Microstructure Evolution and Properties of β -TCP/Mg-Zn-Ca Biocomposite Processed by Hot Extrusion Combined with Multi-Pass ECAP

: To further improve the comprehensive performance of Mg-based alloy, hot extrusion combined with multi-pass equal channel angular pressing (ECAP) was applied to process Mg-3 wt%Zn-0.2 wt%Ca alloy and 1 wt% β -TCP/Mg-3 wt%Zn-0.2 wt%Ca biocomposites. The microstructure evolution, mechanical properties, corrosion behavior, and cell biocompatibility of the experimental specimens were systematically investigated. The average grain size of 13.4 ± 0.6 µ m in MgZnCa alloy and 9.6 ± 0.3 µ m in composites materials can be achieved by six ECAP passes. The uniaxial compressive strength (UCS) of 388.4 ± 7.3 MPa and the strain at failure of 14.3 ± 1.5% were conﬁrmed in MgZnCa alloy, while the UCS of 405.3 ± 7.4 MPa and the strain at failure of 9.8 ± 1.9% were achieved by the addition of β -TCP after six ECAP passes. In spite of different compositions, the minimum corrosion rate of 0.895 mm · Y − 1 and 1.117 mm · Y − 1 can be achieved by two ECAP passes at 593 K. The cytocompatibility evaluation revealed that the experimental materials processed by six ECAP passes had no signiﬁcant cytotoxicity to L929 cells, and the addition of β -TCP improved the cytocompatibility.


Introduction
Biodegradable magnesium (Mg)-based materials attracted significant attention for their potential application in the biomedical industry [1,2]. Compared with conventional metallic biomaterials such as stainless steel [3], CoCr [4], and titanium alloys [5], biodegradable Mg alloys can be safely metabolized after fulfilling their function in the human body, which circumvents secondary operation for the removal of the implant. By using biodegradable Mg-based implant materials, the burden on the body and the economy of the patients could be decreased [6]. Moreover, compared to the bioinert metallic materials, Mg-based materials possess lower densities of 1.74-2.00 g·cm −3 , and a lower elastic modulus of 10-40 GPa [7,8]. It is beneficial to reduce the stress shielding effect after Mg alloys are implanted in the human body [8]. Among various Mg-based alloy systems, the Mg-Zn-Ca alloys have been recognized as promising candidates for orthopedic applications due to their biodegradability and good biocompatibility [9][10][11][12]. Zn as the alloying element could enhance the corrosion resistance of the Mg matrix by forming a compound protective layer on the material's surface, while Ca shows excellent biocompatibility like Mg, and is prone to form calcium phosphate during degradation, which is beneficial to the formation Commercially available high-purity magnesium ingots (99.99 wt%), pure zinc granules (99.99 wt%) with the size of 3-4 mm, magnesium-calcium intermediate alloy (25 wt% Ca) and β-TCP nanoparticles were used as raw materials. β-TCP particles with an average diameter of~150 nm were prepared via a hydrothermal method, which was reported elsewhere [35]. Generally, compared with micro-sized TCP particles, nano-sized counterparts with the same content could result in a higher specific surface area and a higher number density in the Mg matrix, which could possibly induce a more intimate bonding between TCP and Mg matrix, and a more efficient grain refinement effect [36][37][38]. Furthermore, nano-reinforcements can remarkably increase mechanical strength by effectively promoting particle hardening mechanisms [39]. Hence, for the sake of efficient grain refinement and effective strengthening behavior, nano-sized β-TCP particles were chosen as the reinforcement for the Mg-Zn-Ca matrix. According to our experience, to achieve a substantial improvement in mechanical performance and avoid the aggregation of β-TCP particles, the β-TCP content was set at 1 wt%. A schematic diagram of the processing route of Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composite is shown in Figure 1a. A self-developed furnace with the devices for high-speed stirring and ultra-sonication was employed for the material preparation. Under the protective atmosphere of N 2 + SF 6 , the raw materials were melted at 993 K under high-speed stirring and ultra-sonication for 15 min and then cast. Subsequently, the ingot was hot extruded into a square rod with an edge length of 15 mm followed by a multi-pass ECAP process with an internal-channel angle of 120 • and an external angle of 0 • , as shown in Figure 1b. The specimens were preheated for 15 min before each ECAP pass. The ECAP process was carried out in route A (no rotation between subsequent passes) with a cross-head speed of 2 mm·s −1 .
