Corrosion performance of cold deformed austenitic stainless steels for biomedical applications

2.1 Localized corrosion (pitting and crevice corrosion)

Pitting corrosion is one of the most occurring localized corrosion in metallic implants. Studies reported that more than 90% of the failure of implants of 316L SS are due to pitting and crevice corrosion attack (Sivakumar et al., 1995). Pitting corrosion is caused by local dissolution of passive film and the formation of cavities surrounded by an intact passive surface (Zsklarska-Smialowska, 1986). It was suggested that the absorption of the damaging ions on the surface of the passive film of SSs produces mutual repulsion, which lowers the interfacial surface tension, and when the repulsive force is sufficient, the passive film cracks (Zaya, 1984). Pitting usually occurs in solutions containing halide ions in which chloride ion is the most aggressive. Pits almost initiate due to chemical or physical heterogeneity at the surface such as inclusions, second-phase particles, solute-segregated grain boundaries, flaws, mechanical damage, and dislocations (Zsklarska-Smialowska, 1986). Most engineering alloys have many such defects, and a pit tends to initiate at the most susceptible sites first. For example, pits in SSs are often associated with manganese sulfide (MnS) inclusions, which are found in most commercial steels. The role of MnS inclusions in promoting the breakdown and localized corrosion of SSs were recognized (Frankel, 1998; Sidharth, 2009). The initial stage of pitting is generally agreed to be of very short duration, and pit initiation is strongly dependent on the surface conditions of the material (Melchers, 1994).

Crevice corrosion is also a type of localized corrosion similar to pitting, which occurs preferentially in regions on the metal surface where mass transfer is limited, e.g. in narrow crevices or other shielded areas. In the case of bone plates and screws, crevice corrosion can occur at the interface between the tightly contacted bone plate and the screw on the countersink (Buhagiar et al., 2012). To function as a corrosion site, a crevice must be wide enough to permit liquid entry but suitably narrow to maintain the stagnant zone. This type of attack occurs in numerous mediums, although it is usually most intense in ones containing chloride ions. Crevice corrosion can be viewed as a less severe form of localized corrosion when compared with pitting. During crevice corrosion, the attack is localized within shielded areas, while the remaining surface suffers slight or no damage. The presence of crevice corrosion in the countersink portion of the bone plate can induce crack propagation.

2.2 Tribocorrosion/fretting corrosion

Tribocorrosion is defined as the conjoint action of mechanical wear and corrosive attack on a material surface. A special mode of tribocorrosion highly relevant to implants such as hip, knee, and shoulder replacements is fretting corrosion. Fretting corrosion is a form of damage that occurs at the interface of two closely fitted surfaces when they are subjected to slight oscillatory slip and joint corrosion action. Fretting corrosion is very common in all load-bearing metallic orthopedic implants. Fretting corrosion can radically alter the corrosion behavior by mechanically destroying the passive film. Fretting corrosion, alone or in combination with crevice corrosion, was identified as one of the important factors in implant corrosion (Brown et al., 1995; Khan et al., 1999). Fretting can occur even in the absence of a corrosive medium. The medical significance of fretting attack lies in its intensity that may give rise to a high amount of corrosion products in adjacent tissues, or it may be a major factor in crack initiation and fracture failure (fatigue) of an implant (Syrett & Wing, 1978).

2.3 Stress corrosion cracking

The presence of mechanical loading may create further complication of sudden fracture of implants, due to the phenomenon of stress corrosion cracking (SCC), which is believed to be the most dangerous form of corrosion-assisted failures. SCC may be a particular concern for devices having sharp contours such as pins and screws that are used to secure orthopedic implants. Stress corrosion cracks may propagate unobserved to a sudden disastrous failure. Such sudden failures caused by SCC of an implant may have serious results such as difficult removal of the failed device and painful irritation or inflammation of the surrounding tissues. In the past, numerous instances of fracture due to SCC of metallic implants of SSs, titanium alloys, and Co-Cr alloys were reported (Bombara & Cavallini, 1977; Sivakumar & Rajeswari, 1992). Pitting is one of the common SCC initiators, while the localized crack-tip dissolution facilitates crack propagation (Raman & Siew, 2014). Crack initiation takes place due to the localized electrochemical dissolution of the metal. The protective films that form at the crack tip break with sustained tensile stresses, resulting in exposure of the fresh anodic metal to the corrosive medium continuously, and hence, the SCC propagates progressively over a period of time (Panda et al., 2014).

