Biodegradability Engineering of Bio-absorbable Mg alloys: Tailoring the electrochemical property and microstructure of constituent phases

Transcription

1 Supplementary Information Biodegradability Engineering of Bio-absorbable Mg alloys: Tailoring the electrochemical property and microstructure of constituent phases Pil-Ryung Cha, Hyung-Seop Han, Gui-Fu Yang, Yu-Chan Kim, Ki-Ha Hong, Seung-Cheol Lee, Jae-Young Jung, Jae-Pyeong Ahn, Young-Yul Kim, Sung-Youn Cho, Seok-Jo Yang and Hyun-Kwang Seok.

2 1. Microstructures of as-cast and extruded Mg-Ca-Zn alloys Figure S1(a)-(e) shows the microstructures of as cast Mg-5wt%Ca-Xwt%Zn (X= 0, 0.5, 1, 1.5, 3) alloys. As-cast Mg-5wt%Ca is composed of the primary α-mg and eutectic Mg/Mg 2 Ca phases. High contrast represents the Mg 2 Ca phase in the eutectic phase which constructs a well-connected network in the whole specimen. The microstructure was changed significantly by the addition of Zn as shown from Fig. S1 (b)-(e). The network lamella structure almost disappeared and the Mg 2 Ca phase was condensed only by the addition of 1wt.% Zn. Additionally, a new eutectic structure (α-mg + Mg 6 Ca 2 Zn 3 ) appeared and its volume and size was increased with the increase of Zn content. Figure S1(f) and (g) shows the microstucture of extruded Mg-5wt%Ca-Xwt%Zn (X= 1, 3) alloys. The secondary phase with fine lamellar contrasts was observed in the extruded samples and refinement of the microstructure and dispersion of the lamellar phase were observed after extrusion. Each phasees were confirmed by XRD (Fig. S2). Role of Mg 6 Ca 2 Zn 3 in the corrosion Here we present why we concluded that Mg 2 Ca(Zn) phase was responsible for the reduced corrosion rate insead of the ternary intermetallic compound Mg 6 Ca 2 Zn 3. As shown in Fig. 1 of the manuscript, corrosion mainly occurs along Mg 2 Ca phase in the binary Mg-Ca alloy and corrosion rate decreases significantly with the addition of Zn content. Most important fact in this observation is that the corrosion of Mg 2 Ca is reduced sharply with the addition of Zn content. As shown in EDX results on Fig. 2 of the manuscript, corrosion rate shows inverse proportion to the concentration of Zn in Mg 2 Ca phase of the Mg-Ca-Zn alloy. If the ternary intermetallic compound is mainly responsible for the reduction of corrosion rate, higher corrosion rate of the 4wt% Zn alloy compared to the 3wt% Zn one cannot be explained (Fig. 2e of the manuscript) since the phase fraction of ternary intermetallic compound increased as the content of Zn increased (Fig. S3). In particular, the ternary intermetallic compounds were rarely found, and mostly Mg 2 Ca(Zn) phases were observed when the corroded fronts of Mg- Ca-Zn alloys were closely observed using SEM. Based on the above results, we concluded that the low corrosion rate of our Mg-Ca-Zn alloy resulted mainly from the corrosion potential change of the Mg 2 Ca(Zn) phase.

3 Figure S1: Microstructures of alloys: (a-e) as-cast Mg-5wt%Ca-xwt%Zn (x= 0, 0.5, 1, 1.5 and 3); and (f-g) extruded Mg-5wt%Ca-xwt%Zn (x= 1 and 3) (longitudinal (bottom) and radial (top) direction, respectively).

4 Mg Mg 2 Ca Mg 6 Ca 2 Zn 3 f) e) Intensity d) c) a) theta / degree b) Figure S2: XRD patterns obtained from samples of Figure S2. Phase fraction (%) Zn Contents (wt%) Mg 2 Ca Ca 2 Mg 6 Zn 3 Figure S3: Variation of phase fractions of Mg2Ca and Mg6Ca2Zn3 intermetallic compounds with the Zn content in Mg-5wt%Ca-xwt%Zn alloys at 293 K (Pandat thermodynamics software).