effectively promoting particle hardening mechanisms [39]. Hence, for the sake of efficient grain refinement and effective strengthening behavior, nano-sized β-TCP particles were chosen as the reinforcement for the Mg-Zn-Ca matrix. According to our experience, to achieve a substantial improvement in mechanical performance and avoid the aggregation of β-TCP particles, the β-TCP content was set at 1 wt%. A schematic diagram of the processing route of Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composite is shown in Figure 1a. A self-developed furnace with the devices for high-speed stirring and ultra-sonication was employed for the material preparation. Under the protective atmosphere of N2 + SF6, the raw materials were melted at 993 K under high-speed stirring and ultra-sonication for 15 min and then cast. Subsequently, the ingot was hot extruded into a square rod with an edge length of 15 mm followed by a multi-pass ECAP process with an internal-channel angle of 120° and an external angle of 0°, as shown in Figure 1b. The specimens were preheated for 15 min before each ECAP pass. The ECAP process was carried out in route A (no rotation between subsequent passes) with a cross-head speed of 2 mm·s −1 . In this study, the ECAP processing parameters for Mg-3Zn-0.2Ca alloy and the 1β-TCP/Mg-3Zn-0.2Ca composites were summarized in Table 1. The ECAP-processed specimens were named using composition (A means Mg-3Zn-0.2Ca alloy and C means 1β-TCP/Mg-3Zn-0.2Ca composite), ECAP processing times, and ECAP temperature for convenience. For example, A-2-593 K means that Mg-3Zn-0.2Ca alloy ECAP was treated 2 times at 593 K. The effect of ECAP on the microstructure mechanical properties and corrosion resistance of the materials was systematically investigated from the ED of the experimental specimens.

Names
Material ECAP Times Temperature (K) A-as-extruded alloy 0 -C-as-extruded composite 0 - In this study, the ECAP processing parameters for Mg-3Zn-0.2Ca alloy and the 1β-TCP/Mg-3Zn-0.2Ca composites were summarized in Table 1. The ECAP-processed specimens were named using composition (A means Mg-3Zn-0.2Ca alloy and C means 1β-TCP/Mg-3Zn-0.2Ca composite), ECAP processing times, and ECAP temperature for convenience. For example, A-2-593 K means that Mg-3Zn-0.2Ca alloy ECAP was treated 2 times at 593 K. The effect of ECAP on the microstructure mechanical properties and corrosion resistance of the materials was systematically investigated from the ED of the experimental specimens.

Phase Analysis and Microstructure Observation
The phase constitution was analyzed by X-ray diffraction (XRD; SmartLab, Rigaku, Tokyo, Japan) using Cu-Ka radiation (k = 0.154059 nm) with an operating voltage of 40 kV and an operating current of 44 mA. XRD investigation was carried out over the 2 theta range of 20 • to 80 • with a step size of 0.02 • and a dwelling time of 1 s. The test surface was perpendicular to the extrusion direction. The microstructure observations were performed on an optical microscope (GX51, OLYMPUS, Tokyo, Japan). The grain size was measured using an automatic image analyzer (OLYMPUS M3, OLYMPUS, Tokyo, Japan), using the linear intercept method as described in the ASTM standard E112-G6. Transmission electron microscope (TEM) investigations were performed on a JEOL 2100 device (JEOL, Tokyo, Japan) at an accelerated voltage of 200 kV. The TEM specimens were prepared by twin-jet etching by a mixed solution of 2.75 g picric acid, 2.5 mL glacial acetic acid, 5 mL deionized water, and 45 mL anhydrous ethanol. Texture analysis for the C-as-extruded and the C-4-593 K was executed by electron backscatter diffraction (EBSD; Hikari XP, EDAX, Mahwah, NJ, USA). The samples for EBSD were prepared by a typical ion-milling method. The EBSD results were analyzed using the Channel 5 software (Version 3.1, Hobro, Denmark).

Mechanical Properties
The mechanical properties were tested via compressive test. Samples for the compressive tests are processed into the dimension of Φ 4 mm × 8 mm. The compression rate was 0.5 mm·s −1 . The compressive yield strength was determined according to ASTM E111. Three parallel samples were tested for each material.

Electrochemical Measurements
Electrochemical analysis was performed at 310 K using an electrochemical workstation (Zennium Zahner, Sciencetech, London, ON, Canada), which consisted of the electrolytic tank containing simulated body fluid (SBF) and a standard three-electrode system (graphite as the control electrode, saturated calomel electrode as the reference electrode and the sample as the working electrode). After the open circuit potential was recorded for 1800 s, the potential dynamic polarization was performed at a scanning rate of 1 mV·s −1 , and the voltage range was set as self-corrosion potential ± 500 mV.

