The Corrosion Performance and Mechanical Properties of Mg-Zn Based Alloys—A Review

Magnesium alloys have shown great potential for applications as both structural and biomedical materials due to their high strength-to-weight ratio and good biodegradability and biocompatibility, respectively. Among them, Mg-Zn based alloys are attracting increasing interest for both applications. As such, this article provides a review of the corrosion performance and mechanical properties of Mg-Zn based alloys, including the influence of environment and processing on both of them. The strategies for tailoring corrosion resistance and/or mechanical properties by microstructure adjustment and surface treatment are discussed.


Introduction
Magnesium (Mg) alloys have drawn increasing attention during the past two decades for applications in the automotive, aerospace, and electronics industries due to their high specific strength and stiffness, low density, good castability, and machinability [1][2][3][4]. In addition, they have also been considered as attractive candidates for application as biodegradable implant materials because of their mechanical properties that are similar to those of natural bone, good biodegradability, and inherent biocompatibility [5][6][7][8][9]. To expand the application scope of Mg alloys, alloying is one of the most effective strategies for the further enhancement of mechanical properties and corrosion performance [10].
Magnesium-aluminum (Mg-Al) based alloys are the most widely studied and used Mg alloys due to their excellent castability, reasonable mechanical properties at room temperature, and good corrosion resistance [11]. However, the mechanical properties of Mg-Al based alloys significantly decrease as temperature rises above 130 °C, above which softening of the β-Mg17Al12 phases, precipitated at grain boundaries, results in poor creep resistance [12,13]. Additionally, Al has been reported to be a neurotoxicant, and the accumulation of Al can induce various neurological disorders during the degradation of Mg-Al alloys in the human body [14]. Therefore, the development of Al-free Mg alloys is desired.
Zinc (Zn) is the second most common alloying element for commercial Mg alloys [15,16]. Zn alloying additions can lead to a higher free corrosion potential of the MgZn alloy compared to that of pure Mg, reduce the corrosion rate, and improve the mechanical properties through a solid solution hardening mechanism [17,18]. Moreover, in comparison with Mg-Al series alloys, the presence of heat-resistant intermetallics results in the much higher creep resistance and the better tensile properties of Mg-Zn-Al alloys at elevated temperatures [19,20]. Furthermore, Zn is one of the essential trace elements for the human body and is crucial for many biological functions, such as the immune system and sense of smell and taste [21]. Thus, the Mg-Zn system shows great potential as a low-cost Mg alloy alternative with high strength in industrial applications, and has attracted increasing interest for medical applications.
Motivated by the new trend of biomedical applications for Mg-Zn based alloys, the objective of this article is to provide a review of the corrosion performance and mechanical properties of Mg-Zn based alloys developed for both industrial and biomedical applications. Potentials or relative Volta potentials of some secondary phases in Mg-Zn based alloys and the corrosion rates and mechanical properties of Mg-Zn based alloys summarized from publications are compared in Tables A1 and A2 in the Appendix. The influence of environment and processing on the corrosion behavior and mechanical properties of Mg-Zn based alloys are also reviewed. The strategies for improving the properties of Mg-Zn based alloys are discussed based on microstructure design and surface modification. Finally, typical applications are shortly summarized.

Mg-Zn System
The maximum solid solubility of Zn in Mg is considered to be relatively high at high temperatures [22,23], e.g., 2.5 at.% (6.2 wt.%) at 325 °C, as indicated by the binary Mg-Zn phase diagram proposed by Okamoto et al. [24] and shown in Figure 1. However, with decreasing temperature, only 0.6 at.% (1.6 wt.%) Zn remains soluble in the Mg matrix and contributes to solid solution strengthening [22,25,26]. Excess Zn will form intermetallic phases. In the Mg-Zn system, five intermetallic phases, namely Mg51Zn20 (previously denoted as Mg7Zn3 [27]), MgZn, Mg2Zn3 (previously denoted as Mg4Zn7 [28,29]), MgZn2, and Mg2Zn11, exist. The primary Mg7Zn3 phase has a body-centered orthorhombic structure [30]. The crystal structures of MgZn, MgZn2, and Mg2Zn11 are base-centered monoclinic [31], hexagonal [32], and cubic [33], respectively. However, the structure of the Mg4Zn7 phase is under debate, assumed to be a triclinic structure by Gallot and Graf [34], but reported to be a base-centered monoclinic structure by Gao and Nie [31]. It is believed that the MgxZny intermetallics can provide a pronounced age hardening effect [35][36][37], but they are detrimental to the corrosion resistance of the alloy due to accelerated cathodic reaction rates [38][39][40].  [24] (with permission from Springer Nature and Copyright Clearance Center).

Influence of Zn Content on the Corrosion and Mechanical Properties of Binary Mg-Zn Alloys
Addition of Zn as an alloying element can refine the grain size and improve the corrosion resistance and mechanical properties of Mg [10,41]. Compared with pure Mg and high-purity AZ31 alloy, almost one order of magnitude improvement of corrosion resistance can be achieved by small additions of Zn (1 wt.%) [10,42]. Since the electrode potential of Zn (-0.762 V vs. Standard Hydrogen Electrode) is higher than that of Mg (-2.372 vs. Standard Hydrogen Electrode), and Zn is known to have a high hydrogen evolution over-potential, the improved corrosion resistance is presumably attributed to the reduced hydrogen evolution kinetics by Zn in solid solution [42]. It was also suggested that Zn can increase the tolerance limits of impurities (iron (Fe), copper (Cu), nickel, and cobalt) in Mg, which enhances the corrosion resistance of Mg via weakening of the galvanic corrosion between impurities and Mg [1,43,44]. In addition, Gu et al. [45] have reported that the corrosion resistance of as-cast Mg1Zn (in wt.%) alloy is higher compared to several other binary Mg alloys alloyed with the same amount (Ag, Al, In, Mn, Si, Sn, Y, and Zr). Generally, the corrosion resistance and mechanical properties of Mg-Zn binary alloys strongly depend on Zn content. When the added amount of Zn is lower than 5 wt.%, an increase in Zn content enhances the general and localized corrosion resistance of as-cast Mg-Zn alloys in simulated body fluid (SBF) at 37 °C, which is due to the refinement of grain size and the facilitated formation of protective surface film induced by Zn [10,41,46]. However, much higher levels of Zn (up to 7 wt.%) can result in the formation of a significant amount of secondary phases. As a result, severe localized corrosion occurs due to the galvanic effect between intermetallics and matrix, and, consequently, a decrease of the overall corrosion resistance is found [10,47,48]. Opposite results have been reported by Kubàsek et al [42,49], showing that the corrosion rates of Mg-Zn binary alloys increase with increasing Zn amount in the range of 0-6 wt.% when testing is performed in either simple sodium chloride (NaCl) solution or simulated body fluid. This finding is ascribed to the increased galvanic effect due to the increase in volume fraction of secondary phases.
The yield (YS) and ultimate tensile strength (UTS) of binary Mg-Zn alloys increase with increasing Zn content when Zn wt.% is in the range of 0-4 wt.% [10,22,46,50]. This is a combined result of fine grain strengthening [51][52][53], solid solution strengthening [25,26,54], and precipitation hardening effects [54,55]. Nevertheless, when alloying with higher amounts of Zn, plenty of secondary phases will form at the grain boundaries. They will result in the increased dislocation density and act as new crack sources [10]. Consequently, the tensile strength of the alloy can be significantly reduced.
Binary Mg-Zn alloys with much higher amounts of Zn (>10 wt.%) produced by using powder metallurgy have only been studied by Yang et al. [55]. High Zn content (>10 wt.%) causes the formation of large-size secondary phases and undissolved Zn particles, which results in severe pitting and localized corrosion due to the formation of galvanic cells, thus reducing the corrosion resistance of the alloys (corrosion current density: 16.9 µA cm -2 for Mg6Zn alloy, 54.2 µA cm -2 for Mg14.5Zn alloy, 80.3 µA cm -2 for Mg20.3Zn alloy, 132.2 µA cm -2 for Mg40.3Zn alloy). The compression strength increases with increasing Zn content, but reaches the highest value when Zn content reaches 14.5 wt.% (373.5 MPa for M6Zn alloy, 396.5 MPa for Mg14.5Zn alloy, 371.7 MPa for Mg20.3Zn alloy, 354.6 MPa for Mg40.3Zn alloy). Above this value, further addition of Zn is detrimental because large-size secondary phases and Zn particles formed in the alloy could act as crack sources [56].
Considering the aforementioned dependence of a good combination of mechanical properties and corrosion resistance on Zn content, Zhang et al. [22] have suggested that the Zn content of MgZn alloys should be limited to 4 wt.%. Figure 2 depicts the influence of Zn content on the corrosion rate and tensile properties of Mg-Zn binary alloys, summarizing the data provided in the literature. This shows that, as Zn content varies, the corrosion rate of MgZn alloy changes differently between different reports ( Figure 2). One of the reasons might be the fact that other influences besides the Zn content do affect the corrosion rate. The most obvious is the electrolyte composition, which was not the same in all of the experiments (see Table A2 in Appendix). Two other important influence factors are the microstructure of the alloy (e.g., grain size and distribution of secondary phases) and the impurity levels of the tested alloys. Generally, more secondary phases and more impurities would lead to faster corrosion and severer localized corrosion of the alloy. In contrast, an average tendency can be observed for the influence of Zn content on tensile properties. The yield and ultimate tensile strength increase while elongation to fracture decreases with increasing Zn content in the range of 0-4 wt.%. It can be noted that studies about binary Mg-Zn alloys are still quite limited and many inconsistencies exist, suggesting that more systematic studies are needed.

