Advances in Fe-based biodegradable metallic materials

Abstract

In recent years, biodegradable metallic materials, such as Mg-, Fe-, Zn- and W-based materials, have been the focus of many studies. As one of the two most studied types of biodegradable metallic materials, Fe-based materials have aroused a great deal of interest because of their outstanding mechanical properties, which are similar to stainless steel. The processing methods can directly affect the microstructure of the material and influence the mechanical and degradation properties of the material. Furthermore, biocompatibility is directly affected by the degradation properties. Therefore, the processing methods, mechanical properties, degradability and biocompatibility are several of the main concerns in the study of biodegradable Fe-based materials. Here, we systematically summarize recent studies on Fe-based materials and discuss these findings in terms of their processing methods, degradability and biocompatibility.

---

1. Introduction

The use of metallic materials in the medical field has a long history. Traditional biomedical metallic materials, such as Ti (Ti alloy), Co alloy and stainless steel,^1–3 have excellent biocompatibility and mechanical properties. Thus, they occupy a dominant position in medical implant materials. Table 1 gives a summary of the major merits and shortcomings of some frequently studied biomedical metallic materials. Among these materials, Ti (Ti alloy), Co alloy and stainless steel have already been applied commercially. However, these materials are non-degradable and permanently exist in the body as a foreign object, which can cause negative effects.

Table 1 A summary of the main merits and shortcomings of biomedical metallic materials^{1–3,5,6,8–10}

Biomaterials	Typical samples	Merits	Shortcomings
Fe-Based materials	Pure Fe, Fe–Mn, Fe–Fe2O3	Good mechanical properties, appropriate biocompatibility, various processing methods, biodegradable	Low degradation rates, ferromagnetism
Mg/Mg alloys	Pure Mg, Mg–Zn, Mg–Zn–Ca	Excellent biocompatibility, biodegradable, similar mechanical properties with bone	Too fast degradation rates, hydrogen evolution
Ti/Ti alloys	Pure Ti, Ti–6Al–4V	Good mechanical properties, excellent biocompatibility, high specific strength, good corrosion resistance	Non-degradable, toxicity elements (e.g., V)
Stainless steel	316L	Excellent corrosion resistance, high mechanical properties, appropriate biocompatibility	Non-degradable, toxicity elements (e.g., Ni), too high modulus, poor plastic workability
Co alloys	Co–Cr–W–Ni, Co–Cr–Mo	Excellent wear resistance, appropriate biocompatibility, high corrosion resistance	Too high modulus, poor cold workability, toxicity elements (e.g., Ni), non-degradable

---

In recent years, biodegradable metals have aroused a great deal of interest because of their attractive properties, such as the excellent mechanical properties that are similar to traditional biomedical metallic materials, superior biocompatibility and degradability.^4–11 Zheng et al.⁹ defined biodegradable metals: biodegradable metals are metals expected to corrode gradually in vivo, with an appropriate host response elicited by released corrosion products, then dissolve completely upon fulfilling the mission to assist with tissue healing with no implant residues. Over the years, many types of biodegradable metals have been studied, including Mg-,^5,6 Fe-,^8–10 W-^12–15 and Zn-based materials.^16–19 Currently, the study of biodegradable metals mainly focuses on two categories: Fe-based and Mg-based materials. There is relatively less research conducted on other biodegradable metals. Fe is a traditional engineering material that has many advantages, such as excellent mechanical properties, superior machinability and low price. These inherent advantages make Fe-based materials a potentially exceptional source of biodegradable metallic materials for the medical field.

2. Application of biodegradable Fe-based materials

According to reported studies, the applications of biodegradable Fe-based metallic materials mainly consist of three categories: intravascular stents, implant materials for bone surgery and scaffolds for bone tissue engineering.

2.1 Stents

Stents have been widely used for the treatment of cardiovascular disease. However, their applications can produce undesirable complications.²⁰ Restenosis can occur within the stent because of inflammation and neointimal proliferation. Furthermore, stents implanted in children result in a fixed vessel diameter, and subsequent restenosis occurs due to vessel growth at the stent site. Degradable stents can overcome these shortcomings.²⁰ This type of stent can eliminate the continuous interference process between the stent and the vessel wall, and its use as a carrier for drugs can avoid re-dilatation surgery. Another potential advantage of degradable stents is the avoidance of subsequent surgeries on vessels with implanted stents.

An inherent advantage of iron is its excellent mechanical properties, which make it an interesting candidate for use in biodegradable stents.⁸ For example, the higher elastic modulus of iron can produce a high radial strength of the stent, which is beneficial when producing stents with thinner struts. In addition, iron has high ductility, which can be useful when the stent is plastically deformed during the implantation procedure.³ Therefore, the development of degradable stents is one of the main uses for degradable Fe-based materials. Fig. 1 shows examples of Fe and Fe–Mn alloy stents.

Fig. 1 Photographs of a Fe stent (a) and Fe–35Mn alloy stents (b) (reproduced with permission of Elsevier from ref. 20 and 21).

2.2 Implants for bone surgery

The non-degradable metal implants used in bone surgery, such as screws²² and plates,²³ can lead to adverse responses, such as a toxic response to specific ions released from the materials due to their long-term presence in the body. Although the implants can be removed as a second operation, it can result in adverse consequences, such as a second defect at the implantation site, pain, and increased treatment costs. Fe-Based materials are a good candidate to address these shortcomings. First, Fe-based materials have excellent mechanical properties that are similar to those of stainless steel and can satisfy the requirement of high mechanical properties for bone repair. In addition, Fe-based materials have good biocompatibility. Above all, Fe corrodes in the body environment and is expected to be completely dissolved during the lifetime.

2.3 Scaffolds for bone tissue engineering

Scaffolds play an important role in tissue engineering and provide essential conditions for tissue repair. For example, scaffolds provide a surface and space that allows for cell growth and new tissue formation, thus aiding in the remodelling of damaged tissue.²⁴ An ideal scaffold must have reliable degradation and resorption rates that are consistent with the rate of tissue regeneration. Moreover, scaffolds must have appropriate mechanical properties that match those of normal tissues and organs.²⁵ Because metallic materials have excellent mechanical properties, the application of biodegradable metal porous scaffolds is primarily used for the reparation of hard tissue, such as bone. Thus, these novel scaffolds avoid the shortcomings of traditional non-degradable metal scaffolds and overcome the weaknesses of polymer scaffolds.

3. Manufacturing methods and mechanical properties

Many methods, including the manufacturing methods for raw materials and post-treatment methods (Fig. 2), have been used to manufacture biodegradable metallic materials; notably, these methods are similar to those used with traditional metals. The properties of the material, such as mechanical properties and degradability, can be affected by the manufacturing method. Indeed, mechanical properties are most vulnerable. Table 2 summarizes the mechanical properties of the materials after processing with different methods. According to published studies, the methods can be divided into five types: casting, sintering, electrodeposition, surface modification and other methods. Except for surface modification, the other methods refer to the manufacturing or processing of raw materials.

Fig. 2 Diagram of the different manufacturing methods. Initial methods refer to the production methods that transform raw materials into materials that can be used for device manufacturing. Post-processing refers to further treatments after the production of raw materials. Note: surface modification is not an initial method, but it is one of the most studied manufacturing methods.

Table 2 Summary of the manufacturing methods and the mechanical properties of different Fe-based materials^a

Processing methods	Components	Mechanical properties								Ref.
		Tensile				Compress			Hardness
		Ultimate strength (MPa)	Yield strength (MPa)	Elasticity modulus (GPa)	Elongation at break (%)	Yield strength (MPa)	Ultimate strength (MPa)	Elasticity modulus (GPa)
a Note: pure Fe that served as a control is not listed in this table. The error values in the original data from the references are not shown. The values of the same materials manufactured using different processes were expressed as a range.
Casting + forging	Fe–(0.5–6.9)Mn	353–1041	295–814	—	11.5–31.3	—	—	—	—	38
Casting + forging + heat treatment	Fe–10Mn	1300–1400	650–800	—	14	—	—	—	374–428 (HV10)	32
	Fe–10Mn–1Pb	1450–1550	850–950	—	2.0–11.0	—	—	—	376–437 (HV10)
Casting + forging + heat treatment	Fe–21Mn–0.7C(–1Pd)	1255	725	—	38	—	—	—	—	33
(Armco ingot Fe) rolling + annealing	Fe-UD75%	600	593	—	3.5	—	—	—	—	35
	Fe-BD75%	517	507	—	4.0	—	—	—	—
	Fe-UD + annealing	248–283	108–246	—	34.8–48.3	—	—	—	—
	Fe-BD + annealing	232–274	97–229	—	40.2–51.7	—	—	—	—
Casting + forging	Fe–30Mn	632	242	—	94	413	—	—	175 (HBW)	39
Casting + forging	Fe–30Mn	530	—	—	15	—	—	—	175 (HV 5)	40
SPS	Fe–5Pd	—	—	—	—	445	754	—	—	41
	Fe–5Pt	—	—	—	—	503	785	—	—
Sintering	Porous Fe	—	—	—	—	16.1–67.7	—	0.6–4	—	42
Sintering	Fe–35Mn	20.2–30.1	—	—	0.74–0.95	—	—	—	13.58–27.72 (HRH)	43
Sintering + cold rolling + sintering	Fe–35Mn	550	235	—	31	—	—	—	38 (Rockwell A)	44
Sintering + cold rolling	Fe–(20–35)Mn	428–723	234–421	—	4.8–32	—	—	—	38–59 (Rockwell A)	45
Sintering	Fe–HA	—	—	—	—	325	717	—	—	46
	Fe–TCP	—	—	—	—	312	708	—	—
	Fe–BCP	—	—	—	—	312	696	—	—
Sintering	Fe–TCP	—	—	—	—	696	—	—	—	47
Electroforming + annealing	Fe	169–292	130–270	—	18.4–32.3	—	—	—	—	48
Electrodeposition	Fe/Fe–W alloy scaffold	4.01	—	—	—	—	—	—	—	49
3D-printing + sintering	Fe–30Mn	115.53	106.07	32.47	0.73	—	—	—	—	50
ECAP	Fe	470	—	—	—	—	—	—	444 kgf mm^−2	51
Nitriding	Fe	614.35	—	—	—	—	—	—	287.06 (HV)	52
Magnetron sputtering	Fe	343–638	267–612	—	1.3–20	—	—	—	—	53
Vacuum infiltration and dipping	PLGA-infiltrated porous Fe	—	—	—	—	0.20–0.38	0.24–0.42	4.17–8.78 MPa	—	54
	PLGA–coated porous Fe	—	—	—	—	0.27–0.65	0.30–0.71	6.15–14.22 MPa	—

---

3.1 Casting

As a traditional manufacturing method of metals, the advantage of casting is the convenience of adjustments to the alloy components. A certain proportion of raw materials (powders or blocks) are melted in a furnace and then cast to obtain alloys. Many researchers prefer this method of preparing biodegradable iron alloys. However, the poor properties of as-cast alloys are induced by defects (such as segregation, blowholes and shrinkages). Therefore, as-cast alloys must be further processed by forging, rolling and/or heat treatment.