2.4 Fatigue

Biomedical materials that are subjected to cyclic loading and enough stresses in the presence of an aggressive environment fail due to fatigue (Kamachi & Baldev, 2008). The fatigue failure occurs in two stages: (i) the initiation of a fatigue crack and (ii) crack propagation to failure with repeated cyclic loading. Mechanical failure of orthopedic implants is the most common due to fatigue. In some cases, however, the mechanism responsible for crack initiation and crack propagation may be different (Hu et al., 1993). Concerning the fatigue properties, the surface condition is crucial for the development of a crack under cyclic loads. Nucleation of fatigue cracks has been associated to the presence of pits on the surface of metallic materials (Sudarshan et al., 1990). Under fatigue conditions, the aqueous environment can accelerate the initiation of a surface flaw and propagate it to a critical size, leading to fracture. This process is known as corrosion fatigue, indicating the failure of a material under the simultaneous action of cyclic load and chemical attack (Suresh, 2004). The reduction in fatigue life of SS implants under corrosion fatigue was well documented (Giordani et al., 2004).

The fatigue process is found to increase further due to the formation of wear debris leading to fatigue wear. During fatigue, there is disruption of the oxide layer and the inability of the material to repassivate immediately, exposing some regions of the metal to the environment leading to corrosion. The initiation of crack due to fatigue was examined during the measurement of the corrosion potential of CD 316L SS, and it was also observed that the fatigue strength dropped radically when the repassivation was suppressed. This confirms that the oxide layer formation plays a vital role in the determination of the fatigue life of the materials exposed to an aggressive corrosive environment (Park, 1984). In addition, fretting that occurs between an implant and the bone is also found to accelerate the fatigue as the repassivation becomes more difficult in the presence of fretting (Geetha et al., 2010).

2.5 Metal release in a biological environment and their effect

The rupture of the protective surface oxide layer subsequently results in exposure of the nascent surface to corrosive attack and the potential release of metal ions (Shaw & Kelly, 2006). A huge amount of released metal ions could be harmful to human health and may eventually lead to severe complications and failure of the implant system. Metal ion release was frequently recognized as a cause of clinical failure or cutaneous allergic reactions. The correlation between metal ion release and hypersensitivity reaction to Ni, Cr, and Co released from SS and Co-Cr-Mo alloy was observed (Summer et al., 2007). Dermatitis related to Ni toxicity was reported by many researchers. Numerous studies on animals also showed the carcinogenicity of Ni through inhalation, and intravenous and intramuscular parenteral administration of ground Ni complexes. Cr toxicity appears to be closely related to the Cr valence state. Cr(+6), is hemotoxic, genotoxic, and carcinogenic (Barceloux, 1999). When Cr(+6) enters the bloodstream, it damages blood cells by causing oxidation reactions. This oxidative damage can lead to hemolysis and, ultimately, failure of the kidneys and liver (Dayan & Paine, 2001).

The surface oxide of SS is present in complex environments like the body, exposed to different ligands (complexing agents) such as citrate and proteins that influence metal ion release because they are the so-called electrolytes (Hedberg et al., 2017). This induces ligand adsorption, complexation with a surface oxide/hydroxide metal atom, and the possible detachment of the ligand-metal complexes from the surface oxide. Ligand-induced metal release is, thus, adsorption dependent (Carbonaro et al., 2008) but also dependent on pH, stability constants of surface complexes, and other parameters such as temperature (Milosev & Strehblow, 2000; Carbonaro et al., 2008; Hedberg et al., 2011). Depending on the environment, the metal released from steel is also influenced by other processes including active corrosion and protonation (Hedberg et al., 2011, 2013). It was reported that the metallic materials with strong passive films exhibit lower quantities of metal ion release from implants. Therefore, the amount of a metal ion released changes depending on the nature and strength of the metal-oxide bond, structure, role of alloying elements, composition, and thickness of oxide films (Browne & Gregson, 1994). The role of alloying elements in passive films formed on 316L SS and Co-Cr-Mo alloy in a calf serum solution was observed by X-ray photoelectron spectroscopy (XPS). The XPS spectra of the passive films demonstrated the peaks of Cr₂O₃, CoO, Fe oxide, and Mo oxide (Okazaki et al., 1997). Further, XPS examination made clear that metal release from various grades of SS into synthetic body fluids changes depending on the Cr content in the oxide film. The rate of metal release from SSs into artificial lysosomal fluid decreases with the increase in Cr/(Cr + Fe) ratio in the oxide film, which clearly illustrates that Cr is attributed to passive film formation (Herting et al., 2007).

