [9] are reported to be more resistant to corrosion than the original coarse-grained alloys. AlvarezLopez et al. [3] and Birbilis et al. [7] reported the improvement of ECAP in the corrosion resistance of pure Mg and AZ31 alloy, respectively. However, Song et al. [5,6] found that ECAP accelerates the corrosion of pure Mg and AZ91 alloy. Surface mechanical attrition treatment (SMAT) is another SPD method that can obtain a hard layer of nanocrystalline structure in the treated surface. The basic principle of SMAT is the generation of plastic
⁎ Correspondence to: Y.F. Zheng, Department of Materials Science and Engineering, College of Engineering, Peking University, Beijing 100871, China. Tel./fax: + 86 10 62767411. ⁎⁎ Co-corresponding author. E-mail addresses: [email protected] (Y.F. Zheng), [email protected] (Y. Han). 0928-4931/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.msec.2013.11.010
deformation in the surface layer of a bulk material by means of the repeated multidirectional impact of flying balls on the sample surface [10]. It has been successfully applied to a number of alloy systems, such as Fe [11], stainless steel [12], titanium [13], copper [14] and magnesium [10,15–18]. Sun et al. [10] indicated that the nanocrystallization process of AZ91D alloy during SMAT can be divided into three steps: twinning, formation of subgrains and dynamic recrystallization. The nanocrystalline layer showed higher hardness and better wear resistance than the coarse-grained matrix [10,15]. Hoog et al. [16] found that the corrosion resistance of SMATed pure Mg in 0.1 M NaCl showed no improvement, while Laleh and Kargar [18] reported that SMAT with 2 mm balls can improve the corrosion resistance of AZ91D alloy in 3.5 wt.% NaCl. Mg–Ca alloys are promising candidate materials for biodegradable orthopedic implants [19–23]. Calcium is a major component in human bone and is essential in chemical signaling with cells [24]. Magnesium is necessary for the calcium incorporation into the bone [25], which might be expected to be beneficial to the bone healing with the co-releasing of Mg and Ca ions. In a previous study [20], we developed binary Mg–1Ca alloy which did not induce toxicity to L-929 cells and gradually degraded in vivo within 90 days with new bone formed. Since the degradation rate of this Mg–1Ca alloy is too fast to match with the tissue healing rate, we have developed various kinds of surface treatments and coating techniques on Mg–1Ca alloy, such as alkalineheat treatment [26], microarc oxidation [27], chitosan coating [28] and
N. Li et al. / Materials Science and Engineering C 35 (2014) 314–321
315
Fig. 1. Surface morphologies of (a) SMAT-50 min pure Mg, and (b) SMAT-50 min Mg–1Ca alloy.
MgF2 coating [29] to reduce the corrosion rate. In the present study, the effect of SMAT on the microstructure, mechanical properties and corrosion properties of biodegradable Mg–1Ca alloy will be studied, with pure Mg as control group.
peening of zirconia balls (2 mm in diameter) over a range of periods from 40 to 90 min. After that, square samples with dimensions of 10 × 10 × 2 mm3 were cut from the as-SMATed plate. The samples were ultrasonic cleaned in acetone, absolute ethanol and distilled water for 10 min each and dried in air.
2. Materials and methods 2.1. Sample preparation
2.2. Microstructure characterization
The materials used in this investigation were as-cast Mg–1Ca (wt.%) alloy [20] and pure Mg with purity of 99.99 wt.%. Plate samples (100 mm in diameter and 5 mm in thickness) were cut from the asreceived ingot and then ground with SiC paper up to 2000 grit. The SMAT was carried out using an SNC-II surface nanocrystallization testing machine. The sample surface was subjected to multidirectional
An Olympus BX51M optical microscope was used to examine the microstructural development along the cross-section perpendicular to the treated surface of the specimen. Before the observation, the samples were mechanically polished up to 2000 grit and etched with a mixture of 1 g of oxalic acid, 1 ml of acetic acid, 1 ml of nitric acid and 150 ml of water.
Fig. 2. Optical microscopic images of the cross-sections of pure Mg being SMATed for (a) 40 min, (b) 50 min, (c) 60 min, and (d) 90 min. The inset in panel (b) is a high magnification image of the area marked with a square in panel (b).
316
N. Li et al. / Materials Science and Engineering C 35 (2014) 314–321
The average grain size of the top surface layer of the SMATed samples was determined by X-ray diffraction (XRD), which was carried out in a Rigaku DMAX/2400 diffractometer using Cu Kα radiation. The average grain size of the treated samples was calculated according to the Williamson–Hall method (Eq. (1)) [30]. B cos θ ¼ kλ=l þ η sinθ
ð1Þ
where B is the full width at half maximum (FWHM) of the peak, θ is the diffraction angle, k is a constant (0.9), λ is the wavelength of the X-ray (1.54 nm), l is the grain size and η is the microstrain. 2.3. Microhardness test The Vickers microhardness (HV) tests on the cross-section of the experimental samples were carried out on a Shimadzu HMV-2 T microhardness tester with an applied load of 50 g and a loading time of 10 s. The microhardness variation was measured along the depth (up to 2000 μm from the top surface).
2.4. Surface roughness and wettability The average surface roughness (Ra) and of the experimental SMATed samples were measured using a Time TR200 Surface Roughness Tester with a resolution of 0.001 μm. The water contact angle was measured using a Kino SL200B contact angle system. An average of three measurements was taken for each sample. 2.5. Electrochemical test A three-electrode cell was used for electrochemical measurements with a platinum counter-electrode and a saturated calomel electrode (SCE) as the reference electrode. The tests were performed in Hank's solution (NaCl 8.00 g·l−1, KCl 0.40 g·l−1, CaCl2 0.14 g·l−1, NaHCO3 0.35 g·l−1, MgSO4⋅ 7H2O 0.20 g·l−1, Na2HPO4⋅ 12H2O 0.12 g·l−1, KH2PO4 0.06 g·l−1, pH = 7.4) at 37 ± 0.5 °C. A wire lead was attached to one section of each sample and closely sealed with epoxy resin, leaving an exposed area of 1 cm2. After 3600 s open circuit potential (OCP) measurements, potentiodynamic polarization scans were performed at
Fig. 3. Optical microscopic images of the cross-sections of Mg–1Ca alloy being SMATed for (a) 40 min, (b) 50 min, (c) 60 min, (d) 90 min, (e) high magnification image of the crosssections of un-SMATed Mg–1Ca alloy and (f) high magnification image of the cross-sections of SMAT-90 min Mg–1Ca alloy.