Liu et al.²⁶ casted a Fe30Mn6Si ternary alloy with a shape-memory function. The alloy consisted of martensite and austenite and had a smaller grain size and higher ultimate strength when compared with pure Fe. The recovery of this alloy reached 53.7% (casting + solution treated). The authors also produced a series of binary Fe alloys (Fe–Mn/Co/Al/W/Sn/B/C/S) by casting and rolling.²⁷ Here, the addition of Mn, Co, Al, and W had no obvious effect on the grain size of the Fe alloys; however, B notably decreased the grain size. Sn sharply decreased the mechanical properties of the alloy. Mn, Co, W, B, C and S increased the yield and ultimate strength of Fe in the as-rolled alloys; furthermore, these elements increased the magnitude of the difference between the yield strength and ultimate strength of Fe.

Except for casting followed by hot working, more complex technology is typically needed to obtain alloys with specific properties. A Fe–10Mn–1Pd alloy was produced by casting, swaging, solution-heat-treatment and quenching, and the hardening kinetics of a solution-heat-treatment were studied by Moszner et al.²⁸ The authors suggested that the pronounced strengthening was due to the thermally activated formation of coherent plate-like Pd-rich precipitates on 100 matrix planes. The new alloy was somewhat similar to maraging steel, e.g., the initial microstructure, aging behaviour and precipitation; thus, it represents a new high-strength, Ni-free maraging steel. To demonstrate the composition of the Fe–Mn–Pd alloy precipitates, Fe–Mn–Pd alloys containing 10 wt% Mn with varying Pd concentrations (1, 3, and 6 wt%) were created by casting, solution-heat-treatment, quenching, isothermal aging and quenching.²⁹ The authors found that the precipitate was mainly a face-centred tetragonal β1-MnPd phase that consisted of Mn and Pd. The alloys demonstrated remarkable age-hardening after isothermal aging at 500 °C. The mechanism of reverse transformation from martensite to austenite in the Fe–Mn–Pd alloys and the influence of prior precipitation on this process were studied.³⁰ The atomic-scale microstructures of prior austenite grain boundaries in Fe–Mn–Pd alloys were characterized by site-specific atom probe tomography, and the mechanism of embrittlement and de-embrittlement have been discussed.³¹

Schinhammer et al.³² discussed the design strategy of biodegradable Fe-based alloys. The authors suggested that noble intermetallic phases improved the degradation rate and increased the strength of the Fe matrix. For example, the availability of Mn and Pd has been proven. The degradation resistance of the newly developed Fe–Mn–Pd alloys is one order of magnitude less than that of pure iron. Furthermore, both the choice of alloying elements and changes in heat treatment can be used to adjust mechanical properties.

The interaction of recrystallization and precipitation in a biodegradable twinning-induced plasticity (TWIP) steel (Fe–Mn–C–Pd) was studied by Schinhammer et al.³³ The authors found that the formation of a Pd-rich precipitate dramatically impeded recrystallization during the annealing treatment. The impedance was caused by grain boundary pinning by the Pd-rich precipitates and reduced dislocation mobility due to a solute drag effect from the Pd-enrichment of dislocation cores. These alloys showed high strength and superior ductility that exceeded other TWIP steels and the typical alloys used in biomedical implants.

The mechanical properties and corrosion rate of alloys can be affected by the average grain size; thus, these parameters can be optimized by altering the grain size. Obayi et al.³⁴ found that thermo-mechanically treated (cold-rolled and annealed ingot) pure iron had an optimum microstructure with an average grain size of 20–25 μm. The authors also determined that the rolling direction affects the grain size and mechanical properties of pure Fe.³⁵ The grain orientation of unidirectional rolled pure Fe was obvious; however, this orientation was irregular for bi-directional rolled pure Fe. Moreover, the mean grain size and the grain size distribution in bi-directional samples were larger and broader, respectively, when compared with unidirectional samples after annealing. The yield strength and tensile strength of unidirectional samples was larger than bi-directional samples. After annealing, the yield strength and tensile strength were markedly decreased.

Cast Fe–33Mn alloys were further processed with varying methods, including heavy cold-rolling, large-strain machining (LSM) and annealing.³⁶ The grain shape of the alloy using LSM resembled dendrite bands and was approximately 16 dendrite band diameters in length, which is less than that of samples processed with other methods. In order to improve the interface compatibility of implants with bone and orthopaedic soft tissues, a cast Fe–30Mn alloy was further processed by a four-step dealloying process to create a nanoporous surface.³⁷ This surface was designed by the selective leaching of the less noble components (Mn and Zn) from the outer surface layer.

After casting and hot forging, the phase of FeMn30 was mainly composed of γ-austenite and ε-martensite. Its mechanical properties were better than that of iron and 316L stainless steel.³⁹ Manganese remains dissolved in γ-Fe, thus stabilising the austenitic structure at room temperature and causing significant solid-solution hardening and strengthening.⁴⁰ Fe–Mn alloys with lower Mn concentrations (0.5, 2.7, and 6.9 wt% Mn) that were produced by casting and forging exhibited superior mechanical properties.³⁸

3.2 Sintering

Spark plasma sintering (SPS)⁵⁵ is the most advanced sintering technique. SPS has obvious advantages over other sintering techniques, such as lower sintering temperatures and shorter holding times. The technique enables the sintering of nanosized powders and can obtain a nearly full-densification of materials with minimum grain. Cheng et al.⁵⁶ reported that the grain size of the pure iron processed by SPS was far smaller than that of as-cast pure iron. W and CNT, which served as the second phase, were added into the Fe matrix and clearly diminished the grain size; furthermore, the decrease in grain size correlated with increasing second phase fractions. The ultimate compressive strength (UCS) of the Fe matrix composite is greater than pure Fe, but the enhancement of yield strength (YS) for the composite was not significantly different from pure Fe. The YS and UCS of the SPS of pure Fe may be greater than as-cast pure Fe due to the smaller grain size of the former. Similarly, a Fe–Fe₂O₃ composite prepared by SPS⁵⁷ has a fast corrosion rate, good mechanical properties and excellent biocompatibility. Huang et al.⁴¹ prepared Fe-5 wt% Pd and Fe-5 wt% Pt composites using SPS. Because of the second phase strengthening effects, the mechanical properties of the composites, including yield strengths, ultimate compressive strength, rigidity and microhardness, were much higher than that of pure Fe. Furthermore, small grain size improves the mechanical properties of composites. The low ductility of the composites may be due to their relatively low density. Fe–Au and Fe–Ag were also prepared by SPS.⁵⁸ Fe-5 wt% Ag, Fe-2 wt% Au, and Fe-10 wt% Au composites showed improved mechanical properties. This improvement was attributed to the decrease in grain size and second-phase strengthening effects. The enhanced strength of the Fe–Au composite may be due to the ability of Au to partly dissolve into iron and result in solid solution strengthening.

The sintering method is typically used for the manufacturing of porous materials. A polyurethane sponge acted as a template, and the replication method based on powder metallurgy was used to produce an iron matrix composite (Fe–CNTs, Fe–Mg) with open-cell foam structure.^59,60 The pore size of the foam was 35–50 PPI with an apparent density of approximately 24 kg m⁻³. However, there were many closed pores in the foam, which is a disadvantage for tissue penetration. Ammonium bicarbonate served as a space-holder, and porous iron was prepared by vacuum sintering.^42,61 The highest porosity of the samples was 82 vol%. Increased compression pressure decreased porosity, while the use of a finer iron powder and more space-holder material in the initial mixture led to an increase in porosity. Increasing the porosity can decrease the mechanical properties of the sample; however, the use of finer powder increased the porosity and mechanical properties simultaneously. Furthermore, a higher compacting pressure enhanced flexural and compressive properties. The flexural and compressive performance of the sample was similar to that of human bone. A porous Fe–35Mn alloy with an open-cell porosity of 25–31% was prepared by sintering with ammonium bicarbonate as a porogen.⁴³ The mechanical properties were decreased by increasing the amount of porogen used.

The mechanical properties of sintered metals are typically lower; however, they can be improved by rolling. Hermawan et al.⁴⁴ reported that the porosity of a Fe–35Mn alloy prepared by cold-rolling and sintering cycles decreased by 0.3%, and the mechanical properties of this allow were similar to those of 316L stainless steel. This method can also change MnO particles into much smaller particles, which are then distributed in the rolling direction. Furthermore, Fe–Mn alloys with varying Mn content (20, 25, 30, and 35 wt%) were prepared using this method.⁴⁵ The authors reported that alloys with lower Mn content existed mainly in the γ phase with some appearance of the ε phase. The yield strengths were between 234 MPa and 421 MPa, and the elongations were between 7.5% and 32%. Next, the authors processed Fe–35Mn alloy bulk materials produced from a similar method into minitubes with a 1.80 mm outside diameter, 0.15 mm wall thickness and 40 mm length using wire-cut electrical discharge machining and a programmable conventional turning machine.²¹ The minitubes were transformed into stents using laser cutting techniques. The mechanical properties of the stents, including expansion performance and radial force, were satisfactory. However, magnetic stents and those with too rough of a surface were unfavourable.

Using sintering, other elements have also been added into the iron matrix to form alloys. Wegener et al.⁶² prepared Fe–(C, P, B, and Ag) alloys using powder metallurgy. The addition of B and P had a notable effect on the microstructure of the alloy. The addition of P significantly decreased microporosity, while the addition of Ag and B increased microporosity.

Bioceramics have excellent biocompatibility. Therefore, a Fe/bioceramic composite is an interesting biodegradable metallic material. Ulum et al.⁴⁶ prepared Fe/bioceramic (hydroxyapatite (HA), tricalcium phosphate (TCP), and biphasic calcium phosphate (BCP)) composites using sintering. The yield and compressive strength was slightly more than pure Fe. Reindl et al.⁴⁷ prepared a Fe/β-TCP composite with 50 vol% β-TCP using powder injection moulding combined with sintering; here, the maximum compressive yield strength of the composite was 696 MPa, which is markedly greater than the 361 MPa compressive yield strength of pure Fe.

3.3 Electrodeposition

Electrodeposition is a film growth process that consists of the formation of a metallic coating onto a base material via the electrochemical reduction of metal ions in electrolytes.⁶³ Electroforming is an application of electrodeposition that is used for the reproduction of moulds to directly form objects in their final shape.⁶³ The electroforming technique is simple and does not require complex equipment or a large amount of energy consumption. Moravej et al.⁴⁸ reported that electroformed iron had a very fine texture and excellent mechanical properties. Heat treatment decreased the strength of the electroformed iron, whereas elongation was improved. The texture, grain size and grain shape of electrodeposited iron were markedly affected by the current density. The microstructure of the iron film was altered by annealing at 550 °C. This result was due to the recrystallization induced during the annealing process.⁶⁴ Electrodeposition cannot be restricted by shape of substrate, and can even occur on complex surfaces. He et al.⁴⁹ prepared a porous scaffold with a double-layer structured skeleton using a two-step electrodeposition. The structure of the scaffold was similar to that of cancellous bone, and the tensile strength of the scaffold was identical to that of cancellous bone.