4.1 Cardiovascular implants

The use of cardiovascular implants is subjected to its blood compatibility and its integration with the neighboring environment where it is implanted (Agrawal, 1998). Cardiovascular implants should possess unique blood biocompatibility because one of the leading causes of cardiac biomaterial failure is the initiation of thrombosis formation by the blood-contacting devices (Weng et al., 2012). The use of cardiovascular implant applications includes heart valves, endovascular stents, and stent-graft combinations (Jaganathan et al., 2014). One of the important applications of the cardiovascular implant is stents. The 316 SS, Ti and its alloys, and Co-Cr alloys are frequently used metallic materials for stents. The 316L SS is most commonly used as a metal for stents either with or without a coating material. The Co-Cr alloys, which conform to ASTM standards F562 and F90, are also being used for making stents because of their high elastic modulus (210 GPa) (Geetha et al., 2010).

Shape-memory alloys were proposed for cardiovascular stents due to their self-expansion ability. In the medical field, the family of Ni-Ti alloys (nitinol) is the most popular one that has been widely used because of its good biocompatibility, significant resistance to corrosion and fatigue, and the fact that its elastic modulus is close to that of human bone (approximately 35 GPa) (Dai and Ning, 2004). Nitinol constituting 49.5–57.5 at% nickel is used for fabricating self-expanding stents. Self-expanding stents have a smaller diameter at room temperature and are capable of expanding up to their preset diameter at body temperature. After implantation, it regains its original shape and conforms to the vessel wall because of the increase in temperature inside the body (Kurella, 2006). In comparison with other alloys, nitinol exhibits excellent fatigue properties at high strain levels. It forms a passive TiO₂ layer that acts as both a physical barrier to nickel oxidation and protects the bulk material from corrosion. Its corrosion resistance is greater than that of SSs (Ryhanen, 1999). In some cases, Ni or Ti (only a few percent) in Ni-Ti alloys can be partially replaced by Cu, Co, Fe, Nb, or Mo to improve the hysteresis (stress and/or temperature hysteresis), corrosion performance, control of transformation temperatures, fatigue behavior, etc. (Humbeeck et al., 1998).

In comparison to other alloys, nitinol shows excellent fatigue properties at high strain levels.

4.2 Orthopedic implants

One of the most prominent applications of biomaterials is as orthopedic implant devices. The materials used for orthopedic implants, particularly for load-bearing applications, should possess superior corrosion resistance in the body environment, excellent combination of high strength and low modulus, high fatigue and wear resistance, and high ductility. Orthopedic implants include both temporary implants such as plates and screws and permanent implants that are used to replace the hip, knee, spine, shoulder, etc. The corrosion mechanisms that occur in temporary implants are crevice corrosion at shielded sites in the screw/plate interface and beneath the heads of fixing screws and pitting corrosion of the implants made of SS (Yu et al., 1993; Jones, 1996). One of the major causes of the failure of the orthopedic implants is wear, which, in turn, is found to accelerate the corrosion. Hence, high wear-resistant materials such as Co-Cr alloys are often preferred to fabricate orthopedic implants. The femoral components are sometimes coated with cement to have good fixation. Total joint replacement is widely regarded as the major achievement in orthopedic surgery.

4.3 Dental implants

These implants face a more aggressive environment in the mouth because the pH of saliva varies from 5.2 to 7.8. Thus, the major reasons for the corrosion of metallic dental implants and fillings are temperature, quantity and quality of saliva, plaque, pH, protein, and the physical and chemical properties of food and liquids as well as oral health conditions. Chaturvedi (2009) reviewed broadly the corrosion aspect of dental implants. Sometimes, as two metallic components are used together in making dental implants, galvanic corrosion occurs very frequently in dental implants. Titanium and its alloys are resistant enough to pitting corrosion in different in vivo conditions encountered; however, they undergo corrosion in high fluoride solutions in dental cleaning procedures (Probster et al., 1992).