Ternary Mg-Zn-X Alloys
Superior combinations of corrosion resistance and mechanical properties of as-cast Mg-Zn based alloys can often be improved by the addition of a third alloying element or microstructure modification via heat treatment or mechanical processing.

Influence of Microstructure on the Corrosion and Mechanical Properties
Microstructural features, such as grain size, volume fraction, and distribution of secondary phases, generally play an important role in the corrosion behavior of Mg alloys. It is widely accepted that refined grain size is beneficial to the corrosion resistance of Mg alloys in neutral and alkaline sodium chloride (NaCl) corrosive electrolytes [62][63][64][65][66][67]. However, large residual stresses, which can be introduced during grain refinement, can be detrimental to the corrosion resistance of Mg [63,64]. Song et al. have reported that the influence of secondary phases on corrosion behavior of Mg alloys depends on the amount and distribution of the secondary phases [1,68,69]. They can either act as a galvanic cathode to accelerate the corrosion rate, or as a continuous barrier with higher corrosion resistance (compared with the Mg matrix) when exposed to the corrosive environment after the dissolution of the top Mg matrix (Figure 3). Lu et al. [70] have studied the combined influence of secondary phases and grain size on the corrosion of as-cast and heat-treated Mg3Zn0.3Ca (in wt.%) alloy. The results ( Figure 4) indicate that both the volume fraction of secondary phases and grain size are important for controlling the corrosion of the alloy. When the influence of grain size is dominating the corrosion process of the alloy, the large grain size would lead to the increase of the corrosion rate. Vice versa, when the secondary phase volume fraction is high, it overrules the beneficial effect of fine grain size.

Mg-Zn-Ca Alloys
Due to their good biocompatibility, Mg-Zn-Ca alloys have attracted attention in research of Mg alloys for biomedical devices. Calcium (Ca) behaves as an effective grain refiner for MgZn alloys [5,72,73]. The phase formation depends on the contents of Zn and Ca (especially the Zn/Ca ratio), and affects the corrosion performance of the alloy. It has been pointed out that eutectic α-Mg + Mg2Ca + Ca2Mg6Zn3 phase precipitates when Zn/Ca atomic ratio is lower than 1.2, while α-Mg + Ca2Mg6Zn3 eutectic phase is formed when Zn/Ca atomic ratio is higher than 1.2 [5,74]. A new phase Ca2Mg5Zn13 has also been reported in the case of Mg6Zn1Ca (in wt.%) alloy [23]. Mg2Ca and Ca2Mg6Zn3 phases usually distribute along the grain boundary and form interdendritic interstices in as-cast Mg-Zn-Ca alloys. A segregation of Zn may occur in the vicinity of Mg2Ca phase, consequently protecting this area (Mg2Ca phase) against corrosion [75,76]. Mg2Ca is more active than α-Mg matrix, while Ca2Mg6Zn3 is nobler than α-Mg [75,77]. Therefore, Ca2Mg6Zn3 phase with discontinuous distribution can act as a cathode site, accelerating galvanic corrosion of the matrix, but can act as a barrier when the Ca2Mg6Zn3 phase is continuously distributed in the microstructure, e.g., along the grain boundary [75].
When the content of Ca is lower than 1 wt.%, alloying additions of Ca improve the corrosion resistance of Mg-Zn alloys by forming compounds with elemental impurities and thus purifying the melts, while more addition of Ca leads to the formation of secondary phases and thus results in increased pitting corrosion with increasing overall corrosion rate [5,22,59,73,75]. It has been reported that Ca2Mg6Zn3 and Ca2Mg5Zn13 phases have strengthening effects [16,78,79]. Thereby, grain refinement and precipitate strengthening would contribute to an improvement in mechanical properties of Mg-Zn-Ca alloys. Nevertheless, with increasing Ca content, Mg2Ca and Ca2Mg6Zn3 phases continuously precipitate along the grain boundary, which is detrimental to the tensile properties of the alloy, since Mg2Ca is a brittle phase and can produce crack sources that can lead to brittle failure [5,22,80,81].
Increasing the Zn content promotes the corrosion of Mg-Zn-Ca alloys because, as aforementioned, nobler Ca2Mg6Zn3 secondary phases form and thus accelerate galvanic corrosion [7,23,82]. However, the yield and tensile strength of the alloy can be enhanced with the increase of the Zn content when Zn concentration is in the range of 0-4.0 wt.%, which is due to the solid solution hardening and precipitation strengthening effects [7,23]. Nevertheless, more Zn addition would result in decline of mechanical properties of Mg-Zn-Ca alloys and change the fracture type from ductile to brittle [23].
On the basis of the above discussion, the corrosion and mechanical properties of Mg-Zn-Ca alloys can be optimized by adjusting the content of Zn and Ca, influencing the microstructure. Geng et al. studied Mg4ZnCa (in wt.%) alloys with micro-alloying addition of 0.2 wt.% Ca (corrosion and mechanical properties) [6] and 0.5 wt.% Ca (only mechanical property) [16], respectively. The corrosion properties of Mg4Zn0.2Ca alloys are similar to those of high-purity Mg in SBF. In addition, both of these two alloys exhibit good mechanical properties, especially Mg4Zn0.5Ca, which shows a good balance between tensile strength and ductility. The tensile strength and elongation to fracture are 185 MPa and 12.5% for Mg4Zn0.2Ca alloy and 211 MPa and 17% for Mg4Zn0.5Ca alloy, respectively [6,16]. Furthermore, tensile strength and elongation to fracture can be further enhanced by extrusion. The integrity of the alloy with 0.2Ca is still given even after immersion for 30 days in SBF, and the mechanical properties are still high enough (ultimate strength: 220 MPa, yield strength: 160 MPa, ductility: 8.3) for bone fixture applications.

