Qingsheng Dong ,Yng Li ,Huiqin Jing ,Xingxing Zhou ,Hun Liu ,Mengmeng Lu ,Chenglin Chu,Feng Xue,Jing Bi,*

a Jiangsu Key Laboratory for Advanced Metallic Materials,School of Materials Science and Engineering,Southeast University,Nanjing 211189,China

b Institute of Medical Devices (Suzhou),Southeast University,Suzhou 215000,China

c College of Mechanics and Materials,Hohai University,Nanjing 211100,China

d Department of Oral Implantology,Affiliate Hospital of Stomatology,Nanjing Medical University,Nanjing 210029,China

Abstract Mg-based porous materials,as potential bone tissue engineering scaffolds,are considered an attractive strategy for bone repair owing to favorable biodegradability,good biocompatibility and suitable mechanical properties.In this work,3D-cubic interconnected porous MgxZn-0.3Ca (x=0,3,6) scaffolds were prepared to obtain desirable pore structures with a mean porosity up to 73% and main pore size of 400-500μm,which pore structures were close to the human cancellous bone.The structure-property relationships in the present scaffolds were analyzed by experiments and theoretical models of generalized method of cells (GMC).Mg-xZn-0.3Ca scaffolds exhibited good compression properties with a maximum above 5MPa in yield strength and about 0.4GPa in elastic modulus.This was attributed to not only the alloy strengthening but also the large minimum solid area.On the other hand,the scaffolds showed undesirable and relatively serious degradation behavior in Hank’s solution,resulting from Zn addition in Mg-based scaffolds and the high surface area ratio in the pore structure.Therefore,surface modification are worth studying for controlled degradation in the future.In conclusion,this research would explore a novel attempt to introduce 3D-cubic pore structure for Mg-based scaffolds,and provide new insights into the preparations of Mg-based scaffolds with good service performances for bone repair.

Keywords: Mg-based scaffolds;3D-cubic pore;Structure;Mechanical property;Degradation behavior.

1.Introduction

Orthopedic implant materials have been developed from the firs generation of bioinert materials,the second generation of bioactive materials,to the third generation of bone tissue engineering materials.Recently,bone tissue engineering has a great promising application prospect for bone regeneration,regarded as a synthetic subject combining materials science and biomedical science [1].Bone tissue engineering makes demands on biomaterials performances and structures,and bone tissue engineering materials have been designed as scaffolds with interconnected porous structures that provide convenience for tissue ingrowth and transportation of nutrients [2].Besides,bone tissue engineering scaffolds have been developed with rising demands on biomaterials in terms of mechanical properties,bioactivity,and biodegradability [3].

The ideal biomaterials will play a significan role in bone tissue engineering.Up to now,a variety of biomaterials have been developed for bone tissue engineering [3-5],including bioactive ceramics (e.g.TCP [6],HA [7]),biodegradable polymers (e.g.PLA [8],PCL [9],collagen [10]) and biodegradable metals (e.g.Mg [11],Zn [12]).Noteworthy,Mg and its alloys have attracted wide attention owing to special features in mechanical properties [13,14] and biological functions [15,16],such as suitable elastic modulus and density close to human cortical bone,favorable biodegradability,and good biocompatibility.Moreover,Mg is an essential macro-element for biological activities in human bodies,which can promote osteoblast growth,proliferation,and differentiation so as to accelerate bone repair [17].Thus,the interconnected porous Mg-based scaffolds have been considered as potentially promising materials for bone tissue engineering[18-20].