5 2. Mechanical and corrosion properties of the Mg-Ca-Zn alloys with different Zn contents Element alloying technology is one of the effective measurements to improve the corrosion resistance as well as mechanical properties of magnesium. However, some alloying elements can cause a high risk of side effects during in vivo service so the alloying elements must be biocompatible. Among the possible elements, calcium was proposed by many researchers to alloy with magnesium to achieve greater mechanical properties. The addition of at least 5wt. % Ca is needed to obtain the suitable properties of Mg alloys for use in field of orthopedics. Unfortunately, the biodegradation rate of Mg-Ca binary alloy was accelerated seriously compared to that of pure Mg when the addition of Ca content was more than 1wt. % [S1]. In order to improve the bio-corrosion resistance of Mg-Ca alloy without sacrificing the benefits of calcium element, zinc element was considered as a third alloying element because zinc has the potential to improve the corrosion resistance after aluminum. Zhang et al. developed Mg- Ca-Zn-Mn alloys and implied that the corrosion resistance had close relationship with the atomic ratio of Zn/Ca [S1]. Furthermore, Zn is highly essential microelement and component of many proteins, nucleic acid synthetase of human body. Biodegradation behavior of Mg-5wt.%Ca-xwt.%Zn alloys was investigated by immersion test in Hank s solution. The corrosion amount of Mg-5wt.%Ca-xwt.%Zn (x=0, 1, 1.5, 3, 5) alloys was evaluated by measuring the amount of hydrogen evolution during the immersion test. Figure S4 shows that the amount of hydrogen evolution for each as-cast alloy as a function of immersion time. For Mg-5%Ca binary alloy, the amount of hydrogen evolution was 3.47 ml/cm 2 during 3.37 hours immersion. However, after the addition of Zn content, for all the Mg-5wt.%Ca-xwt.%Zn (x=1, 1.5, 3, 5) ternary alloys, the amount of hydrogen evolution was less than 0.02ml/cm 2 during the same immersion time. The addition of Zn affected the biodegradation behavior not only for short immersion time but also for long time immersion. The hydrogen evolution of Mg-5wt.%Ca-xwt.%Zn (x=1, 1.5, 3, 5) ternary alloys during 300 hours immersion was still less than that of Mg-5wt.%Ca binary alloy during 3.37 hours immersion. So it was obvious that the biodegradation behavior of Mg-5wt.%Ca alloy can be controlled effectively by the addition of the Zn content.

6 Hydrogen amount (ml/cm 2 ) Mg5Ca Mg5Ca1Zn Mg5Ca1.5Zn Mg5Ca3Zn Mg5Ca5Zn Immersion Time [hr] Figure S4: Hydrogen evolution amount of as-cast Mg-5%Ca-x%Zn alloy. In order to show the Zn effects on the biodegradation behavior from quantitative calculation, the immersion time of each sample for one fixed hydrogen evolution amount was compared with each other. According to Figure S4, to reach the corroded state corresponding to 0.4ml/cm 2 hydrogen gas, the immersion time for each alloy was required as follows: 1.02 hours for Mg-5wt.%Ca, hours for Mg-5wt.%Ca-1wt.%Zn, hours for Mg- 5wt.%Ca-1.5wt.%Zn, hours for Mg-5wt.%Ca-3wt.%Zn, and hours for Mg- 5wt.%Ca-5wt.%Zn. It showed that the biodegradation rate was decreased significantly when the addition of Zn content was increased from 1 wt.% to 1.5 wt. % and it was almost maintained when the addition of Zn content was increased from 1.5wt.% to 3wt.%. However, the biodegradation rate was increased slightly when the addition of Zn content was more than 3wt.% in Mg-Ca alloy.

7 Figure S4: Hydrogen evolution rate of extruded Mg-5wt.%Ca-xwt.%Zn alloy. Strength (MPa) Y.S U.C.S El Mg5Ca Mg5Ca0.5Zn Mg5Ca1Zn Mg5Ca1.5Zn Mg5Ca3Zn 194 Figure S5: Mechanical properties of Mg alloys with different Zn contents Figure S5 shows the variation of yield stress (YS), ultimate compressive stress (UCS) and elongation (El) in several different Mg alloys. YS and UCS increase with increasing Zn contents while elongation property stays similar.