Cell Biocompatibility Evaluation
The test was determined pursuant to ISO 10993-5 standard [40]. The medium was RPMI-1640 (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% inactivated fetal bovine serum, 100 U·mL −1 penicillin, and 100 U·mL −1 streptomycin. Samples (7 mm × 7 mm × 3 mm) were sterilized by ultraviolet rays, and then added into RPMI-1640 culture medium according to the ratio of sample surface area to culture volume of 3 cm 2 ·mL −1 , and placed in an incubator with 5% CO 2 at 310 K for leaching for 7 days to obtain the leaching stock solution. Mouse fibroblasts (L-929 cells) were grown in a cell culture incubator with 5% CO 2 at 310 K. Four parallel samples were set for each material in this test. The negative control group was 10% serum +90% RPMI-1640 culture medium. 100 mL cells were stained with 0.4% trypan blue. According to the concentration of 5000 cells·well-1, the cells were injected into a 96-well culture plate and cultured at 310 K for 24 h, then washed with phosphate buffer solution (PBS). After 1 day, 3 days, and 5 days of culture, the cells were taken out, and 20 mL 3-(4, 5-Dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) with a concentration of 5 mg·mL-1 was added to continue to culture for 3 h. After MTT fully reacted with the cells, the supernatant of the culture medium was removed, and 150 mL dimethyl sulfoxide was added. After the crystals were completely dissolved, the absorbance (OD value) was measured with a microplate reader at a wavelength of 570 mm. The relative proliferation rate of the cells was calculated as RGR (average OD of the experimental group × 100%/mean OD of the negative control group). Figure 2 shows the XRD profiles of as-extruded Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composite. For A-as-extruded and C-as-extruded, the main constitution was identified as the α-Mg, the second phases were composed of Ca 2 Mg 6 Zn 3 and MgZn phases. However, no diffraction peak of TCP was detected by XRD in the C-as-extruded. It may be because the β-TCP content is lower than the sensitive value detected by XRD. No diffraction peak of the intermediate phase between the β-TCP and the matrix alloy was identified in the C-as-extruded, indicating that no significant chemical reaction occurred between β-TCP and the matrix alloy.

Phase Analysis
days to obtain the leaching stock solution. Mouse fibroblasts (L-929 cells) were grown in a cell culture incubator with 5% CO2 at 310 K. Four parallel samples were set for each material in this test. The negative control group was 10% serum +90% RPMI-1640 culture medium. 100 mL cells were stained with 0.4% trypan blue. According to the concentration of 5000 cells·well-1, the cells were injected into a 96-well culture plate and cultured at 310 K for 24 h, then washed with phosphate buffer solution (PBS). After 1 day, 3 days, and 5 days of culture, the cells were taken out, and 20 mL 3-(4, 5-Dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) with a concentration of 5 mg·mL-1 was added to continue to culture for 3 h. After MTT fully reacted with the cells, the supernatant of the culture medium was removed, and 150 mL dimethyl sulfoxide was added. After the crystals were completely dissolved, the absorbance (OD value) was measured with a microplate reader at a wavelength of 570 mm. The relative proliferation rate of the cells was calculated as RGR (average OD of the experimental group × 100%/mean OD of the negative control group). Figure 2 shows the XRD profiles of as-extruded Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composite. For A-as-extruded and C-as-extruded, the main constitution was identified as the α-Mg, the second phases were composed of Ca2Mg6Zn3 and MgZn phases. However, no diffraction peak of TCP was detected by XRD in the C-as-extruded. It may be because the β-TCP content is lower than the sensitive value detected by XRD. No diffraction peak of the intermediate phase between the β-TCP and the matrix alloy was identified in the C-as-extruded, indicating that no significant chemical reaction occurred between β-TCP and the matrix alloy.  Figure 3 shows the microstructure of as-cast, as-extruded and different ECAP passes (0, 2, 4, 6, and 8) at 593 K of Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composites. Compared with the as-cast microstructure of Mg-3Zn-0.2Ca alloy as shown in Figure 3a, the as-extruded alloy exhibits significant grain refinement. Equiaxed morphology was observed in as-extruded and ECAP-processed specimens. When the ECAP times up to  Figure 3 shows the microstructure of as-cast, as-extruded and different ECAP passes (0, 2, 4, 6, and 8) at 593 K of Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composites. Compared with the as-cast microstructure of Mg-3Zn-0.2Ca alloy as shown in Figure 3a, the as-extruded alloy exhibits significant grain refinement. Equiaxed morphology was observed in as-extruded and ECAP-processed specimens. When the ECAP times up to two, the Mg-3Zn-0.2Ca alloy exhibits apparent grain refinement compared with the as-extruded.