3.4 Surface modification

Surface modification is a method typically used to improve the biocompatibility or degradability of biodegradable metal. To prevent the premature loss of mechanical stability of Fe stents by corrosion, oxide or nitride layers were prepared on the surface of pure iron using different methods. For example, a Fe–O thin film was prepared using plasma immersion ion implantation,⁶⁵ a La₂O₃ film was prepared with a metal vapour vacuum arc,⁶⁶ and a nitride layer was created on the surface of pure Fe using plasma nitriding.⁶⁷ Feng et al.⁵² modified a pure Fe stent with plasma nitriding. The Fe₄N particles generated by nitriding were uniformly dispersed in the Fe matrix, which enhanced its mechanical properties, including tensile strength, microhardness, radial strength and stiffness. In addition, metal ions can be implanted in the pure Fe matrix. For example, Ag ions were implanted into the surface of pure iron using a metal vapour vacuum arc (MEVVA) to form an approximately 60 nm gradient layer that consisted of Ag₂O in the outermost layer and Ag atoms in the inner layer of the Fe matrix.⁶⁸ Zn was also implanted into the surface of pure Fe by MEVVA, and the implantation depth reached to about 60 nm. In the outer layer, Zn element existed as ZnO, and Zn existed as Fe–Zn solid solution beneath the ZnO layer.⁶⁹ Some micro-patterned noble metals arrays, such as Au⁷⁰ and Pt,⁷¹ were coated on the surface of pure Fe to improve its degradation performance. Lin et al.⁷² designed a novel iron-based drug-eluting stent, which was created by electroplating Zn and coating sirolimus-carrying PDLLA on the surface of the stents. Nitride iron materials have outstanding comprehensive mechanical properties, thus stents made with this material show excellent device performance. The Zn barrier layer of the stent allows for ultrathin struts, and the thick PDLLA coating maintains adequate scaffolding at 3 months after implantation. A calcium phosphate/chitosan coating was prepared on the surface of iron foam using electrophoretic deposition.⁷³ After immersion in PBS and m-SBF, apatite coatings were formed on the composite surface. Plasma polymerisation was employed to deposit an allylamine coating onto pure Fe samples, and the results showed that an ultra-thin, pinhole-free, polymer-like, and amine-rich layer with a thickness of ∼250 nm was completely and uniformly coated on the samples.⁷⁴

3.5 Other methods

Three dimensional (3D)-printing technology has been used to produce versatile scaffolds with complex shapes that are capable of homogeneous cell distribution; in addition, this technique shows high precision and accuracy for creating intricately-detailed biomimetic 3D structures.⁷⁵ Fe–30Mn scaffolds were processed using inkjet 3D-printing combined with sintering.⁵⁰ The scaffolds had an open porosity of 36.3%. The tensile mechanical properties were very similar to those of natural bone. Equal channel angular pressing (ECAP) is a processing method for nanocrystalline (NC) metallic materials, and it was used to obtain NC pure Fe.^51,76 After eight passes, the mean grain size decreased from approximately 50 μm to 80–200 nm. Accordingly, the micro-hardness and tensile strength greatly increased, and these results conform to the Hall–Petch rule. Furthermore, the contact angle of the NC pure Fe was smaller than that of microcrystalline (MC) pure Fe. This result shows that NC pure Fe has a lower surface energy relative to MC pure Fe. The surface roughness of NC-Fe is larger than MC-Fe, which may be beneficial for protein adsorption, cell spread and attachment. Magnetron sputtering was used for produce pure iron foils.⁵³ The foils showed a preferential orientation, a fine-grained structure and an enhanced mechanical strength when compared with cast Fe. A series of biodegradable wires composed of Fe, Mn, Mg and Zn were manufactured by cold drawing technology.⁷⁷ Because of the fine microstructures and fibre textures produced during cold drawing, these wires demonstrated excellent mechanical properties. Among them, an Fe35Mn–Mg composite wire showed more than a 7% increase in elasticity, a similar fatigue strength, an ultimate strength of more than 1.4 GPa and a toughness exceeding 35 mJ mm⁻³ when compared with 316L stainless steel. Vacuum infiltration and dipping methods were employed to prepare PLGA–porous pure iron composites, including PLGA-infiltrated and coated pure porous iron.⁵⁴ The compressive strength of PLGA–coated porous pure iron was higher than PLGA-infiltrated porous pure iron and similar to that of cancellous bone.

4. Degradation properties

The corrosion rates of different Fe-based materials are summarized in Table 3. An ideal biodegradable metal device should have an appropriate degradation rate that matches with the repair rate at the implant site. For stents, a complete degradation is expected to occur after the vessel remodelling phase, which typically takes 90–120 days.⁹ During this phase, a very low degeneration rate provides sufficient mechanical support to the injured vessel. Depending on the clinical conditions, the mechanical support from bone implants should be sustained for 12–24 weeks, as shown in Fig. 3.⁹ For biodegradable Fe materials, a relatively low degradation rate is currently a major problem. For example, pure iron stents were implanted in blood vessels, and although there were signs of degradation after 1 year, a large portion of the stent remained intact, which can produce negative effects similar to those caused by permanent stents, such as restenosis.²⁰ Thus, improving the degradation rate of Fe is a problem that needs to be resolved by further research into biodegradable Fe-based materials. A variety of methods have attempted to improve the degradation performance of Fe, such as alloying, adding a second phase, surface modification and other processing methods. Fig. 4 summarizes the previous reports on the regulation methods used to control the corrosion rate of Fe-based materials.

Table 3 Summary of the corrosion rates of different Fe-based materials^a

Materials	Corrosion environment	Corrosion rates from different methods				Ref.
		Electrochemical	Static immersion	Dynamic immersion	In vivo
a Note: pure Fe that served as a control is not listed in this table. The error values in the original data from the references are not shown. The following units were changed to g m^−2 d^−1 or mm per year: ^a, μg m^−2 h^−1; ^b, mg cm^−2 d^−1; ^c, mg m^−2 d^−1; ^d, g m^−2 h^−1; ^e, μm per year; and ^f, μg cm^−2 h^−1.
Fe (rolling)	Modified Hank's solution	0.172–0.244 mm per year	0.120–0.146 mm per year	—	—	34
Fe (rolling)	Modified Hank's solution	0.209–0.243 mm per year	0.115–0.144 mm per year	—	—	35
Electroforming Fe	Modified Hank's solution	0.85 mm per year	—	—	—	48
Electroforming Fe, annealed		0.51 mm per year	—	—	—
Nanocrystalline Fe	Hank's solution	^a1.896 g m^−2 d^−1	—	—	—	51
Nanocrystalline Fe	Gas-flow physiological saline	0.009–0.100 mm per year	—	—	—	76
Microcrystalline Fe		0.039–0.127 mm per year	—	—	—
Magnetron sputtered Fe	Modified HBSS	0.06–0.10 mm per year	—	—	—	53
As-Electroformed Fe	Hank's solution	—	0.4 mm per year	—	—	83
E-Fe annealed		—	0.25 mm per year	—	—
CTT-Fe annealed		—	0.14 mm per year	—	—
Fe	SBF	—	—	^f4.896 g m^−2 d^−1	—	84
Fe	Hank's solution	0.105 mm per year	—	—	—	85
Fe–Mn/Co/Al/W/B/C/S	Hank's solution	1.863–3.991 g m^−2 d^−1	0.028–0.361 g m^−2 d^−1	0.678–3.456 g m^−2 d^−1	—	27
Fe–30Mn	HBSS	0.73 mm per year	—	—	—	50
Fe–33Mn (LSM)	Osteogenic media	0.8354 mm per year	—	—	—	36
Hot forged Fe–30Mn	SBF	^b18.91 g m^−2 d^−1	0.028 mm per year	—	—	39
Fe–35Mn	Modified Hank's solution	0.44 mm per year	—	—	—	44
Fe–(20–35)Mn quenched	Modified Hank's solution	0.4–1.3 mm per year	—	—	—	45
Cold rolled		0.5–0.7 mm per year	—	—	—
Fe–35Mn	5% NaCl and SBF	3.72–8.28 mm per year	—	—	—	43
Fe–20Mn (casting)	Osteogenic media	0.9427 mm per year	—	—	—	81
Fe–20Mn (cold rolling)		0.5397 mm per year	—	—	—
Fe–35Mn	SBF	0.51 mm per year	—	—	—	82
Fe–21Mn–0.7C–1Pd	SBF	—	0.21 mm per year	—	—	86
Fe–W	Hank's solution	1.604–3.025 g m^−2 d^−1	0.560–0.663 g m^−2 d^−1	—	—	56
Fe–CNT		2.108–2.419 g m^−2 d^−1	0.884–1.028 g m^−2 d^−1	—	—
Fe–(2–50)Fe2O3	Hank's solution	0.107–2.407 g m^−2 d^−1	0.273–0.668 g m^−2 d^−1	—	—	57
Fe–CNTs	Hank's solution	—	^c0.009058 g m^−2 d^−1	—	—	59
Fe–Mg		—	^c0.02362 g m^−2 d^−1	—	—
Fe–HA	SBF	^d4.776 g m^−2 d^−1	^d1.008 g m^−2 d^−1	—	—	46
Fe–BCP		^d4.608 g m^−2 d^−1	^d1.488 g m^−2 d^−1	—	—
Fe–TCP		^d4.344 g m^−2 d^−1	^d2.16 g m^−2 d^−1	—	—
Fe–TCP	0.9% NaCl	—	^e0.196 mm per year	—	—	47
HA/PCL–Fe	SBF	0.002 mm per year	—	—	—	87
HA–Fe		0.003 mm per year	—	—	—
Nitrided Fe stent	PBS	0.225 mm per year	—	—	—	52
Ag ion implanted Fe	Hank's solution	^b1.01 g m^−2 d^−1	^b0.55 g m^−2 d^−1	—	—	68
Micro-patterned Au arrays coated Fe	Hank's solution	2.338–3.174 g m^−2 d^−1	1.134–1.417 g m^−2 d^−1	—	—	70
Pt disc patterned Fe	Hank's solution	^b4.4285–4.7927 g m^−2 d^−1	^b3.4565–3.8324 g m^−2 d^−1	—	—	71
Fe/Fe-W alloy scaffolds	Hank's solution	—	0.149–0.264 g m^−2 d^−1	—	—	49
PLGA–porous Fe composite	PBS	0.42–0.72 mm per year	0.76–6.42 mm per year	—	—	54
Fe–PPAam	PBS	0.0386 mm per year	—	—	—	74
Fe stent	Descending aorta of minipigs	—	—	—	After 1 year, a large portion of the stent remains complete	20
Fe–10Mn–1Pd Fe–21Mn–0.7C–1Pd	Femoral mid-diaphyseal region of male Sprague-Dawley rat	—	—	—	No statistically significant loss in volume	88
Fe–(0.5–6.9)Mn	Underneath the subcutis of mice	—	—	—	No significant corrosion after 9 months	38
Fe wires	Wall and luminal implant in abdominal aorta of male Sprague-Dawley rats	—	—	—	Substantial biocorrosion by 22 days in wall and minimal biocorrosion at 9 months	89

---

Fig. 3 The schematic diagram of degradation behaviour and changes in the mechanical integrity of biodegradable metal implants during the healing process: (a), vascular; and (b), bone (reproduced with permission of Elsevier from ref. 9).

Fig. 4 Summary diagram of the regulation methods for the corrosion rates of Fe-based materials. Blue boxes represent regulation methods, green boxes represent specific measures, and purple boxes represent the role of the specific measures.

4.1 Alloying

Alloying with other elements is a preferred method to improve the corrosion rate of Fe. There are many alloying elements in industrial steels; however, the effects of these elements on the corrosion resistance of Fe can differ. To determine which alloying elements could be used to produce biodegradable Fe alloys, Liu et al.²⁷ studied the feasibility of numerous elements, including Mn, Co, Al, W, Sn, B, C and S. The authors reported that the corrosion rates of pure Fe and its corresponding Fe–X binary alloys were of the same order of magnitude. With regards to mechanical properties, corrosion rates and biocompatibility, the authors suggested that Co, W, C and S are feasible alloying elements for biodegradable Fe alloys.