4.4 Other implants

Evolution of cochlear implant technology provides a sense of sound to a person with severe to profound sensorineural hearing loss. Binaural cochlear implantation was used in children as well as in adults. The implant consists of an external portion that sits behind the ear and a second portion that is surgically placed under the skin. Because of their development, perimodiolar electrodes, implantable microphones, and rechargeable batteries, etc., are used for fully implanted devices (Balkany et al., 2002). Solid biocompatible implantable devices for sustained or controlled intravitreal drug delivery to the posterior segment of the eye were developed employing diverse approaches and including osmotic mini-pumps, nonbioerodible and bioerodible drug-loaded pellets, configured capillary fibers, biodegradable scleral plugs, scleral discs, polymeric matrices, and scaffolds of a variety of geometries providing unique mechanisms of drug release for the delivery of drugs (Choonara et al., 2010).

In all types of corrosion of metallic implants, the surface passive film plays a crucial role in corrosion resistance, and the film determines in vivo metal ion release (Hanawa, 2004). The function of the passive film on the surface of SSs shall be discussed in the coming section of this review. The special mode of corrosion of SSs used as implants is pitting corrosion in the presence of body fluids due to the presence of a higher extent of chloride ions. Therefore, this review will focus especially on pitting corrosion in a chloride-containing solution.

6.1 Material parameters

This includes the factors that are directly related to the material itself. The following factors are under this category.

6.2 Environmental parameters

This includes the surrounding environment of the implants. The following factors are under this category.

7.1 Change in mechanical properties

The material that will be selected for a particular biomedical application is decided by mechanical properties. Hence, the mechanical behavior of an implant material is a very essential factor. Some of the properties that are of primary concern are hardness, tensile strength, modulus, and elongation. It is well known that austenitic SSs cannot be hardened by heat treatments. On the other hand, cold or warm working (drawing, rolling, forging, etc.) can make such SSs harden. A number of studies were performed on the effect of CD on the structural and mechanical behavior of SSs. Bhav Singh et al. (2009) reported that increasing the percentage of cold reduction increases the strength and hardness with loss of ductility for nitrogen-alloyed steels. It is known that austenitic SSs are metastable, which means that they can undergo a diffusionless phase transformation from austenite to martensite by deep cooling or by plastic deformation or by cyclic loading (Takaki et al., 2004; Hou et al., 2011). Strain-induced α′-martensite produced in metastable austenitic SSs during cold rolling led to significant increase in their strength. Formation and the amount of strain-induced martensite depend on the austenite stability (chemical composition and initial austenite grain size) and the rolling conditions (extent of deformation given, temperature, and rolling speed). The amount of deformation-induced martensite also depends upon numerous other factors such as plastic strain, strain rate, stress state, deformation mode, grain size, and grain orientation (Sato et al., 1989; Nakada et al., 2010). The martensite formation is also largely dependent on the steel composition and stacking fault energy (SFE) (Opiela et al., 2009).

When the austenite stability and the deformation temperature are low, or the degree of deformation is high, the martensite content increases (Murata et al., 1993). The increase in α′ martensite formed by plastic deformation causes an alteration in the physical properties of austenitic SSs (Huang et al., 2007). CD can modify the physical properties even when the martensite phase is not formed. The hardness was found to increase with an increase in the percent reduction in thickness, which could be attributed to the work-hardening effect. The Vickers hardness number of 304SS was found to double at the initial stages of cold work (up to 25%) (Roger, 1985). Further, cold working of this type of steel (up to 50%) raised the hardness value only by 7%. Milad et al. (2008) reported that the tensile strength, yield strength, and hardness of 304SS were found to increase with the increase in cold rolling up to 45%. It was found that the ratio of tensile strength to Vickers hardness is constant and is approximately three times up to 50% of cold rolling. Figure 1 demonstrates the engineering stress-strain curves for annealed and cold-worked 304 steel. It may be seen that yield strength (0.2%) and ultimate tensile strength increased, and elongation decreased with an increase in the degree of cold working (Milad et al., 2008). It was demonstrated for 316L and HNS that the annealed materials exhibit large strain before fracture and reveal the largest area under the stress-strain curve and, therefore, possess a high capacity for energy absorption (Talha et al., 2015). The ability for energy absorption decreased with an increase in cold reduction. Lula (1986) also reported that the CD of 304SS caused a substantial increase in the yield and tensile strengths. The ductility was found to decrease sharply at the beginning of rolling (up to 30%); after that, the rate of decrease was observed to be lower. A detailed discussion related to the effect of CD on the mechanical properties of austenitic SSs is beyond the scope of this review. CD has significant impact on the corrosion behavior of austenitic SS, which shall be discussed in the next section.

