Laser additive manufacturing of magnesium alloys and its biomedical applications

Biomedical magnesium (Mg) alloy with unique biodegradability and excellent biocompatibility is one of the most sought after materials in medical field for orthopedics applications. Nevertheless, the high corrosion rate and inadequate mechanical properties hinder its development. Apart from that, to obtain the best surgical result, the size and shape of the fixation implant need to be adapted to the individual case. Thus, additive manufacturing (AM) processes, such as laser powder bed fusion (LPBF), are used to overcome these issues. This work reviews the recent advancements in biodegradable Mg-based alloys prepared by LPBF for biomedical applications. The influence of feedstock features and manufacturing parameters on the formability and quality is delineated in detail. The mechanical performances, degradation behaviors, and biological behavior of the LPBF-processed parts are discussed. Furthermore, we also made some suggestions for the challenges of Mg alloys in LPBF processing and applications in biomedical.


Potential of magnesium in biomedical applications
With the increase in elderly population and the frequent traffic accidents, more and more people are suffering from bone defect [1] . At present, statistics shows that the cumulative number of cases exceeds 20 million in China, with an increment of 3 million new cases every year [2] . The regeneration and reconstruction of damaged tissues have great strategic importance [3] . Medical metal materials, such as titanium alloys, tantalum alloys, and stainless steel, are commonly used as implant to replace the damaged parts in bones due to their excellent mechanical properties and good biocompatibility [4] . https://doi.org/10.18063/msam.v1i4.24

Materials Science in Additive Manufacturing
However, their elastic modulus and mechanical strength are much higher than natural bone, which cause the "stress occlusion" effect after implantation, thus leading to the atrophy of natural bone tissue and even migration failure [5] . Moreover, they might release some toxic metal ions, which induce inflammation. Unfortunately, the implants do not easily degrade, and thus, they need to be removed in secondary surgery after implantation [6] . In this case, this could the patients will undoubtedly have to endure huge pain and shoulder extra economic burden. Therefore, it is imperative to develop new metal scaffolds for repairing bone defect [7] .
Recently, biodegradable metals such as magnesium (Mg), zinc (Zn), and iron (Fe) alloys have become a hot research topic in tissue repair because of their mechanical strength, unique degradation properties, and good biocompatibility [8,9] . After implantation, these biodegradable metals have the sufficient mechanical strength to provide support during the healing process [10] . Importantly, they can be completely degraded in vivo, and their degradation products will be metabolized by the human body without any toxic side effects [11] . Compared to Zn and Fe alloys, Mg alloy has received wide attention from experts and scholars since it has the following three significant advantages: (i) Mechanical compatibility. Mg alloy is a light alloy with low density and high specific strength. Its density (1.8 -2.1 g/cm 3 ) and elastic modulus (40)(41)(42)(43)(44)(45) are similar to those of human bone, indicating that they can effectively reduce the stress shielding effect [12,13] . It has excellent application prospects in bone tissue defect repair. (ii) Biodegradability. The standard electrode potential of Mg is −2.37 V. Mg alloy can completely degrade the body by self-corrosion. The degradation products have no apparent side effects on the human body and can be excreted through human metabolic processes [14,15] . (iii) Biocompatibility. Mg, which is involved in synthesizing a variety of proteases and nucleic acids, is one of the essential nutritional elements for humans. Mg ion is conducive to the dilatation of blood vessels and the healing of bone tissue. Therefore, biodegradable Mg alloy is hailed as the new-generation medical metal materials. The comparison between Mg alloy and other bone implants is shown in Table 1.

Development of Mg-based implants
Mg and its alloys, as the promising implants, are mainly used in bone fixation devices, cardiovascular scaffolds, and tissue engineering scaffolds [20][21][22] . The advantages of Mg implants and its clinical applications are presented in Figure 1. The development of Mg-based implants has gone through two stages, with the 21 st century as the dividing line. Before the 21 st century, Mg alloys were mainly used in bone fixation devices. In 1900, Payr et al. [23] pioneered the use of Mg as a bone connector for the fixation of injured bone, and the Mg plate produced no adverse effects in animals. In 1906, Lambotte extended the animal in vivo experiment to a human clinical study, using Mg plates and steel nails to fix the calf fracture of patients [24] . Still, the Mg plate and steel nails formed galvanic corrosion, which produced a large amount of hydrogen. The implantation site appeared to have a gas swelling phenomenon. In 1938, McBride et al. [25] used a bone fixation device made of Mg-Al-Mn alloy to treat multiple fractures, and no negative effects were observed in the fracture tissue after implantation. However, due to the rapid degradation rate of Mg alloy in the human body, the clinical research of Mg alloy as a bone fixation device has stalled in the late 20 th century [26] .
Due to the rapid development of science and technology after the 21 st century, some Mg alloy bone fixation devices have passed clinical trials and obtained regional certification. In 2008, the United States established a research center to study biodegradable biomaterials. In 2013, Syntellix AG successfully developed degradable Mg alloy screws, which passed clinical trials and obtained CE certification [27] . In 2014, the China Food and Drug Administration certified the Mg-Zn-Ca bone nail developed in South Korea because of its relatively slow degradation rate [14] . In recent years, Mg alloys have gradually been used to manufacture cardiovascular and tissue engineering scaffolds. Chaya et al. [28] used Mg bone plates and screws and titanium alloy plates and screws to repair bone cracks in rabbits. The results showed that the degradation of Mg bone plate was beneficial to bone healing. Zhao et al. [29] used Mg screws to fix vascularized bone grafts with femoral head necrosis and found that released Mg ions could stimulate the generation of fresh bone. Wang et al. [30] used Mg-Zn-Y-Nd alloy scaffolds to treat esophageal cancer. The data showed that Mg alloy had good biocompatibility and degradation performance and could kill the esophageal cancer cells.