N. Li et al. / Materials Science and Engineering C 35 (2014) 314–321
317
Fig. 5. Variation curves of microhardness of (a) pure Mg and (b) Mg–1Ca with distance from the topmost surface to the substrate. Fig. 4. XRD patterns of the (a) un-SMATed and SMATed pure Mg and (b) un-SMATed and SMATed Mg–1Ca alloy.
a scanning rate of 1 mV/s and the initial potential was 300 mV below the OCP.
3. Results and discussion 3.1. Microstructures Fig. 1 shows surface morphologies of SMAT-50 min pure Mg and Mg–1Ca alloy at macroscopic scale. Due to the high energy of impacts of balls on the experimental sample surfaces, the SMATed samples presented typical crater and dimple surface morphology feature. The cross-sectional microscopic images of pure Mg SMATed for different times were shown in Fig. 2, featured by a deformed top layer consisted of fine grains. In the deformed layer, the grain size increased with the increase of the distance from top surface. The grains in the
lower undeformed matrix were as large as several millimeters. For instance, the average grain size of SMAT-50 min pure Mg 50 μm below the top surface was 6.05 ± 0.51 μm, while that 200 μm below the top surface was 10.70 ± 1.08 μm. The gradient microstructure resulted from a gradual decrease in the applied strain with increasing depth of the deformed layer during SMAT. The thickness of the deformed layer increased with the extension of SMAT treatment time from 350–500 μm to 650–750 μm. For the SMATed Mg–1Ca alloy samples, although grain refinement cannot be clearly observed from the optical images (Fig. 3(a)–(d)), there is a clear evidence of plastic deformation near the SMATed surface. Fig. 3(e) and (f) shows high magnification images of the cross-sections
Table 1 Average grain sizes and microstrain of the SMATed pure Mg and Mg–1Ca alloy calculated from XRD patterns. Treating time (min)
40 50 60 90
Grain size (nm)
Microstrain (%)
Pure Mg
Mg–1Ca
Pure Mg
88.8 57.4 58.2 95.0
71.1 65.0 72.9 68.4
0.015 0.044 0.055 0.022
± ± ± ±
3.2 1.4 1.2 4.1
± ± ± ±
1.9 1.7 2.2 2.1
± ± ± ±
Mg1–Ca 0.004 0.004 0.003 0.005
0.046 0.025 0.044 0.042
± ± ± ±
0.004 0.004 0.004 0.004
Fig. 6. Surface roughness of the un-SMATed and SMATed pure Mg and Mg–1Ca alloy.
318
N. Li et al. / Materials Science and Engineering C 35 (2014) 314–321
the finest grain, whereas the grain size of SMATed Mg–1Ca alloy didn't change much with the treating time. 3.2. Microhardness
Fig. 7. Surface wettability of the un-SMATed and SMATed pure Mg and Mg–1Ca alloy.
of un-SMATed and SMAT-90 min Mg–1Ca alloy. Part of the Mg2Ca phase along the grain boundaries in the deformed layer dissolved into the matrix after SMAT, which was marked by arrows in Fig. 3(f). Fig. 4 shows the XRD patterns of the SMATed and un-SMATed experimental samples. The broadening of the diffraction peaks and a slight shift in the centroid position can be observed both in the XRD patterns of pure Mg and Mg–1Ca alloy after SMAT. This might be attributed to grain refinement, microstrain development and micro-distortions of the crystalline lattice. Average grain sizes and microstrain of the SMATed pure Mg and Mg–1Ca alloy calculated from XRD patterns were shown in Table 1. The average grain sizes in the top layer of all the SMATed samples were less than 100 nm. The SMAT-50 min and SMAT-60 min pure Mg showed
Fig. 5 shows the variation of microhardness from the SMATed surface to the substrate beneath, which indicates that the hardness of the samples was enhanced within 1400 μm from the surface for the pure Mg and 1000 μm for the Mg–1Ca alloy after SMAT. The microhardness of the SMATed sample gradually decreases along the depth and approaches the value of the matrix. The microhardness didn't change with the SMAT treating time. It can be seen that the improvement in microhardness of SMATed Mg–1Ca alloy is more obvious than that of SMATed pure Mg. The top most layer of the SMATed Mg–1Ca alloy is at least twice harder than that of its undeformed matrix. The increase of the surface hardness of the SMATed pure Mg can be attributed to the both effects of the grain refinement and work-hardening. As for the SMATed Mg–1Ca alloy, the re-dissolution of the Mg2Ca phase is another reason for microhardness increase. The result indicates that the effect of solution strengthening out-performs grain refinement and work-hardening on the surface. 3.3. Surface roughness and wettability Surface roughness of pure Mg and Mg–1Ca alloy was significantly increased after SMAT, as shown in Fig. 6. Correspondingly, the SMATed samples with rougher surfaces exhibited better wettability (Fig. 7). The enhanced hydrophilicity favors the adsorption of adhesive proteins. Many previous works showed that implant materials with higher roughness promoted in vitro cell response [31–35] and in vivo
Fig. 8. Hydrogen evolution behavior of the SMATed (a) pure Mg and (b) Mg–1Ca alloy, and the pH value variation of the Hank's solution incubating the SMATed (c) pure Mg and (b) Mg–1Ca alloy as a function of the immersion time, with the un-SMATed pure Mg and Mg–1Ca alloy as controls.