The standard potential of reaction for Mn → Mn²⁺ + 2e⁻ is −1.18 V, which is lower than the standard potential of reaction for Fe → Fe²⁺ + 2e⁻, which is −0.440 V.⁷⁸ Because Fe and Mn can form a solid solution that exists as a less noble state, the standard potential of the Fe–Mn alloy was decreased by increasing Mn content.³² Moreover, Mn is an essential trace element that plays an important role in the growth, development and maintenance of healthy bones.⁷⁹ Therefore, Fe–Mn alloy aroused a great deal of interest, and Mn was added into the Fe matrix as a preferred alloying element. The corrosion properties of Fe–Mn alloys with varying Mn content, including Fe20Mn,^45,80,81 Fe25Mn,^45,80 Fe30Mn,^39,45,50,80 Fe33Mn,³⁶ Fe35Mn^{43–45,80,82} and lower Mn concentration Fe–Mn alloys (0.5, 2.7, and 6.9 wt% Mn),³⁸ have been extensively studied.

Hermawan et al.^45,80 found that the average corrosion rate of Fe–Mn alloys was approximately 520 μm per year, which is two times the rate of pure iron. In these alloys, the corrosion rate of the double-phase (γ + ε) alloys was higher than that of the single phase (γ) alloys. A similar result was reported for a 3D printed Fe–30Mn alloy consisting of two phases (martensite and austenite).⁵⁰ After cold-rolling, the corrosion rate of Fe30Mn and Fe35Mn slightly increased, whereas the rate for Fe20Mn and Fe25Mn declined.⁴⁵ The degradation products mainly consist of Ca/P compounds and iron hydroxides. Because the degradation products are not completely soluble, the release of iron and manganese ions into solution was limited.⁸⁰

Čapek et al.³⁹ found that the free corrosion potential of the hot-forged FeMn30 alloy was lower than that of iron, whereas the corrosion rate was faster than that of iron. However, the corrosion rate derived from a semi-static immersion test was remarkably lower than that of iron. Heiden et al.⁸¹ reported that the oxidation layer formed on the surface of the Fe–Mn alloy strongly inhibited the corrosion rate. Machining affected the morphology of oxidation layer on the surface of the samples, and the structure and morphology of oxidation layer impacted the diffusion of metal ions and the corrosion rate of the alloy. Compared with cast Fe–20Mn, the instantaneous corrosion rate of cold-rolled Fe–20Mn was lower. After quenching, cast samples have a finer grain and larger grain boundary, which promotes corrosion. After large-strain machining (LSM), the corrosion rate of the cast Fe–33Mn alloy showed a 140% increase when compared with the cast alloy without LSM.³⁶

Zhang et al.⁴³ reported that the degradation rate of Fe–35Mn prepared by sintering was remarkably increased when compared with compact samples. The degradation rate of approximately 2–8 mm per year meets the required timeframe for degradation. The degradation mechanisms include uniform corrosion and crevice corrosion.⁴³ A Fe–35Mn alloy prepared by sintering-cold rolling-sintering included more dispersed MnO fine particles and micropores, which provided more microsites with a different potential and larger surface; thus, the corrosion rate was greatly increased for this alloy.⁴⁴ Because of the surface alterations caused by the stent fabrication process, the degradation rate of stent prototypes produced from a Fe35Mn alloy was higher than that of the original alloy. The expanded stents showed a lower polarisation resistance when compared with nonexpanded stents, which highlights the effect of the expansion procedure on degradation.⁸²

Fe–Mn alloys with lower Mn content (0.5–6.9 wt%) showed suitable in vitro corrosion properties. However, no remarkable corrosion was observed during in vivo tests, which may be due to the inhibition of corrosion by phosphate passivation layers.³⁸ Moreover, different manganese concentrations had no detectable influence on the biodegradation rate.

The addition of Si increased the content of austenite. The electrical resistivity of martensitic is higher than that of austenite; therefore, the Fe30Mn6Si alloy showed a higher corrosion rate in electrochemical tests.²⁶ However, immersion tests revealed that the corrosion rate of Fe30Mn6Si was lower than pure Fe because of the influence of the corrosion products on the surface. An in vivo study⁹⁰ of 1.5Fe–Mn–Si implants showed generalized corrosion, and the corrosion products consisted of hydroxide layers on the surface. Except for implants with basic elements, P, C, Ca, S and K compounds were formed on the surface of subcutaneous implants, whereas P, C, Ca and Na but not Mn compounds were found in tibia implants.⁹⁰

Small amounts of noble alloying elements added into the Fe matrix can enhance its corrosion susceptibility. Moreover, the addition of noble alloying elements can produce small and finely dispersed intermetallic phases, which serve as cathodic sites in the Fe matrix and cause microgalvanic corrosion.³² Therefore, noble elements, such as Pd^32,86,88 and Ag,⁶² were added to the Fe matrix to improve its corrosion rate. Schinhammer et al.³² reported that the addition of Pd decreased the polarisation resistance and increased the degradation rate of the alloy. Furthermore, heat treatment generated homogeneously distributed intermetallic phases, which facilitates macroscopically homogeneous degradation behaviour in Fe–Mn–Pd alloys. The authors also suggested that Pd deposited on the metal surface and formed a macroscopic, short-circuited primary cell that accelerated metal degradation. In vitro tests showed that the addition of only 1 wt% of Pd to Fe–21Mn–0.7C effectively improved the degradation rate.⁸⁶ In vivo tests⁸⁸ did not produce similar results. Cylindrical Fe, Fe–10Mn–1Pd and Fe–21Mn–0.7C–1Pd pins were implanted in the femurs of Sprague-Dawley rats. The results showed that during the implantation phase of 52 weeks and after, the volume and surface of all of the sample pins showed no obvious statistical changes. The samples showed a slight, non-significant decrease in weight.⁸⁸ A lower oxygen concentration at and near the implant site produces a very slow degradation rate of the alloy. In addition, degradation products can hinder oxygen diffusion to the sample surface and limit further degradation. In Fe alloys fabricated with powder metallurgy,⁶² the addition of C, Ag, P and B increased the corrosion rate relative to pure Fe. However, the C and P content as well as P and B content had no significant effect on corrosion rates.

4.2 Adding a second phase

Adding a second phase into the Fe matrix is another method used to improve the corrosion rate of Fe composites. Cheng et al.^56,57 found that the addition of second phases led to a decrease in corrosion potential; however, the corrosion current densities increased. An immersion test showed that W did not affect the long-term corrosion rate of the composite, whereas the addition of CNTs and Fe₂O₃ markedly increased the corrosion rate. A more uniform corrosion mode occurred on the composite. Another study showed that the degradation rate of Fe-based composite scaffolds decreased after the addition of CNTs and immersion in Hank's solution for 8 weeks, whereas the addition of Mg greatly increased the degradation rate of the scaffolds.⁵⁹ A similar result for electrochemical measurements was reported.⁶⁰

Huang et al.^41,58 selected noble metal elements (Pd, Pt, Ag and Au) to serve as second phases in the Fe matrix. In Hank's solution, Fe-5 wt% Pd and Fe-5 wt% Pt composites showed uniform corrosion.⁴¹ Pd- and Pt-rich areas act as cathodes and have a very high corrosion potential. The iron acted as an anode and abundant small galvanic cells were formed, thus, markedly enhancing the corrosion rate of the iron matrix. The corrosion rate of Fe–Ag and Fe–Au composites was improved because pure Ag and Fe–Au solid solutions, which have relatively high corrosion potential, act as a cathode and the iron substrate acts as an anode to form galvanic cells on the surface of the material when immersed in Hank's solution.⁵⁸

Bioceramics are frequently used for second phases. Bioceramics (e.g., HA, TCP, and BCP) are soluble; therefore, the addition of bioceramics to the Fe matrix increased the corrosion rate and enhanced the biocompatibility of the Fe matrix. Ulum et al.⁴⁶ reported that the corrosion rate of the Fe/bioceramic (HA, TCP, BCP) composite tested with an electrochemistry method was lower than pure Fe; however, the immersion test yielded contrasting results. The authors suggested that the time required for the electrochemistry test procedure was shorter than the immersion test, and the bioceramics did not have enough time to go into solution. However, during the long immersion period, bioceramics can enter the solution and form a degradation layer. Immersion medium can also affect the corrosion rate;⁴⁷ for instance, the corrosion rate of the Fe/β-TCP composite in a 0.9% NaCl solution was faster than that of SBF. The corrosion rate of the Fe/β-TCP composite containing 40 vol% of β-TCP was 196 μm per year, which was 28% more than pure Fe.

4.3 Surface modification

Surface modification can improve degradability and biocompatibility of implants during the initial implantation phase.

The oxide film deposited on the surface remarkably improved the corrosion resistance of pure Fe in SBF,^65,66 and a nitride layer improved the corrosion resistance of pure Fe in a 0.9% NaCl aqueous solution.⁶⁷ These layers may prevent the premature loss of mechanical stability in Fe stents due to corrosion. Feng et al.⁵² reported that the effect of nitriding on corrosion potential is limited, but it markedly enhanced the corrosion current density of pure Fe stents. Twelve months after implantation, the nitrided pure Fe stent showed severe corrosion with a visible bulk decline (Fig. 5), and the strut thickness of the stent decreased from 120 μm prior to implantation to approximately 60 μm. Lin et al.⁷² reported that nitrided iron-based drug-eluting coronary stents had remarkably shortened degradation periods, almost completely corroded by 13 months after implantation, and produced a small amount of degradation products, as shown in Fig. 6.

Fig. 5 Histological sections of nitrided iron stents at 3 (a), 6 (b), and 12 (c) months after implantation (reproduced with permission of Springer from ref. 52).

Fig. 6 The images of the nitrided iron-based drug-eluting coronary stent after (a) 3 days, (b) 3 months, (c) 6 months, and (d) 13 months after implantation in the rabbit abdominal aorta (reproduced with permission of Elsevier from ref. 72).

Noble metal elements improved the degradability of the Fe matrix when used as alloying elements or second phases. Similar results were obtained when these elements were used for surface modification. Pure iron implanted with Ag ions showed faster and more uniform corrosion.⁶⁸ Micro-patterned noble metal arrays coated on the surface of pure Fe modulated the corrosion behaviour of pure Fe. Cheng et al.⁷⁰ reported that depositing a micro-patterned Au disc film on the surface of pure iron effectively increased the corrosion rate. Au served as an anode and the iron acted as the cathode, thus micro-galvanic corrosion occurred and accelerated the corrosion of the iron matrix. Another advantage of the Au disc coating was macroscopically uniform corrosion. This was due to the wide and uniform distribution of Au which formed numerous active sites for corrosion reactions on the surface of the matrix. Similar results were reported with the coating of the pure Fe surface with Pt disc arrays.⁷¹

PLGA–porous Fe composites showed a faster degradation because of a stronger interfacial interaction between PLGA and the surface of porous Fe.⁵⁴ The hydrolysis of PLGA accelerated the degradation of composites. Daud et al.⁸⁷ reported that the degradation rate of porous iron scaffold coated with HA and HA/PCL was approximately 10 times lower than samples without a coating. The coating may have inhibited the degradation of the samples. Farack et al.⁹¹ reported that unmodified iron foams demonstrated the fastest corrosion rate and were accompanied by high concentrations of H₂O₂. The iron foam coated with a brushite layer was highly protected and showed almost no corrosion. HA-coated iron foams showed significant iron release, hydrogen peroxide formation and oxygen depletion. The brushite coating was a remarkably more effective means of corrosion control than the HA coating. Furthermore, the corrosion behaviour was affected by the type of the incubation medium. DMEM induced a more powerful corrosion when compared with McCoy's medium.