Advantages of laser powder bed fusion (LPBF) for the fabrication of implants
To meet the requirements of clinical applications, bone implants should possess not only appropriate mechanical strength and good biocompatibility but also personalized shape to match different damaged parts. The traditional preparation methods of Mg alloy components are mainly casting and powder metallurgy. Although the formed Mg alloy components have good mechanical properties, https://doi.org/10.18063/msam.v1i4.24 LPBF of Mg and its bio-applications Materials Science in Additive Manufacturing the process has an extended processing cycle and low material utilization. Unfortunately, these methods cannot easily control the pore size to obtain complex geometric shapes. Additive manufacturing (AM), commonly known as three-dimensional (3D) printing, is a manufacturing technology that integrates computer-aided design, material processing, and molding technologies and uses digital model files as a basis. Meanwhile, through software and CNC systems, special materials are stacked layer by layer in accordance with extrusion, sintering, melting, light curing, or spraying to fabricate block parts [31] . Compared with traditional manufacturing technology, AM could provide a reliable way to obtain personalized complex 3D structures, which can efficiently and reliably replicate anatomical morphology related to tissues and organs. It could prepare precisely controlled pore structure to meet the personalized customization needs of patients. In particular, LPBF is commercially known as selective laser melting (SLM), which uses metallic and non-metallic powders as the raw material [32] . In contrast to other AM technologies, SLM can process a very wide variety of materials. In addition, it is possible to recycle and reuse unmelted metal powder, which allows the efficient use of the material. Due to its characteristics, SLM can be effectively applied as a prospective production technique for valuable materials and components, through cutting down the cost and lead time of fabrication and reducing the loss of material. Therefore, it has attracted increasing interest and attention in fabricating biodegradable Mg-based implants.
In 2011, Ng et al. [33] used the SLM process to prepare pure Mg, which was the first report about the Mg alloy for custom biomedical implants. Since then, Mg and its alloys for the degradable implants has become hotspot research in biomedical field. In this review, we present a systematic analysis and discussion on the recent literature on LPBF-processed Mg alloy. In addition, the effect of the comprehensive powder properties, parameters optimization, and post-treatment on their mechanical and degradable properties will be highlighted, along with the current challenges in LPBF-processed Mg alloys. The review also presents insights into the future of Mg alloys and their use in biomedical applications.

LPBF-processed Mg alloys
The fabrication of Mg and its alloys through LPBF is deemed to be extremely challenging. On the one hand, due to the inherent inflammability and explosiveness, the preparation conditions of Mg powders are extremely demanding. On the other hand, low evaporation temperature and high vapor pressure of Mg tend to trigger micro-crack during LPBF processing, causing poor structural integrity of parts. Thus, until now, there is no relevant report on LPBFprocessed ultrapure Mg (>99.9%). A great deal of current research on Mg prepared by LPBF has focused on its alloys. This is due to their good workability and low risk.

Preparation of the feedstock
It is well known that LPBF is a typical powder metallurgy technology. Powder properties are an essential part of the LPBF industry chain. The powder should possess a

Water atomization
Water atomization is an atomization method that uses water as the atomization medium to break the molten metal. The advantages of water atomization include simple equipment structure, high efficiency, and low atomization cost. However, the disadvantage of water atomization is that the prepared powder has high impurity and poor sphericity. It was due to the fact that the reaction between active metal and atomization medium at high temperature increases oxygen content. In addition, the metal droplets formed by water atomization breaking in the rapid solidification stage are irregular due to the large specific heat capacity of water.