Figure 1:

The stress-strain curves of the deformed and the undeformed 304 SS (Milad et al., 2008), reproduced with permission from Elsevier).

7.2 Change in corrosion resistance

SSs are exposed to different levels of cold work during the final manufacturing stages of components for biomedical applications. Moreover, work hardening was a common process to further enhance the strength of austenitic SSs, and in many cases, they are used at a cold-worked state. CD is known to modify the corrosion resistance properties of SSs by increasing the dislocation density, which may help in increasing the diffusion of certain species, phase transformation, nucleation, etc. (Hamdy et al., 2007; Dwivedi et al., 2017). Both the thickness and composition of passive films are likely to be altered in many ways by CD (Navai and Debbouz, 1999; Phadnis et al., 2003). CD results in the formation of martensite and residual stresses on the surface, which affects the localized corrosion resistance (Elayaperumal et al., 1972).

As discussed above, the austenite phase in SSs exists as a metastable phase. It can be transformed into a different phase such as α′ or ε martensite, during the CD process (Nagy et al., 2004; Shirdel et al., 2015). Figure 2 shows the image quality (IQ) maps and kernel average misorientation (KAM) maps obtained from electron backscatter diffraction (EBSD) for non-deformed and deformed 304 SS samples (Luo et al., 2017). As revealed in Figure 2A, the samples without deformation have irregular austenite grains with only annealing twins. After 5% deformation, the steel possessed the same polygonal microstructure as the non-deformed samples. Slight CD did not alter the grain size or angles of the grain boundaries. After 20% deformation reduction, slip bands and deformation twins were found in some grains. As shown in KAM maps, the non-deformed samples exhibited defect-free grains. It is clear from the KAM maps in Figure 2C₂ that the changes in each grain were not uniform after 20% deformation, and the degree of non-uniformity is increased with the increase in the deformation reduction. Numerous low-angle boundaries and CD-induced martensite phases were observed in the austenite matrix with an increase in the degree of CD, and the presence of martensite also affected the uniformity of the grains. Figure 3 reveals he IQ maps, IQ maps with grain boundaries, and KAM maps after 40%, 60%, and 80% CD of 304 SS (Luo et al., 2017). The majority of the twin boundaries disappeared during the deformation process, and numerous low-angle boundaries instead of high-angle boundaries were observed after 40% deformation. Martensite phases were generated at the slip lines or stacking faults. The allocation of the martensite phase was heterogeneous, and some of the remaining austenite phase that did not transform was evidently observed in the microstructure. When the CD increased from 60% to 80%, as shown in Figure 3B, B₁, C, and C₁, the content of the martensite phase increased with the level of deformation, and the originally large-sized grains were replaced by small-sized grains. The KAM maps (Figure 3A₂–C₂) exhibit that the grains were non-uniformly deformed, and there was more rigorous distortion in the martensite phase grains. As the deformation reduction increased to 80%, the grain boundaries disappear, and the average grain size could not be accurately measured. It was demonstrated that the grain boundary diffusion and corrosion resistance strongly depend on the crystallographic structure of the grain boundary (Shimada et al., 2002; Kobayashi et al., 2016).

Figure 2:

EBSD results of image quality maps, image quality maps with grain boundaries (red lines represent low-angle boundary (2° 5°), green lines represent low-angle boundary (5°–15°), blue lines represent high-angle boundary above 15°), and the KAM maps of 304L SS with different deformations (A), (A₁), (A₂) 0% deformation reduction; (B), (B₁), (B₂) 5% deformation reduction; (C), (C₁), (C₂) 20% deformation reduction (Luo et al., 2017), reproduced with permission from Elsevier).