Recently,lots of preparation methods for Mg-based scaffolds have been developed,including negative salt pattern molding [18,19],titanium wire space holder [20],powder metallurgy [21],laser perforation [22,23] and so on.The negative salt pattern molding method is a representative fabrication technology for interconnected porous materials.Herein,NaCl particles were often served as space holder templates because of a higher melting point than Mg and easier removal in post-process treatments [24].Besides,pore structures in scaffolds are precisely designed based on the controllability of NaCl templates in size and geometry [25].Jia et al.[18] prepared two kinds of spherical and irregular polyhedral porous Mg scaffolds via the negative salt pattern molding method which was reported as precise fabrication.In addition,the pore structure is a typical feature of porous Mg-based scaffolds,and plays a crucial role in service performances including mechanical properties,degradation behavior and biological features.Herein,the porous materials,with the porosity in excess of 50% and the average pore size of more than 300μm,can provide a desirable condition for tissue ingrowth and bone reconstruction [20,26].Witte et al.[27] reported that human cancellous bone has a porosity of about 75%and a pore size of 200-300μm.The increased porosity is beneficia to cell adhesion,proliferation,differentiation and tissue ingrowth,leading to an enhanced bone regeneration effect [28],but would weaken corrosion behavior and mechanical properties [18,19].As mentioned above,it seems a contradiction between porosity and physical-chemical properties.How to obtain suitable Mg scaffolds is still a research focus.

Recently,it has been reported that there were obvious differences in mechanical properties and degradation behavior between 3D-spherical and 3D-irregular-polyhedral interconnected porous Mg scaffolds [18,19].Apart from 3D sphere and 3D polyhedron,3D cube is also a typical and regular pore geometry.If the 3D-cubic templates are arranged in order,the ideal porosity of Mg scaffolds will reach the maximum value of 100% in theory [29].Therefore,the present work investigated the 3D-cubic interconnected porous Mg-based scaffolds for bone repair.Besides,the nontoxic Zn and Ca alloying elements were introduced to alter physical-chemical properties.In addition,the structure-property relationship in Mg-xZn-0.3Ca scaffolds was evaluated through experiments and theoretical analysis.This research would introduce new insights into the relationship between 3D-cubic pore structure and physical-chemical properties,and provide further opportunities to develop 3D-cubic porous Mg-based scaffolds for bone repair.

2.Materials and methods

2.1.Materials

The cubic NaCl templates (from Nanjing Jiayi Sunway Chemical Co.Ltd.),with a size of 400-600μm,were selected as space holder templates.The Mg-xZn-0.3Ca (x=0,3,6,wt%) alloys were cast with pure Mg (99.95wt%),pure Zn(99.95wt%) and Mg-30Ca (wt%) ingots.Mg-Zn-Ca alloys were considered as representative biodegradable Mg alloys owing to great mechanical property and corrosion resistance[30,31].

2.2.Preparation of porous Mg-based scaffolds

The porous Mg-based scaffolds were prepared by negative salt pattern molding.The preparation process is illustrated in Fig.1.Firstly,NaCl templates were put into casting mold(Φ56 mm×260mm).NaCl templates were pre-loaded with a pressure of approximately 0.2MPa,and the pre-loaded pressure can not only increase the porosity of Mg scaffolds but also make cube orientation in order.The casting mold containing NaCl templates was heated to 580 °C,and kept in heat preservation.Besides,the raw materials were fully molten and healed to 740-750 °C and kept heating for 10min under a mixed protective gas of CO2and SF6.Then the molten metals were infiltrate into the casting mold containing NaCl templates based on a negative pressure of 0.1MPa.During the casting process,a mixed gas of CO2and SF6was introduced to retard the oxidation of the molten metals.After solidifi cation,the Mg ingots containing NaCl templates were prepared,which the average size was aboutΦ56mm×100mm.In order to remove NaCl templates,the mixed ingots were ultrasonic-cleaned in 1mol/L NaOH solution to dissolve NaCl templates.Herein,NaOH solution was selected and renewed every 5 h in order to protect Mg scaffolds against corrosion.NaCl templates were removed until the NaCl particles cannot be found in the cross-section of the cast ingot,indicating that porous Mg scaffolds were prepared completely.Then the Mg-based scaffolds were washed in 1vol.%nitric acid alcohol solution for 30s in order to remove the surface layer.Finally,the porous Mg-based scaffolds were dried,and vacuum packaged.