Gas atomization
The aerosolization method uses high-speed airflow to crush the liquid metal stream, forming tiny droplets, and then quickly condensing to obtain the formed powder. Compared with water atomization, gas atomization uses inert gas (i.e., argon and nitrogen) as atomization medium. The metal prepared by this method has the advantages of small powder (particle size <150 μm), uniform composition, and high sphericity. Gas atomization is suitable for producing most metals and alloys, which is the primary production method for spherical metal powders in AM processing. At present, most of the powder is produced and prepared by the gas atomization method, accounting for 30%-50% of the powder produced in the world [29] . The density and microstructure of 17-4PH stainless steel parts prepared from gas-atomized and water-atomized powders were comprehensively investigated by Irrinki et al. [34] . The morphology of water-atomized powder and gas-atomized powder is shown in Figure 2.

Plasma atomization
Plasma atomization was reported by Entezarian et al. in 1996 for the production of highly spherical powders with an average particle size of 40 μm [36] . This approach uses the focused argon plasma jet generated by a plasma torch to melt the wire as initial material, which then results in the formation of tiny metal droplets. The plasma torch produces an extremely high-temperature argon plasma, which can inhibit the rapid solidification of molten and superheated metal droplets into irregular shapes [36] . Therefore, the droplets spend sufficient time in the superheated state and adopt an equilibrium shape (ideal spheres) driven by the surface tension during the falling process. The atomization and condensation process of the metal powder is carried out in an inert atmosphere.

Plasma rotating electrode atomization
Plasma rotating electrode (PRE) atomization is a centrifugal atomization process. It uses the plasma arc generated by a plasma gun as a high-temperature heat source to melt the raw material, which is in the form of an electrode rod and rotate at approximately 15,000 rpm. Meanwhile, centrifugal force separates the molten material from the rod, which then results in the formation of tiny metal droplets. The metal droplets solidify into spherical powders before they hit the chamber walls. This method was invented to reduce the volume of gas consumed to prepare each unit weight of atomized powder, as well as to address the generally low yield of fine powders prepared by conventional gas atomization processes. Notably, the yield for the process was closely related to this centrifugal force, that is, the rotation speed. In addition, the process can eliminate the risk of ceramic contamination due to melting of reactive metals in ceramic crucibles, thus obtaining highpurity spherical metal powder. Zheng et al. [37] reported that Mg-15Gd-0.4Zr (wt%) alloy powders with the size ranging from 75 to 250 μm were prepared by PRE atomization. As compared with aluminum (Al) powder prepared by gas atomization, it exhibited relatively high sphericity, as shown in Figure 3. The Mg alloy powder was adopted as the base material, and corresponding samples were successfully prepared by laser-directed energy deposition.
Based on the above description, the powder preparation methods are capable of producing spherical powders tailored for AM process. Recently, the rapid development of AM significantly increased the demand for spherical Figure 2. The SEM image of (A) water-atomized powders and (B) gasatomized powders [35] .

LPBF of Mg and its bio-applications
Materials Science in Additive Manufacturing metal powders in the market. The efficient preparation of high-quality spherical metal powders is vital to promote AM technology upgrades and industrial applications. Therefore, the existing powder preparation methods need to be improved. For the gas atomization method, optimizing the structure of the atomizer and improving the kinetic efficiency of the atomized gas are the most effective means to improve the quality and production efficiency of metal powders. For the PRE atomization method, to efficiently prepare high-quality spherical powders, the rotation speed of the electrode bar is increased, while the reasonable plasma heat source and power are also fed.

Parameters optimization
The performance of SLM prepared product strongly depends on the processing parameters, such as laser power, scanning speed, hatch spacing, and layer thickness. The previous studies have shown that the process parameters are considered appropriate when the following requirements are fulfilled in scanning tracks and layers [38] : (π) The scanning tracks should be continuity, (θ) each layer has to be high enough to build up the part cumulatively, (ρ) every layer should be high enough, and (σ) the connection angle between two adjacent layers should be close to 90°.

Laser power
High laser power can improve the wettability of the molten pool and provide more power for powder consolidation. Ji et al. [39] reported an analytical model to predict the grain size of the part after LPBF process. Results show that the maximum surface temperature of the component is not influenced by the scanning speed. In addition, the average grain size reduced with the increasing of laser power. He et al. [40] prepared AZ61 alloy with a relative density of 98%. The forming zone of AZ61 with input laser power from 60 W to 90 W was determined. The results show that the increase in laser power helps improve the densification of the material and the formation of equiaxed grains, thus improving the resistance to degradation and microhardness. However, too high laser power can lead to grain coarsening and a decrease of Al solid solution in the Mg matrix, resulting in increased mass loss and decreased microhardness.