8 3. Mechanical and corrosion properties of the extruded Mg-Ca-Zn alloys with different Ca contents The related studies on biodegradable Mg alloys have been mostly focused on relatively low contents of Ca (less than < 2 3wt%). This originates from the consensus developed in the paper [S2] stating that the increase of Ca content to more than 3wt% resulted in the increase of mechanical strength accompanied by steep decline in elongation and sharp increase of corrosion rate. This is due to the brittleness of Mg2Ca phase. Under this consensus, various Mg-Ca-Zn alloy systems with less than 2-3wt% Ca have been published. Currently, due to this high brittleness, there are only few reported studies showing the tensile test results (yield strength and elongation) of Mg-Ca/Mg-Ca-Zn alloys with high Ca content (>2wt%Ca). The brittleness also prevented microstructural modification of alloys with high Ca contents by using mechanical methods such as extrusion and ECAP. However, enough ductility was achieved for extrusion of Mg-Ca/Mg-Ca-Zn cast alloys in this study. The microstructural modification by extrusion provided high yield strength for ductile Mg-Ca-Zn alloys as shown in Fig. S6. YS and ultimate tensile strength (UTS) increased with increasing Ca contents whereas the elongation decreased. Mg-5wt%Ca-1wt%Zn alloy deeply investigated in our manuscript shows high YS (~ 200MPa) and good ductility (~10%). Strength (MPa) Y.S U.T.S El Mg0.8Ca1Zn Mg1Ca1Zn Mg3Ca1Zn Mg5Ca1Zn Mg7Ca1Zn

9 Figure S6: Mechanical properties of extruded Mg-Ca-Zn alloys with different Ca contents The corrosion rate of the various alloys with different Ca contents were investigated via hydrogen evolution as shown in Fig. S7. The result showed that the corrosion rate of Mg- 5wt%Ca-1wt%Zn alloy was very close to that of extremely pure Mg. The reduced corrosion rate of Mg-5wt%Ca-1wt%Zn is mainly due to the Mg 2 Ca(Zn) phase as mentioned in our manuscript Hydrogen amount (ml/cm 2 ) E-pure Mg E-Mg-0.8Ca-1Zn E-Mg-1Ca-1Zn E-Mg-3Ca-1Zn E-Mg-5Ca-1Zn E-Mg-7Ca-1Zn Time (hr) Figure S7: Hydrogen evolution amount over time of the Mg-xwt.%Ca-1wt.%Zn alloys (x=0.8, 1, 2, 3, 5, 7). In summary, the optimized electro-chemical tailoring was performed to make the corrosion rate of our Mg alloys close to that of extremely pure Mg as much as possible by modifying corrosion potential of two constituent phases. Furthermore, clinically acceptable corrosion property (no observation of bubble formation in vivo) was achieved through the optimized microstructural modification. As shown in Fig. S8, the application of both electro-chemical tailoring and microstructural modification produced low corrosion rate as extruded pure Mg with far superior mechanical properties.

10 Figure S8: Tensile yield strength versus bio-corrosion rate map of various Mg alloys 4. Weight Loss Measurements The weight loss measurement is a commonly practiced method to measure the corrosion rate of magnesium, and the result can be compared with measurements done via hydrogen evolution to validate the general trend. The weight loss measurements for this study were performed using the same method from previously reported studies [S3, S4]. All the samples were washed with distilled water after immersion in Hanks solution. The immersion at room temperature in chromic acid cleaning solution containing 200gL-1 CrO3 + 10gL-1 AgNO3 followed to remove the corrosion products without removing any metallic Mg alloys. Samples were then washed with distilled water and dried with high-pressure air to be dried in the desiccator for 1-2 days. The weight was measured, and the weight loss (mg cm -2 d -1 ) was evaluated as: W = W b W a AA where W a was the weight after the immersion test and W b was the weight before the immersion test. A was the surface area of sample (cm 2 ) and t was the time of immersion (d). The average corrosion rate, P W (mm y -1 ) was calculated by converting the weight loss using