2.3.Structure characterizations

The size distribution of NaCl templates was identifie by a laser particle size analyzer (Mastersizer 3000E).For Mg scaffolds,the pore structure characterizations were observed by an optical microscope (OM),and the internal configuration were re-established by computed tomography (CT,YXLON CT Precision).The pore size distribution in the as-prepared scaffolds was collected by the Image-Pro Plus program based on the color deviation,in which the pore size of less than 50μm was neglected in order to reduce errors.The surface morphologies and relevant elemental analysis were characterized by a scanning electron microscope (SEM,Sirion 200)equipped with an energy dispersive X-ray spectrometer (EDS,Oxford X-Max).The phase composition was analyzed by an X-ray diffractometer (XRD,Bruker D8-Discover) at a scanning rate of 8°/min ranging from 10°to 90°.

Fig.1.Schematic diagram of negative salt pattern molding for Mg-xZn-0.3Ca scaffolds.

2.4.Mechanical properties

The compression performances were evaluated by a uniaxial testing machine (SANS CMT4503).The specimens for compression tests were cut into a dimension ofΦ10 mm×10mm,and the loading direction was along with the pre-loaded pressure direction on NaCl templates.The compression tests were carried out at a compression speed of 1mm/min at room temperature,and repeated three times in each case.

2.5.Degradation behavior

The specimens for corrosion tests were cut into a dimension ofΦ12 mm×1.5mm,and the exposed surface area was about 5.4 cm2.The corrosion tests were composed of hydrogen evolution and mass loss measurements,in which three samples in each group were evaluated to ensure repeatability.The degradation behavior was evaluated in Hank’s solution at 37±0.5 °C [32],in which Hank’s solutions were composed of NaCl (8.0g/L),KCl (0.4g/L),NaHCO3(0.35g/L),MgCl2·6H2O (0.1g/L) MgSO4·7H2O (0.06g/L),CaCl2(0.14g/L),Na2HPO4(0.06g/L),KH2PO4(0.06g/L)and glucose (1g/L)).The ratio of the surface area to Hank’s solution volume was about 1 cm2:50mL.The escaped hydrogen was collected by an inverted buret.Besides,the morphologies and phase structures of Mg-based scaffolds after immersion tests were characterized by SEM,EDS and XRD.In addition,the specimens before and after immersion were weighted to evaluate the weight loss during immersion tests.Prior to weight loss evaluation,the corrosion products were removed when immersed in a boiled solution containing 200g/L CrO3and 10g/L AgNO3.Based on the weight before and after immersion tests,the average degradation rate was calculated based on Eq.(1) [33]:

WhereWbefore(unit:mg) andWafter(unit:mg) are the weight of specimens before and after immersion tests,respectively.Pmis the degradation rate (unit:mm/year),ΔWis weight loss (unit:mg/cm2/d),Ais the total surface area(unit:cm2),tis the immersion time (unit:d),andρis the density of Mg alloys (1.739g/cm3for Mg-0.3Ca,1.901g/cm3for Mg-3Zn-0.3Ca,2.063g/cm3for Mg-6Zn-0.3Ca).

3.Results and discussion

3.1.Structure characterizations

Fig.2 shows the characterizations of NaCl templates for the negative salt pattern molding method.As shown in Fig.2(a),NaCl particles are typical 3D cubes with the edges of 400-600μm.The size distribution of NaCl templates is displayed in Fig.2(b),mainly ranging from 300μm to 800μm (D50=542μm).