Scanning speed
The appropriate scanning rate can increase the heating time of metal powder. The fabrication of compact component can be achieved with low scanning speed, as it allows longer contact time between powder and laser, thereby increasing the rate of energy transfer to the powder bed. With the scanning speed increasing, less energy is transferred to powder, which always results in incomplete melting of the particles. Shuai et al. [41] reported the influence of SLM processing parameters on the corrosion performance of ZK60 alloy. At an energy density of 600 J/mm 3 , the final as-built part with a relative density of 97.3% was obtained. Spierings et al. [42] analyzed the impact of different laser scanning speeds on the static mechanical properties of SLM-treated scandium (Sc) and zirconium (Zr)-modified Al-Mg alloys. The results showed that the scanning speed could affect the hardness and mechanical performance of the alloy, while yield strength was barely affected by the laser scanning speed.

Laser energy density
Although the variety of laser powers or scanning speeds can significantly influence the forming quality and performance of AM-processed part, it is difficult to describe their effects individually. An empirical formula for evaluating the input energy of laser additive using laser energy density [43] is given below: Where, E v represents energy density (J/mm 3 ), P is the laser power (W), V is the scanning speed (mm/s), H is the hatch spacing (mm), and D is the layer thickness (mm). After forming, the forming quality of samples is measured by density and surface roughness. In general, the surface of the sample with high density is relatively flat and has a low surface roughness, as shown in Figure 4. In the case of low laser energy density, the temperature in the molten pool is also relatively low so that the metal powder cannot be fully melted, and noticeable unmelted powder particles and a large number of pores can be observed between the adjacent molten layers, which result in low density. When the laser energy density is too high, the temperature in the molten pool easily exceeds the boiling point of magnesium powder because the melting and boiling point of magnesium powder is relatively close, resulting in a large amount of powder evaporation, which leads to the formation of local vapor pressure, powder splashing phenomenon, and formation of a large number of pores, as displayed in Figure 5.

Materials Science in Additive Manufacturing
Esmaily et al. [45] reported the influence of laser energy density on the forming quality of WE43 magnesium alloy by changing the laser power and scanning rate and obtained the sintering process window diagram. WE43 has many pores and unmelted metal powders in the low-energy density region, resulting in a very low material density and poor mechanical properties. In the high-energy density area, due to the high temperature, there is apparent burning and evaporation phenomenon, resulting in the keyhole and other phenomena, and the forming quality is poor. The stable molten pool can only be obtained under the appropriate laser energy density in the forming area, which ensures good forming quality and dimensional accuracy and achieves 98.3% density.
Wei et al. [46] studied the influence of different laser energy densities on the forming properties, microstructure, and mechanical performance of AZ91D samples. They found that the forming parts with high density and no obvious macroscopic defects could be obtained between 83 and 167 J/mm 3 . When the laser energy input is over 214 J/mm 3 , the sample cannot be deposited due to solid evaporation, and there is a severe burning phenomenon. When the laser energy input is <77 J/mm 3 , the powder cannot be completely melted, resulting in many holes in the sample. Therefore, only in the appropriate processing window can high-density parts be prepared. The optimization parameters obtained from the current literature are collated in Table 2.  [44] .  [44] .

Post treatments
SLM, as a typical AM technology, can reliably prepare Mg-based products, which especially own complex geometries and do not require molds and accessories [54] . Besides, it also possesses these advantages, such as small machining allowances and high material utilization. Notably, Mg is susceptible to oxidation due to its high chemical reactivity and relatively low melting and boiling points to cause high evaporation during melting [55,56] . Therefore, SLMprocessed Mg alloy parts exhibit many defects such as high surface roughness, porosity, residual stresses, anisotropy, and undesired microstructures, which can greatly reduce the overall performance of the parts. The removal of the defects is of great practical importance, where some post-treatments are commonly performed, especially hot isostatic pressing (HIP) and calcium hydrogen phosphate dihydrate (DCPD).

Hot isostatic pressing (HIP)
HIP is a frequently used thermomechanical treatment method that eliminates the porosity and relieves the residual stresses, thus improving ductility, fatigue resistance, and microstructure. The method is performed under hightemperature and -pressure conditions, in which the temperature and pressure usually reach 1000 -2000°C and 200 MPa, respectively. The working pressure generated by a high-pressure inert gas in a closed vessel is close to the yield point of the as-built parts, thus causing plastic deformation. The parts are pressed evenly in all directions with high temperature and pressure, which eliminates these defects to form a dense and uniform microstructure. Therefore, the treated parts can show high density, good uniformity, and excellent performance. Esmaily et al. [45] processed WE43 Mg alloy through SLM and found that HIP treatment is an effective method to eliminate the processing defects, thereby obtaining almost fully dense parts, as shown in Figure 6. Gangireddy et al. [57] showed that at higher initial porosity, HIP treatment was beneficial for the densification of LPBF formed WE43 magnesium alloys, but could not improve the densification of samples with smaller porosity due to the closed nature of the pores. The cooling rate of laser AM is much higher than that of traditional casting, and an excessively fast cooling rate may be detrimental to the precipitation of strengthening phases. A large amount of residual stress caused by excessive cooling rate, texture, and mechanical anisotropy generated along the direction of heat flow can be eliminated by heat treatment [58] .