Mg-Zn-Y Alloys
The solubility of Zn in the Mg matrix is greatly decreased when Y is incorporated as an alloying addition. This decrease is believed to be a result of the interaction between Zn and Y [53]. Ternary MgZnY phases would precipitate firstly at grain boundaries because of the higher eutectic temperature of ternary phases compared with that of binary phases [83]. The phase constituency in Mg-Zn-Y alloys is strongly dependent on the weight ratio of Zn to Y for both low (2 wt.%) and high (≥ 3 wt.%) Zn-containing systems [53,83,84]. When the Zn/Y weight ratio is lower than 1.5, Mg12ZnY phase precipitates (X-phase or LPSO phase-long period stacking ordered phase). When the Zn/Y weight ratio is 1.5-2, Mg3Zn3Y2 phase (W-phase, dendritic phase) forms in the interdendritic region. When the Zn/Y ratio increases to 2-2.5, Mg3Zn6Y phase (I-phase, icosahedral quasicrystal phase) begins to form and co-exists with W-phase. When the Zn/Y ratio is 5-7, only I-phase forms in the Mg-Zn-Y alloy. However, with further increase of the Zn/Y ratio (~10), the composition is close to that of the binary Mg-Zn system; thus, mainly binary MgZn phases form [84,85]. Those secondary phases are electrochemically nobler than the α-Mg matrix and can thus be effective cathode sites when the alloys are exposed to corrosive environments, resulting in pitting corrosion at the secondary phase/α-Mg interface. With similar sizes and volume fractions of intermetallic particles, MgZnY alloys with a single secondary phase (e.g. Mg3Zn6Y I-phase) exhibit better corrosion resistance than those with two secondary phases (e.g. Mg3Zn3Y2 W-phase and Mg3Zn6Y I-phase) [53,86]. Generally, with the increase of volume fraction of secondary phases with discontinuous distribution, the corrosion properties of the alloys would deteriorate, since the effective cathode area increases [87][88][89]. However, continuous distribution and moderate volume fraction of secondary phases can hinder the corrosion propagation [87,88]. For example, Mg97Zn1Y2 alloy (continuous distribution, volume fraction of secondary phases: 30.4%) has higher corrosion resistance compared with Mg98.5Zn0.5Y1 alloy (discontinuous distribution, volume fraction of secondary phases: 11.5%), Mg94Zn2Y4 alloy (continuous distribution, volume fraction of secondary phases: 55.4%) and Mg88Zn4Y8 alloy (continuous distribution, volume fraction of secondary phases:, 63.2%) [87]. Li et al. [88] have studied the influence of volume fraction of LPSO phases on the corrosion resistance of Mg0.9Zn1.6Y, Mg2.1Zn5.2Y, and Mg3.1Zn7.6Y (all in wt.%) alloys in 0.1 M NaCl solution. It has been revealed that the Volta potential difference at the LPSO phase/α-Mg interface could be up to 250 mV (Figure 5b). As a result, severe micro-galvanic corrosion preferentially occurs at the LPSO/α-Mg interfaces. Then, the corrosion progresses along the boundary of LPSO phases instead of in the α-Mg matrix (Figures 5c and 5d). A new phase (currently cannot be indexed) has been found in the Raman spectra of the corrosion products forming on LPSO phases besides magnesium hydroxide (Mg(OH)2).
The mechanical properties of Mg-Zn-Y alloys also largely rely on the phase constituency. The presence of stable I-phase is beneficial for the mechanical properties of Mg-Zn-Y alloys, especially at elevated temperatures [90][91][92], due to the low interfacial energy between the I-phase and α-Mg matrix. The yield and tensile strengths of the alloys increase with increasing volume fraction of I-phase [84,90,93]. W-phase also contributes to the strength, but not as effectively as I-phase. However, W-phase has a better softening effect than I-phase, which is beneficial for the ductility of the alloys [53,94]. Many reports have demonstrated that the LPSO phases are effective for strengthening of Mg alloys [88,[95][96][97][98][99][100][101][102]. Kawamura et al. have firstly reported a Mg97Zn1Y (in at.%) alloy with high yield strength of 610 MPa, usable ductility (5%), and a yield strength of 300 MPa at 473 K due to the presence of LPSO phase [103]. The strengthening mechanisms of LPSO phase were attributed to four mechanisms: 1) increased critical resolved shear stress (CRSS) of the basal plane, 2) activated non-basal slip, 3) kinking bands on LPSO formed during the deformation process, and 4) a coherent LPSO/Mg interface along the basal and prismatic planes [104][105][106]. However, the improvement of strength is slight (the ultimate strength is only increased from 141 MPa to 148 MPa) when the volume fraction of LPSO phases exceeds 20.3%, and even declines with much higher LPSO phase concentration because of the large length and thickness of the LPSO phases [88,96]. The influences of different phases on the mechanical properties of Mg-Zn-Y alloys are summarized in Table 1. In addition to the abovementioned influence of phase constituency, conditions of the alloys should also be taken into serious consideration during the optimization of mechanical properties.

Mg-Zn-Mn Alloys
It is well known that manganese (Mn) is effective in combining with some heavy metal impurities, for example, Fe, thereby decreasing the galvanic corrosion between impurities and the Mg matrix [44]. When Mn is micro-alloyed (Mn wt.% < 1.0 wt.%) with the Mg-Zn system, Mn exists in a soluble state, and some binary MgZn phases precipitate with increase of Zn content in the range of 0 to 3 wt.% [109][110][111][112]. In addition, Mn in solid solution could stabilize the corrosion product layer formed when alloys corrode in corrosive environments by incorporating manganese as an oxide into the Mg(OH)2 layer [51,112]. Rosalbino et al. [51] have found that Mg2Zn0.2Mn exhibits the best corrosion resistance in comparison with Mg2Zn0.2Ca, Mg2Zn0.2Si (all in wt.%), and AZ91 alloys, revealing a four-fold increase in polarization resistance over that of AZ91 alloy after exposure for 168 h in Ringer's solution at 37 °C. Moreover, it has been reported by Abidin et al. [113] that the steady state corrosion rate of Mg2Zn0.2Mn (wt.%) alloy tested in Hank's solution at 37 °C is much lower than that of high-purity Mg. However, this has been ascribed to a lower content of Fe impurity particles in the Mg2Zn0.2Mn alloy. In addition to MgZn2 phase, additions of silicon (Si, 0-2 wt.%) into Mg-Zn-Mn alloys can induce the formation of Mg2Si and Mn5Si3 phases, but only Mg2Si phase influences the corrosion resistance of the alloy significantly [114]. The role of Mg2Si phase on the corrosion property of the alloy depends on its morphology and volume fraction in the microstructure ( Figure 6). When Mg2Si is in Chinese script type (0.5 wt.% Si), the corrosion rate is higher than that in the polygonal type. However, when the morphology of Mg2Si phase changes from Chinese script type to polygonal type with increasing Si content (1 wt.% and 2 wt.%), the corrosion resistance of the alloy increases again [114][115][116].

Representative Mg-Zn-RE and Mg-Zn-Zr Based Alloys
Generally, the addition of rare earth (RE) elements is beneficial for the mechanical properties of Mg alloys at elevated temperatures, and zirconium (Zr) is considered to behave as an effective grain refiner [117,118]. Thus, many efforts have been devoted to the development of RE/Zr-containing Mg alloys with fine grains. Representatively, Mg-Zn-RE (ZE) and Mg-Zn-Zr (ZK) alloys have been widely studied for automotive/aerospace and biomedical applications, especially commercial ZE41 and ZK60 alloys.
The corrosion resistance of ZE41 alloy is poor, with a corrosion rate twice as high as that of AZ91D and about thirteen times higher than that of pure Mg in 1 M NaCl solution [119]. Mg7Zn3RE (T-phase) and Mg12RE phases with nobler corrosion potential can precipitate along the grain boundaries, as well as inside of the grains, thereby inducing severe galvanic corrosion. As a result, the Mg matrix adjacent to the secondary phases will dissolve preferentially; then, pitting and intergranular corrosion can occur [118,120]. Caution should be paid during mass loss measurements because the T-phase could be dissolved by chromium-trioxide-based corrosion product removal solutions [121], thus leading to misleading results of corrosion rate. Furthermore, with micro-additions of Zr into ZE41 alloy, a so-called Zr-rich interaction zone can form within the grains and Zr-rich particles, together with some Zn and Fe precipitates can be found in those Zr-rich regions (shown in Figure 7) [121][122][123]. Consequently, the Zr-rich zones are also favorable sites for galvanic-driven localized corrosion attack (galvanic couple between the Zr-rich particles and the surrounding matrix). ZK60 alloy has wide engineering applications due to its good combination of tensile strength (≥ 250 MPa) and uniform elongation (≥ 15%) [124]. However, a piece of φ 28 mm × 5 mm specimen can degrade completely after 12 weeks of immersion in Hank's solution due to the galvanic corrosion between binary MgZn phases (Mg7Zn3 or MgZn2) and the Mg matrix [125]. This fast degradation rate cannot meet the requirements of biomedical applications [126]. By optimizing the content of Zn (Mg3Zn0.6Zr, in wt.%) in Mg-Zn-Zr alloy, the corrosion rate of Mg-Zn-Zr alloy can be comparable to that of WE43 alloy, which exhibits good corrosion behavior relative to AZ31, AZ91, and LAE42 alloys [125,127]. The further addition of Y and/or neodymium (Nd) into Mg-Zn-Zr alloys results in the precipitation of less noble T-phase and/or W-phase, and leads to the formation of more a compact corrosion product layer with higher corrosion resistance [128,129]. It is quite often observed for Mg alloys that the protective film is denser when the overall corrosion rate is low. Too much and fast Mg(OH)2 formation results in more compact films with coarse needle-like structures [130] The tensile properties of Mg-Zn-Zr alloys further alloyed with some rare earth elements (Nd, Gd, etc.) strongly depend on the microstructure of the alloy, i.e., the types and shapes of secondary phases [131][132][133][134][135]. For example, the addition of 2 wt.% Nd into Mg5Zn0.6Zr (in wt.%) alloy leads to the formation of a continuous network of intergranular phases, which significantly deteriorates the ultimate strength and elongation to fracture. However, the presence of discontinuous phases in Mg5Zn0.6Zr2Nd0.5Y (in wt.%) favors the ultimate strength and elongation to fracture [136]. Gadolinium (Gd) has a large solubility in Mg (23.49 wt.% at 548 ºC and 3.82 wt.% at 200 °C), but binary MgGd (Mg5Gd or Mg3Gd) phase and/or ternary MgZnGd (Mg3Gd2Zn3) phase precipitate in the matrix when Gd is alloyed to the Mg-Zn system. Similar to the Mg-Zn-Y alloys, the phase constitutions of Mg-Zn-Gd alloys also strongly depend on the Zn/Gd atomic ratio [137]. Small amounts of Gd addition (1 wt.%) are beneficial to the corrosion resistance of the Mg1Zn (in wt.%) system. Furthermore, both Gd and Zn enrich in the Mg matrix surrounding the secondary phases, improving the corrosion resistance of the matrix [49]. Yang et al. [138,139] have studied the influence of varying Gd content (from 0.5 up to 5 wt.%) on the microstructures and mechanical properties of the Mg4.5Zn (in wt.%) system. It has been revealed that Gd can refine the grains when additions are less than 2 wt.%, and further increase of Gd content would cause increasing grains. The strength of the Mg4.5Zn alloy is improved with increasing Gd addition up to 2 wt.%, but deteriorates with further increase of Gd content. This is related to the effects of grain size, solid solubility of Zn and/or Gd, and the size and volume fraction of secondary phases.