A plasma allylamine polymerized coating remarkably decreased the corrosion rate and the formation of biodegradable Fe products.⁷⁴ Polypyrrole film was added to the surface of an iron electrode using electrochemical polymerisation. This film increased the corrosion resistance of the iron. The redox and electrical properties of the film can be tailored by applying appropriate potentials during synthesis.⁹² A phosphate coating (CaZn₂(PO₄)₂·2H₂O) on the Fe surface effectively improved the corrosion resistance of pure iron and provided effective protection during the initial implantation stages.⁹³ Because of the inhibition of corrosion products, the corrosion rates of Fe/Fe–W alloy double-layer scaffolds gradually decreased with increasing immersion time, and the final corrosion rates of different samples were approximated during static immersion.⁴⁹

4.4 Degradation rate of pure Fe

The degradation rates of pure Fe fabricated by different methods showed obvious differences. Typically, different fabrication methods produce materials with different microstructures. Thus, the corrosion performance is changed.

Obayi et al.³⁴ reported that the corrosion rates of cold-rolled and annealed pure Fe slightly declined with a decrease of grain size and grain size distribution. However, the authors also reported that the effect of mean grain size on the corrosion rates was not significant.³⁵ Annealing may reduce the corrosion rate by reducing defect density. The corrosion rates decreased with rising annealing temperatures due to more heterogeneity and a higher dislocation density at lower annealing temperatures. The corrosion rate of UD samples was slightly faster than BD samples due to more residual stress. UD samples exhibited more grain boundary corrosion than BD samples. This result shows that the corrosion of cross-rolled samples is more uniform.

Electroformed Fe showed uniform degradation with a moderate rate,⁸³ and the corrosion rate of this composite in Hank's solution was faster than that of iron produced by casting and thermomechanical treatment.⁴⁸ The current density can affect the corrosion rate of electrodeposited Fe film.⁶⁴ Because of larger grains size and texture, electrodeposited Fe film at 2 A dm⁻² presented with the lowest corrosion rate during electrochemical tests. Furthermore, the degradation morphology of this film showed uniform corrosion during a static immersion test because of its strong texture. A degradation layer that decreased the degradation rate was formed on the sample surface; therefore, static degradation tests demonstrated lower corrosion rates for iron. Microscopic pits were found for Fe film with an electrodeposition at 1 A dm⁻², 5 A dm⁻² and 10 A dm⁻².

Nie et al.⁵¹ reported that the corrosion rate of pure Fe gradually decreased as ECAP times increased. Corresponding to the decreased corrosion rate, the corrosion mode became milder, as evidenced by the change of larger and deeper holes to only small dimples on the corroded surfaces. Thus, the NC distinctly reduced rigorous corrosion. This effect benefits the design of degradable stents. To simulate the corrosion of stents in the body, electrochemical measurements for NC and MC pure Fe were completed in physiological saline solutions with varying dissolved oxygen concentrations.⁷⁶ NC-Fe exhibited higher corrosion resistance when compared with MC-Fe, and the corrosion resistance was improved by decreasing the oxygen concentration in the physiological saline. In an oxygen-free solution, corrosion was attenuated by a hydrogen adsorption mechanism. This mechanism was dominated by an oxygen-consuming mechanism in oxygen-containing solutions.

The degradation properties of other pure Fe composites were studied. Magnetron-sputtered pure iron foils showed a comparable low degradation rate that could be increased by annealing and grain coarsening.⁵³ The in vivo corrosion behaviour of an iron wire was assessed by Pierson et al.⁸⁹ The iron wire wrapped in the extracellular matrix of the arterial wall underwent substantial corrosion by 22 days, but a large corrosion product remained in the vessel wall for 9 months. However, iron wire implanted in the artery lumen showed minimal corrosion. This phenomenon was caused by a metallic surface that was passivated in ion-rich blood and an inhibited corrosion reaction; however, the wires placed in the artery wall may not be protected due to minor levels of passivation. Zhu et al.⁸⁴ reported that pure iron showed uniform corrosion and no obvious pitting corrosion during a dynamic corrosion test. The mean corrosion rate was 20.4 μg (cm⁻² h⁻¹).

5. Biocompatibilities

5.1 In vitro

Table 4 shows the results of in vitro tests of Fe-based materials. In general, the results of these in vitro tests were directly affected by the ion concentration in the culture medium; furthermore, the ion concentration correlated with the corrosion rate and composition of the material.

Table 4 Summary of the results of in vitro biocompatibility studies on different Fe-based materials

Materials	Methods	Cells	Results	Ref.
a Alloy powders.b Disc specimens.
Fe	Extraction medium culture	L929, ECV304	No signs of cytotoxicity	85
Nanocrystalline pure Fe	Extraction medium culture	L929, ECV304, VSMC	Inhibited growth of VSMCs but promoted growth of ECs and improved cytocompatibility with L929 cells	51
Electroformed Fe	Indirect contact^b	SMC	No inhibition of metabolic activity or cell proliferation	83
Ag-implanted pure Fe	Extraction medium culture	L929, VSMC, EA.hy926	Viabilities were decreased slightly	68
Phosphating pure Fe	Extraction medium culture	MSC	Relative growth rate was higher than 90% and exhibited no toxicity to cells	93
Nitride iron stent	Extraction medium culture/direct cell culture	L929, SMC, EC	Iron ions in high concentrations show no cytotoxicity to L929 cells	94
Fe–O film	Direct cell culture	HUVEC	Good adhesion and proliferation of HUVECs	65
Fe–30Mn	Extraction medium culture/ direct cell culture	MC3T3-E1	Direct culture did not show reduced cell viability, and high live cell attachment; indirect cultures show good cytocompatibility	50
Fe–Mn	Indirect contact^a	3T3 mouse fibroblast cell line	Low inhibition of fibroblasts' metabolic activities	80
Fe-30Mn	Extraction medium culture	L929	Metabolic activity is lower than iron but higher than 70% limit	39
Fe–Mn/Co/Al/W/Sn/B/C/S alloy	Extraction medium culture	L929, ECV304, VSMC	Except for the Fe–Mn alloy, no remarkable cytotoxicity to ECV304 cells but reduced the viability of L929 cells and VSMCs	27
Fe–21Mn–0.7C(–1Pd)	Extraction medium culture	HUVEC	Cytocompatibility was related to degradation rate	96
Fe–(0.5–6.9)Mn	Direct cell culture	SMC, EC	No significant inhibition zones after 24 and 72 h; inhibition zones around the discs were observed for all tested discs in both cell types after 144 and 240 h	38
Fe–(C/P/B/Ag)	Direct cell culture	Fibroblast	Monolayer cell culture test revealed cell death after 24 h; dynamic perfusion chamber revealed cytotoxicity was greatly reduced	62
Fe-Matrix wires	Direct cell culture/metal chloride solutions	EC, SMC	Significant EC attachment, good EC coverage and proliferation; SMC migration differed depending on ion species (Fe^2+, Fe^3+, Mn^2+ and Mg^2+)	97
Fe/Fe-W alloy porous scaffolds	Extraction medium culture	MC3T3-E1	A higher corrosion rate led to lower cell viability	49
Fe–(2–50)Fe2O3	Extraction medium culture	L929, VSMC, ECV304	After culturing for 4 days, the cell viabilities were 80–90% for L929 cells, decreased for VSMCs, and approximately 90% for ECV304 cells	57
Fe–5Pd/Pt	Extraction medium culture	L929, VSMC, ECV304	Slight cytotoxicity to L929 and ECV304 cells and inhibited VSMCs	41
Fe–Ag/Au	Extraction medium culture	L929, VSMC, HUVEC	No cytotoxicity to HUVECs or L929 cells; suppressed cell viability of VSMCs	58
Fe–W/CNTs	Extraction medium culture	L929, ECV304, VSMC	Fe–W showed no significant cytotoxicity to L929 and ECV304 cells; all composites showed mild cytotoxicity to VSMCs	56
Fe–CNTs/Mg	Direct cell culture	Fibroblast, MC3T3	Fibroblasts died after 24 h of incubation; nuclei morphology in MC3T3 cells showed significant changes and cell viability was inhibited	59 and 60
Fe–HA/BCP/TCP	Direct cell culture/ extraction medium culture	RSMC	Cellular activity increased compared to pure iron	46
(Fe, Fe–HA, Fe–brushite) foams	Direct cell culture	SaOs-2, hMSC	Perfusion cell culture enhanced proliferation and osteogenic differentiation of hMSCs	91
Fe–HA, Fe/HA–PCL	Direct cell culture	HSF1184, hMSC	Cell viability was high; cells preferably attached and grew actively	87
Pt disc patterned pure Fe	Extraction medium culture	EA.hy926, VSMC	No toxicity to human umbilical vein endothelial cells; inhibited the proliferation of VSMCs	71
PLGA–porous Fe composite	Direct cell culture	HSF1184	Accelerated degradation enhanced cell viability during an early degradation period of up to 72 h	54
Plasma polymeric allylamine coated Fe	Direct cell culture	HUVEC	Enhanced EC attachment, spreading, and proliferation	74
Heparin modified Fe	Extraction medium culture/direct cell culture	ECV304, VSMC	Increased the viability and number of adhered VSMCs and ECV304 cells	98

---

Pure Fe consists of only one element; therefore, its biocompatibility is only affected by Fe²⁺, Fe³⁺, Fe hydroxide and Fe particles. These products have been used in medical applications because of their advantageous properties, such as drug loading and disease diagnoses.⁹ In vitro studies of pure Fe showed excellent biocompatibility. Electroformed pure Fe did not inhibit the metabolic activity of primary rat SMCs. When used as a cardiovascular stent, electroformed Fe produced a beneficial inhibition of cell proliferation.⁸³ Cheng et al.⁸⁵ reported that industrial pure Fe showed no signs of cytotoxicity on L929 or ECV304 cells; in addition, the haemolysis percentage was less than 5%, and platelets showed a normal round shape. These results suggest that industrial pure Fe has excellent biocompatibility and haemocompatibility. Eighth pass ECAPed bulk nanocrystalline pure Fe distinctly inhibited the growth of VSMCs, markedly enhanced the growth of ECs and the cytocompatibility of L929 cells, and resulted in a haemolysis percentage of less than 5%; thus, this material showed superior haemocompatibility and biocompatibility.⁵¹ The in vitro cytotoxicity of nitrided iron stents was studied by Lin et al.⁹⁴ These authors suggested that the extraction medium and available oxygen are very important during the processes of extraction and incubation. Extracts with high iron ion concentrations showed no cytotoxicity to L929 cells. The in vitro cytotoxicity was produced by the size effect of corrosion particles rather than the material itself.

The biocompatibility of Fe alloys can be affected to varying degrees due to the addition of alloying elements. A study on the filtering of alloying elements used to create Fe alloys showed that, except for the Fe–Mn alloy, the extracts of Fe–Co/Al/W/Sn/B/C/S binary alloys showed no remarkable cytotoxicity to ECV304 cells; however, these extracts reduced the cell viabilities of L929 cells and VSMCs. The haemolysis percentage of the alloys was less than 5%.²⁷ A decrease in the metabolic activity of L929 cells was caused by a hot-forged FeMn30 alloy; however, the decrease was less than the tolerable 70% limit.³⁹ The higher release of ions from the Fe30Mn6Si alloy extracts decreased cell viability.²⁶ However, the viability of ECV304 cells began to increase after two days, and a visible inhibition of VSMCs was observed. Moreover, the haemolysis percentage of the Fe30Mn6Si alloy was less than 2%, thus showing no haemolysis.