Figure 3:

EBSD results of image quality maps, image quality maps with grain boundaries (red lines represent low-angle boundary (2°–5°), green lines represent low-angle boundary (5°–15°), blue lines represent high-angle boundary above 15°), and the KAM maps of 304L SS with different deformations (A), (A₁), (A₂) 40% deformation reduction; (B), (B₁), (B₂) 60% deformation reduction; (C), (C₁), (C₂) 80% deformation reduction (Luo et al., 2017), reproduced with permission from Elsevier).

Figure 4 reveals XRD patterns and EBSD phase maps with different deformation reductions of 304 SS (Luo et al., 2017). A single fcc-phase was attained without CD and no phase change was observed after 5% CD. After 20% CD, a weak peak representing the α′-martensite (110) phase was detected in the XRD patterns. However, there is no evidence of formation of martensite up to 20% CD for 316LVM and HNSs (Talha et al., 2014a). As shown in Figure 4, the intensity of the austenite phase peaks decreased, and the martensite phase peaks increased, with an increase in the deformation reduction. The amount of martensite formed was associated with the degree of CD, and the amount of the martensite phase increases with the CD level as shown in the phase maps of the EBSD (Figure 2). Thus, the SSs are more susceptible to martensite formation when deformed at room temperature. It was also reported that during the plastic deformation process, two types of martensite could be formed: α′-martensite (bcc) and ε-martensite (hcp), depending on the chemical compositions, deformation variables, and the stacking-fault energies (Tavares et al., 2003).

Figure 4:

XRD patterns and EBSD of phase maps of 304L SS with different CD reductions (blue in the phase map stands for the austenite phase, and the red stands for the martensite phase) (Luo et al., 2017), reproduced with permission from Elsevier).

Many studies have reported the influence of deformation level on the passive behavior of SS in different solutions (Phadnis et al., 2003; Feng et al., 2016; Ren et al., 2016). The influence of CD, in particular, the amount of α′ martensite formation, of AISI 304 SS in 3.5 wt.% NaCl was studied by Xu and Hu (2004). They observed that when the content of martensite was less than 6% and more than 22%, the pitting sensitivity increased. However, when the martensite content was between 6% and 22%, the sensitivity decreased with increasing degree of deformation. Fang et al. (1997) reported that the corrosion potential of the martensite phase was more negative than that of austenite, and this is the primary reason for the selective corrosion of martensite. In the case of high chloride concentration, the degree of pitting increases linearly with increasing volume fraction of martensite (Fang et al., 1997). It was observed that the intergranular SCC of 316L SS and a nickel-based alloy were associated with the increase in the fraction of the low coincidence site lattice (CSL) boundaries (West and Was, 2009). It was demonstrated by Luo et al. (2017) that when the deformation increases, the fraction of low CSL decreases, which increase corrosion in acid solution. The related passive current in acidic media is found to increase with deformation by several orders of magnitude (Barbucci et al., 2002), but an opposite result was reported in neutral chloride-containing medium (Phadnis et al., 2003). For cold-worked 18Cr-10Ni-2Mo steel, the number of pits generally increases with an increase in deformation in NaCl solution, except in the 15% deformation where low pits count and large average pit sizes were reported (Szklarska-Smialowska, 1971). Forchhammer and Engell (1969) studied the effect of CD on the pitting corrosion behavior of different austenitic SSs, i.e. 18Cr-10Ni, 25Cr-10Ni, 17Cr-10Ni-2.4Mo, and 16Cr-14Ni-Mo. They established that in 30% NaCl solution, pitting potential was not greatly affected by CD, but for the deformed specimens, the number of pits was higher, and the pits were smaller. According to Randak and Trautes (1970) the quasi-martensite produced in 18Cr-8Ni steels by cold working does not change the pitting potential value in chloride solutions, but pit density increases with increasing CD level.