Other Mg-Zn-X Alloys
Addition of strontium (Sr) can also refine the microstructure of Mg-Zn alloys in spite of its limited solubility in Mg (0.11 wt.% at 858 K), and the refinement effect increases with increasing Sr content (0.1-1 wt.%) due to precipitate formation [140,141]. Although nobler secondary phases (compared with Mg matrix) precipitate, the corrosion behavior of Mg4Zn1Sr (in wt.%) is superior to that of pure Mg [142]. Nevertheless, the corrosion rate of Mg-Zn-Sr alloy increases with the increase of either Zn or Sr contents due to the enhanced galvanic corrosion induced by the increased amount of secondary phases. In comparison, the strength of Mg-Zn-Sr alloys is improved when both Zn (2-6 wt.%) and Sr (0.2-1 wt.%) contents increase, because of the refined grains and homogeneously distributed secondary phases [140,143].
A few studies have also been carried out to investigate the influence of alloying with silver (Ag), Cu, germanium (Ge), etc. on the corrosion and/or mechanical properties of an Mg-Zn system. For example, Ben-Hamu et al. [144] have revealed that addition of Ag into an Mg6Zn (in wt.%) system is harmful to the corrosion resistance due to the galvanic coupling between Ag17Mg54 phases and the α-Mg matrix, but contributes to the hardness of the alloy. Minor additions of Cu (< 0.5 wt.%) into the Mg2Zn (in wt.%) system improves the corrosion resistance and tensile properties of the alloy, but higher addition levels are detrimental to both corrosion and tensile performance due to the increased volume fraction of secondary phases [145]. Liu et al. [146] and our recent study [147] have demonstrated that a small addition of Ge (≤ 0.5 wt.%) significantly improves the corrosion resistance and tensile strength of binary Mg-Zn alloys. Furthermore, the presence of Ge suppresses the kinetics of cathodic reactions and facilitates the incorporation of alloying elements into the corrosion product layer, which are beneficial for the corrosion resistance of Mg-Zn alloys. However, the amount of related work is too small to give a comprehensive overview.