Microstructure Observation
With the further increase in ECAP passes, the grain refinement effect of Mg-3Zn-0.2Ca alloy is enhanced. The A-6 P-593 K possesses the smallest grain size. When the ECAP times were more than six (eight in this study), it is apparent that some grains grow excessively. The average grain sizes of Mg-3Zn-0.2Ca alloy components with 0, 2, 4, 6, and 8 ECAP passes are approximately 17.1 ± 1.3 mm, 15.6 ± 1.0 mm, 14.4 ± 0.8 mm, 13.4 ± 0.6 mm and 14.2 ± 0.8 mm, respectively. as-extruded. With the further increase in ECAP passes, the grain refinement effect of Mg-3Zn-0.2Ca alloy is enhanced. The A-6 P-593 K possesses the smallest grain size. When the ECAP times were more than six (eight in this study), it is apparent that some grains grow excessively. The average grain sizes of Mg-3Zn-0.2Ca alloy components with 0, 2, 4, 6, and 8 ECAP passes are approximately 17.1 ± 1.3 mm, 15.6 ± 1.0 mm, 14.4 ± 0.8 mm, 13.4 ± 0.6 mm and 14.2 ± 0.8 mm, respectively.   Similarly, compared with the as-cast microstructure of 1β-TCP/Mg-3Zn-0.2Ca composite as shown in Figure 3(b), the as-extruded composite also shows significant grain refinement. The average grain sizes of 1β-TCP/Mg-3Zn-0.2Ca composites after 0, 2, 4, 6 and 8 ECAP passes processing are approximately 13.4 ± 1.1 mm, 11.6 ± 0.9 mm, 10.8 ± 0.5 mm, 9.6 ± 0.3 mm and 10.4 ± 0.5 mm, respectively. It is worth noting that 1β-TCP/Mg-3Zn-0.2Ca composite possessed smaller average grain sizes than the Mg-3Zn-0.2Ca alloy processed by the same ECAP parameters. This is mainly due to the role of β-TCP particles in facilitating the dynamic recrystallization and hindering the grain growth during the ECAP process. It is worth noting that the dispersion properties of β-TCP were improved after the ECAP process, due to the redistribution of β-TCP caused by significant shear deformation. The microstructure evolution of the experiment materials with multiple ECAP passes can be partially attributed to the dynamic recrystallization and the redistribution of second phases and β-TCP particles. Under the strong shear force derived from the ECAP process, the Mg-3Zn-0.2Ca alloy undergoes severe plastic deformation in each ECAP processing. The severe plastic deformation caused Similarly, compared with the as-cast microstructure of 1β-TCP/Mg-3Zn-0.2Ca composite as shown in Figure 3b, the as-extruded composite also shows significant grain refinement. The average grain sizes of 1β-TCP/Mg-3Zn-0.2Ca composites after 0, 2, 4, 6 and 8 ECAP passes processing are approximately 13.4 ± 1.1 mm, 11.6 ± 0.9 mm, 10.8 ± 0.5 mm, 9.6 ± 0.3 mm and 10.4 ± 0.5 mm, respectively. It is worth noting that 1β-TCP/Mg-3Zn-0.2Ca composite possessed smaller average grain sizes than the Mg-3Zn-0.2Ca alloy processed by the same ECAP parameters. This is mainly due to the role of β-TCP particles in facilitating the dynamic recrystallization and hindering the grain growth during the ECAP process. It is worth noting that the dispersion properties of β-TCP were improved after the ECAP process, due to the redistribution of β-TCP caused by significant shear deformation. The microstructure evolution of the experiment materials with multiple ECAP passes can be partially attributed to the dynamic recrystallization and the redistribution of second phases and β-TCP particles. Under the strong shear force derived from the ECAP process, the Mg-3Zn-0.2Ca alloy undergoes severe plastic deformation in each ECAP processing. The severe plastic deformation caused dynamic recrystallization and promoted grain refinement. In the meantime, the redistribution of the second phase would occur in each ECAP processing due to the strong shear force of ECAP. Figure 4 shows TEM images and corresponding EDS results for the second phase which should be the Ca 2 Mg 6 Zn 3 phase. However, when the ECAP times were beyond six (eight in this study), the factor of recovery dominated the microstructure evolution of A-8 P-593 K, thereby facilitating the grain growth, resulting in higher grain size than that of A-6 P-593 K specimens. In the case of 1β-TCP/Mg-3Zn-0.2Ca biocomposites, except for the grain refinement and redistribution of Ca 2 Mg 6 Zn 3 s phase, the dispersion properties of the β-TCP particles were also improved under the substantial deformation derived from the ECAP process. The improved dispersion properties of β-TCP are also beneficial to grain refinement by hindering the movement of grain boundary and acting as a potential nucleation site. Therefore, from the viewpoint of grain refinement, six ECAP passes at 593 K should be the optimal ECAP parameter for Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composite.