N. Li et al. / Materials Science and Engineering C 35 (2014) 314–321
319
Fig. 9. Open circuit potential of (a) un-SMATed and SMATed pure Mg, (b) un-SMATed and SMATed Mg–1Ca alloy, and potentiodynamic polarization curves of (c) un-SMATed and SMATed pure Mg and (d) un-SMATed and SMATed Mg–1Ca alloy.
osteointegration [34,35]. However, the increasing roughness also resulted in an accelerated corrosion rate [36,37]. 3.4. Corrosion behavior Fig. 8 shows the hydrogen evolution rates of the experimental samples and the pH value variation of the Hank's solution during the immersion testing. All the SMATed samples generated more hydrogen gas and the incubated Hank's solution exhibited much higher pH value in comparison with the un-SMATed counterparts. The corrosion developed rapidly as soon as the SMATed samples touched the solution and the pH value was over 9 (some even over 10) after 1 h, thereafter the hydrogen evolution rates slowed down and the pH values gradually stabilized at 11–13. Results of the electrochemical testing further confirmed the poor corrosion resistance of the SMATed samples (Fig. 9 and Table 2). The
corrosion rates of the SMATed sample calculated from the potentiodynamic polarization curves were two to three orders of magnitude higher than those of the un-SMATed samples. By conventional definition, the more negative the Ecorr, the less noble the material is. However, it's interesting that the SMATed samples exhibited nobler OCP and Ecorr than the un-SMATed samples. From the potentiodynamic polarization curves (Fig. 9(c) and (d)), it can be seen that SMAT appears to lower the anodic kinetics, but detrimentally increases the cathodic kinetics as compared to the un-SMATed counterparts. Similar phenomenon is also reported by Hoog et al. [16]. SMAT process is believed to induce an extremely high density of strain-induced crystalline defects in the near-surface region, including grain boundaries (subgrain boundaries) and dislocations. The defects are susceptible to corrosion and passive film forms rapidly on the surface, which could explain the elevated initial OCP and Ecorr values of SMATed samples in comparison to the un-SMATed one. For easy-
Table 2 The electrochemical data obtained from potentiodynamic polarization measurements. Icorr (μA/cm2)
Ecorr (V)
Pure Mg
Mg–1Ca alloy
Un-SMATed SMAT-40 min SMAT-50 min SMAT-60 min SMAT-90 min Un-SMATed SMAT-40 min SMAT-50 min SMAT-60 min SMAT-90 min
Vcorr (mm/yr)
Original surface
After grinding
Original surface
After grinding
Original surface
After grinding
−1.78 −1.49 −1.43 −1.37 −1.47 −1.62 −1.48 −1.51 −1.55 −1.52
−1.78 −1.59 −1.55 −1.57 −1.55 −1.62 −1.60 −1.56 −1.52 −1.53
1.56 2.65 1.29 3.03 1.91 8.45 3.36 2.35 6.14 1.59
1.56 2.28 2.39 3.70 4.65 8.45 6.17 31.7 45.3 50.5
3.52 × 10−2 59.8 29.1 6.83 43.1 1.91 × 10−1 78.8 53.0 13.8 35.9
3.52 5.14 5.39 8.34 1.05 1.91 1.39 7.15 1.02 1.14
× × × ×
103 103 102 103
× × × ×
103 103 102 103
× 10−2
× × × ×
10−1 10−1 10−1 10−1
320
N. Li et al. / Materials Science and Engineering C 35 (2014) 314–321
Fig. 10. Open circuit potential of (a) un-SMATed and SMATed pure Mg, (b) un-SMATed and SMATed Mg–1Ca alloy, and potentiodynamic polarization curves of (c) un-SMATed and SMATed pure Mg and (d) un-SMATed and SMATed Mg–1Ca alloy after the SMATed surfaces were ground off 100 μm.
passive metallic materials, such as titanium alloys, the passive films are compact, so the SMATed titanium alloy surface layer composed by a high fraction of passive layers would reduce the corrosion rates [38]. However, the magnesium hydroxide film is only partially protective, and it allows electrolyte to contact the underlying magnesium and its alloy matrix. Since the high density of crystalline defects drastically activates corrosion reaction and the resulting Mg(OH)2 films present no improved protection, the SMATed pure Mg and Mg–1Ca alloy are more erodible than the un-SMATed counterparts. To eliminate the influence of the surface roughness and reduce the density of defects, 100 μm of the top layer of the SMATed samples was ground off with a 2000 grade carbide abrasive paper. The resulting new surfaces were still in the deformed region with grain size in micrometer scale. Fig. 10 showed that after grinding, the OCP and Ecorr of the SMATed samples shifted to the negative direction and the corrosion rate got close to the un-SMATed samples. Due to the poor corrosion resistance, the SMAT method might not be suitable for slowing down the corrosion rate of biodegradable Mg alloy implants. 4. Conclusions Strain-induced grain refinement layers with grain size at the nanometer scale in the topmost surface were prepared on pure Mg and Mg–1Ca alloy using SMAT. The microhardness of the SMATed samples was improved within 1400 μm depth beneath the surface for the pure Mg and 1000 μm for the Mg–1Ca alloy. The surface roughness and wettability of SMATed samples were greatly enhanced in comparison to the un-SMATed samples. Due to high density of strain-induced crystalline defects in the near-surface region, the corrosion resistance of the SMATed samples was deteriorated. The variation of treating time from 40 min to 90 min didn't influence the properties of the SMATed pure
Mg and Mg–1Ca alloy except the thickness of the deformed layer. The SMAT is proved to be ineffective for biodegradable magnesium and its alloys to enhance its corrosion resistance. Acknowledgments This work was supported by the National Basic Research Program of China (973 Program) (Grant Nos. 2012CB619102 and 2012CB619103), National Natural Science Foundation of China (No. 31170909), National Science Fund for Distinguished Young Scholars (Grant No. 51225101), Research Fund for the Doctoral Program of Higher Education under Grant No. 20100001110011, Project for Supervisor of Excellent Doctoral Dissertation of Beijing (20121000101), Natural Science Foundation of Heilongjiang Province (ZD201012) and Guangdong Province innovation R&D team project (No. 201001C0104669453). References [1] H. Wang, Y. Estrin, H. Fu, G. Song, Z. Zuberova, Adv. Eng. Mater. 9 (2007) 967–972. [2] H. Wang, Y. Estrin, Z. Zuberova, Mater. Lett. 62 (2008) 2476–2479. [3] M. Alvarez-Lopez, M.D. Pereda, J.A. del Valle, M. Fernandez-Lorenzo, M.C. Garcia-Alonso, O.A. Ruano, M.L. Escudero, Acta Biomater. 6 (2010) 1763–1771. [4] X.N. Gu, N. Li, Y.F. Zheng, F. Kang, J.T. Wang, L.Q. Ruan, Mater. Sci. Eng. B 176 (2011) 1802–1806. [5] D. Song, A. Ma, J. Jiang, P. Lin, D. Yang, J. Fan, Corros. Sci. 52 (2010) 481–490. [6] D. Song, A.B. Ma, J.H. Jiang, P.H. Lin, D.H. Yang, J.F. Fan, Corros. Sci. 53 (2011) 362–373. [7] N. Birbilis, K.D. Ralston, S. Virtanen, H.L. Fraser, C.H.J. Davies, Corros. Eng. Sci. Technol. 45 (2010) 224–230. [8] J.H. Gao, S.K. Guan, Z.W. Ren, Y.F. Sun, S.J. Zhu, B. Wang, Mater. Lett. 65 (2011) 691–693. [9] B.R. Sunil, A.A. Kumar, T.S. Sampath Kumar, U. Chakkingal, Mater. Sci. Eng. C 33 (2013) 1607–1615. [10] H.Q. Sun, Y.N. Shi, M.X. Zhang, K. Lu, Acta Mater. 55 (2007) 975–982. [11] A. Amanov, O.V. Penkov, Y.S. Pyun, D.E. Kim, Tribol. Int. 54 (2012) 106–113.