Influence of Environment on the Corrosion Behavior
In addition to the galvanic coupling between secondary phases and the Mg matrix (influence of microstructure), the corrosion behavior of Mg alloys is also governed by surface product layers and their protective properties, which are significantly affected by the environment. Fang et al. [73] have studied the corrosion behavior of Mg1ZnxCa (x = 0.2, 0.5, 0.8 1.0; wt.%) alloys in SBF and 3.5 wt.% NaCl solutions, which reveals that Mg1Zn0.8Ca alloy is the most corrosion-resistant alloy in 3.5 wt.% NaCl solution, while Mg1Zn0.5Ca alloy has the best corrosion properties in SBF, demonstrating that the corrosion properties of the alloys are significantly influenced by the compositions of the corrosive media.
It is well known that the surface film formed on Mg alloys in aqueous solution is mainly composed of magnesium oxide (MgO) and Mg(OH)2, and the film can be easily broken down by the penetration of chloride ions (Cl -) [1]. Zhao et al. [119] have studied the corrosion behavior of ZE41 alloy in NaCl solutions at different pH and Clconcentrations. The incubation period to the onset of corrosion is shortened as pH decreases or Clconcentration increases. This is because corrosion predominantly happens at the uncovered regions or defects of the surface film, and the fraction of film-free surface of the alloy increases with decreasing pH and increasing Clconcentration. This phenomenon has been further confirmed by Johnston et al. [148] and Taltavull et al. [149], who investigate the corrosion behavior of ZE41 alloy in Hanks' solution at different pH and Clconcentrations buffered with bicarbonate. A more stable and thicker Mg(OH)2 layer could be formed with increasing pH, since higher pH values mean that Mg(OH)2 can precipitate at a lower magnesium ion (Mg 2+ ) concentration. The breakdown of the surface layer induced by Clsubsequently accelerates galvanic corrosion between the constituencies in the alloy.
Very few corrosion studies of Mg-Zn based alloys have been carried out in other salt solutions, such as sodium sulfate (Na2SO4) and ammonium nitrate (NH4NO3) solutions. The limited studies show that the corrosion rate of ZE41 alloy in Na2SO4 solution increases with increased concentration of Na2SO4 at each pH, and decreases with increased pH at each concentration of Na2SO4 [150]. In mixed solutions of NaCl and Na2SO4, the increased concentration of Na2SO4 also leads to a higher corrosion rate of ZE41 alloy [117]. Mg2.03Zn0.83Mn (in wt.%) alloy is more corrosion-resistant in 0.01 M Na2SO4 solution than in 0.01 M NH4NO3 solution [151]. As such, much more research is needed in order to get more general influence of those salt solutions on the corrosion behavior of Mg-Zn based alloys, and the application background should be specified when the corrosion of Mg-Zn based alloys is discussed or investigated.
In addition to immersion tests in salt solutions, salt spray tests are also commonly used to evaluate the corrosion resistance of Mg alloys. Under salt spray conditions, weight loss of Mg alloys is lower, and the pits formed on the sample surface are fewer and grow more slowly [152]. Generally, salt spray atmosphere is less deleterious in terms of general and pitting corrosion compared with solution immersion environments. The differences between salt spray tests and immersion tests come from the fact that a thin solution layer can be formed on the sample surface in salt spray tests, which can be considered as a stagnation, and subsequently results in rapid increase of local pH as Mg dissolves, whereas the immersion environment favors the diffusion of dissolved Mg ions, as well as the formation and growth of pits [152][153][154]. Therefore, interrupted salt spray and alternate immersion tests are suggested to comment on the normal corrosion of Mg-Zn alloys [57].
Direct comparison between the corrosion rates obtained from the two tests would not always be possible, since different concentrations of salt are usually used for each testing environment. Cao et al. [57] have tested the corrosion behavior of as-cast and heat-treated Mg5Zn (in wt.%) alloys in 3.5 wt.% NaCl solution saturated with Mg(OH)2 at room temperature, as well as in spray of 5 wt.% NaCl solution at 35 °C. The general trend of corrosion rates is the same in the two kinds of tests. Corrosion rates of Mg5Zn alloys measured by weight loss after immersion and salt spray tests are comparable, and both of those two tests have demonstrated that heat treatment can improve the corrosion resistance of Mg5Zn alloy. Moreover, the corroded morphologies of Mg5Zn alloy are similar after immersion and salt spray tests. Zhao et al. [155] have studied the corrosion of ZE41 alloy in interrupted salt spray and solution immersion tests with 3 wt.% NaCl. When the interrupted salt spray period is 1 min for salt spray and 119 min for drying at 20 % humid conditions (duration for one week), the corrosion rate of ZE41 alloy is similar to the steady state corrosion rate in the solution immersion test. However, when the procedure is changed to be 15 min for salt spray and 105 min for drying (duration for 1 d), the corrosion rate of ZE41 alloy measured in interrupted salt spray tests is significantly lower than that measured in solution immersion for 1 day, revealing the importance of salt spray intervals to the corrosion tests of alloys.
In contrast to simple salt solutions, a variety of different corrosion products (e.g., calcium carbonate (CaCO3), carbonated calcium phosphates (Ca10(PO4)6(OH)6 or Ca3(PO4)2·3H2O), hydroxyapatite (HA), etc.) are formed on the surfaces of Mg alloys exposed to complex electrolytes, such as SBF. Usually, a simulated body solution contains appropriate inorganic components, a buffering system, and/or organic ingredients. Ringer's solution, Hanks' solution, phosphate-buffered saline (PBS), SBF, Dulbecco's modified Eagles medium (DMEM), Earle's balanced salt solution (EBSS), and minimum essential media (MEM) are frequently used for the in vitro investigation of the corrosion behavior of Mg alloys [46,52,113,126]. The typical chemical compositions of those simulated body fluids are listed in Table 2. The corrosion behavior of Mg alloys differs when different simulated body fluids are used due to the different compositions of the fluids and different corrosion products formed. In particular, the use and type of buffering agent plays a particularly significant role in determining the type of corrosion product formed [9,156]. For example, the corrosion resistance of ZK60 alloy in Hanks' solution is higher than in DMEM and DMEM + fetal bovine serum (FBS). In DMEM+FBS, the corrosion resistance of ZK60 alloy is the lowest compared to that in Hanks' solution and DMEM. The lower content of hydrocarbonate, as well as the absence of organic components in Hanks' solution, results in less protective corrosion products on the surface and the consequent faster corrosion occurring on ZK60 alloy. Additionally, in the presence of FBS, the interaction between proteins and corrosion products may affect the equilibrium of dissolution and regeneration of corrosion products, thus accelerating the corrosion of the alloy [126]. Jamesh et al. [124] have reported that the corrosion products formed on ZK60 alloy surface are Ca10(PO4)6(OH)6, Ca3(PO4)2·3H2O, and Mg(OH)2 in SBF, while CaCO3 and Mg(OH)2 predominate in the case of Ringer's solution. In addition, much more severe pitting and localized corrosion occurs on ZK60 samples immersed in SBF. Zander et al. [75] have further demonstrated that the influence of Zn content on the corrosion resistance of Mg-Zn-Ca alloys is more significant in Hank's solution than in PBS, as indicated by the anodic polarization curves measured in Hanks' solution.  Normally, the corrosion rate of Mg alloy in vivo is around 1-5 times lower than that obtained in vitro [159]. For example, the corrosion rate of as-cast Mg3Zn1.34Ca (in wt.%) alloy in MEM is 4.718 mm/year, whilst that measured in vivo implanted in Lewis rats is only 0.786 mm/year after immersion or implantation for 7 d [157]. However, opposite results have also been reported. Extruded Mg6Zn corrodes faster in vivo (2.32 mm/year) than in SBF (average at 0.14 mm/year) [160]. As indicated in Table 2, the lower concentration of Clin blood plasma (100 mM/L), compared to simulated physiological solutions such as Hank's solution (145 mM/L), generally contributes to slower corrosion of Mg alloy in vivo than in vitro. Additionally, the possible interaction between the complicated components in vivo (e.g., cells and proteins) and Mg may also contribute to the lower corrosion rate measured in vivo [157,159,161]. As already suggested by many researchers [157,[162][163][164], much caution should be paid when choosing the appropriate simulated body fluid to evaluate the corrosion rates of Mg-Zn alloys in vitro to get results comparable to those in vivo.
The behavior of Mg alloys in aggressive environments under dynamic loading, including stress corrosion cracking (SCC) (applied tensile mechanical loading) and corrosion fatigue (CF) (applied cyclic loading), is also a critical concern for their applications. The combined action of stress and corrosive electrolytes can result in an unexpected premature failure of alloys at a loading below the designed value for static loading service conditions [165]. Generally, the corrosion rates of Mg alloys are significantly higher in corrosion fatigue tests than those measured in static immersion. The corrosion product layers can be broken due to the cyclic loading or the strain induced, which consequently promotes the penetration of corrosive electrolytes into the underlying metals and the occurrence of corrosion [166]. Furthermore, the corrosion rates of Mg alloys increase with increasing cyclic loading. Bian et al. [167] have studied the corrosion behavior of extruded Mg2Zn0.2Ca (in wt.%) alloy in SBF under static conditions and cyclic loading, revealing the effect of cyclic loading on the corrosion rate of the alloy mentioned above (Figure 8b). Moreover, it has been pointed out that a material that corroded more slowly in static immersion may exhibit a faster corrosion rate under cyclic loading conditions. This finding is often attributed to the stress-accumulating role of intermetallics due to the evolution of hydrogen at these locations. SCC of Mg alloys, which can occur in either the presence or absence of Cl - [168], is often associated with hydrogen embrittlement. Under tensile loading in a corrosive environment, the rupture of surface film can be induced by an applied stress or localized corrosion, and can facilitate the initiation of cracks. This, in turn, allows the diffusion of hydrogen into the material [169,170]. The strain rate plays a critical role in the failure of alloys. Fast strain rate results in failure via mechanical overload as opposed to SCC. However, low strain rate increases the contact time of the alloy with the corrosive medium, thus facilitating the penetration of the electrolyte and the ingress of hydrogen [171]. Additionally, SCC can even occur in the presence of residual stress that accumulates during thermal-mechanical processing without externally applied stress. Jadari et al. [172] have observed that the fracture morphologies of extruded Mg2Zn1Zr (in wt.%) alloy, immersed in modified simulated body fluid (m-SBF) and then strained in air, are similar to those strained directly in m-SBF. The SCC of the pre-immersion sample is ascribed to the residual stress developed during the extrusion processing.