11 [S3]: P W = 2.1 W While there is a linear relationship between these two measurements, the corrosion rate evaluated via examination of the hydrogen evolution rate would be less than the measurement done by the weight loss evaluation as summarized in the elaborate corrosion work done by N.I. Z Abidin et al [S3]. Several reported evaluation of high purity (HP) Mg (99.99%) corrosion via weight loss and hydrogen evolution is available [S4-S7] and they are summarized in the table below. Table S1: Corrosion rate of HP Mg (99.99%) obtained from weight loss and hydrogen evolution measurments. (mm/y) Fe Content (ppm) P W (3d) P W (30d) P H (30d) HP-Mg (99.99) by G. Song Et al (2007) HP-Mg (99.99) by S. Zhang Et al (2009) HP-Mg (99.99) by S. Zhang Et al (2010) HP Mg used for these studies included 45~70 ppm of Fe whereas HP Mg used for our study contained only ~14 ppm of Fe. The corrosion rate of HP Mg with lower Fe contents (15 ppm) was much lower than that of the above HP Mg (45~70 ppm of Fe). Fe impurity of the Mg greatly affects the corrosion rate and this phenomenon is evident from the result of our study. Table S2: Corrosion rates of extruded HP Mg (99.98%) and extruded Mg-5wt%Ca-1wt%Zn alloy obtained from weight loss and hydrogen evolution measurements. Bold characters represent the values obtained from weight loss meansurement. (mm/y) P H (4d) P H (8d) P H (14d) P W (4d) P W (8d) P H (4d) Hanks' Hanks' Hanks' Hanks' Hanks' 3%NaCl HP-Mg(14ppm) Mg-5wt%Ca-1wt%Zn (27ppm) As shown in Table S2, newly developed ternary Mg alloy showed significantly lower corrosion rate when compared to HP Mg used for this study. Weight loss evaluation showed

12 higher corrosion rate when compared to the evaluation via hydrogen evolution but the trend of the corrosion rate was the same in both measurements. 5. Work functions of pure Mg and Mg2Ca The microstructure of Mg-Ca alloys is composed of pro-eutectic Mg and Mg/Mg 2 Ca eutectic phases. The rapid corrosion of Mg-Ca alloys in Hank s solution is accompanied by the local selective corrosion of Mg 2 Ca phase in the eutectic microstructure (Fig. 1 b in the manuscript). The local corrosion of Mg 2 Ca phase is typical behavior of Galvanic corrosion where we can conjecture that Mg and Mg2Ca have different corrosion potentials in Hank s solution. We conducted the first principle calculations to obtain the information for the corrosion potentials of Mg and Mg2Ca and to investigate the effect of Zn. Which of material properties can give us the information for the corrosion potential of each phase? There is a clue in the prevailing equipment used in the measurement of the corrosion potential. Scanning Kelvin Probe (SKP), one of the corrosion potential measurement equipments, measures the work function of a material to obtain its corrosion potential. It has been reported for several materials that measured work function bears a proportionate relationship to the corrosion potential [S8]. The proportional relationship between work function and corrosion potential may be explained as follows. The Fermi level of electron in a material means its chemical potential [S9]. Based on the Fermi levels of materials, it can be approximately determined the electron transfer from one material to the other one in two materials junction because electron moves from higher chemical potential area to lower one. As relative heights of the Fermi levels of materials between materials are correlated to their work functions, the work function can be conjectured to be related to the corrosion potential. We performed the first principle calculations to obtain the work functions of pro-eutectic Mg and Mg 2 Ca phases and to investigate the effect of Zn. Density functional theory (DFT) calculations using the VASP program packages [S10] were used. The plane wave basis expansions with an energy cutoff of 300 ev and the generalized gradient approximation (GGA) with the PW91 exchange-correlation functional were used. The core-valence interaction is described by the projector-augmented wave (PAW) method [S11]. We constructed Mg 2 Ca slab structure to calculate work function as shown in Fig. 9, which are

13 constructed with 96 Mg atoms and 48 Ca atoms. Vacuum sizes are given larger than 20 Å to minimize interaction between slabs. We first obtained the bulk Mg 2 Ca structure applying Monkhorst-Pack sampling with a k-point grid and then constructed a slab structure with the resulting lattice constant and a k-point grid. We fully optimized the atomic configurations except 8 bottom atomic layers of which positions are fixed to mimic bulk properties. Figure S9 shows the atomic configuration of Mg 2 Ca and the top view of its (0001) surface used in our first principle calculations. Mg 2 Ca has hexagonal structure and its (0001) surface has minimum surface energy. On (0001) surface Ca atoms are located at higher positions compared to Mg atoms. The work function of Mg 2 Ca in the most stable (0001) surface is 3.16 ev that is lower than that of pure Mg, say, 3.7 ev. This means that the Fermi level of Mg 2 Ca is higher than that of pro-eutectic Mg and that the Galvanic circuit forms from Mg 2 Ca with higher electrochemical potential to pro-eutectic Mg. Figure S9: Atomic configurations of Mg 2 Ca used in the work function calculation. The upperright figure shows the top view of the calculation domain that corresponds to (0001) surface. To calculate the effect of Zn on the work function of Mg 2 Ca, we replaced Ca atoms protruding from (0001) surface with Zn atoms. The change of work function with the number