were beyond six (eight in this study), the factor of recovery dominated the microstructure evolution of A-8 P-593 K, thereby facilitating the grain growth, resulting in higher grain size than that of A-6 P-593 K specimens. In the case of 1β-TCP/Mg-3Zn-0.2Ca biocomposites, except for the grain refinement and redistribution of Ca2Mg6Zn3 s phase, the dispersion properties of the β-TCP particles were also improved under the substantial deformation derived from the ECAP process. The improved dispersion properties of β-TCP are also beneficial to grain refinement by hindering the movement of grain boundary and acting as a potential nucleation site. Therefore, from the viewpoint of grain refinement, six ECAP passes at 593 K should be the optimal ECAP parameter for Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composite. On the other hand, ECAP temperature also obviously impacts the microstructure of 1β-TCP/Mg-3Zn-0.2Ca composites. Figure 5 shows the microstructure of the as-extruded 1β-TCP/Mg-3Zn-0.2Ca composites processed by 4 ECAP passes at different temperatures (553 K, 573 K, 593 K, and 613 K). When the 1β-TCP/Mg-3Zn-0.2Ca composite was processed at 553 K, coarse grains and small grains coexisted to form a heterogeneous microstructure, as shown in Figure 5a. With the further increase in ECAP temperature, the volume fraction of refined grains increased. When the ECAP temperature was up to 593 K, the C-4 P-593 K displayed the most uniform microstructure. According to Figure 5d, when the ECAP temperature was up to 613 K, abnormal grain growth was observed due to the excessive temperature. The average grain sizes of C-4 P-553 K, C-4 P-573 K, C-4 P-573 K, and C-4 P-613 K are approximately 9.8 ± 0.9 mm, 10.3 ± 0.8 mm, 10.8 ± 0.5 mm and 11.1 ± 0.7 mm, respectively. The increase in ECAP temperature was beneficial to the dislocation slip and plastic deformation of the 1β-TCP/Mg-3Zn-0.2Ca composite, which means that the uniform microstructure could be achieved. On the other hand, the increase in ECAP temperature accelerated grain growth of 1β-TCP/Mg-3Zn-0.2Ca composites [40], which was adverse to grain refinement. Hence, due to the competitive effect of grain growth and recrystallization, the C-4 P-593 K displayed the most uniform microstructure among all specimens with different ECAP temperatures.
From the viewpoint of microstructure, 593 K should be the optimal temperature of the ECAP process for the 1β-TCP/Mg-3Zn-0.2Ca composites. On the other hand, ECAP temperature also obviously impacts the microstructure of 1β-TCP/Mg-3Zn-0.2Ca composites. Figure 5 shows the microstructure of the as-extruded 1β-TCP/Mg-3Zn-0.2Ca composites processed by 4 ECAP passes at different temperatures (553 K, 573 K, 593 K, and 613 K). When the 1β-TCP/Mg-3Zn-0.2Ca composite was processed at 553 K, coarse grains and small grains coexisted to form a heterogeneous microstructure, as shown in Figure 5a. With the further increase in ECAP temperature, the volume fraction of refined grains increased. When the ECAP temperature was up to 593 K, the C-4 P-593 K displayed the most uniform microstructure. According to Figure 5d, when the ECAP temperature was up to 613 K, abnormal grain growth was observed due to the excessive temperature. The average grain sizes of C-4 P-553 K, C-4 P-573 K, C-4 P-573 K, and C-4 P-613 K are approximately 9.8 ± 0.9 mm, 10.3 ± 0.8 mm, 10.8 ± 0.5 mm and 11.1 ± 0.7 mm, respectively. The increase in ECAP temperature was beneficial to the dislocation slip and plastic deformation of the 1β-TCP/Mg-3Zn-0.2Ca composite, which means that the uniform microstructure could be achieved. On the other hand, the increase in ECAP temperature accelerated grain growth of 1β-TCP/Mg-3Zn-0.2Ca composites [40], which was adverse to grain refinement. Hence, due to the competitive effect of grain growth and recrystallization, the C-4 P-593 K displayed the most uniform microstructure among all specimens with different ECAP temperatures.  Figure 6 provides the inverse pole figures, the misorientation angle distributions, and the grain size distributions of the C-as-extruded and the C-6 P-593K specimens. According to Figure. 