Pore structure characterizations of as-prepared porous Mgbased scaffolds are displayed in Fig.3.Corrosion is unavoidable during the removal of NaCl template,but the pores mainly remain cube-like shape as same as NaCl templates,as shown in Fig.3(a,b).3D model of Mg-based scaffolds was reconstructed by CT,and the pore structure characterizations was obtained from CT results,as exhibited in Fig.3(c)and Table 1.The porosity reaches 73.4±2.5%,close to the porosity in human cancellous bone(75%)[27].The CT reconstructed cross-section morphology is displayed in Fig.3(d).Fig.3(e) is the pore size distribution curve in Mg scaffolds,collected from Fig.3(d) by Image-Pro Plus program.In addition,the open-pore structures are composed of main pores(400-500μm) and interconnected pores (50-150μm),which is close to that of human cancellous bone (300-400μm) [27].The porous scaffolds with an average pore size of more than 300μm are able to provide a better physiological condition for cell proliferation and tissue growth [20,26].Herein,the main pores are uniformly distributed while the interconnected pores provide good connectivity in Mg-based scaffolds.The pore size displays a Gaussian-like distribution trend similar to NaCl templates (in Fig.2(b)).However,the results from the image analysis method might be smaller than that from laser particle analyzer.It is because that the image analysis method identifie the cross-section pore size while the laser particle analyzer detects the projector size of 3D-cubic NaCl particles.Based on the characterizations on pore structure,the as-prepared Mg-xZn-0.3Ca scaffolds are well-matched with human cancellous bone in terms of pore structures.

Fig.2.Characterizations of NaCl templates for the negative salt pattern molding:(a) the morphologies of NaCl templates,and (b) the size distribution curves.

Fig.3.Pore structure characterizations of as-prepared porous Mg-based scaffolds:(a) OM,(b) SEM,(c) CT reconstructed 3D model (red regions represent pores),(d) CT reconstructed cross-section morphology,(e) pore size distribution.

Table 1 Pore structure features of as-prepared 3D-cubic porous Mg-based scaffolds.

Fig.4.Summary on porosity of Mg-based scaffolds via typical preparation methods (reproduced from the data in References [18-21,24,34-38,40-43]).

Fig.4 summarizes the porosity of various porous Mgbased scaffolds via typical preparation methods,including powder metallurgy [21,34-38],negative salt pattern molding[18,19,24,39],titanium wire space holder [20,40],fibe deposition hot pressing [41],laser perforation [42],and additive manufacturing [43].According to the as-reported researches,Mg-based scaffolds with higher porosity were realized through negative salt pattern molding.The 3D-cubic porous Mg-based scaffolds exhibit a porosity of 73.4±2.5%,which is superior to the most data on porosity,and very close to human cancellous bone [27].

Herakovich et al.[29] proposed a generalized method of cells (GMC) to evaluate the influenc of pore geometry and porosity on equivalent compression strength of porous materials.The porous materials with 3D cube were reported a wide range of porosity ranging from 0% to 100%.Based on GMC,3D-cubic pores and 3D-spherical pores are discussed for comparison.Assuming that each main pore is located in a cell of 1×1×1,the 3D-cubic porosity is calculated according to the following equation:

Wherearepresents the edge length of a cubic pore in the cell,ranging from 0 to 1;Pcubicis the porosity of 3D-cubic porous material.In Eq.(2),0

By contrast,3D-spherical porosity (Pspherical) is calculated based on the following equation.

Wherebrepresents the diameter of 3D-spherical pore,ranging from 0% to 100%;Psphericalis the 3D-sphere porosity,ranging from 0 toπ/6 (52.36%).When the pores are formed in porous materials arranged in face-centered cubic order,the porosity might reach a maximum of 74.05%.In other words,(rrepresents the radius of 3D-sphere pore),thus(≈74.05%).Compared with 3D-cubic pore,Mg-based scaffolds with 3D-spherical pore have a limited designed porosity range in theory.

There is no denying that the theoretical porosity evaluated by GMC model might be amended in the present experiments,but the above discussion is also an available evaluation for the design of porous materials.Besides,the porosity in interconnected pore regions could be regarded as 100%.The interconnected pores are formed in the porous materials,which actually increase the porosity.The pore structures of open-pore materials are composed of main pores determined by NaCl templates and interconnected pores where the porosity is define as 100%.Owing to interconnected pores,Jia et al.[18,19] prepared a spherical porous Mg scaffold with a porosity of 75.14±0.35%,very close to the theoretical porosity (74.05%) in face-centered cubic order,which is a considerable porosity in the reported Mg scaffolds via negative salt pattern molding methods (in Fig.4).In addition,3D-cubic porous Mg-based scaffolds have a wide designable porosity range in theory,and the as-prepared Mg-based scaffolds exhibit a high porosity close to the previous researches on 3D-spherical porous Mg scaffolds [18,19].All in all,3Dcubic porous Mg-based scaffolds have a great potential to upgrade in porosity.