Calcium hydrogen phosphate dihydrate (DCPD)
The surface roughness of as-built parts affects the mechanical properties and degradation rate. Coatings can modulate the degradation behavior and improve the biological properties. Wang et al. [59] applied DCPD to the surface coating of JDBM porous scaffolds with helical tetrahedral structural units, as shown in Figure 7. DCPD treatment slowed down the degradation rate of the scaffolds and improved their biocompatibility. Dou et al. [60] used the solgel impregnation method to prepare 45S bioactive ceramic coatings on AZ31 Mg alloy substrates and found that the corrosion resistance was significantly improved. Rojaee et al. [61] synthesized hydroxyapatite coating on AZ91 alloy by electrophoretic deposition process, which significantly improved its corrosion resistance and biological properties.
Razavi et al. [62] prepared nanostructured magnesite and diopside coatings by electrophoretic deposition, which also improved the corrosion resistance and biological activity of magnesium alloys. Kopp et al. [63] used plasma electrolytic oxidation (PEO) to treat the surfaces of WE43 Mg alloy scaffolds with different pore sizes prepared by Table 2. Additive manufacturing process parameters for magnesium and its alloy

Materials Science in Additive Manufacturing
LPBF technique. The results showed that PEO significantly reduced the degradation rate of the scaffolds at the initial stage of DMEM soaking. After soaking for 21 days, the stability of the PEO-modified surface decreased, and the weak parts were eroded and cracked.

Friction stir processing
In recent years, severe plastic deformation (SPD), which is an emerging technology, has been widely used to the posttreatments of SLMed parts. Friction stirring processing (FSP) is a typical SPD technique, with its basic principle derived from friction stir welding (FSW), as shown in Figure 8. It is able to simultaneously use high strains, high strain rates, and high temperatures to fully move the material, which results in grain refinement, second phase fragmentation, homogenization, and densification within the stirring zone [64] . Huang et al. [65] used FSP to reduce the porosity and

Materials Science in Additive Manufacturing
homogenize the microstructure of the SLMed Ti 6 Al 4 V alloy, the results showed that fracture strain increased from 0.21 to 0.65 after FSP. In addition, FSP improves the ductility and fatigue life of the SLMed-AlSi 10 Mg alloy, due to the spheroidization of the eutectic phase, the homogenization of the microstructure, and the reduction in porosity [66] . However, to the best of our knowledge, there are only a very limited number of studies relating to the adjustment of the microstructure and enhanced mechanical properties of SLMed-Mg alloys by FSP. Deng et al. [67] found that FSP treatment of SLMed-Mg alloys resulted in a reduction of porosity from 0.779% to 0.015%, disappearance of melt pool boundaries, columnar to equiaxed transformation, and grain refinement, which significantly improved the mechanical properties of the Mg-Gd-Zr alloy.

Laser beam polishing
Laser polishing is an efficient, non-contact, and fully automated post-treatment technology that provides excellent performance in reducing the surface roughness of SLMed parts. In the past two decades, laser polishing has been widely used to reduce the surface roughness of polymers, metals, ceramics, and other materials. In contrast to conventional polishing processes, laser polishing smoothens rough surfaces using thermal energy to melt a thin layer of material, as shown in Figure 9. In laser polishing, a precisely controlled laser beam is guided onto the surface of the part for polishing. By controlling the energy of the laser beam, the peak on the surface is just melted, and the molten metal is generated due to the multidirectional surface tension and gravity, and then, it is filled into the valley. In this case, there is no material loss during laser polishing. The laser power, spot size, and scanning speed are the main control parameters to obtain a smooth surface [68] . At present, laser polishing has demonstrated its ability to polish various materials, from reflective materials such as aluminum to high strength materials such as Inconel and titanium alloys [69] . Unfortunately, there are no relevant reports on laser polishing for SLMed Mg alloys.