6a and 6b, a heterogeneous microstructure was formed in an as-extruded composite, including the coarse grains whose grain sizes are larger than 60 From the viewpoint of microstructure, 593 K should be the optimal temperature of the ECAP process for the 1β-TCP/Mg-3Zn-0.2Ca composites. Figure 6 provides the inverse pole figures, the misorientation angle distributions, and the grain size distributions of the C-as-extruded and the C-6 P-593K specimens. According to Figure 6a,b, a heterogeneous microstructure was formed in an as-extruded composite, including the coarse grains whose grain sizes are larger than 60 um and fine grains whose grain sizes are under 5 um. After six passes of ECAP processing, almost all elongated grains in as-extruded specimens were transformed into fine recrystallized grains, reducing the average grain size. Grain size distribution results reveal that a relatively narrow distribution was achieved in C-6 P-593 K compared with as-extruded counterparts, which means the ECAP process contributes to the homogenization of microstructure, as shown in Figure 6d,e. This is mainly due to ECAP-induced dynamic recrystallization, and the redistribution of the second phase and β-TCP particles. As shown in Figure 6c,f, the proportion of high angle grain boundaries (HAGBs) of the C-6 P-593 K increases compared with that of the C-as-extruded 1β-TCP/Mg-3Zn-0.2Ca composite, which is a key character of dynamic recrystallization. Figure 7 shows the (0 0 0 1) pole Figures of the C-as-extruded and the C-6 P-593 K. The C-as-extruded possessed a single (0 0 0 1) texture, in which the (0 0 0 1) basal plane paralleled to ED-axis. The texture is typical for magnesium alloys processed by hot extrusion [33]. The texture of the C-6 P-593 K changed and was divided into two kinds of texture. One, which corresponds to area A in Figure 7b, indicates the (0 0 0 1) basal plane parallel to the ND plane. The other one, which corresponds to area B in Figure 7b, indicates that the angle between the normal direction of (0 0 0 1) basal plane and the ED-axis is 60 degrees. It is supposed that the substantial shear stress, which also possessed an angle of 60 degrees with the ED-axis, contributes to the formation of the second texture type during the ECAP process.  Figure 7 shows the (0 0 0 1) pole Figures of the C-as-extruded and the C-6 P-593 K. The C-as-extruded possessed a single (0 0 0 1) texture, in which the (0 0 0 1) basal plane paralleled to ED-axis. The texture is typical for magnesium alloys processed by hot extrusion [33]. The texture of the C-6 P-593 K changed and was divided into two kinds of texture. One, which corresponds to area A in Figure 7b, indicates the (0 0 0 1) basal plane parallel to the ND plane. The other one, which corresponds to area B in Figure 7b, indicates that the angle between the normal direction of (0 0 0 1) basal plane and the ED-axis trusion [33]. The texture of the C-6 P-593 K changed and was divided into two kinds of texture. One, which corresponds to area A in Figure 7b, indicates the (0 0 0 1) basal plane parallel to the ND plane. The other one, which corresponds to area B in Figure 7b, indicates that the angle between the normal direction of (0 0 0 1) basal plane and the ED-axis is 60 degrees. It is supposed that the substantial shear stress, which also possessed an angle of 60 degrees with the ED-axis, contributes to the formation of the second texture type during the ECAP process.   Figure 8 shows the compressive stress-strain curves of Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca with different ECAP process parameters. The results of the compressive test (UCS, CYS, and strain at failure (σ)) are shown in Table 2.   For Mg-3Zn-0.2Ca alloy, it was found that both strength and strain at failure of Mg-3Zn-0.2Ca alloy increase first and then decrease with the increasing ECAP passes. The A-6 P-593K exhibits the optimized mechanical performance in Mg-3Zn-0.2Ca alloy groups, which includes the UCS of 388.4 ± 7.3 MPa, and the CYS of 94.5 ± 7.2 MPa, and the strain at failure of 14.3 ± 1.5%. For 1β-TCP/Mg-3Zn-0.2Ca composites groups, the C-6 P-593K specimens showed the best comprehensively mechanical performance, including the UCS of 411.7 ± 6.8 MPa, the CYS of 123.6 ± 5.5 MPa and the strain at failure of 10.1 ± 2.8%. Thus, the optimal ECAP parameter for Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composite should be six ECAP passes at 593 K from the viewpoint of mechanical properties. It is worth mentioning that 1β-TCP/Mg-3Zn-0.2Ca composites Figure 8. Compressive stress-strain curves of (a) Mg-3Zn-0.2Ca alloy and (b) 1β-TCP/Mg-3Zn-0.2Ca composites with different ECAP process parameters. Table 2. Summary of mechanical properties of the experimental specimens.