N. Li et al. / Materials Science and Engineering C 35 (2014) 314–321 [12] H.W. Zhang, Z.K. Hei, G. Liu, J. Lu, K. Lu, Acta Mater. 51 (2003) 1871–1881. [13] K.Y. Zhu, A. Vassel, F. Brisset, K. Lu, J. Lu, Acta Mater. 52 (2004) 4101–4110. [14] Y.M. Wang, K. Wang, D. Pan, K. Lu, K.J. Hemker, E. Ma, Scr. Mater. 48 (2003) 1581–1586. [15] H.Q. Sun, Y.N. Shi, M.X. Zhang, Surf. Coat. Technol. 202 (2008) 2859–2864. [16] C. op't Hoog, N. Birbilis, Y. Estrin, Adv. Eng. Mater. 10 (2008) 579–582. [17] Y.-h. Wei, B.-s. Liu, L.-f. Hou, B.-s. Xu, G. Liu, J. Alloys Compd. 452 (2008) 336–342. [18] M. Laleh, F. Kargar, J. Alloys Compd. 509 (2011) 9150–9156. [19] C.D. Yfantis, D.K. Yfantis, J. Anastassopoulou, T. Theophanides, M. Staige, The 5th WSEAS International Conference on Environment, Ecosystess and Development, Venice, Italy, 2006, pp. 186–189. [20] Z.J. Li, X.N. Gu, S.Q. Lou, Y.F. Zheng, Biomaterials 29 (2008) 1329–1344. [21] Y. Wan, G. Xiong, H. Luo, F. He, Y. Huang, X. Zhou, Mater. Des. 29 (2008) 2034–2037. [22] A. Drynda, T. Hassel, R. Hoehn, A. Perz, F.W. Bach, M. Peuster, J. Biomed. Mater. Res. A 93A (2010) 763–775. [23] N. Erdmann, N. Angrisani, J. Reifenrath, A. Lucas, F. Thorey, D. Bormann, A. Meyer-Lindenberg, Acta Biomater. 7 (2011) 1421–1428. [24] J.Z. Ilich, J.E. Kerstetter, J. Am. Coll. Nutr. 19 (2000) 715–737. [25] C.M. Serre, M. Papillard, P. Chavassieux, J.C. Voegel, G. Boivin, J. Biomed. Mater. Res. A 42 (1998) 626–633.
321
[26] X.N. Gu, W. Zheng, Y. Cheng, Y.F. Zheng, Acta Biomater. 5 (2009) 2790–2799. [27] X.N. Gu, N. Li, W.R. Zhou, Y.F. Zheng, X. Zhao, Q.Z. Cai, L. Ruan, Acta Biomater. 7 (2011) 1880–1889. [28] X.N. Gu, Y.F. Zheng, Q.X. Lan, Y. Cheng, Z.X. Zhang, T.F. Xi, D.Y. Zhang, Biomed. Mater. 4 (2009). [29] N. Li, Y.D. Li, Y.B. Wang, M. Li, Y. Cheng, Y.H. Wu, Y.F. Zheng, Surf. Interface Anal. 45 (2013) 1217–1222. [30] G.K. Williamson, W.H. Hall, Acta Metall. 1 (1953) 22–31. [31] N. Nishimura, T. Kawai, J. Mater. Sci. Mater. Med. 9 (1998) 99–102. [32] J.C. Keller, C.M. Stanford, J.P. Wightman, R.A. Draughn, R. Zaharias, J. Biomed. Mater. Res. 28 (1994) 939–946. [33] K.T. Bowers, J.C. Keller, B.A. Randolph, D.G. Wick, C.M. Michaels, Int. J. Oral Maxillofac. Surg. 7 (1992) 302–310. [34] D. Buser, R.K. Schenk, S. Steinemann, J.P. Fiorellini, C.H. Fox, H. Stich, J. Biomed. Mater. Res. 25 (1991) 889–902. [35] K.A. Thomas, S.D. Cook, J. Biomed. Mater. Res. 19 (1985) 875–901. [36] R. Walter, M.B. Kannan, Mater. Des. 32 (2011) 2350–2354. [37] T.L. Nguyen, A. Blanquet, M.P. Staiger, G.J. Dias, T.B. Woodfield, J. Biomed. Mater. Res. B Appl. Biomater. 100 (2012) 1310–1318. [38] R. Huang, Y. Han, Mater. Sci. Eng. C 33 (2013) 2353–2359.
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Effect of surface mechanical attrition treatment on biodegradable Mg–1Ca alloy N. Li a,f, Y.D. Li b, Y.X. Li c, Y.H. Wu d, Y.F. Zheng a,c,d,f,⁎, Y. Han e,⁎⁎ a
State Key Laboratory for Turbulence and Complex System, College of Engineering, Peking University, Beijing 100871, China DongGuan EONTEC Co., Ltd, Yin Quan Industrial District, Qing Xi, DongGuan 523662, China Center for Biomedical Materials and Engineering, Harbin Engineering University, Harbin 150001, China d Center for Biomedical Materials and Tissue Engineering, Academy for Advanced Interdisciplinary Studies, Peking University, Beijing 100871, China e State-key Laboratory for Mechanical Behavior of Materials, Xi'an Jiaotong University, Xi'an 710049, China f Department of Materials Science and Engineering, College of Engineering, Peking University, Beijing 100871, China b c
a r t i c l e
i n f o
Article history: Received 2 September 2013 Received in revised form 24 October 2013 Accepted 2 November 2013 Available online 15 November 2013 Keywords: Biomedical Mg alloys Mg–Ca SMAT Grain refinement Corrosion resistance
a b s t r a c t Surface mechanical attrition treatment (SMAT) is considered to be an effective approach to obtain a nanostructured layer in the treated surface of metals. In this study, we evaluated the effect of SMAT on the microstructure, mechanical properties and corrosion properties of biodegradable Mg–1Ca alloy, with pure Mg as control. Grain refinement layers with grain size at the nanometer scale in the topmost surface were successfully prepared on Mg–1Ca alloy using SMAT technique, similar to pure Mg. The SMAT not only refined the surface layer of Mg–1Ca alloy, but also promoted the re-dissolution of the Mg2Ca phase into the matrix. As a result, the microhardness of the SMATed samples in the near-surface region was considerably enhanced, and the surface roughness and wettability of the SMATed samples were increased. However, the SMAT led to high density of crystalline defects such as grain boundaries (subgrain boundaries) and dislocations, which severely weakened the corrosion resistance of Mg–1Ca alloy, same as pure Mg. © 2013 Elsevier B.V. All rights reserved.