Mechanical properties
An ideal orthopedic implant should have sufficient mechanical strength and excellent stress conduction capability, matches with the surrounding bone tissue, and provides mechanical support during the healing process of bone tissue. Processing method directly affects the mechanical properties of Mg alloys. Compared with cast Mg alloys, due to the influence of grain refinement, solid solution strengthening, and other factors, the mechanical properties of Mg alloys manufactured by laser AM have been improved to varying degrees in various aspects. The mechanical properties of bone are closely related to age, bone health, and bone type. Zhang et al. [70] prepared porous Mg scaffolds with porosity of 33 -54% using fiber deposition hot pressing technology, and their Young's modulus and compressive strength were comparable to those of human cancellous bone.
Hyer et al. [71] used LPBF to prepare a dense WE43 alloy. Under the compressive action, the average yield strength of the newly built WE43 structure was 224 MPa, the compressive strength was 417 MPa, and the failure strain was 9.5%. Under tension, the newly constructed LPBF WE43 has an average yield strength of 215 MPa, a tensile strength of 251 MPa, and a failure strain of 2.6%. A study by Fang et al. [72] on the mechanical properties of hot rolled and WAAM-GTA prepared AZ31 samples showed that the performance of WAAM-GTA AZ31 along the direction of travel was always better than that along the build direction. Whether there were large pores within the specification length of the processed tensile samples, its mechanical properties will vary. Deng et al. [53] systematically analyzed the effect of different solution conditions and aging heat treatment on the microstructure and mechanical properties of SLM GZ112K alloy. Appropriate post-heat treatment is an effective measure to improve the microstructure and mechanical properties of SLMed alloy. SLM-T6 alloy meets the requirements of high tensile strength, SLM-T4 alloy meets the requirements of high ductility, and SLMed alloy achieves a good combination of tensile strength and ductility. The ultra-high yield strength of SLM-T6 GZ112K alloy mainly comes from the following four aspects: Finegrain strengthening with average grain size <3.1 μm, precipitation strengthening of β΄-aging precipitation, secondary phase strengthening of long-period stackingordered (LPSO) structure and X phase, and the extra composite strengthening from the coexistence of basal and prismatic precipitates. Wang et al. [73] studied the effect of geometric design on the dynamic response of additively fabricated Mg scaffolds, and the results showed that geometric design had a significant effect on  Table 3.

Degradation performance
The previous studies have shown that AM prepared Mg and its alloys possess fine grain size and homogeneous microstructure, which can obtain lower degradation rates, mainly due to improved passivation properties and reduced micro-galvanic corrosion [77] . Despite that, their corrosion rate in the body fluid environment is still unable to meet the needs of bone implants. For an ideal biodegradable bone metal implant, they have a corrosion rate of <0.5 mm/year and need to provide mechanical support for 12 -24 weeks to meet the clinical requirements [78,79] . This means that the AM-processed biodegradable Mg and its alloys implants need to be regulated. The corrosion resistance of the additively manufactured Mg and its alloys in the current studies is summarized in Table 4.

Scavenging effect
Many impurities, such as iron, nickel, and copper, in Mg and its alloy commonly exhibit relatively high potential. This causes the micro-galvanic corrosion with matrix and thus increases their degradation rate. In this case, the corrosion resistance of Mg alloys can be significantly improved by adopting high purification methods or reducing the impurity concentration by improving the processing technology. Han et al. [80] found that the degradation rate of high purity Mg in vivo was much lower than that of Mg-containing iron impurities. Cao et al. [81] slowed down the corrosion rate of Mg alloys by adding a certain amount of Zr to the molten Mg alloy to remove the impurity iron. Studies have shown that Zr and iron can easily form a precipitate phase, which precipitate at the bottom of the melt, thereby achieving purification effect. Peng et al. [82] used a zone solidification method to prepare Mg alloys and found that the corrosion rate of the alloys purified by this method was lower than that of conventionally cast Mg alloys.

Alloying treatment
Alloying treatment is an effective way to improve the corrosion resistance of Mg alloys by changing the microstructure and the type of precipitates. At present, researchers have developed a series of biological Mg alloys by alloying methods, and their properties have been studied. Shuai et al. [83] found that with the increase of Al content, α-Mg dendrites and intermetallic compounds

Materials Science in Additive Manufacturing
were effectively refined. Li et al. [84] reported that Zn addition refined the grain size to promote the formation of passivation films on the substrate, thus providing effective protection for the Mg substrate. Luo et al. [85] found that the alloying of rare earths significantly reduced the proportion of β phase and promoted the formation of γ phase with a larger active potential, which reduced the micro-galvanic corrosion with the Mg matrix. In addition, the introduction of alloying elements improves the stability or structural integrity of corrosion product layer on the substrate surface with strong protective ability. Leleu et al. [86] introduced alloying element Y into Mg alloys, which formed a dense surface film after immersion in chloride rinsing solution and played a significant protective role. Willbold et al. [87] added rare earth elements (La, Nd, and Ce) to Mg matrix and found that the rare earth oxides formed on the surface of the Mg alloy to improve the passivation ability of the surface film. Notably, most of rare earth elements have low solid solubility in Mg matrix, thus excessive addition of rare earths can cause galvanic corrosion to accelerate the degradation and bring about cytotoxicity problems.