Thus, the optimal ECAP parameter for Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composite should be six ECAP passes at 593 K from the viewpoint of mechanical properties. It is worth mentioning that 1β-TCP/Mg-3Zn-0.2Ca composites possess higher strength but lower strain at failure than the Mg-3Zn-0.2Ca alloy components processed by the same ECAP parameters. This may be due to the improved dispersion properties of β-TCP particles induced by the severe shear force during the ECAP process. Furthermore, β-TCP particles could facilitate nucleation during the ECAP process, thereby contributing to the recrystallization and grain refinement in the ECAP-process composites. Above all, the improved mechanical performance of ECAP-processed 1β-TCP/Mg-3Zn-0.2Ca composites is associated with enhanced second phase strengthening and grain refinement strengthening caused by uniformly dispersed β-TCP particles and β-TCP-induced recrystallization.
The enhanced mechanical properties of the ECAP-processed 1β-TCP/Mg-3Zn-0.2Ca composites could be explained as follows. According to the Hall-Petch relationship, grain refinement induced by the ECAP process has a positive effect on both strength and ductility. Moreover, the ECAP process also led to the redistribution of the second phases, as shown in Figure 4, which enhanced the role of the second phases in hindering dislocation migration and grain boundary migration, resulting in a increased second phase strengthening. Similarly, the dispersion properties of the β-TCP particle was improved by the ECAP-induced redistribution, which could contribute to an increased strengthening efficiency of β-TCP in the ECAP-processed 1β-TCP/Mg-3Zn-0.2Ca composites. Furthermore, texture plays a vital role in the mechanical properties of materials with strong textures [41]. According to the Figure 7, the single texture with the (0 0 0 1) basal plane paralleled to ED-axis was confirmed in the C-as-extruded. This kind of texture is not conducive to basal slip during the compressive deformation, but the basal slip possessed the lowest activated energy in the slip system of magnesium alloy, thereby resulting in a relatively low ductility [42]. On the contrary, after the ECAP process, a new texture component was confirmed in Figure 7b. The new texture component indicates the normal direction of the (0 0 0 1) basal plane inclines to the ED-axis by about 60 degrees. Such a texture component facilitated the activation of the basal slip system resulting in improved ductility.
Consequently, the evolution in the mechanical properties of Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composites with multiple ECAP passes depend on the synergistic effect factors, including grain refinement, second phases, β-TCP particles, and texture transformation. With the further increasing ECAP passes, the effect of grain refinement and second phases redistribution had a much higher impact on the resulting mechanical properties than the texture transformation, resulting in the increase in both strength and strain at failure of Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca within six (including six) ECAP passes and reached peak values at six passes. However, the ECAP process resulted in excessive grain growth of the C-8 P-593K, which deteriorated the comprehensive mechanical properties of Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composites. Figure 9 shows the potentiodynamic polarization curves of Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composite with different ECAP passes (0, 2, 4, 6, and 8) at 593 K, and the corrosion parameters of the polarization test are shown in Table 3.

Electrochemical Analysis
According to Table 3, the A-2 P-593 K possesses higher corrosion resistance than that of A-as-extruded, while the corrosion resistance of Mg-3Zn-0.2Ca alloy decreases with the further increased ECAP passes. Similarly, according to the electrochemical analysis results of the ECAP-processed 1β-TCP/Mg-3Zn-0.2Ca composites, the C-2 P-593 K exhibited enhanced corrosion resistance compared to the C-as-extruded. Nevertheless, when the ECAP passes are higher than two, the corrosion resistance was trending downward.
hensive mechanical properties of Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composites. Figure 9 shows the potentiodynamic polarization curves of Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composite with different ECAP passes (0, 2, 4, 6, and 8) at 593 K, and the corrosion parameters of the polarization test are shown in Table 3.    The evolution in corrosion resistance of Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composite with multiple ECAP passes could be explained from two aspects. The grain refinement has a significant impact on corrosion resistance. A protective surface film composed of MgO formed during corrosion, which could improve the corrosion resistance of matrix alloy. However, the free volume mismatch between surface film and metal matrix caused tensile stress in the surface film, thus increasing its cracking tendency. The grain refinement reduced the mismatch between the protective surface film and the underlying matrix alloy, thereby reducing the crack potential. Moreover, the ECAP process also contributes to the redistribution of the second phases and β-TCP particles in the experimental materials, which ameliorate the homogeneity and corrosion resistance of ECAP-processed materials [43]. A similar phenomenon was reported by several studies about the ECAP-processed Mg-based alloys [44,45].