1. Introduction In the past decade, severe plastic deformation (SPD) approaches, such as equal channel angular pressing (ECAP) [1–7], high-pressure torsion (HPT) [8] and groove pressing (GP) [9], have been applied to produce ultra-fine grained (UFG) magnesium and its alloys. The resulting UFG Mg materials usually exhibit enhanced mechanical strength due to the grain refinement, however different corrosion resistances were found with different SPD methods: HPT treated Mg–Zn–Ca alloy [8] and groove pressed AZ31 alloy [9] are reported to be more resistant to corrosion than the original coarse-grained alloys. AlvarezLopez et al. [3] and Birbilis et al. [7] reported the improvement of ECAP in the corrosion resistance of pure Mg and AZ31 alloy, respectively. However, Song et al. [5,6] found that ECAP accelerates the corrosion of pure Mg and AZ91 alloy. Surface mechanical attrition treatment (SMAT) is another SPD method that can obtain a hard layer of nanocrystalline structure in the treated surface. The basic principle of SMAT is the generation of plastic
⁎ Correspondence to: Y.F. Zheng, Department of Materials Science and Engineering, College of Engineering, Peking University, Beijing 100871, China. Tel./fax: + 86 10 62767411. ⁎⁎ Co-corresponding author. E-mail addresses: [email protected] (Y.F. Zheng), [email protected] (Y. Han). 0928-4931/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.msec.2013.11.010
deformation in the surface layer of a bulk material by means of the repeated multidirectional impact of flying balls on the sample surface [10]. It has been successfully applied to a number of alloy systems, such as Fe [11], stainless steel [12], titanium [13], copper [14] and magnesium [10,15–18]. Sun et al. [10] indicated that the nanocrystallization process of AZ91D alloy during SMAT can be divided into three steps: twinning, formation of subgrains and dynamic recrystallization. The nanocrystalline layer showed higher hardness and better wear resistance than the coarse-grained matrix [10,15]. Hoog et al. [16] found that the corrosion resistance of SMATed pure Mg in 0.1 M NaCl showed no improvement, while Laleh and Kargar [18] reported that SMAT with 2 mm balls can improve the corrosion resistance of AZ91D alloy in 3.5 wt.% NaCl. Mg–Ca alloys are promising candidate materials for biodegradable orthopedic implants [19–23]. Calcium is a major component in human bone and is essential in chemical signaling with cells [24]. Magnesium is necessary for the calcium incorporation into the bone [25], which might be expected to be beneficial to the bone healing with the co-releasing of Mg and Ca ions. In a previous study [20], we developed binary Mg–1Ca alloy which did not induce toxicity to L-929 cells and gradually degraded in vivo within 90 days with new bone formed. Since the degradation rate of this Mg–1Ca alloy is too fast to match with the tissue healing rate, we have developed various kinds of surface treatments and coating techniques on Mg–1Ca alloy, such as alkalineheat treatment [26], microarc oxidation [27], chitosan coating [28] and
N. Li et al. / Materials Science and Engineering C 35 (2014) 314–321
315
Fig. 1. Surface morphologies of (a) SMAT-50 min pure Mg, and (b) SMAT-50 min Mg–1Ca alloy.
MgF2 coating [29] to reduce the corrosion rate. In the present study, the effect of SMAT on the microstructure, mechanical properties and corrosion properties of biodegradable Mg–1Ca alloy will be studied, with pure Mg as control group.
peening of zirconia balls (2 mm in diameter) over a range of periods from 40 to 90 min. After that, square samples with dimensions of 10 × 10 × 2 mm3 were cut from the as-SMATed plate. The samples were ultrasonic cleaned in acetone, absolute ethanol and distilled water for 10 min each and dried in air.
2. Materials and methods 2.1. Sample preparation
2.2. Microstructure characterization
The materials used in this investigation were as-cast Mg–1Ca (wt.%) alloy [20] and pure Mg with purity of 99.99 wt.%. Plate samples (100 mm in diameter and 5 mm in thickness) were cut from the asreceived ingot and then ground with SiC paper up to 2000 grit. The SMAT was carried out using an SNC-II surface nanocrystallization testing machine. The sample surface was subjected to multidirectional
An Olympus BX51M optical microscope was used to examine the microstructural development along the cross-section perpendicular to the treated surface of the specimen. Before the observation, the samples were mechanically polished up to 2000 grit and etched with a mixture of 1 g of oxalic acid, 1 ml of acetic acid, 1 ml of nitric acid and 150 ml of water.
Fig. 2. Optical microscopic images of the cross-sections of pure Mg being SMATed for (a) 40 min, (b) 50 min, (c) 60 min, and (d) 90 min. The inset in panel (b) is a high magnification image of the area marked with a square in panel (b).
316
N. Li et al. / Materials Science and Engineering C 35 (2014) 314–321
The average grain size of the top surface layer of the SMATed samples was determined by X-ray diffraction (XRD), which was carried out in a Rigaku DMAX/2400 diffractometer using Cu Kα radiation. The average grain size of the treated samples was calculated according to the Williamson–Hall method (Eq. (1)) [30]. B cos θ ¼ kλ=l þ η sinθ
ð1Þ
where B is the full width at half maximum (FWHM) of the peak, θ is the diffraction angle, k is a constant (0.9), λ is the wavelength of the X-ray (1.54 nm), l is the grain size and η is the microstrain. 2.3. Microhardness test The Vickers microhardness (HV) tests on the cross-section of the experimental samples were carried out on a Shimadzu HMV-2 T microhardness tester with an applied load of 50 g and a loading time of 10 s. The microhardness variation was measured along the depth (up to 2000 μm from the top surface).