Influence of Processing
The influence of thermal and mechanical processing on corrosion and mechanical properties of Mg-Zn-X alloys is closely associated with the grain size and the distribution of precipitates within the microstructure. This results in significant differences in the performance compared to as-cast alloys.
Solid solution heat treatment is beneficial to the corrosion resistance of Mg-Zn-X alloys, especially when conducted at high temperatures or for relatively long time periods. Some secondary phases-for example, MgZn binary phase-could dissolve into the Mg matrix during solution treatment. Thus, dendrite structures would disappear, resulting in the formation of a homogenized microstructure [70,173,174]. With the decrease of volume fraction of secondary phases, micro-galvanic corrosion is reduced; thus, the corrosion resistance of the alloys is improved. However, solution treatment can increase grain size, which could decrease corrosion resistance of Mg-Zn-X alloys when the effect of grain size overwhelms that of secondary phases on corrosion of Mg-Zn-X alloys [70]. In contrast, aging treatments facilitate the precipitation of secondary phases [58,142,173,175]. For example, rod-like metastable MgZn2 secondary phases would precipitate both at the grain boundaries and inside the grains, inducing enhanced strength of the alloy due to the strengthening effect of this phase. With the increase of ageing time, the average volume fraction of precipitates increases. Guan et al. [142] have studied the development of microstructure and mechanical properties of rolled Mg4Zn1Sr (in wt.%) alloy with different aging times (0-16 h). The hardness, ultimate tensile strength, and elongation to fracture of Mg4Zn1Sr alloy increased as aging time increased, and reached their maximum after aging for 8 h at 175 ºC. This was attributed to the strengthening effect of the precipitated rod-like transition MgZn2 phase. However, with further increase in aging time, the volume of rod-like MgZn2 phases decreases, while the amount of particle-like flaky MgZn2 phase and stable MgZn phase, which weaken the strengthening effect, increases. Therefore, mechanical properties of the alloy start to decrease. Ibrahim et al. [176] have also confirmed the influence of aging time on the microstructure and mechanical properties of Mg-Zn-Ca alloy. They have studied the microstructural, mechanical, and corrosion characteristics of Mg1.2Zn0.5Ca (in wt.%) alloy after solid solution treatment and with further aging treatment for different time. It has been demonstrated that when the age hardening duration increases to 2-5 h, both the mechanical properties and corrosion resistance of Mg1.2Zn0.5Ca alloy strengthen because of the refined microstructure and finely distributed Mg6Ca2Zn3 precipitates. Further increases in aging time did not contribute to further enhancement of either mechanical or corrosion properties.
Wrought processing can introduce several effects, such as grain refinement and texture, conferring specific improved properties to Mg alloys. Extrusion treatment induces severe plastic deformation, simultaneously refines α-Mg grains by dynamic recrystallization, and produces highly dispersed precipitates. The refined microstructure is beneficial for the improvement of the corrosion resistance of Mg-Zn-X alloys [6,126]. Moreover, the tensile strength and ductility of Mg-Zn-X alloys can be enhanced by grain refinement strengthening, solid solution strengthening, and secondary phase strengthening during the extrusion process [6,16,96,160,177]. Extrusion parameters, such as extrusion temperature, speed, and ratio, have a significant influence on the microstructure and mechanical properties of Mg-Zn-X alloys (Figure 9). In general, with the increase of extrusion temperature, ratio, and speed, average grain size and area fraction of dynamic recrystallized (DRXed) grains increase. This is because the deformation heating, which occurs during the extrusion process, results in rise of temperature at the deformation zone. Concomitantly, the ultimate strength and yield strength decrease due to the coarsened microstructure [86,173,[178][179][180]. Formation of precipitates also increases with elevated extrusion temperature because the diffusion of solute atoms is faster at higher temperatures [173,181]. However, the influence of extrusion temperature on the plastic deformation of Mg-Zn-X alloys is still under debate. Park et al. [179] and Zeng et al. [180] have reported increased elongation to fracture of MgZn(Mn)Ce/Gd and MgZnYZr alloys with increased extrusion temperature, while the relationship was reversed in the case of MgZnHo [178] and MgZnGd alloys [173]. Park   It is believed that a weaker or non-basal texture is beneficial to the improvement of ductility of Mg sheets at room temperature, since the majority of grains in weaker basal or non-basal textures favor dislocation glide on the basal plane [182][183][184]. To control the texture, many processing technologies have been applied to Mg alloys. Rolling processes have been shown to be more effective in developing weaker or non-basal textures in Mg alloys compared with continuous extrusion, torsion extrusion, and equal channel angular extrusion [185]. The refined microstructure and uniformly dispersed fine secondary phases, formed during the rolling processing, contribute to enhanced yield strength and elongation to fracture of Mg-Zn-X alloys [101]. Lee et al. [186] have compared the recrystallization behavior of hot-rolled Mg3Zn0.5Ca alloy with that of hot-rolled Mg3Al1Zn alloy (both in wt.%). The results showed a totally different twins formation mode in Mg3Zn0.5Ca and Mg3Al1Zn alloys. Both compressive and secondary twins are generated in Mg3Zn0.5Ca alloy, while only tensile twins are observed in Mg3Al1Zn alloy. The presence of compressive and secondary twins leads to a weaker basal texture evolution in the Mg3Zn0.5Ca sheet. As a result, the hot-rolled Mg3Zn0.5Ca alloy exhibits much better formability than hot-rolled Mg3Al1Zn alloy. The influences of rolling on the corrosion properties of Mg-Zn-X alloys are not widely studied. The increase of grain boundary area due to significantly refined microstructure after rolling can decrease the corrosion resistance of Mg3.0Zn (in wt.%) alloy in SBF [187].
Equal channel angular pressing (ECAP), which induces severe plastic deformation (SPD) by introducing an extremely large strain, is one of the most effective processing techniques for fabricating ultrafine-grained (UFG) Mg alloys. Compared with conventionally processed Mg alloys, ECAPed Mg alloys exhibit improved corrosion resistance and excellent mechanical properties [188][189][190][191][192][193]. Jiang et al. [188] have studied the influence of ECAP pass number on the corrosion resistance of as-cast ZE41A alloy in NaCl solution. Higher ECAP passes endow UFG ZE41A alloy with better corrosion resistance. The corrosion product layers formed on ECAPed ZE41A alloy, which has undergone 60 passes, are free of cracks compared to those formed on the ECAPed alloy surfaces with fewer passes. This can be explained by the decreased residual internal stress of the alloy, resulting from the complete DRX of the deformed microstructure formed after a high number of ECAP passes. ECAP processing of extruded ZK60 alloy is also beneficial for the reduction of the corrosion rate of the alloy in PBS solution [189]. The refinement and redistribution of precipitates during the ECAP process lead to a remarkable change in corrosion behavior from a localized to a more uniform mode. However, the influence of ECAP processing on the mechanical properties of Mg-Zn-X alloys differs with the initial conditions of the alloys. Zheng et al. [190] have compared the mechanical properties of as-cast and extruded Mg5Zn0.9Y0.2Zr (in wt.%) alloys after different passes of ECAP processing. Both the strength and elongation to fracture of as-cast Mg5Zn0.9Y0.2Zr alloy are significantly enhanced after ECAP processing, especially after an increased number of ECAP passes. This is attributed to the uniformly dispersed I-phase and refined grains formed during recrystallization. In contrast, for extruded Mg5Zn0.9Y0.2Zr alloy, the grain size is further refined after ECAP and the ductility is improved, while both yield strength and ultimate strength are reduced. This trend, also confirmed by other researchers [189,191,192], does not follow the Hall-Petch relation. It is explained by the intensive development of texture during ECAP. ECAP processing routes (shown in Figure 10), regarding the rotation of the billet between each pass, also have significant influence on the microstructure and mechanical properties of Mg alloys [193]. Route Bc (Figure 10b) is the most effective way to refine grains, while route A is the least effective one. The different strain paths can also result in different textures. Processing as route C would lead to the strongest texture of the alloys, with high-intensity basal planes inclining about 45º to the extrusion direction. As a result of the competition between grain refinement strengthening and texture softening effect, both yield strength and ultimate tensile strength of extruded Mg5.25Zn0.6Ca (in wt.%) alloy processed by ECAP with routes A and Bc increase, but decrease with route C.  [193,194] (with permission from Elsevier and Copyright Clearance Center).

Strategy to Improve Properties of Mg-Zn Based Alloys
In the previous chapters, it was demonstrated how alloying additions and thermal-mechanical treatments can be used to meet the requirements of different applications and expand the application field. However, microstructure adjustment by phase composition and unconventional fabrication approaches can also improve the corrosion performance and mechanical properties of Mg-Zn based alloys. Furthermore, surface treatment is effective for tailoring surface properties and corrosion resistance.

Magnesium-Based Metal Matrix Composites
Magnesium-based metal matrix composites (MMCs) have high specific stiffness and strength both at room and elevated temperature, as well as improved damping capacity and wear resistance compared to the conventional alloys. The fabrication of MMCs has been considered as an approach to tailor the corrosion rate of Mg alloys with enhanced mechanical properties. Therefore, they have attracted considerable attention as high-performance structural materials for applications in the automotive and aerospace industries [195][196][197]. Nunez-Lopez et al. [198,199] have studied the corrosion behavior of Mg-Zn-Cu (ZC71)/SiCp (silicon carbide particles) metal matrix composite in salt spray tests at 25 °C, particularly the micro-galvanic corrosion between the reinforcement and the matrix. Compared with conventional ZC71 alloy, the local penetration rate recorded for ZC71/SiCp composites is about three times higher, which is due to the formation of less protective corrosion products. However, the maximum depth of corrosion attack in the salt spray test is the greatest for extruded ZC71 alloy, which is believed to be a result of the more uniform distribution of eutectic Mg(Zn, Cu)2 phases in the composite. This uniform distribution is ascribed to the effective nucleation sites of SiCp for precipitation of eutectic Mg(Zn, Cu)2 phases. Additionally, no galvanic corrosion is detected in the vicinity of the reinforcement particles. Difficulty in forming protective corrosion product layers on SiCp/ZK80A and SiCp/ZK60A composites has also been revealed by Zucchi et al. using electrochemical impedance spectroscopy [200]. Nevertheless, dense corrosion product (Mg(OH)2) layers can be formed on the surfaces of calcium-polyphosphate-reinforced ZK60A composites [201]. Compared with the porous and plate-like crystalline morphology of the Mg(OH)2 formed on the surface of conventional ZK60A alloy, dense and plate-like Mg(OH)2 is formed on the surface of the composite. This dense layer significantly slows down the corrosion of the composite. Nieh et al. [202] have reported superplasticity of a ZK60A composite reinforced by 17 vol.% SiC particles, with an elongation to failure of up to 350% at a very high strain rate of about 1 s -1 , which is due to the refined grain size of about 0.5 µm. The presence of fine SiC particles contributes to the refinement and stabilization of the microstructure of the composites at elevated temperature. Hu et al. [203] have also fabricated ZK51A matrix composites reinforced by SiC whiskers (SiCw) using a two-step squeeze casting process. The modulus and mechanical strength increase compared with unreinforced ZK51A alloy. Particularly, the increase in 0.2% offset yield strength and modulus of the composite is linearly proportional to the volume fraction of SiCw in the range of 0-0.2, demonstrating the load transfer strengthening mechanism. The adjustment of mechanical properties and corrosion rates of MMCs by choosing appropriate composites, such as HA [204,205], β-tricalcium phosphate (β-TCP) [206], and poly-L-lactic acid (PLLA) [207], also confers great potential as biomaterial candidates on MMCs. Yu et al. [208] have investigated the in vitro and in vivo degradation behavior of β-Ca3(PO4)2/Mg6Zn (wt.%) composites in SBF and rabbits, respectively. The corrosion rate of 10% β-Ca3(PO4)2/Mg6Zn composites significantly decreases compared with Mg6Zn alloy fabricated with the same processing route. This is explained by the decomposition of β-Ca3(PO4)2 that favored the formation of protective insoluble phosphates and carbonates on the composite surfaces during immersion. Furthermore, in vivo tests indicate that the composites improve the concrescence of pre-broken bone tissues and exhibit good biocompatibility with vital organs like the hearts, livers, and kidneys of rabbits, even in the presence of hydrogen bubbles.