Electrochemical Analysis
Therefore, the evolution in corrosion resistance of Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composite with multiple ECAP passes depends on the synergistic effect of the above three factors (grain size, redistribution of second phases and β-TCP particles, lattice defects). For the specimens processed by ECAP with a few passes (less than two passes), the positive effect of grain refinement and redistribution of the second phase β-TCP particles on the corrosion resistance were dominant factors and exceeded the opposite effect of the increased residual stress. With the increase in ECAP passes, the negative effect of the lattice defects became the dominant factor for the corrosion behavior of experimental specimens. Hence, thus, the more ECAP passes correlate with the lower corrosion resistance.

Cell Biocompatibility
From the viewpoint of mechanical performance, A-6 P-593 K and C-6 P-593 K samples with optimized mechanical properties in each group were selected for cell biocompatibility evaluation. Figure 10 shows the relative proliferation rate (RGR) of mouse fibroblasts (L-929 cells) cultured in extracts of the A-6 P-593 K and the C-6 P-593 K specimens for 1 day, 3 days, and 5 days. The result shows that A-6 P-593 K and C-6 P-593 K exhibit non-cytotoxicity to L-929 cells. Moreover, it is worth mentioning that C-6 P-593 K exhibited higher RGR compared to the A-6 P-593 K with the same incubation times, indicating the positive effect of β-TCP addition on the cytocompatibility of the experimental composites.
Metals 2022, 12, x FOR PEER REVIEW 14 of 17 Figure 10. The relative growth rate (RGR) of L929 cells cultured with extracts of A-6 P-593 K and C-6 P-593 K specimens.

Conclusions
In summary, this study reported the microstructure, mechanical performance, corrosion behavior, and cytotoxicity of Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composites processed by hot extrusion combined with subsequent multi-pass ECAP process. The main findings are summarized as follows: (1) From the viewpoint of grain refinement, hot extrusion combined with six ECAP passes at 593 K is the optimized route, resulting in an average grain size of 13.4 ± 0.6 μm and 9.6 ± 0.3 μm in Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composite, respectively. The redistribution of second phases, as well as β-TCP particles induced by the ECAP process, were also confirmed.
(2) Due to the synergistic effect of grain refinement, redistribution of second phases and β-TCP particles, and texture transformation, C-6 P-593 K displays the best mechanical properties, including the UCS of 405.3 ± 7.4 MPa, UYS of 118.9 ± 7.0 MPa and strain at failure of 10.1 ± 2.8%.
(3) For Mg-3Zn-0.2Ca alloy, the A-2 P-593 K exhibited the lowest degradation rate of 0,895 mm·Y −1 . The improvement of corrosion resistance is associated with the grain refinement and the uniform distribution of the second phases after the ECAP process.
(4) The cell biocompatibility result shows that the A-6 P-593 K and C-6 P-593 K had no significant cytotoxicity to L929 cells, and the addition of β-TCP could improve the cell biocompatibility. C-6 P-593 K exhibited great potential for applications in the fields of bone repair and bone replacement. Relative growth rate, RGR/% Control A-6 P-593 K C-6 P-593 K Culture time, t/d Figure 10. The relative growth rate (RGR) of L929 cells cultured with extracts of A-6 P-593 K and C-6 P-593 K specimens.

Conclusions
In summary, this study reported the microstructure, mechanical performance, corrosion behavior, and cytotoxicity of Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composites processed by hot extrusion combined with subsequent multi-pass ECAP process. The main findings are summarized as follows: (1) From the viewpoint of grain refinement, hot extrusion combined with six ECAP passes at 593 K is the optimized route, resulting in an average grain size of 13.4 ± 0.6 µm and 9.6 ± 0.3 µm in Mg-3Zn-0.2Ca alloy and 1β-TCP/Mg-3Zn-0.2Ca composite, respectively. The redistribution of second phases, as well as β-TCP particles induced by the ECAP process, were also confirmed.
(2) Due to the synergistic effect of grain refinement, redistribution of second phases and β-TCP particles, and texture transformation, C-6 P-593 K displays the best mechanical properties, including the UCS of 405.3 ± 7.4 MPa, UYS of 118.9 ± 7.0 MPa and strain at failure of 10.1 ± 2.8%.
(3) For Mg-3Zn-0.2Ca alloy, the A-2 P-593 K exhibited the lowest degradation rate of 0.895 mm·Y −1 . The improvement of corrosion resistance is associated with the grain refinement and the uniform distribution of the second phases after the ECAP process.
(4) The cell biocompatibility result shows that the A-6 P-593 K and C-6 P-593 K had no significant cytotoxicity to L929 cells, and the addition of β-TCP could improve the cell biocompatibility. C-6 P-593 K exhibited great potential for applications in the fields of bone repair and bone replacement.