Excess manganese can produce intoxication and neurotoxicity;⁹⁵ therefore, the biocompatibility of Fe–Mn alloys has been extensively studied. Compared to pure manganese, the Fe–Mn alloy showed a low inhibition effect on the metabolic activities of 3T3 fibroblast cells. Increasing the alloy concentration in the cellular medium increased the inhibition effect. A 50% inhibition effect was reported at a concentration of 6 mg ml⁻¹, and a 100% inhibition effect was reached with concentrations of 16 mg ml⁻¹ or greater.⁸⁰ In vitro cell cultures incubated with an inkjet 3D printed Fe–30Mn extract showed a cell viability that was 80% of control after 3 days (cultured in cell culture media). After direct culturing of the samples for 3 days, the attachment level of live cells and the cell density was higher than culture day 1. This result demonstrates that the material had good in vitro cytocompatibility and that the cells permeated through the pores.⁵⁰ SMCs and ECs were cultured on Fe–(0.5–6.9)Mn alloy samples and showed excellent biocompatibility.³⁸ After 24 and 72 h, no significant inhibition zones were observed. After 144 and 240 h, inhibition zones were observed around the sample discs for all of the tested samples in both cell types. For SMCs, the inhibition zones decreased with increasing amounts of manganese in the alloys, whereas this produced larger inhibition zones for ECs. Multicomponent Fe alloys were also studied. Schinhammer et al.⁹⁶ investigated the cytocompatibility of the extract media of an Fe–Mn–C(–Pd) alloy with HUVECs. The authors showed that cell viability and metabolic activity were closely related to the concentration of the extract media. The addition of Pd did not impact the cytocompatibility of the alloy. The authors suggested that as long as the release rate of ions is within tolerance level, Fe-based alloys will be cytocompatible. The release rate of ions depends on the degradation rate; therefore, the degradation rate of Fe-based alloys can influence cytocompatibility. Cell viability closely correlated with the corrosion rate of a Fe–W alloy.⁴⁹ A higher corrosion rate led to lower cell viability. In a monolayer cell culture test,⁶² cells died after 24 h. An in vitro study of Fe alloys performed cytotoxic tests in a dynamic perfusion chamber and found a marked decrease in cytotoxicity. This effect was likely due to the continuous removal of cumulative cytotoxic agents within the dynamic perfusion chamber. P in the Fe alloy did not significantly influence cytotoxicity. A large number of ECs attached to Fe-matrix wires, and all Fe-matrix wires showed good EC coverage and proliferation.⁹⁷ The migration of SMCs showed different tendencies depending on the ion species (Fe²⁺, Fe³⁺, Mn²⁺ and Mg²⁺).

Fe matrix composites prepared by SPS, including Fe–W, Fe–CNT,⁵⁶ Fe–Fe₂O₃,⁵⁷ Fe–Pd and Fe–Pt,⁴¹ had no obvious cytotoxicity on L929 or ECV304 cells, except for Fe–Pd and Fe–Pt. This effect may be due to the faster corrosion rate of these two composites, which led to a higher ion concentration in the extract media.⁴¹ Pd showed mild cytotoxicity in eukaryotic cells, and Pt showed a relatively higher cytotoxicity. However, their content was very low in the extract media, which suggests they had a negligible impact. All of the composites showed inhibition of VSMCs. The haemolysis percentage for all of the composites was less than 5% and showed good haemocompatibility. These results suggest that the composites are promising candidates for degradable stents. Fe–Ag and Fe–Au composites showed almost no cytotoxicity to EA.hy926 and L929 cells; however, the cell viability of VSMCs was remarkably suppressed.⁵⁸ The haemolysis percentages of these composites were lower than 5%, and a similar number of round platelets adhered on the surface of the samples. Oriňák et al.⁵⁹ studied in vitro static cell cultures with porous Fe-based scaffolds and reported that the fibroblasts died after 24 h; furthermore, the nuclei of MC3T3 cells showed significant changes in morphology, and cell viability was inhibited.⁶⁰ However, the scaffolds showed good blood compatibility. The authors suggest that under the conditions of the static culture, the degradation products of the scaffold inhibited the activity of the cell. Therefore, the cytotoxicity assessment should be performed in a dynamic system to avoid excessive accumulation of corrosion products. Ulum⁴⁶ et al. reported that Fe/bioceramic composite increased the cell viability of RSMCs when compared with pure Fe. The fast degradation of PLGA–porous pure Fe and its interaction with PLGA resulted in good fibroblast cell viability during the early and most active period of degradation.⁵⁴

In general, biocompatibility at the beginning phases of implantation can be improved by surface modification. This is very important for the initial phase of implantation, which requires good biocompatibility and slow corrosion rates. HA- or HA/PCL-coated samples showed high viability for HSF cells and hMSCs, which indicates that HA improves the cytocompatibility of the surface.⁸⁷ In comparison to uncoated specimens, plasma polymeric allylamine-coated Fe resulted in higher cell viability, as shown by the enhancement of EC attachment, spreading, and proliferation.⁷⁴ The oxide films deposited on the surface of pure iron reduced the number of adhered platelets and inhibited the aggregation and activation of platelets;^65,66 in addition, these composites showed good adhesion and proliferation of HUVECs.⁶⁵ The anti-haemolysis properties and cell compatibility of pure iron coated with phosphate were notably enhanced, whereas the anti-coagulant properties slightly decreased.⁹³ After modification with heparin, the hydrophilicity of iron markedly increased. The blood clotting coagulation time was extended; therefore, the risk of thrombosis was reduced. VSMC proliferation was reduced, whereas ECV304 cells were not dramatically affected.⁹⁸ Ag ions implanted into pure Fe slightly decreased the viabilities of L929 cells, EA.hy926 cells and VSMCs. The haemolysis percentage was lower than 2% and demonstrated good haemocompatibility; however, an increased the risk of thrombosis was revealed by platelets adhesion tests.⁶⁸ Pt patterned discs of pure iron showed almost no toxicity to HUVECs, but exhibited a remarkable inhibition of VSMC proliferation. The haemolysis percentage was lower than 1%. The number of adhered platelets on the samples was less than that of the uncoated samples.⁷¹ Farack et al.⁹¹ suggested that the continuous supply of fresh medium and removal of cytotoxic corrosion products in perfusion cell cultures enhanced the proliferation and osteogenic differentiation of hMSCs on calcium iron foams coated with phosphate.

5.2 In vivo

In vivo testing is the most straightforward method to estimate the biocompatibility and degradability of Fe-based materials. Table 5 summarizes the results of in vivo tests of Fe-based materials. Degradable stents are one of the most important applications of Fe-based materials; thus, in vivo studies of stents were the first to be completed. The first in vivo testing of pure iron stents was published by Peuster et al.⁹⁹ The authors implanted pure iron stents into the native descending aorta of New Zealand white rabbits. During the follow-up of 6 to 18 months, no thromboembolic complications or adverse events occurred. No remarkable neointimal proliferation, inflammatory response or systemic toxicity was observed (Fig. 7). Next, the authors implanted pure iron stents into the descending aorta of minipigs for 1–360 days. There was no occurrence of Fe overload or Fe-related organ toxicity; in addition, there was no local toxicity caused by corrosion products, as shown in Fig. 8.²⁰

Table 5 Summary of the results of in vivo tests of different Fe-based materials

Implants	Animals	Site of implantation	Results	Ref.
Fe stent	Minipigs	Descending aorta	No local or systemic toxicity	20
Fe stent	New Zealand white rabbit	Native descending aorta	No thromboembolic complications or adverse events	99
Fe stent	Juvenile domestic pigs	Coronary arteries	No stent particle embolisation or thrombosis; no traces of excess inflammation or fibrin deposition	100
Fe stent	New Zealand white rabbit	Abdominal aorta	No adverse events or adverse effects to local tissues	72
Fe stent	Juvenile pigs	Iliac arteries	Mild luminal loss and relative stenosis after 12 months	52
Fe stent	Mini-swine	Coronary arteries	Mild intimal hyperplasia and intimal coverage appears more complete after 28 days; did not cause inflammation or other adverse reactions	101
Fe wire	Wistar rats	Subcutaneous loose connective tissue in the dorsal thoracic region	The Fe implant was encapsulated with granulation tissue; emigration of macrophages, neutrophils and multinucleated giant cells; dilation of blood vessels	102
Fe–21Mn(–0.7C)–1Pd (cylindrical pins)	Sprague-Dawley rats	Femoral mid-diaphyseal region	All animals showed good general health; the implants were well-integrated and sheathed by a narrow capsule of connective tissue; no signs of inflammation or local toxicity	88
Fe–(0.5–6.9)Mn (cylindrical plate)	NMRI mice	Underneath the subcutis resting on the fascia of the gluteal muscle	No adverse effects or signs of infection	38
1.5Fe–Mn–Si (small piece)	Wistar rats	Subcutaneous/tibia crest	Subcutaneous implant showed very good biocompatibility; no local reaction of any kind; adverse biological reactions were not induced by bone implants	90
Fe–HA/BCP/TCP (small slice)	Indonesian thin tailed sheep	Below the radius periosteum membrane of radial forelegs on medio proximal region	Positive tissue response for up to 70 days	46
Fe–HA/BCP/TCP (small slice)	Indonesian thin tailed sheep	Radial bones of leg	Minimal tissue response; normal dynamic change in blood cellular response; no stress effects; lower inflammatory giant cell counts	103

---

Fig. 7 Histological sections of degradable iron stent 18 months after implantation in rabbit aorta. A stent strut is covered by neointima (N). There is moderate infiltration of macrophages along the adventitial side (arrows) (reproduced with permission of MBJ Publishing Group Ltd from ref. 99).

Fig. 8 Histological sections of iron and 316-L (control) stent struts during the course of follow-up (reproduced with permission of Elsevier from ref. 20).

In a short-term study, Fe stents were implanted into juvenile pig coronary arteries for 28 days to assess safety and efficacy.¹⁰⁰ No particle embolisation or thrombosis was observed, and histomorphometry revealed no excess inflammation or fibrin deposition at 28 days after implantation. The tissue in close proximity to the stent turned a brown colour, which was caused by the absorption of the ferric salts generated during stent degradation by adjacent tissue. In a similar study, nitrided Fe stents were implanted into mini-swine coronary arteries.¹⁰¹ Twenty-eight days after implantation, the Fe stent produced intimal hyperplasia similar to that of the control Co–Cr alloy stent, whereas the intimal coverage appeared more complete than the control. Furthermore, the degradation of the stent did not lead to an inflammatory response of the surrounding tissue or remote organs. No other negative effects were observed, suggesting that the Fe stent was safe and effective.

Feng et al.⁵² implanted nitrided Fe stents into the iliac arteries of juvenile pigs. One month after implantation, the stent was covered by endothelial cells, and the inflammatory response had decreased. After 12 months, intimal hyperplasia led to mild luminal loss; in addition, the piglets' growth resulted in stenosis of the vessel at the implanted stent site. Thrombosis or local tissue necrosis was not observed. The nitrided Fe stent showed acceptable biocompatibility. Next, the nitrided iron stents were underwent Zn electroplating and were coated with a PDLLA carrying sirolimus.⁷² The drug-eluting coronary iron stents were implanted into New Zealand white rabbits. No adverse events or adverse effects on the local tissue were observed. After implantation for up to 13 months, there were no identified biological problems.