14 of replaced Zn atoms is listed in Table S3. Table S3 shows the variation of work function in (0001) surface with the number of Zn atoms on the surface. In case of bare surface without Zn, the WF is 3.16 ev and increases with Zn surface coverage such that it reaches 3.48 ev at four surface Zn atoms. Therefore, the Galvanic potential between Mg and Mg 2 Ca decreases with increasing Zn contents. Table S3: Variation of the work function with Zn contents Number of surface Zn atoms Work function (ev) Cytotoxicity test of excessive Mg, Ca and Zn ions To investigate the effects of metal ions, cytotoxicity tests using HeLa and L929 cell line were performed. The HeLa and L929 cells were maintained in the Dulbecco s Modified Eagle s Medium (DMEM, Welgene, Korea) supplemented with 10% fetal bovine serum (FBS, Welgene), 100 U/ml penicillin and 100ug/ml streptomycin (Welgene) at 37 in a humidified atmosphere of 5% CO 2, the media being changed every three days. For cytotoxicity test of three metal ions, the Magnesium, calcium and zinc chloride (Sigma-Aldrich, St. Louis, MO) was used. 5 x 10 3 of cells were seeded in 96 well plate and pre-cultured for 24 hours to allow cell adhesion. Cell culture media was changed with fresh media containing up to 32mM of magnesium, calcium ions and up to 1mM of zinc ions. Cells were incubated for 24 hours and cell viability was measured by the CCK-8 kit (Dojindo, Japan). Figure S10 show the effect of Mg, Ca and Zn ions on cell viability. Zinc ion showed the most toxic effect on both cell models. Though it was slightly depend on the cell type, the cell viability started to be reduced from about more than 60µM of zinc ion. In both magnesium and calcium ions, which had no significant differences in cytotoxicity, a decline of the cell viability peered from about more than 30mM.

15 Cell Viability (%) Cytotoxicity of metal ions (HeLa) Concentration (um) C(-) Mg Ca Zn ppm ppm Cell Viability (%) Cytotoxicity of metal ions (L929) Concentration (um) C(-) Mg Ca 120 Zn

16 Figure S10: The cytotoxicity for Mg, Ca and Zn ions in HeLa and L929 cell lines. References [S1] E. Zhang, L. Yang, Mater. Sci. and Eng.A 497, 111 (2008) [S2] Z. Li, X. Gu, S. Lou and Y. Zheng, Biomaterials, 29 (2008), pp [S3] N.I.Z. Abidin, A.D. Atrens, D. Martin, A. Atrens, Corrosion Science, 53 (2011) [S4] Z. Shi, A. Atrens, Corrosion Science 53 (2011) ; [S5] S. Zhang, X. Zhang, C. Zhao, J. Li, Y. Song, C. Xie, H. Tao, Y. Zhang, Y. He, Y. Jiang, Y. Bian, Acta Biomaterialia 6 (2010) [S6] G. Song, Corrosion Science 49 (2007) [S7] S. Zhang, J. Li, Y. Song, C. Zhao, X. Zhang, C. Xie, Y. Zhang, H. Tao, Y. He, Y. Jiang, Y. Bian, Materials Science and Engineering C 29 (2009) [S8] S. Yee, R.A. Oriani, M. Stratmann, J. Electrochem. Soc., 138 (1991) 55 [S9] C. Kittel, Introduction to solid state physics, John Wiley & Sons, Inc [S10] Kresse, G.; Hafner, J. Phys. Rev. B 1994, 49, 14251; Kresse, G.; Furthmuller, J. Comp. Mater. Sci. 1996, 6, 15. [S11] Blochl, P. E. Phys. Rev. B 1994, 50, 17953