2.4. Surface roughness and wettability The average surface roughness (Ra) and of the experimental SMATed samples were measured using a Time TR200 Surface Roughness Tester with a resolution of 0.001 μm. The water contact angle was measured using a Kino SL200B contact angle system. An average of three measurements was taken for each sample. 2.5. Electrochemical test A three-electrode cell was used for electrochemical measurements with a platinum counter-electrode and a saturated calomel electrode (SCE) as the reference electrode. The tests were performed in Hank's solution (NaCl 8.00 g·l−1, KCl 0.40 g·l−1, CaCl2 0.14 g·l−1, NaHCO3 0.35 g·l−1, MgSO4⋅ 7H2O 0.20 g·l−1, Na2HPO4⋅ 12H2O 0.12 g·l−1, KH2PO4 0.06 g·l−1, pH = 7.4) at 37 ± 0.5 °C. A wire lead was attached to one section of each sample and closely sealed with epoxy resin, leaving an exposed area of 1 cm2. After 3600 s open circuit potential (OCP) measurements, potentiodynamic polarization scans were performed at
Fig. 3. Optical microscopic images of the cross-sections of Mg–1Ca alloy being SMATed for (a) 40 min, (b) 50 min, (c) 60 min, (d) 90 min, (e) high magnification image of the crosssections of un-SMATed Mg–1Ca alloy and (f) high magnification image of the cross-sections of SMAT-90 min Mg–1Ca alloy.
N. Li et al. / Materials Science and Engineering C 35 (2014) 314–321
317
Fig. 5. Variation curves of microhardness of (a) pure Mg and (b) Mg–1Ca with distance from the topmost surface to the substrate. Fig. 4. XRD patterns of the (a) un-SMATed and SMATed pure Mg and (b) un-SMATed and SMATed Mg–1Ca alloy.
a scanning rate of 1 mV/s and the initial potential was 300 mV below the OCP.
3. Results and discussion 3.1. Microstructures Fig. 1 shows surface morphologies of SMAT-50 min pure Mg and Mg–1Ca alloy at macroscopic scale. Due to the high energy of impacts of balls on the experimental sample surfaces, the SMATed samples presented typical crater and dimple surface morphology feature. The cross-sectional microscopic images of pure Mg SMATed for different times were shown in Fig. 2, featured by a deformed top layer consisted of fine grains. In the deformed layer, the grain size increased with the increase of the distance from top surface. The grains in the
lower undeformed matrix were as large as several millimeters. For instance, the average grain size of SMAT-50 min pure Mg 50 μm below the top surface was 6.05 ± 0.51 μm, while that 200 μm below the top surface was 10.70 ± 1.08 μm. The gradient microstructure resulted from a gradual decrease in the applied strain with increasing depth of the deformed layer during SMAT. The thickness of the deformed layer increased with the extension of SMAT treatment time from 350–500 μm to 650–750 μm. For the SMATed Mg–1Ca alloy samples, although grain refinement cannot be clearly observed from the optical images (Fig. 3(a)–(d)), there is a clear evidence of plastic deformation near the SMATed surface. Fig. 3(e) and (f) shows high magnification images of the cross-sections
Table 1 Average grain sizes and microstrain of the SMATed pure Mg and Mg–1Ca alloy calculated from XRD patterns. Treating time (min)
40 50 60 90
Grain size (nm)
Microstrain (%)
Pure Mg
Mg–1Ca
Pure Mg
88.8 57.4 58.2 95.0
71.1 65.0 72.9 68.4
0.015 0.044 0.055 0.022
± ± ± ±
3.2 1.4 1.2 4.1
± ± ± ±
1.9 1.7 2.2 2.1
± ± ± ±
Mg1–Ca 0.004 0.004 0.003 0.005
0.046 0.025 0.044 0.042
± ± ± ±
0.004 0.004 0.004 0.004
Fig. 6. Surface roughness of the un-SMATed and SMATed pure Mg and Mg–1Ca alloy.
318
N. Li et al. / Materials Science and Engineering C 35 (2014) 314–321
the finest grain, whereas the grain size of SMATed Mg–1Ca alloy didn't change much with the treating time. 3.2. Microhardness
Fig. 7. Surface wettability of the un-SMATed and SMATed pure Mg and Mg–1Ca alloy.
of un-SMATed and SMAT-90 min Mg–1Ca alloy. Part of the Mg2Ca phase along the grain boundaries in the deformed layer dissolved into the matrix after SMAT, which was marked by arrows in Fig. 3(f). Fig. 4 shows the XRD patterns of the SMATed and un-SMATed experimental samples. The broadening of the diffraction peaks and a slight shift in the centroid position can be observed both in the XRD patterns of pure Mg and Mg–1Ca alloy after SMAT. This might be attributed to grain refinement, microstrain development and micro-distortions of the crystalline lattice. Average grain sizes and microstrain of the SMATed pure Mg and Mg–1Ca alloy calculated from XRD patterns were shown in Table 1. The average grain sizes in the top layer of all the SMATed samples were less than 100 nm. The SMAT-50 min and SMAT-60 min pure Mg showed
Fig. 5 shows the variation of microhardness from the SMATed surface to the substrate beneath, which indicates that the hardness of the samples was enhanced within 1400 μm from the surface for the pure Mg and 1000 μm for the Mg–1Ca alloy after SMAT. The microhardness of the SMATed sample gradually decreases along the depth and approaches the value of the matrix. The microhardness didn't change with the SMAT treating time. It can be seen that the improvement in microhardness of SMATed Mg–1Ca alloy is more obvious than that of SMATed pure Mg. The top most layer of the SMATed Mg–1Ca alloy is at least twice harder than that of its undeformed matrix. The increase of the surface hardness of the SMATed pure Mg can be attributed to the both effects of the grain refinement and work-hardening. As for the SMATed Mg–1Ca alloy, the re-dissolution of the Mg2Ca phase is another reason for microhardness increase. The result indicates that the effect of solution strengthening out-performs grain refinement and work-hardening on the surface. 3.3. Surface roughness and wettability Surface roughness of pure Mg and Mg–1Ca alloy was significantly increased after SMAT, as shown in Fig. 6. Correspondingly, the SMATed samples with rougher surfaces exhibited better wettability (Fig. 7). The enhanced hydrophilicity favors the adsorption of adhesive proteins. Many previous works showed that implant materials with higher roughness promoted in vitro cell response [31–35] and in vivo
Fig. 8. Hydrogen evolution behavior of the SMATed (a) pure Mg and (b) Mg–1Ca alloy, and the pH value variation of the Hank's solution incubating the SMATed (c) pure Mg and (b) Mg–1Ca alloy as a function of the immersion time, with the un-SMATed pure Mg and Mg–1Ca alloy as controls.