Kraus et al.⁸⁸ implanted Fe, Fe–10Mn–1Pd, and Fe–21Mn–0.7C–1Pd pins into the femur of Sprague-Dawley rats. Over the course of one year, the authors found that the animals showed no severe pathological characteristics. After surgery, the wound showed slight swelling for 1–2 days, which conforms with clinical features, then showed no signs of inflammation and healed at the expected rate. All animals were healthy after surgery. The histological results showed that the Fe degradation products in the tissue near the implants were mainly Fe³⁺ and a relatively low amount of Fe²⁺. No inflammation or local toxicity was observed, and the degradation process did not harm the tissue adjacent to the implant, as shown in Fig. 9.

Fig. 9 Detection of (a) Fe³⁺/Fe²⁺, (b) Fe²⁺ and (c) Fe³⁺ in consecutive sections of pure Fe samples after 36 weeks (reproduced with permission of Elsevier from ref. 88).

Fe–(0.5–6.9)Mn cylindrical plates were implanted underneath the subcutis on the fascia of the gluteal muscle in NMRI mice. During implantation, no adverse effects or signs of infection were observed. All animals survived the implantation procedure and reached the assigned follow-up period.³⁸ The 1.5Fe–Mn–Si implants showed excellent biocompatibility, and no local reactions were observed. Adverse biological reactions were not induced by the bone implants.⁹⁰

In soft tissue implantation studies,¹⁰² Fe showed the least toxicity when compared with the same concentration of Ni or Cu. The damage to surrounding tissue caused by the Fe implant was relatively minimal, and the implant was encapsulated by granulation tissue.

Ulum et al.¹⁰³ implanted Fe–bioceramic composites into the medio-proximal region of the radial leg bones of male Indonesian thin tailed sheep to thoroughly investigate their bioactivity. Inflammation disappeared 35 days after implantation. B-Mode ultrasonography showed minimal tissue response during the wound healing process. A normal dynamic change in blood cellular response and no effects of stress were observed. The number of inflammatory giant cells was lower than that of SS316L. The composites showed enhanced bioactivity and were beneficial to wound healing. X-ray radiography revealed that the composite showed a consistent degradation progress and good tissue response.⁴⁶

6. Concluding remarks

As one of the main types of biodegradable metallic materials, the Fe-based materials are remarkable because of their excellent mechanical properties and biocompatibility. Previous researches already proved that the Fe-based biomedical materials can be successfully manufactured by many techniques. The biocompatibility studies, both in vitro and in vivo, together with other studies like mechanical research, showed that the Fe-based materials are potentially useful for applications in stents, orthopedics and bone tissue engineering.

Nevertheless, on the way to practical applications, there are still some problems to be solved. First, the degradation rate of Fe-based materials and its effect on biocompatibility require more researches. The degradation rate is relatively slow and should be improved and controlled. The degradation rate can impact the biocompatibility of the material. A lower degradation rate yields better biocompatibility and vice versa. The excessive degradation products produced under higher degradation rates can lead to a decline in biocompatibility. Therefore, fast degradation rates are not necessarily better. The ideal degradation rate is the one with sufficient biocompatibility and complete degradation by an expected time. Effort to make the degradation rate controllable should be an important research direction for these materials.

Second, corrosion mode is another important issue that directly relates to safety and implant efficacy. Uniform corrosion is an ideal corrosion mode. Non-uniform corrosion (such as point corrosion, intergranular corrosion, crevice corrosion, and so on) can lead to the premature failure of implants. Corrosion mode depends not only on the microstructure of the materials, but also on the implantation environment. The two factors need to be considered in the control of the corrosion mode. For the Fe-based materials, a homogeneous phase and fine microstructure are conductive for the corrosion to occur uniformly. Some reports^32,35,64 have shown that relatively uniform corrosion can be attained by controlling microstructure of the Fe-based materials. The composition and the distribution of the body fluid can also affect the corrosion mode of the implant, because they are different at different implant positions. So far the effects of implantation position on corrosion mode of Fe-based materials have not been reported. Therefore, a systematic research needs to be considered in the future.

The above two problems, i.e., controlling the degradation rate and corrosion mode, can be solved via controlling the composition and structure of the materials. To realize the controllability, the composition and structure of the materials should be designed according to the specific requirements in the applications. For example, the corrosion rate of the implant should not be constant. On the contrary, it should be varying against the time (as shown in Fig. 3). In the beginning after implantation, the corrosion rate should be slow, and the slow degradation rate should remain for a period of time during which remodelling or repair can happen. After that the materials should be degraded at an increasing rate so that a complete degradation will finish within an expected period of time. A multilayer structure may be a feasible solution to this requirement. In the outer layer, a material with low corrosion rate can be used to satisfy the requirement in the remodelling or repair phase. In the inner layer, a material with high corrosion rate can be used to make sure that the implant will be degraded completely within expected time. More advanced materials for different purposes can be similarly designed and manufactured.

In summary, biodegradable Fe-based materials are very promising and further researches on controlling the degradation process are key to promote the applicability.

Acknowledgements

This work was supported by the National Basic Research Program of China (973 Program) (Grant No. 2011CB710905), the National Natural Science Foundation of China (Grant No. 31170816, 31200551), the Fundamental Research Funds for the Central Universities (Grant No. 3102014KYJD019, 3102016ZY039, 3102015ZY083), China Postdoctoral Science Foundation (Grant No. 2013T60890), and the Natural Science Foundation of Shannxi Province, China (Grant No. 2016JM3012).