Amorphous alloy
Amorphous alloy, also known as metallic glass, is a nonequilibrium metal material with excellent corrosion resistance. Wang et al. [88] prepared Mg 60 Zn 33 Ca 7 amorphous alloy by melt spin quenching method. The corrosion potential shifts positively and the corrosion current density decreases, thus showing excellent corrosion resistance. Chen et al. [89] reported the research on Mg-based amorphous alloys, which were processed by traditional copper mold casting process. The amorphous alloys possess great glass forming ability, high fracture strength and surface hardness, and good plasticity. Zhang et al. [90] prepared Mg-based amorphous alloys by rapid cooling. The results show that the corrosion resistance of the amorphous alloy is obviously better than that of the crystalline alloy with the same composition. Zberg et al. [91] also studied the degradation process of Mg-Zn-Ca amorphous alloy as a biodegradable medical material and found that the corrosion resistance was significantly improved. Moreover, there was no obvious hydrogen evolution reaction during the degradation process.

Biological behavior
Medical implants need to have excellent biocompatibility to avoid toxic effects on the human body [93] . At present, there are little reports on the biocompatibility of additively manufactured metal implants. The biocompatibility evaluation of Mg-based degradable metals for AM is still at the cellular and in vitro levels. The factors affecting biocompatibility are mainly its chemical properties and degradation products. Ouyang et al. [94] reported that the large pores of the metal scaffolds were favorable for nutrient supply, while the small pores were favorable for cell growth.
Bioactive ceramics have excellent osteoconductive and bioactivity. Rojaee et al. [95] synthesized hydroxyapatite coating on AZ91 alloy by electrophoretic deposition process, and its corrosion resistance and biological properties were significantly improved. Razavi et al. [62] prepared nanostructured magnesite and diopside coatings by electrophoretic deposition, which also improved the corrosion resistance and bioactivity of magnesium alloys. Tian et al. [96] used ammonium bicarbonate particles as a pore-forming agent, and then prepared porous Mg scaffolds by powder metallurgy process, and coated bioactive ceramics on the surface of Mg scaffolds in a low vacuum environment. The results showed that the coated Mg scaffolds have obvious biological activity, and the coating effectively delays the degradation rate of magnesium stents and improves its mechanical integrity.
Rahimi et al. [97] successfully prepared chitosan and nanofiber coatings on the surface of AZ31 Mg alloy by anodizing combined with electrospinning. The coating not only has good corrosion resistance but also has good cell adhesion and proliferation ability. However, due to the large differences in physical properties and mismatched degradation rates, the surface coating is easy to crack or even falls off after implantation, making it difficult to achieve long-term effective protection. Dutta et al. [98] prepared Mg/bioglass composites by microwave sintering, and the results showed that the corrosion resistance, mechanical properties, and biocompatibility were improved. Obviously, the introduction of bioactive ceramics can not only induce the formation of calcium and phosphorus layers, providing long-term stable protection, but also greatly improve the biological performance.
Recently, the use of mesoporous bioglass as a reinforcing phase to prepare Mg-based composites for bone repair has been proposed. Mesoporous bioglass has uniform and ordered mesoporous channels (2 -50 nm) and high specific surface area (500 -800 m 2 /g) [99] . More importantly, as a silicon-containing active ceramic, a large number of silanol groups will be formed on the interface at the initial stage of degradation, thereby forming a negatively charged silica gel layer. Under alkaline conditions, the silica gel layer acts electrostatically to adsorb Ca 2+ and HPO 4 2in the solution, thereby inducing in situ deposition of apatite. In vitro degradation tests showed that this in situ deposited calcium-phosphorus layer effectively enhanced the biological activity of Mg alloy substrates.