N. Li et al. / Materials Science and Engineering C 35 (2014) 314–321
319
Fig. 9. Open circuit potential of (a) un-SMATed and SMATed pure Mg, (b) un-SMATed and SMATed Mg–1Ca alloy, and potentiodynamic polarization curves of (c) un-SMATed and SMATed pure Mg and (d) un-SMATed and SMATed Mg–1Ca alloy.
osteointegration [34,35]. However, the increasing roughness also resulted in an accelerated corrosion rate [36,37]. 3.4. Corrosion behavior Fig. 8 shows the hydrogen evolution rates of the experimental samples and the pH value variation of the Hank's solution during the immersion testing. All the SMATed samples generated more hydrogen gas and the incubated Hank's solution exhibited much higher pH value in comparison with the un-SMATed counterparts. The corrosion developed rapidly as soon as the SMATed samples touched the solution and the pH value was over 9 (some even over 10) after 1 h, thereafter the hydrogen evolution rates slowed down and the pH values gradually stabilized at 11–13. Results of the electrochemical testing further confirmed the poor corrosion resistance of the SMATed samples (Fig. 9 and Table 2). The
corrosion rates of the SMATed sample calculated from the potentiodynamic polarization curves were two to three orders of magnitude higher than those of the un-SMATed samples. By conventional definition, the more negative the Ecorr, the less noble the material is. However, it's interesting that the SMATed samples exhibited nobler OCP and Ecorr than the un-SMATed samples. From the potentiodynamic polarization curves (Fig. 9(c) and (d)), it can be seen that SMAT appears to lower the anodic kinetics, but detrimentally increases the cathodic kinetics as compared to the un-SMATed counterparts. Similar phenomenon is also reported by Hoog et al. [16]. SMAT process is believed to induce an extremely high density of strain-induced crystalline defects in the near-surface region, including grain boundaries (subgrain boundaries) and dislocations. The defects are susceptible to corrosion and passive film forms rapidly on the surface, which could explain the elevated initial OCP and Ecorr values of SMATed samples in comparison to the un-SMATed one. For easy-
Table 2 The electrochemical data obtained from potentiodynamic polarization measurements. Icorr (μA/cm2)
Ecorr (V)
Pure Mg
Mg–1Ca alloy
Un-SMATed SMAT-40 min SMAT-50 min SMAT-60 min SMAT-90 min Un-SMATed SMAT-40 min SMAT-50 min SMAT-60 min SMAT-90 min
Vcorr (mm/yr)
Original surface
After grinding
Original surface
After grinding
Original surface
After grinding
−1.78 −1.49 −1.43 −1.37 −1.47 −1.62 −1.48 −1.51 −1.55 −1.52
−1.78 −1.59 −1.55 −1.57 −1.55 −1.62 −1.60 −1.56 −1.52 −1.53
1.56 2.65 1.29 3.03 1.91 8.45 3.36 2.35 6.14 1.59
1.56 2.28 2.39 3.70 4.65 8.45 6.17 31.7 45.3 50.5
3.52 × 10−2 59.8 29.1 6.83 43.1 1.91 × 10−1 78.8 53.0 13.8 35.9
3.52 5.14 5.39 8.34 1.05 1.91 1.39 7.15 1.02 1.14
× × × ×
103 103 102 103
× × × ×
103 103 102 103
× 10−2
× × × ×
10−1 10−1 10−1 10−1
320
N. Li et al. / Materials Science and Engineering C 35 (2014) 314–321
Fig. 10. Open circuit potential of (a) un-SMATed and SMATed pure Mg, (b) un-SMATed and SMATed Mg–1Ca alloy, and potentiodynamic polarization curves of (c) un-SMATed and SMATed pure Mg and (d) un-SMATed and SMATed Mg–1Ca alloy after the SMATed surfaces were ground off 100 μm.
passive metallic materials, such as titanium alloys, the passive films are compact, so the SMATed titanium alloy surface layer composed by a high fraction of passive layers would reduce the corrosion rates [38]. However, the magnesium hydroxide film is only partially protective, and it allows electrolyte to contact the underlying magnesium and its alloy matrix. Since the high density of crystalline defects drastically activates corrosion reaction and the resulting Mg(OH)2 films present no improved protection, the SMATed pure Mg and Mg–1Ca alloy are more erodible than the un-SMATed counterparts. To eliminate the influence of the surface roughness and reduce the density of defects, 100 μm of the top layer of the SMATed samples was ground off with a 2000 grade carbide abrasive paper. The resulting new surfaces were still in the deformed region with grain size in micrometer scale. Fig. 10 showed that after grinding, the OCP and Ecorr of the SMATed samples shifted to the negative direction and the corrosion rate got close to the un-SMATed samples. Due to the poor corrosion resistance, the SMAT method might not be suitable for slowing down the corrosion rate of biodegradable Mg alloy implants. 4. Conclusions Strain-induced grain refinement layers with grain size at the nanometer scale in the topmost surface were prepared on pure Mg and Mg–1Ca alloy using SMAT. The microhardness of the SMATed samples was improved within 1400 μm depth beneath the surface for the pure Mg and 1000 μm for the Mg–1Ca alloy. The surface roughness and wettability of SMATed samples were greatly enhanced in comparison to the un-SMATed samples. Due to high density of strain-induced crystalline defects in the near-surface region, the corrosion resistance of the SMATed samples was deteriorated. The variation of treating time from 40 min to 90 min didn't influence the properties of the SMATed pure