3.2.Compression properties

Fig.5 shows the compression test results of three 3D-cubic porous Mg-based scaffolds.The compression stressstrain processes are composed of three stages [18],including elastic stage,yield stage and densificatio stage,as illustrated in Fig.5(a).During the firs elastic stage,the stress is proportional to the strain,and the compression stress-strain curve is linear.Besides,the pore structure keeps basically stable.With the increase in the compression load,the compression process enters into the yield stage,during which the plastic deformation results in that the pore walls are broken down.Furthermore,a large compression force causes a serious deformation so that the pore structures are collapsed and the Mg-based scaffolds become densified which is corresponding to the densificatio stage.In this stage,the compression stress as a function of the strain gets a steep rise,and the curve slope is rapidly increased.Fig.5(b) displays the compression properties of the as-prepared Mg-xZn-0.3Ca scaffolds.The porous Mg-xZn-0.3Ca scaffold exhibits compression elastic modulus and compressive strength (yield strength),which are similar to the human cancellous bone,satisfying the demand of clinic applications for bone tissue engineering.These mechanical parameters are much lower than bulk Mg-based alloys [44],avoiding stress shielding effects in cancellous bone healing process.Among three kinds of Mg-xZn-0.3Ca alloys,the porous Mg-3Zn-0.3Ca scaffold shows a superior mechanical property,and the compression elastic modulus is 0.39±0.05GPa and the compression strength is 5.2±0.3MPa.For cancellous bone,the compressive strength is located in 2-20MPa,and Young’s modulus is between 0.1 and 2GPa [45].Thus these porous Mg-based scaffolds could satisfy the demand of mechanical properties for cancellous bone repair.

Fig.5.Typical compression stress-strain curves (a) as well as the relevant compression properties (b) of three 3D-cubic porous Mg-based scaffolds.

Fig.6.Summary on compression properties of Mg-based scaffolds (reproduced from the data from References [18-21,24,34-38,40-43]) (a,b),and fMSA curves as a function of porosity (c).

The ideal scaffolds in bone tissue engineering are desired for both favorable pore structure and approximate mechanical properties.Pore size has been reported no obvious effects on yield strength and elastic modulus [46].However,the mechanical properties are inversely proportional to the porosity [11].It is a research topic on how to obtain enhanced mechanical properties based on the non-decreased porosity.Both pore geometry and material performance are significan factors to the mechanical properties of porous scaffolds.For comparison,Fig.6(a) summarizes the compression strength of Mg-based scaffolds via typical preparation methods[18-21,24,34-38,40-43].With the increase in porosity,the compression strength shows a decreasing trend.The asreported compression strength as a function of porosity is fitte with an exponential formula,as shown in Eq.(4).

Wherea=4.72405±0.06851,b=-0.05574±0.00273,and the adjusted R square in fittin is 0.85018.Rice [47] reported the commonly used exponential relationship,and the property ratio of Mg scaffolds to the bulk one is given by e-cP(c is a parameter determined by the character of the porosity).In addition,an interesting point to consider is that Mg-based scaffolds in the present work exhibit a slightly higher compression property compared with the fittin formula.

On the one hand,the alloying with Ca and Zn is a strengthening method for mechanical properties [15].Mg-0.3Ca alloy is composed ofα-Mg and Mg2Ca phases while Mg-3Zn-0.3Ca and Mg-6Zn-0.3Ca alloys consist ofα-Mg,Ca5Mg55Zn43,and Ca2Mg5Zn5phases.The alloying with Zn exceeding 4.0wt%would tend to the formation of Zn-rich eutectic phases,leading to coarse microstructure [30].In conclusion,the porous Mg-3Zn-0.3Ca scaffold displays better mechanical properties via the alloying with an appropriate Zn content.The detailed descriptions of microstructure are provided inSupplementary material.