Challenges of LPBF-processed Mg Alloys
Due to the inherent characteristics of Mg alloys such as low evaporation temperatures, high vapor pressures, and a high propensity to oxidize, the manufacture of degradable Mg-based implants through AM presents a great number of challenges.
(i) The preparation of Mg powder that can be used for AM processing and degradable Mg-based implants is difficult. The preparation conditions of Mg powders are extremely demanding and the slightest inadvertence can lead to explosive accidents. Moreover, in the current market, the Mg alloy powders commonly used in AM processing are pure Mg, AZ91D, and WE43 powders. Due to the biological toxicity of Al element, AZ91D alloy contains 9% (mass fraction) of Al; therefore, only pure Mg and WE43 powders are suitable for degradable Mg-based implants. (ii) The AM processing for Mg alloys always produces severe powder splashes due to low evaporation temperature and high vapor pressure of Mg alloys, and this phenomenon is very different from AM processing steel, Ti, or Al. Powder spattering can significantly reduce the stability of the Mg alloy during AM processing, as some Mg powder is removed by steam along the scan path, where defects are likely to occur in subsequent scan passes. In this case, a strategy of powder replenishment is necessary for Mg alloys during AM processing. However, there are no relevant studies on the interaction between Mg powder evaporation, gas flow, and laser input. (iii) The quality of components prepared by LPBF is difficult to maintain consistently. Due to the complexity of the process chain, many potential fluctuations may occur during manufacturing process, which leads to variable quality of LPBF parts. Recently, a large number of scholars have explored the use of machine learning (ML) algorithms to overcome this obstacle using datasets obtained at various stages of the LPBF process chain [100] . Before LPBF, ML algorithms can be used for part design and document preparation. During the LPBF process, ML can be applied for process parameter optimization and in situ monitoring [101] . In addition, ML can also be integrated into postprocessing. Therefore, in the future, it is promising to attempt to integrate ML algorithms into different stages of the LPBF process chain to better control the quality of LPBF Mg alloys.

Challenges of Mg alloys in biomedical applications
In the early stage, most of the medical Mg alloys were in the basic research stage, and the types of alloys that can be clinically applied are rare. At present, there are only high-purity Mg and WE43. Other Mg alloys still face great challenges in clinical applications, including the following problems.
(i) The degradation rate is too fast. Since the electrode potential of magnesium is −2.37 V, it usually appears as an active anode. Corrosion reaction occurs in the body fluid environment, and more Mg(OH) 2 is generated on the Mg matrix. The corrosion layer has a loose structure and poor corrosion resistance. In addition, bodily fluids contain a large amount of Cl -, which will further react with Mg(OH) 2 to form the more soluble MgCl 2 , thereby accelerating the degradation rate. (ii) The mechanical strength and toughness are insufficient.
For materials used in bone fixation and support, high strength and moderate plasticity are required, such as yield strength ≥300 MPa and elongation ≥10%; and for materials used in coronary stents and balloon dilators, high plasticity and medium strength are required, such as elongation ≥20% and yield strength ≥150 MPa. Mg alloys are difficult to enhance plasticity with increasing strength. (iii) Biocompatibility verification is insufficient. Mg has good biocompatibility, but other alloying elements are inevitably added in the smelting process, which is potentially toxic to the human body. For example, Al can cause chronic neurotoxicity and lead to Alzheimer's disease; some rare earth elements (Y, Nd, Pr, etc.) are potentially toxic after implantation. The corrosion process will be accompanied by the production of a large amount of OHand H 2 , which can easily trigger an inflammatory response. https://doi.org/10.18063/msam.v1i4.24

Materials Science in Additive Manufacturing
Usually, the regeneration cycle of bone tissue is very long, and the implant needs to remain in the body for about 6 -12 months to fully the reconstruct bone tissue. Bone healing goes through three necessary stages: Inflammation, repair, and bone remodeling. In the first two stages, the bone defect site cannot bear weight, and implants are needed to provide sufficient support to prevent secondary injury. Therefore, in the early stage of healing, Mg alloy implants are required to have high mechanical strength to provide sufficient support; at the same time, a low degradation rate must be maintained to maintain the integrity of the mechanical structure and prevent mechanical failure due to excessive degradation. In addition, it can avoid problems such as swelling and local alkalinity caused by the production of a large amount of H 2 , which results in inflammation of the implantation site. Therefore, improving the corrosion resistance of Mg alloys has become a top priority.

Conclusions
This paper reviews the research progress in the field of laser AM of Mg alloys, and summarizes and compares their powder preparation, process parameters, and post-processing. Biological Mg alloys show great application potential in the field of tissue engineering due to their good mechanical properties, natural degradability, and biocompatibility. However, under the complex human physiological environment, the corrosion rate cannot be effectively controlled. The follow-up research can start from the following aspects: (i) Further studying the influence of alloying elements on the mechanical properties of Mg alloys, analyzing the changing trend of the mechanical properties of magnesium alloys during the corrosion process, and establishing the alloying criteria for biological Mg alloys.
(ii) Carrying out an in-depth study on the corrosion properties of biological Mg alloys in simulated body fluids and blood to find out the influencing factors, corrosion mechanism and corrosion laws in the corrosion process, and to provide scientific basis for the development and clinical applications of biological Mg alloys.
(iii) Setting up dynamic simulation environment, simulating the degradation behavior of Mg alloys in vivo, and collecting and analyzing in vivo experimental data.