Mg and Mg–1Ca alloy except the thickness of the deformed layer. The SMAT is proved to be ineffective for biodegradable magnesium and its alloys to enhance its corrosion resistance. Acknowledgments This work was supported by the National Basic Research Program of China (973 Program) (Grant Nos. 2012CB619102 and 2012CB619103), National Natural Science Foundation of China (No. 31170909), National Science Fund for Distinguished Young Scholars (Grant No. 51225101), Research Fund for the Doctoral Program of Higher Education under Grant No. 20100001110011, Project for Supervisor of Excellent Doctoral Dissertation of Beijing (20121000101), Natural Science Foundation of Heilongjiang Province (ZD201012) and Guangdong Province innovation R&D team project (No. 201001C0104669453). References [1] H. Wang, Y. Estrin, H. Fu, G. Song, Z. Zuberova, Adv. Eng. Mater. 9 (2007) 967–972. [2] H. Wang, Y. Estrin, Z. Zuberova, Mater. Lett. 62 (2008) 2476–2479. [3] M. Alvarez-Lopez, M.D. Pereda, J.A. del Valle, M. Fernandez-Lorenzo, M.C. Garcia-Alonso, O.A. Ruano, M.L. Escudero, Acta Biomater. 6 (2010) 1763–1771. [4] X.N. Gu, N. Li, Y.F. Zheng, F. Kang, J.T. Wang, L.Q. Ruan, Mater. Sci. Eng. B 176 (2011) 1802–1806. [5] D. Song, A. Ma, J. Jiang, P. Lin, D. Yang, J. Fan, Corros. Sci. 52 (2010) 481–490. [6] D. Song, A.B. Ma, J.H. Jiang, P.H. Lin, D.H. Yang, J.F. Fan, Corros. Sci. 53 (2011) 362–373. [7] N. Birbilis, K.D. Ralston, S. Virtanen, H.L. Fraser, C.H.J. Davies, Corros. Eng. Sci. Technol. 45 (2010) 224–230. [8] J.H. Gao, S.K. Guan, Z.W. Ren, Y.F. Sun, S.J. Zhu, B. Wang, Mater. Lett. 65 (2011) 691–693. [9] B.R. Sunil, A.A. Kumar, T.S. Sampath Kumar, U. Chakkingal, Mater. Sci. Eng. C 33 (2013) 1607–1615. [10] H.Q. Sun, Y.N. Shi, M.X. Zhang, K. Lu, Acta Mater. 55 (2007) 975–982. [11] A. Amanov, O.V. Penkov, Y.S. Pyun, D.E. Kim, Tribol. Int. 54 (2012) 106–113.
N. Li et al. / Materials Science and Engineering C 35 (2014) 314–321 [12] H.W. Zhang, Z.K. Hei, G. Liu, J. Lu, K. Lu, Acta Mater. 51 (2003) 1871–1881. [13] K.Y. Zhu, A. Vassel, F. Brisset, K. Lu, J. Lu, Acta Mater. 52 (2004) 4101–4110. [14] Y.M. Wang, K. Wang, D. Pan, K. Lu, K.J. Hemker, E. Ma, Scr. Mater. 48 (2003) 1581–1586. [15] H.Q. Sun, Y.N. Shi, M.X. Zhang, Surf. Coat. Technol. 202 (2008) 2859–2864. [16] C. op't Hoog, N. Birbilis, Y. Estrin, Adv. Eng. Mater. 10 (2008) 579–582. [17] Y.-h. Wei, B.-s. Liu, L.-f. Hou, B.-s. Xu, G. Liu, J. Alloys Compd. 452 (2008) 336–342. [18] M. Laleh, F. Kargar, J. Alloys Compd. 509 (2011) 9150–9156. [19] C.D. Yfantis, D.K. Yfantis, J. Anastassopoulou, T. Theophanides, M. Staige, The 5th WSEAS International Conference on Environment, Ecosystess and Development, Venice, Italy, 2006, pp. 186–189. [20] Z.J. Li, X.N. Gu, S.Q. Lou, Y.F. Zheng, Biomaterials 29 (2008) 1329–1344. [21] Y. Wan, G. Xiong, H. Luo, F. He, Y. Huang, X. Zhou, Mater. Des. 29 (2008) 2034–2037. [22] A. Drynda, T. Hassel, R. Hoehn, A. Perz, F.W. Bach, M. Peuster, J. Biomed. Mater. Res. A 93A (2010) 763–775. [23] N. Erdmann, N. Angrisani, J. Reifenrath, A. Lucas, F. Thorey, D. Bormann, A. Meyer-Lindenberg, Acta Biomater. 7 (2011) 1421–1428. [24] J.Z. Ilich, J.E. Kerstetter, J. Am. Coll. Nutr. 19 (2000) 715–737. [25] C.M. Serre, M. Papillard, P. Chavassieux, J.C. Voegel, G. Boivin, J. Biomed. Mater. Res. A 42 (1998) 626–633.
321
[26] X.N. Gu, W. Zheng, Y. Cheng, Y.F. Zheng, Acta Biomater. 5 (2009) 2790–2799. [27] X.N. Gu, N. Li, W.R. Zhou, Y.F. Zheng, X. Zhao, Q.Z. Cai, L. Ruan, Acta Biomater. 7 (2011) 1880–1889. [28] X.N. Gu, Y.F. Zheng, Q.X. Lan, Y. Cheng, Z.X. Zhang, T.F. Xi, D.Y. Zhang, Biomed. Mater. 4 (2009). [29] N. Li, Y.D. Li, Y.B. Wang, M. Li, Y. Cheng, Y.H. Wu, Y.F. Zheng, Surf. Interface Anal. 45 (2013) 1217–1222. [30] G.K. Williamson, W.H. Hall, Acta Metall. 1 (1953) 22–31. [31] N. Nishimura, T. Kawai, J. Mater. Sci. Mater. Med. 9 (1998) 99–102. [32] J.C. Keller, C.M. Stanford, J.P. Wightman, R.A. Draughn, R. Zaharias, J. Biomed. Mater. Res. 28 (1994) 939–946. [33] K.T. Bowers, J.C. Keller, B.A. Randolph, D.G. Wick, C.M. Michaels, Int. J. Oral Maxillofac. Surg. 7 (1992) 302–310. [34] D. Buser, R.K. Schenk, S. Steinemann, J.P. Fiorellini, C.H. Fox, H. Stich, J. Biomed. Mater. Res. 25 (1991) 889–902. [35] K.A. Thomas, S.D. Cook, J. Biomed. Mater. Res. 19 (1985) 875–901. [36] R. Walter, M.B. Kannan, Mater. Des. 32 (2011) 2350–2354. [37] T.L. Nguyen, A. Blanquet, M.P. Staiger, G.J. Dias, T.B. Woodfield, J. Biomed. Mater. Res. B Appl. Biomater. 100 (2012) 1310–1318. [38] R. Huang, Y. Han, Mater. Sci. Eng. C 33 (2013) 2353–2359.