Notes and references

1. M. Niinomi, M. Nakai and J. Hieda, Acta Biomater., 2012, 8, 3888–3903.
2. M. Geetha, A. K. Singh, R. Asokamani and A. K. Gogia, Prog. Mater. Sci., 2009, 54, 397–425.
3. G. Mani, M. D. Feldman, D. Patel and C. M. Agrawal, Biomaterials, 2007, 28, 1689–1710.
4. M. Peuster, P. Beerbaum, F.-W. Bach and H. Hauser, Cardiol Young, 2006, 16, 107–116.
5. M. P. Staiger, A. M. Pietak, J. Huadmai and G. Dias, Biomaterials, 2006, 27, 1728–1734.
6. F. Witte, N. Hort, C. Vogt, S. Cohen, K. U. Kainer, R. Willumeit and F. Feyerabend, Curr. Opin. Solid State Mater. Sci., 2008, 12, 63–72.
7. A. Purnama, H. Hermawan, J. Couet and D. Mantovani, Acta Biomater., 2010, 6, 1800–1807.
8. M. Moravej and D. Mantovani, Int. J. Mol. Sci., 2011, 12, 4250–4270.
9. Y. F. Zheng, X. N. Gu and F. Witte, Mater. Sci. Eng., R, 2014, 77, 1–34.
10. A. Francis, Y. Yang, S. Virtanen and A. R. Boccaccini, J. Mater. Sci.: Mater. Med., 2015, 26, 138.
11. J.-M. Seitz, M. Durisin, J. Goldman and J. W. Drelich, Adv. Healthcare Mater., 2015, 4, 1915–1936.
12. T. J. Butler, R. W. Jackson, J. Y. Robson, R. J. Owen, H. T. Delves, C. E. Sieniawska and J. D. Rose, Br. J. Radiol., 2000, 73, 601–603.
13. M. Peuster, C. Fink, C. von Schnakenburg and G. Hausdorf, Cardiol Young, 2002, 12, 229–235.
14. M. Peuster, C. Fink, P. Wohlsein, M. Bruegmann, A. Günther, V. Kaese, M. Niemeyer, H. Haferkamp and C. v. Schnakenburg, Biomaterials, 2003, 24, 393–399.
15. M. Peuster, C. Fink and C. von Schnakenburg, Biomaterials, 2003, 24, 4057–4061.
16. H. Gong, K. Wang, R. Strich and J. G. Zhou, J. Biomed. Mater. Res., Part B, 2015, 103, 1632–1640.
17. J. Kubásek, D. Vojtěch, E. Jablonská, I. Pospíšilová, J. Lipov and T. Ruml, Mater. Sci. Eng., C, 2016, 58, 24–35.
18. H. F. Li, X. H. Xie, Y. F. Zheng, Y. Cong, F. Y. Zhou, K. J. Qiu, X. Wang, S. H. Chen, L. Huang, L. Tian and L. Qin, Sci. Rep., 2015, 5, 10719.
19. X. Liu, J. Sun, K. Qiu, Y. Yang, Z. Pu, L. Li and Y. Zheng, J. Alloys Compd., 2016, 664, 444–452.
20. M. Peuster, C. Hesse, T. Schloo, C. Fink, P. Beerbaum and C. von Schnakenburg, Biomaterials, 2006, 27, 4955–4962.
21. H. Hermawan and D. Mantovani, Acta Biomater., 2013, 9, 8585–8592.
22. P. Tengvall, B. Skoglund, A. Askendal and P. Aspenberg, Biomaterials, 2004, 25, 2133–2138.
23. S. Radin and P. Ducheyne, Biomaterials, 2007, 28, 1721–1729.
24. G. F. Muschler, C. Nakamoto and L. G. Griffith, J. Bone Jt. Surg., Am. Vol., 2004, 86, 1541–1558.
25. D. L. Shi, Introduction to Biomaterials, Tsinghua University Press, World Scientific Publishing Co. Pte. Ltd., Beijing, Singapore, 2006.
26. B. Liu, Y. F. Zheng and L. Q. Ruan, Mater. Lett., 2011, 65, 540–543.
27. B. Liu and Y. F. Zheng, Acta Biomater., 2011, 7, 1407–1420.
28. F. Moszner, A. S. Sologubenko, M. Schinhammer, C. Lerchbacher, A. C. Hänzi, H. Leitner, P. J. Uggowitzer and J. F. Löffler, Acta Mater., 2011, 59, 981–991.
29. F. Moszner, S. S. A. Gerstl, P. J. Uggowitzer and J. F. Löffler, J. Mater. Res., 2014, 29, 1069–1076.
30. F. Moszner, E. Povoden-Karadeniz, S. Pogatscher, P. J. Uggowitzer, Y. Estrin, S. S. A. Gerstl, E. Kozeschnik and J. F. Löffler, Acta Mater., 2014, 72, 99–109.
31. F. Moszner, S. S. A. Gerstl, P. J. Uggowitzer and J. F. Löffler, Acta Mater., 2014, 73, 215–226.
32. M. Schinhammer, A. C. Hänzi, J. F. Löffler and P. J. Uggowitzer, Acta Biomater., 2010, 6, 1705–1713.
33. M. Schinhammer, C. M. Pecnik, F. Rechberger, A. C. Hänzi, J. F. Löffler and P. J. Uggowitzer, Acta Mater., 2012, 60, 2746–2756.
34. C. S. Obayi, R. Tolouei, A. Mostavan, C. Paternoster, S. Turgeon, B. A. Okorie, D. O. Obikwelu and D. Mantovani, Biomatter, 2014, 6, e959874.
35. C. S. Obayi, R. Tolouei, C. Paternoster, S. Turgeon, B. A. Okorie, D. O. Obikwelu, G. Cassar, J. Buhagiar and D. Mantovani, Acta Biomater., 2015, 17, 68–77.
36. M. Heiden, A. Kustas, K. Chaput, E. Nauman, D. Johnson and L. Stanciu, J. Biomed. Mater. Res., Part A, 2015, 103, 738–745.
37. M. Heiden, D. Johnson and L. Stanciu, Acta Mater., 2016, 103, 115–127.
38. A. Drynda, T. Hassel, F. W. Bach and M. Peuster, J. Biomed. Mater. Res., Part B, 2015, 103, 649–660.
39. J. Čapek, J. Kubásek, D. Vojtěch, E. Jablonská, J. Lipov and T. Ruml, Mater. Sci. Eng., C, 2016, 58, 900–908.
40. D. Vojtěch, J. Kubásek, J. Čapek and I. Pospíšilová, Mater. Tehnol., 2015, 49, 877–882.
41. T. Huang, J. Cheng and Y. F. Zheng, Mater. Sci. Eng., C, 2014, 35, 43–53.
42. J. Čapek and D. Vojtěch, Mater. Sci. Eng., C, 2014, 43, 494–501.
43. Q. Zhang and P. Cao, Mater. Chem. Phys., 2015, 163, 394–401.
44. H. Hermawan, H. Alamdari, D. Mantovani and D. Dubé, Powder Metall., 2008, 51, 38–45.
45. H. Hermawan, D. Dubé and D. Mantovani, J. Biomed. Mater. Res., Part A, 2010, 93, 1–11.
46. M. F. Ulum, A. Arafat, D. Noviana, A. H. Yusop, A. K. Nasution, M. R. A. Kadir and H. Hermawan, Mater. Sci. Eng., C, 2014, 36, 336–344.
47. A. Reindl, R. Borowsky, S. Hein, J. Geis-Gerstorfer, P. Imgrund and F. Petzoldt, J. Mater. Sci., 2014, 49, 8234–8243.
48. M. Moravej, F. Prima, M. Fiset and D. Mantovani, Acta Biomater., 2010, 6, 1726–1735.
49. J. He, F.-L. He, D.-W. Li, Y.-L. Liu and D.-C. Yin, Colloids Surf., B, 2016, 142, 325–333.
50. D.-T. Chou, D. Wells, D. Hong, B. Lee, H. Kuhn and P. N. Kumta, Acta Biomater., 2013, 9, 8593–8603.
51. F. L. Nie, Y. F. Zheng, S. C. Wei, C. Hu and G. Yang, Biomed. Mater., 2010, 5, 065015.
52. Q. Feng, D. Zhang, C. Xin, X. Liu, W. Lin, W. Zhang, S. Chen and K. Sun, J. Mater. Sci.: Mater. Med., 2013, 24, 713–724.
53. T. Jurgeleit, E. Quandt and C. Zamponi, Adv. Mater. Sci. Eng., 2015, 2015, 294686.
54. A. H. M. Yusop, N. M. Daud, H. Nur, M. R. A. Kadir and H. Hermawan, Sci. Rep., 2015, 5, 11194.
55. L. J. Wang, J. F. Zhang and W. Jiang, Int. J. Refract. Met. Hard Mater., 2013, 39, 103–112.
56. J. Cheng and Y. F. Zheng, J. Biomed. Mater. Res., Part B, 2013, 101, 485–497.
57. J. Cheng, T. Huang and Y. F. Zheng, J. Biomed. Mater. Res., Part A, 2014, 102, 2277–2287.
58. T. Huang, J. Cheng, D. Bian and Y. F. Zheng, J. Biomed. Mater. Res., Part B, 2016, 104, 225–240.
59. A. Oriňák, R. Oriňáková, Z. Králová, A. Turoňová, M. Kupková, M. Hrubovčáková, J. Radoňák and R. Džunda, J. Porous Mater., 2014, 21, 131–140.
60. R. Orinakova, A. Orinak, L. M. Buckova, M. Giretova, L. Medvecky, E. Labbanczova, M. Kupkova, M. Hrubovcakova and K. Koval, Int. J. Electrochem. Sci., 2013, 8, 12451–12465.
61. J. Čapek, D. Vojtěch and A. Oborná, Mater. Des., 2015, 83, 468–482.
62. B. Wegener, B. Sievers, S. Utzschneider, P. Muller, V. Jansson, S. Rossler, B. Nies, G. Stephani, B. Kieback and P. Quadbeck, Mater. Sci. Eng., B, 2011, 176, 1789–1796.
63. Y. D. Gamburg and G. Zangari, Theory and Practice of Metal Electrodeposition, Springer, New York, 2011.
64. M. Moravej, S. Amira, F. Prima, A. Rahem, M. Fiset and D. Mantovani, Mater. Sci. Eng., B, 2011, 176, 1812–1822.
65. S. Zhu, N. Huang, L. Xu, Y. Zhang, H. Liu, Y. Lei, H. Sun and Y. Yao, Surf. Coat. Technol., 2009, 203, 1523–1529.
66. S. Zhu, N. Huang, H. Shu, Y. Wu and L. Xu, Appl. Surf. Sci., 2009, 256, 99–104.
67. C. Z. Chen, X. H. Shi, P. C. Zhang, B. Bai, Y. X. Leng and H. Nan, Solid State Ionics, 2008, 179, 971–974.
68. T. Huang, Y. Cheng and Y. Zheng, Colloids Surf., B, 2016, 142, 20–29.
69. T. Huang, Y. Zheng and Y. Han, Regener. Biomater., 2016, 205–215.
70. J. Cheng, T. Huang and Y. F. Zheng, Mater. Sci. Eng., C, 2015, 48, 679–687.
71. T. Huang and Y. Zheng, Sci. Rep., 2016, 6, 23627.
72. W.-J. Lin, D.-Y. Zhang, G. Zhang, H.-T. Sun, H.-P. Qi, L.-P. Chen, Z.-Q. Liu, R.-L. Gao and W. Zheng, Mater. Des., 2016, 91, 72–79.
73. Z. H. Wen, L. M. Zhang, C. Chen, Y. B. Liu, C. J. Wu and C. S. Dai, Mater. Sci. Eng., C, 2013, 33, 1022–1031.
74. P. Qi, Y. Yang, S. Zhao, J. Wang, X. Li, Q. Tu, Z. Yang and N. Huang, Mater. Des., 2016, 96, 341–349.
75. A.-V. Do, B. Khorsand, S. M. Geary and A. K. Salem, Adv. Healthcare Mater., 2015, 4, 1742–1762.
76. F. L. Nie and Y. F. Zheng, J. Biomed. Mater. Res., Part B, 2012, 100, 1404–1410.
77. J. E. Schaffer, E. A. Nauman and L. A. Stanciu, Metall. Mater. Trans. B, 2012, 43, 984–994.
78. R. W. Revie and H. H. Uhlig, Corrosion and Corrosion Control: An Introduction to Corrosion Science and Engineering, 4th edn, John Wiley & Sons, Inc., Hoboken, 2008.
79. J. H. Beattie and A. Avenell, Nutr. Res. Rev., 1992, 5, 167–188.
80. H. Hermawan, A. Purnama, D. Dube, J. Couet and D. Mantovani, Acta Biomater., 2010, 6, 1852–1860.
81. M. Heiden, E. Walker, E. Nauman and L. Stanciu, J. Biomed. Mater. Res., Part A, 2015, 103, 185–193.
82. N. B. Sing, A. Mostavan, E. Hamzah, D. Mantovani and H. Hermawan, J. Biomed. Mater. Res., Part B, 2015, 103, 572–577.
83. M. Moravej, A. Purnama, M. Fiset, J. Couet and D. Mantovani, Acta Biomater., 2010, 6, 1843–1851.
84. S. F. Zhu, N. Huang, L. Xu, Y. Zhang, H. Q. Liu, H. Sun and Y. X. Leng, Mater. Sci. Eng., C, 2009, 29, 1589–1592.
85. J. Cheng, B. Liu, Y. H. Wu and Y. F. Zheng, J. Mater. Sci. Technol., 2013, 29, 619–627.
86. M. Schinhammer, P. Steiger, F. Moszner, J. F. Löffler and P. J. Uggowitzer, Mater. Sci. Eng., C, 2013, 33, 1882–1893.
87. N. Mohd Daud, N. B. Sing, A. H. Yusop, F. A. Abdul Majid and H. Hermawan, J. Orthop. Transl., 2014, 2, 177–184.
88. T. Kraus, F. Moszner, S. Fischerauer, M. Fiedler, E. Martinelli, J. Eichler, F. Witte, E. Willbold, M. Schinhammer, M. Meischel, P. J. Uggowitzer, J. F. Loeffler and A. Weinberg, Acta Biomater., 2014, 10, 3346–3353.
89. D. Pierson, J. Edick, A. Tauscher, E. Pokorney, P. Bowen, J. Gelbaugh, J. Stinson, H. Getty, C. H. Lee, J. Drelich and J. Goldman, J. Biomed. Mater. Res., Part B, 2012, 100, 58–67.
90. M. Fântânariu, L. C. Trincă, C. Solcan, A. Trofin, Ş. Strungaru, E. V. Şindilar, G. Plăvan and S. Stanciu, Appl. Surf. Sci., 2015, 352, 129–139.
91. J. Farack, C. Wolf-Brandstetter, S. Glorius, B. Nies, G. Standke, P. Quadbeck, H. Worch and D. Scharnweber, Mater. Sci. Eng., B, 2011, 176, 1767–1772.
92. K. Cysewska, S. Virtanen and P. Jasinski, J. Electrochem. Soc., 2015, 162, E307–E313.
93. H. Chen, E. Zhang and K. Yang, Mater. Sci. Eng., C, 2014, 34, 201–206.
94. W. J. Lin, G. Zhang, P. Cao, D. Y. Zhang, Y. F. Zheng, R. X. Wu, L. Qin, G. Q. Wang and T. Y. Wen, J. Biomed. Mater. Res., Part B, 2015, 103, 764–776.
95. M. Aschner, T. R. Guilarte, J. S. Schneider and W. Zheng, Toxicol. Appl. Pharmacol., 2007, 221, 131–147.
96. M. Schinhammer, I. Gerber, A. C. Hänzi and P. J. Uggowitzer, Mater. Sci. Eng., C, 2013, 33, 782–789.
97. J. E. Schaffer, E. A. Nauman and L. A. Stanciu, Acta Biomater., 2013, 9, 8574–8584.
98. X. Xu, M. Li, Q. Liu, Z. Jia, Y. Shi, Y. Cheng, Y. Zheng and L. Q. Ruan, Prog. Nat. Sci.: Mater., 2014, 24, 458–465.
99. M. Peuster, P. Wohlsein, M. Brugmann, M. Ehlerding, K. Seidler, C. Fink, H. Brauer, A. Fischer and G. Hausdorf, Heart, 2001, 86, 563–569.
100. R. Waksman, R. Pakala, R. Baffour, R. Seabron, D. Hellinga and F. Tio, J. Intervent. Cardiol., 2008, 21, 15–20.
101. C. Wu, H. Qiu, X. Y. Hu, Y. M. Ruan, Y. Tian, Y. Chu, X. L. Xu, L. Xu, Y. Tang and R. L. Gao, Chin. Med. J., 2013, 126, 4752–4757.
102. M. Uo, F. Watari, A. Yokoyama, H. Matsuno and T. Kawasaki, Biomaterials, 2001, 22, 677–685.
103. M. F. Ulum, A. K. Nasution, A. H. Yusop, A. Arafat, M. R. A. Kadir, V. Juniantito, D. Noviana and H. Hermawan, J. Biomed. Mater. Res., Part B, 2015, 103, 1354–1365.

---