It is well established that the main factor that is responsible for these degradations in austenitic SS is the martensitic transformation of γ (fcc) to α (bcc) by deformation or irradiation (Niffenegger & Leber, 2008; Xu et al., 2017). In the past decades, the effect of CD on pitting corrosion resistance of SSs was extensively studied. Almost all the researchers suggested that severe CD (beyond 20% cold work) would result in a sharp decrease in pitting corrosion resistance for conventional SSs. Additionally, it can be concluded that the deformation microstructures including deformation bands (or dislocation pile-ups) (Mudali et al., 2002; Peguet et al., 2007), martensite formation (Barbucci et al., 2002; Hamdy et al., 2007), or strain-induced residual stress (Barbucci et al., 2002; Hamdy et al., 2007; Kumar et al., 2007) would weaken the pitting corrosion resistance.

It was established that pitting and crevice corrosion resistances of ferritic SSs were not changed by cold working up to 20% in 0.5-m NaCl solution irrespective of the defect structure containing deformation bands (Mudali et al., 2002). Alternatively, it was also found that corrosion resistance properties of 316L and 316LVM steels increased with an increase in CD up to 20% deformation in simulated body fluid (Talha et al., 2014b). The calculated corrosion rates obtained from the weight loss study of these steels after 24 weeks of immersion in Hank’s and 3.5 wt.% NaCl solutions demonstrated that corrosion rate decreased with increasing degree of CD in both the solutions. Figure 5 shows the polarization curves for annealed and deformed 316L SSs in Hank’s solution at 37°C. It was established that the values of corrosion current decreased with increasing CD up to 20%, indicative of greater protection of the passive films as a result of CD (Talha et al., 2014b). One important requirement for selecting an implant for biomedical application is breakdown potential (E_bd), i.e. the potential at which the passive film breaks and the anodic current significantly increases. The E_bd was also found higher for CD samples than for annealed samples and increased with the degree of deformation. Figure 6 shows the scanning electron microscopy (SEM) micrographs of the different corroded surfaces of SSs (annealed and cold work) after the polarization tests in Hank’s solution (Talha et al., 2014b). Pits were found on annealed as well as on cold deformed steel surfaces, but shallower and smaller pits were associated with deformed steels (up to 20%) compared to annealed steels. Similar results were also observed for HNSs (Navai, 1995). Figure 7 illustrates the change in critical pitting potential with the degree of CD for different austenitic SSs in neutral chloride medium at room temperature (Mudali et al., 2002). It is clear from the figure that the critical pitting potential increased up to 20% CD, beyond which the values decreased for 30% and 40% cold working for the 316L and two more nitrogen-containing steels. A potentiodynamic anodic polarization study by Mudali et al. (2002) also revealed that CD up to 20% enhanced the pitting resistance, and thereafter, a sudden decrease in pitting resistance was noticed for these steels. Pit initiation is closely related to passive film stability and to non-metallic inclusions acting as triggers for future pits. The XPS studies by Talha et al. (2014b) indicated that the Cr oxide:Cr hydroxide ratio and the Fe oxide:Fe hydroxide ratio were higher for deformed (up to 20%) 316LVM SS than for the annealed steel, indicating that the passive films on deformed surfaces are more stable and protective, which improves the corrosion resistance. Phadnis et al. (2003) also attributed the enhanced passivation characteristics of rolled 304 SS to the occurrence of a thicker passive film with a higher Cr/Fe ratio.

Figure 5:

Polarization curves for stainless steels in Hank’s solution at 37°C for 316L SS: (A) annealed, (B) 10% C.W., (C) 20% C.W. (Talha et al., 2014b).

Figure 6:

SEM micrographs of SSs after the polarization tests in Hank’s solution: (A) 316L (annealed), (B) 316L (10% C.W.), (C) 316L (20% C.W.), (D) 316LVM (annealed), (E) 316LVM (10% C.W.), (F) 316LVM (20% C.W.) (Talha et al., 2014b).

Figure 7:

Critical pitting potential values for the alloys (A, B, and C) against cold working levels: (A) N=0 in 316L, (B) 0.1 wt.% in 316LN, and (C) 0.22 wt.% in 36LHN (Mudali et al., 2002), reproduced with permission from Elsevier).