Magnesium Bulk Metallic Glasses
Magnesium bulk metallic glasses (BMGs), especially Mg-Zn-Ca glassy alloys, have been extensively studied as biodegradable materials. Their single-phase and chemically homogenous microstructures are beneficial for the improvement of both corrosion resistance and mechanical properties of Mg alloys [209]. The specific strength of Mg-Zn-Ca alloy in glass state (250-300 MPa cm 3 /g) can be about 40% higher than those of traditional crystalline Mg alloys (around 220 MPa cm 3 /g for die-casting AZ91D alloy) [210]. Gu et al. [211] have assessed the corrosion behavior of Mg66Zn30Ca4 and Mg70Zn25Ca5 (both in at.%) BMGs, and have revealed that the corrosion rate decreases compared to some previously reported Mg alloys, such as as-cast AZ91 alloy and as-rolled pure Mg. This was attributed to the absence of secondary phases in BMGs, which can induce galvanic corrosion and the dense corrosion products containing zinc oxide/hydroxide forming on Mg-Zn-Ca BMGs. Due to the absence of crystal slip systems in Mg-Zn-Ca BGMs, the single-phase and chemically homogeneous nature also results in a three-times-higher compressive strength compared to as-rolled pure Mg. Moreover, the cytocompatibility is improved, and good adherence and growth of MG63 cells occur on the Mg-Zn-Ca BMGs surface. However, glassy Mg-Zn-Ca alloys are extraordinarily brittle at room temperature, and the fabrication of BMGs usually requires high-purity raw materials to guarantee glass-forming ability (GFA). Wang et al. [209] have successfully synthesized Mg69Zn27Ca4 (in at.%) glassy alloy reinforced by ductile Fe particles by using industrial raw metallic elements. The GFA decreases due to the addition of Fe with the formation of ductile α-Mg and dendrite MgZn phases. Nevertheless, the compressive strength and fracture strain are enhanced.

Squeeze Casting
Squeeze casting is a metal fabrication process in which the solidification of molten metal is accelerated by applying high pressure. It combines permanent mold casting with die forging. Compared with conventional casting process, such as die-casting, gravity casting or sand casting, squeeze casting processing can produce Mg alloys with finer microstructure, less porosity, longer die-life, and reduced microshrinkage. These merits confer both better corrosion resistance and mechanical properties on alloys fabricated by squeeze casting [212,213]. Mo et al. [214] have compared the microstructures and mechanical properties of Mg12Zn4Al0.5Ca (in wt.%) alloys fabricated by gravity and squeeze casting. The applied pressure used during squeeze casting limits the formation of shrinkage porosity. The refined microstructure obtained during the squeeze-cast process contributes to improved mechanical properties of components compared to those produced by gravity casting. The influence of squeeze casting processing parameters, such as pouring temperature, applied pressure, and dwell time, has also been investigated. Increasing applied pressure and dwell time (just before the ending of solidification) refine microstructure and promote cast densification, which consequently improves the mechanical properties of the alloy. However, as pouring temperature increases, cast densification is promoted and grain size is coarsened, which leads first to an increase, then to a decrease of the mechanical properties. The study of Doležal et al. [215] on Mg3Zn2Ca (in wt.%) alloy has also confirmed the grain refinement and the enhanced mechanical properties of Mg-Zn alloys produced by squeeze casting compared with those produced by gravity casting. Moreover, subsequent aging treatment can further improve the ultimate tensile strength and elongation to fracture due to the homogenization of microstructure.

Twin-Roll Casting
Twin-roll casting (TRC) processing reduces several production steps compared to the conventional production techniques by combining casting and hot rolling into a single step. This reduces manufacturing costs for commercial weight-saving applications of wrought Mg alloys. It has been revealed that TRC processing can provide much faster solidification speed, refine gain size, homogenize the microstructure, and reduce, or even remove, segregation of alloying elements [216][217][218][219][220]. Moreover, the basal texture can be reduced when Mg alloys are fabricated by TRC processing compared to the strong basal texture developed in hot rolling. Little attention has been paid to the corrosion behavior of Mg alloys fabricated by twin-roll casting. Oktay et al. [221] have compared the corrosion behavior of AZ31 sheets produced by twin-roll casting and conventional direct chill casting + rolling in 0.01 M NaCl and 0.5 M Na2SO4 solutions. The higher solidification rate during twin-roll casting results in much smaller and more homogenously distributed secondary phases in the alloy, which contributes to a slightly higher corrosion resistance of TRC AZ31 alloy in both Na2SO4 and NaCl solutions. Mg alloys produced by TRC have comparable or better mechanical properties to those of Mg alloys sheets produced by conventional ingot casting and subsequent hot rolling, but a better combination of yield strength and fracture elongation [222,223]. Park et al. [224] have compared the high-temperature superplastic deformation behavior of TRC-prepared Mg6Zn1Mn1Al (ZMA611) (in wt.%) and ingot cast AZ31 alloys. The TRC ZMA611 alloy exhibits larger fracture elongation than ingot cast AZ31 alloy at all testing temperatures and strain rates, which is caused by the presence of thermally stable Al8Mn5 dispersoid particles. Hou et al. [225] have compared the degradability and mechanical integrity of a TRC Mg0.7Zn0.6Ca alloy (in wt.%) before and after annealing. The results show that the residual strain and deformation twins restored in the as-rolled alloy lead to a slightly higher corrosion rate than that of the annealed one. For both alloys, sufficient mechanical support still exists after immersion in α-MEM + 10 % FBS for up to 42 days, with a residual YS and UTS that are more than 85% those of the pristine states. With further minor (0.1 wt.%) addition of Zr [226,227], the sheet exhibits excellent stretch formability at room temperature, slow corrosion rate (<0.25 mm/year, determined by a neutral salt spray test at room temperature), and a widely adjustable strength level (YS = 130-260 MPa, UTS = 210-300 MPa), which shows great potential in applications as both structural and biomedical materials.

Rapid Solidification
Compared with conventional ingot metallurgy (I/M), rapid solidification (RS) processing confers better corrosion and mechanical properties to Mg parts due to the refinement of grain and precipitate sizes, the extended solid solubility of alloying elements, and the homogeneous microstructure [228,229]. Zhang et al. [111] have compared the corrosion resistance ( Figure 11) and mechanical properties of Mg6Zn1Mn (in wt.%) alloy fabricated by conventional casting and rapid solidification. In comparison to the formation of coarse dendrite cells (about 200 µm) in as-cast Mg6Zn1Mn alloy, extremely small dendrites with an average size of about 25 µm are formed in rapid solidification. The refinement of the dendritic microstructure consequently decreases the corrosion rate of Mg6Zn1Mn alloy in 3.5 wt.% NaCl solution by approximately one order of magnitude, and increases the ultimate strength from 335 MPa to 460 MPa. Xu et al. [230] have studied the influence of solidification rate on the microstructure and corrosion resistance of Mg20Zn1Ca (in wt.%) alloy. They have found that the grain size and precipitate size decrease with increased cooling rate. A continuous 3D network of secondary phases is formed under a cooling rate of 3000 ºC/min. Moreover, as a result of fast cooling, the contents of Zn and Ca in the Mg matrix are much higher than their equilibrium solid solubility in Mg. The supersaturations of Zn and Ca elements and the dispersed alloying elements due to homogeneous microstructure enhance the corrosion resistance of Mg20Zn1Ca alloy in PBS solution.