On the other hand,pore geometry is also an important factor in mechanical properties.Rice [47] put forward the Minimum Solid Area (MSA) model to infer the relative compression strength.In this model,the relative compression strength is proportional to the MSA fraction in the porous materials along with the loading direction,as shown in the following equation.

Whereσ*is the equivalent compression strength,σrepresents the compression strength of bulk materials,andfMSAis the MSA fraction.

For 3D-cubic porous materials,

WherePcubic=a3as shown in Eq.(2).

For 3D-sphere porous materials,

In order to schematically evaluate the mechanical properties of porous Mg-based scaffolds,the relationship betweenfMSAandPis demonstrated in Fig.6(c).The 3D-cubic porous Mg-based scaffold shows a higherfMSAthan 3D-spherical porous scaffolds at a constant porosity,demonstrating superior compression strength (in Fig.6).Hence,the design of pore geometry provides a potential research approach to synergistically improve both porosity and mechanical performance.

3.3.Degradation behavior

Fig.7 shows the curves of pH variation,hydrogen evolution and residual mass percentage of three porous Mg-based scaffolds immersed in Hank’s solution.The pH variation and hydrogen evolution are related to the degradation of Mg alloys in the physiological environment,as summarized in Reactions(1,2) [48].

The pH curve as a function of immersion time is consistent with the hydrogen evolution curve(in Fig.7(a,b)).Among the three Mg-based scaffolds,Mg-0.3Ca alloy exhibits the lowest corrosion rate.Besides,the corrosion rate of Mg-xZn-0.3Ca alloys increases with the addition of Zn content (in Fig.7(b)).Fig.7(c) exhibits weight loss curves during immersion tests,and the fully-degradation period seems to be less than 1 week for the present experimental samples.The average corrosion rate,calculated based on weight loss[33],is 1.68mm/year for Mg-0.3Ca,1.98mm/year for Mg-3Zn-0.3Ca,2.39mm/year for Mg-6Zn-0.3Ca,respectively.Corrosion often originates from surface,and it is no doubt that a larger surface ratio of 12.0±1.3 cm2/cm3enlarges total degradation behavior.In addition,the bone fracture healing process undergoes up to 12-24 weeks [15] so that the present Mg-based scaffolds cannot meet clinical requirements.Therefore,it is worth investigating surface modification for enhanced corrosion resistance in the future.

Fig.8(a-c) shows the corrosion morphologies of porous Mg scaffolds.During immersion in Hank’s solution for 36h,lots of corrosion products are gradually formed on porous Mg scaffolds.Mg-6Zn-0.3Ca alloy is seriously corroded,and the pores are mostly covered by corrosion products,and these results correspond to those from Fig.7.Furthermore,the corrosion products are identifie by XRD and EDS,as shown in Fig.9.The corrosion products are composed of Mg(OH)2(JCPDS No.07-0239) and few hydroxyapatite (JCPDS No.09-0432) phases.Herein,there are relatively weaker diffuse peaks of Mg(OH)2and hydroxyapatite phases detected on the Mg-0.3Ca scaffold,which is corresponding to fewer corrosion products as shown in Fig.8(a).Fig.8(d-f) are the morphologies of porous Mg-based scaffolds immersed 36h after removing corrosion products.Corrosion behavior might accelerate the heteromorphism of pore structure [19],but 3D-cubic pore structure characteristics still remain in the immersed Mgscaffolds,as shown in Fig.8(d-f).The average corrosion rate of Mg-xZn-Ca alloys is 1.68-2.39mm/year,and bulk Mgbased implants seem to be served for long time.However,the large surface ratio in porous Mg scaffolds might seriously shorten the degradation process.