The effect of deformation on the corrosion of SS is not similar in acid solution than in neutral simulated body fluid or neutral chloride solution. It is prominent that rolling results in a preferential texture of high-index, low-density planes along the rolling direction of austenitic SSs. The effect of this texture results reduced the binding energy of the atoms in this less closed-packed plane, which can result in enhanced dissolution in an acidic chloride medium, where the metal dissolution step plays a main role in the rate control mechanism. This is not the case in a neutral chloride solution, where oxygen diffusion is the rate-controlling step. Enhanced surface diffusion could have contributed to the stable passive film formation at a low degree of cold working, leading to enhancement in pitting resistance (Talha et al., 2014b). Additionally, the chromium content in the passive film formed on a deformed steel surface is richer than that formed on heat-treated samples, and the higher chromium content improved the corrosion resistance of passive films (Phadnis et al., 2003). The results up to now demonstrated that a small amount of CD is significant to the corrosion resistance, but a high deformation is damaging in a neutral solution containing chloride ions (Fu et al., 2009). This is attributed to the fact that the formation of a significant amount of deformation twins because of a higher level of CD in the alloys, which show a different potential than the matrix (Glavatskaya, 1999) and the formation of strain-induced α′ martensite phase as discussed above. A higher degree of CD introduced a high defect density into the matrix, resulting in a less protective passive film as a result of reduced corrosion resistance for heavily cold-worked SS (Fu et al., 2009). This can be illustrated as shown in Figure 8.

Figure 8:

Schematic illustration of the stability of passive film on surface and martensite formation in SSs during CD.

For HNSs, the findings are even more interesting. It is well known that HNSs are more corrosion resistant due to the presence of nitrogen. Actually, rigorous CD significantly weakened the pitting corrosion resistance of HNS with relatively low nitrogen content (Fu et al., 2008). However, for HNS with high nitrogen content, the detrimental effect was found to be mitigated (Ren et al., 2016). Recently, Wang et al. (2017) surprisingly found that the detrimental effect of cold working (up to 60%) could be completely eliminated for HNS with very high nitrogen content (0.92 wt.%). This means that HNS could acquire the ability to heal the damage of passive film caused by cold working. Nitrogen enrichment at the oxide/metal interface was found by Olsson (1995). Although nitrogen is believed to play a significant role for the exceptional corrosion resistance of HNS, the form of nitrogen in passive film remains controversial (Baba et al., 2002; Ningshen et al., 2007). Wang et al. also found that severe CD significantly weakened the stability of passive film on the surface of HNS with 0.76 wt.% N, but had almost no detrimental effect on steels having 0.92 wt.% N. After severe CD, both passive films on steels having 0.76 and 0.92 still retained their special structures. Severe CD had a very strong influence on the enrichment layers of passive film on steel having 0.76 wt.% nitrogen, but almost had no influence on that of passive film on steels with 0.92 wt.% N. According to Wang et al. (2017), the doped nitrogen concentration in the passive film on 0.92 wt.% N was higher than that on 0.76 wt.% N. Severe CD significantly decreased the doped nitrogen concentration in the passive film on steel with 0.76 wt.% N, but this doped nitrogen concentration was not influenced in passive film on steel with 0.92 wt.% N. In other words, the doping effect of nitrogen enrichment in 0.92 N containing steel was high enough to compensate the detrimental effect of the severely deformed microstructure on passive film and could maintain its excellent pitting corrosion resistance at severely cold-worked state. Twins and deformation bands are their main structures, and no martensite was detected from X-ray diffraction (XRD) and EBSD studies in cold-worked microstructures of steels having 0.76 N and 0.92 N. Also, no difference in the behavior of inclusions was found between these two steels at cold-worked state. They conclude on the basis of previous study that the deformation bands (Mudali et al., 2002; Peguet et al., 2007) and residual stress (Barbucci et al., 2002; Hamdy et al., 2007) might be the main reason for the decay of pitting corrosion resistance for severely CD steel with 0.76 wt.% N as discussed above.

One can conclude from the above discussion that CD has no detrimental effect up to 20% CD, and it would result in a sharp decrease in pitting corrosion resistance for conventional SSs beyond 20% cold work in neutral chloride solutions, but it decreases regularly in acid solution even at slight cold work state. This effect is not obtained in steels having a very high content of nitrogen in neutral chloride solution. The main reason of the decreasing corrosion resistance is the formation of martensite phase, deformation bands, and residual stress after severe cold working. The work done by previous researchers and their outputs associated to the effect of cold working on the corrosion behavior of SSs in solutions containing chloride ions is summarized in Table 2.