Surface Treatments
Modification of the surface is another very popular strategy to improve the (corrosion) properties of Mg alloys in general and Mg-Zn based alloys in particular. However, the field is wide; here, it is only intended to give a short overview of the topic. For a deeper study of the topic, a range of reviews about surface treatments of Mg alloys are available [231][232][233][234][235][236][237][238][239].
There are a couple of different approaches and concepts to either improving the substrate by itself (e.g., surface alloying), to coating it with a more resistant material, and/or to separating the substrate material from the surrounding. However, no coating is perfect or will survive in service forever without defects. From this point of view, the substrate corrosion resistance is still an important factor. Good substrate corrosion resistance is a requirement for minimizing reactions at the interface when diffusion of ions and water through the layer or defects take place. The formation of corrosion products (mainly hydroxide formation) can cause severe volume increase and thus stresses at the interface, leading to blistering and flaking-off of the coatings. As-cast Mg-Zn based alloys are, from this point of view, not the best choice, but the newly developed wrought alloys with much lower Zn contents and additional micro-alloying elements (e.g., Ca, Zr, etc.) do offer substrates with good corrosion resistance and uniform corrosion without localized attack [147,226]. To guarantee good corrosion resistance, especially for wrought products, a sufficient cleaning (material removal) is the pre-requirement for any successful coating on Mg alloys. Removal of the undefined and contaminated surface oxide layer by etching has to be done, replacing it afterwards with a defined "passive" layer (conversion). The etching of Zn-rich alloys can result in Zn enrichment at the surface with problems to obtain such a uniform conversion layer. The latter is not a problem for most alloys with lower alloy content or the alloying elements in solid solution (Zn and other alloying elements <1 wt.% each).
After cleaning, depending on the application and the required properties, a wide set of treatments/coatings are available, ranging from metals to ceramics and polymers. To deposit them on Mg, a range of coating techniques is available: • Chemical conversion coatings -Chromates (restricted in use in most countries) [240] -Chromate-free alternatives: Phosphate-permanganate, vanadates, molybdates, stannates, tungstates, fluorozirconate/titanate, and potassium-permanganate [241][242][243][244][245][246][247][248][249][250][251][252] • Electro-chemical coatings -Electroplating (Zn, Cu, Ni, Cr) [253][254][255][256] -Anodizing [257,258] -Plasma electrolytic oxidation (PEO) [259,260] • Polymeric coatings/paints (e.g., Powder and E-Coat) and other non-metallic coatings (e.g., varnish, wax) [261][262][263] • Physical techniques -Physical vapor deposition [264][265][266] -Plasma/laser surface treatments (alloying, cladding) [267][268][269][270] -Thermal/plasma spraying [271] Currently, the most industrially used processes for corrosion protection are conversion coatings in combination with paint or polymer top coats. The conversion treatment is the base for paint application, and should create a strong interface with good adhesion to the paint. There are chromate-free conversions available, and if more adhesion and better barrier properties are required, the chemical conversion layer can be replaced by anodizing/PEO coatings. The paint/polymer top coats can range from single-to multi-layer coatings according to the requirements of the applications and severity of the environment. Such combinations are sufficient for most of today's transport applications. The main challenge in coating development today is the integration of inhibitors into the coatings, offering release on demand, self-healing ability, and additional functionalities.

Applications of Mg-Zn Based Alloys
Mg-Zn based alloys have a long tradition in aerospace applications in helicopter, aircraft, and aeroengine components [272]. For example, ZE41 alloy is applied in the main transmission of Sikorsky UH60 Family (Blackhawk) and the turbofan of Pratt and Whitney Canada PW305. In addition, the tray of Rolls Royce is made from ELEKTRON ZRE1 alloy ( Figure 12) [273]. In automotive applications, Mg4.2Zn1.2RE0.7Zr (ZE41A) and Mg6Zn2.7Cu0.25Mn (ZC63A) (both in wt.%) alloys have been used for functional prototypes due to the good pressure tightness of ZE41A alloy and the good elevated temperature properties of ZC63A alloy [274]. Moreover, ZK60 extruded at an extremely low extrusion rate is used for wheels and stems of racing cars and bicycles [275]. IMRA America Inc. [276] has developed a series of Mg8Zn5AlCa (in wt.%) alloys with a Ca content varying from 0.2 to 1.2 wt.% for automotive powertrain applications. The developed alloys showed a comparable or slightly better corrosion resistance than that of AZ91D alloy. The creep performance at 150 ºC and 35 MPa was decreased by almost one order of magnitude, while the tensile properties were comparable to or slightly better than those of AZ91D alloy. Volkswagen AG, Helmholtz-Zentrum Geesthacht, and Posco collaborated together to develop a new sheet alloy based on TRC Mg-Zn-Ca-Zr. The TRC Mg-Zn-Ca-Zr sheet reduced the forming temperatures from over 220 °C for a conventional AZ31 sheet to 160 °C, which means saving more energy and significantly simplified temperature control in a series-forming process. A tailgate was successfully produced as a demonstrator [227,277]. Up to now, Mg-Zn based alloys have not been commercialized for clinical/medical applications, but a relatively high number of in vivo and in vitro studies have been carried out to investigate the feasibility of Mg-Zn based alloys as biomedical materials. He et al. [278,279] have comprehensively studied the in vitro and in vivo degradation behavior of extruded Mg6Zn (in wt.%) alloy. It exhibited better cell attachment, mineralization ability, and improved mRNA expression compared with PLLA. More importantly, considerable new bone formation is observed surrounding the implanted Mg6Zn alloy, and no adverse effect induced by hydrogen evolution is detected during the in vivo experiment. In addition, pathological analysis demonstrated that the in vivo degradation of Mg6Zn alloy did not harm the vital organs (heart, liver, kidney, and spleen) [278]. Guo et al. [60] have also studied the in vitro and in vivo histocompatibility of urinary implants made of Mg6Zn and pure Mg. The Mg6Zn alloy degrades faster than pure Mg in SBF, but exhibits better histocompatibility in the bladder compared to pure Mg after implantation for two weeks. Kraus et al. [280] have utilized online microfocus computed tomography (µCT) (Figure 13) to observe the interaction of implant and tissue performance by implanting Mg5Zn0.25Ca0.15Mn (in wt.%) pins into a growing rat skeleton. During degradation in a physiological environment, large amounts of Mg 2+ are released into the surrounding tissue. As a result, magnesium hydroxide is formed, which consequently facilitates the accumulation of calcium phosphates and the formation of new bone around the implanted material. Although the Mg5Zn0.25Ca0.15Mn alloy starts to degrade immediately after the implantation and generates a massive volume of hydrogen gas (about 270 mm 3 /d), the gas can be rapidly resorbed, and the bone recovers restitution after the complete degradation of the implanted pin. In addition, no inflammation is observed. These studies suggested that Mg-Zn based alloys have quite excellent in vivo biocompatibility and are promising candidates as biodegradable medical devices.

Summary
This review presents recent developments of Mg-Zn based alloys for both industrial and biomedical applications. Various attempts have been made to improve the corrosion resistance and mechanical properties of Mg-Zn based alloys, including alloying addition, processing, development of composites, etc. The new developments focus on wrought alloys rather than as-cast alloys.
Obvious composition-microstructure-properties relationships can be observed, especially for Mg-Zn-Y alloys. Some Mg-Zn based alloys show good potential for structural or biomedical applications-for example, Mg-Zn-Ca alloys as implants and new sheet materials for industrial wrought applications. On the other hand, inconsistent results between different studies can still be found, especially for the binary Mg-Zn system, which seems to offer no real potential for further developments. However, a good understanding of the binary Mg-Zn system is the basis for any further development of more complicated Mg-Zn-X alloys. Moreover, a satisfactory combination of high strength and good corrosion properties is difficult to be achieved, since more element addition usually results in a higher volume fraction of secondary phases, which is usually unfavorable for the corrosion performance. The combination of micro-alloying and wrought processing seems to be a promising way to solve this problem.
In summary, it can be stated that the current driving force for the development of Mg-Zn based alloys comes from two directions, both of which aim to prevent strong localized corrosion and to keep the corrosion rate of the alloy as low as possible. The first direction is the demand for new sheet material, mainly for industrial transport applications, and the second one is that of biomedical applications. Thus, the clear trend for the current ternary Mg-Zn-X and quaternary Mg-Zn-X-Y systems is to reduce the Zn content (less than 1 wt.%) and to improve strength by micro-alloying with ternary/quaternary alloying addition and wrought processing. In particular, the newly developed Mg-Zn-Ca-Zr sheet alloys [227,277] do have the potential to replace the currently most used AZ31 sheet alloys in many applications. To assure the biocompatibility and cytotoxicity, careful selection of bio-safe alloying elements and controlling of release rates of alloying elements within the safety range are of fundamental importance. However, more research is still needed for a better understanding of the Mg-Zn system and its interaction with other alloying elements to develop new Mg-Zn-X-Y alloys with better combinations of corrosion performance and mechanical properties.  35 mV * The value was the highest potential compiled from the line-profile analysis of the secondary phase because no average value was afforded in the reference.