Fig.7.Variation curves of pH (a),hydrogen evolution (b),and residual mass percentage (c) of three porous Mg-based scaffolds immersed in Hank’s solution.

Fig.8.Surface morphologies of porous Mg scaffolds immersed in Hank’s solution for 36h before (a-c) and after (d-f) removing corrosion products:(a,d)Mg-0.3Ca,(b,e) Mg-3Zn-0.3Ca and (c,f) Mg-6Zn-0.3Ca.

Fig.9.XRD patterns (a) and EDS elements analysis (b) of Mg-xZn-0.3Ca scaffolds immersed in Hank’s for 36h.

Degradation behavior is also ascribed to both material performance and pore geometry.On the one hand,Ca2Mg5Zn phase is prone to form in Mg-3Zn-0.3Ca and Mg-6Zn-0.3Ca alloys.Owing to the obviously different electrochemical behavior between Zn-rich second phases andα-Mg phases [49],the relevant galvanic corrosion would accelerate the degradation rate.The degradation rate of bulk Mg-xZn-0.3Ca alloys is dependent on microstructure.But it seems unreasonable that the total degradation behavior of porous Mg-based scaffolds is evaluated by degradation rate (unit:mL/cm2or mm/year) of bulk Mg alloys.One the other hand,the pore geometry determines the surface area exposed in corrosive electrolytes,and the large surface ratio would aggravate total degradation behavior [50].Herein,the surface area ratio (R)is discussed and calculated by GMA,as follows.

For 3D-cubic porous materials,

WherePcubic=a3as shown in Eq.(2)

For 3D-sphere porous materials,

Fig.10 demonstratesRas a function of porosity(P),based on Eqs.(10,13).The surface area ratio (in Table 1) is wellfitte with the relationship curve in Fig.10.Compared with the 3D-sphere porous materials,the porous materials with 3Dcube pores display a largerRat a constant porosity,leading to a larger amount of hydrogen evolution and weight loss (in Fig.7(a,b)).However,regardless of this serious degradation behavior,the largeRprovides a compatible platform for cell adhesion,proliferation and differentiation [20,26].In conclusion,3D-cubic pore structure demonstrated many favorable features in mechanical properties and biological functions.Therefore,surface modification are required to improve biocorrosion resistance in our further research.

Fig.10.Schematic curves of surface area ratio as a function of porosity.

4.Conclusion

In this work,3D-cubic interconnected porous Mg-xZn-Ca(x=0,3,6) scaffolds were prepared to obtain the porosity of 73.4±2.5% and pore size of 400-500μm,which is matched with human cancellous bone.The Mg-based scaffolds exhibited compression strength (yield strength) of 3-5.9MPa and elastic modulus of 0.24-0.39GPa,which meet the requirements of mechanical properties for cancellous bone repair.The optimized mechanical properties were attributed to not only alloying strengthening but also large minimum solid area.Besides,the degradation behavior depended on both material characteristics and exposed surface area in physiological solution.Zn alloying in Mg-xZn-0.3Ca alloys worsened corrosion resistance,and a high surface area ratio aggravated total degradation.However,the high surface area ratio is a favorable feature for biocompatibility by providing a large platform for cell growth,proliferation,and differentiation.Therefore,surface modification are worth investigating in the future.In conclusion,3D-cubic pore structure exhibits many special characteristics in chemical-physical properties.This research would provide new insights into 3D-cubic interconnected porous Mg-based scaffolds for bone repair.

Declaration of Competing Interest

None.

Acknowledgment

This work was supported by the National Key Research and Development Program of China (No.2016YFC1102402),the National Natural Science Foundation of China (No.51771054,No.51971062),the Science and Technology Project of Jiangsu Province (No.BE2019679) and the Fundamental Research Funds for the Central Universities (No.2242018K3DN03,No.2242019K40057).

Supplementary materials

Supplementary material associated with this article can be found,in the online version,at doi:10.1016/j.jma.2020.05.022.