Introduction

Corrosion resistance is commonly assessed using standard electrochemical methods that may not adequately represent the effects of service environments or material behavior that are important to long-term performance. For example, alloys are better suited for biomedical applications than ceramics or polymeric materials due to their superior load-bearing properties that provide high mechanical strength and fracture toughness1,2,3,4. Research on biomedical alloys during the past two decades has focused on corrosion behavior and biocompatibility in response to the accumulation of metal debris that has been observed in soft tissues (metallosis) and can lead to the death of tissues and cells in an organ due to the resulting lack of blood supply (necrosis). Durability tests during the development of those alloys failed to indicate susceptibility to corrosion during applications5,6,7. High corrosion resistance in the range of hostile environments that occur within the human body is a crucial property of metal implants, which can be exposed to fluids having significantly different pH and dissolved oxygen concentrations that vary locally8. Understanding the effects of environmental variables on performance is crucial for selecting appropriate materials for implant devices.

Materials used extensively for biomedical applications include titanium and cobalt-based alloys. Titanium and its alloys are known to undergo spontaneous oxidation in air and aqueous environments to form tenaciously adherent and passivating surface oxide films9. These passive states are dynamic, and breakdown and reforming of passive films is known to occur in some environments10,11.

Although these films are very chemically resistant and typically less than 10 nm thick12,13, titanium has been observed to accumulate in tissues near titanium implants14,15, which indicates in vivo localized corrosion does occur. Cobalt-based alloys are used in orthopedic implants because of their excellent mechanical and corrosion properties. The cobalt contents in CoCrMo alloys used in biomedical applications can be as low as 60 wt %, and the chromium contents are ~28 wt%, which is sufficient to spontaneously form a chromium-rich passive oxide film (Cr2O3) on the surface16; Co- and Mo-oxides can also form on these alloys17,18. Localized corrosion of these implants will release toxic species that will lead to inflammation and cause adverse tissue reactions in the body19,20,21.

Environments in the human body can have high concentrations of aggressive chloride ions that are highly corrosive to most alloys. For example, fluids near an implant typically have pH values between 5.3 and 5.622, but can range between pH 3 and pH 9 in response to accidents, diseases, infections, and following surgery. Different areas of an implant can be subjected to different local pH values, which can promote galvanic interactions that enhance localized corrosion. Despite the formation of passive films that protect materials from corrosion, localized attacks can occur in the presence of aggressive ions.

Measurements were made to assess the effect of passivation on the responses of commercially pure titanium (CP-Ti) and CoCrMo alloy in standard methods that are commonly used to determine the susceptibility of implant materials to localized corrosion. The standard potentiodynamic (PD) scan is one of the most widely accepted methods used to measure the corrosion resistance of alloys, where the measured current indicates the rate of corrosion. The cyclic potentiodynamic polarization (CPP) scan is used to determine corrosion susceptibility, assess passivity, oxide film breakdown, and repassivation after pitting, and determine the corrosion rates of alloys. For example, the propensity for localized corrosion is commonly thought to be revealed if the reverse (cathodic) scan does not overlay the forward (anodic) scan and, instead, a positive hysteresis loop with higher currents develops23,24. In standardized methods, two features of the CPP scans quantify the response: (1) the breakdown or transpassive potential occurring in the forward anodic scan and (2) the potential at which the hysteresis loop in the reverse cathodic polarization scan is completed, which is referred to as the repassivation or protection potential. In general, once initiated, localized corrosion is expected to propagate at any potential that is more electropositive than protection potential.

The effects of the immersion time under open circuit (OC) conditions prior to a CPP scan and the vertex potential used in the CPP scan on characteristic values were assessed. Measurements with two biomedical alloys with different corrosion behaviors and property values were used to evaluate the effects on the electrochemical responses: CP-Ti shows an extensive passive region, and CoCrMo alloy displays a short passive region with a distinct breakdown or transpassive potential. Phosphate buffer saline (PBS) solution was used as the electrolyte to simulate the human body environment. Electrodes made from these materials were immersed in PBS solution at room temperature, and open circuit potential (OCP) measurements were made over several durations. Electrochemical impedance spectroscopy (EIS) was used to measure changes in the electrical properties of the surfaces after each immersion time at OC to evaluate the surface stability. The CPP scans were made immediately after the EIS measurements. In addition, CPP scans of freshly polished electrode surfaces of CP-Ti and CoCrMo alloy not subjected to prior immersion were run from the same starting potential to three different vertex potentials to determine the effect of the vertex potential on the reverse CPP scan profile and the characteristic values derived from that profile. Finally, cathodic polarization scans with freshly polished CP-Ti and CoCrMo alloy were performed starting at the vertex potential used in the CPP scans to assess the effects of the anodic scan on the characteristics of the cathodic return scan.

Results and discussion

Metallurgical characterization of CP-Ti and CoCrMo

The microstructure of commercially pure titanium is α single-phase with grain sizes ranging between 5 and 20 μm. Optical and field-emission scanning electron microscope (FESEM) images of the overall microstructures of chemically etched (using Kroll’s reagent: 10 mL H20 + 2 mL HF + 5 mL HNO3) CP-Ti electrode surfaces are shown in Fig. 1a, b, respectively. Figure 2 is a high-magnification FESEM photomicrograph showing precipitated impurities located within the grain and at grain boundaries. The numbers within the figure show locations at which energy-dispersive X-ray emission spectroscopy (EDS) analyses were performed. The results of those analyses are given in Table 1 and show that the matrix is pure Ti, and the grain boundary and matrix precipitates are rich in iron25,26.

Fig. 1: Representative microstructures of chemically etched CP-Ti electrode surfaces.
figure 1

a Light optical microscope and b field-emission SEM.

Full size image
Fig. 2
figure 2

A high-magnification FESEM micrograph shows precipitated impurities located within the grain and at grain boundaries of CP-Ti.

Full size image
Table 1 Results of EDS analyses of CP-Ti (wt. %).
Full size table

The CoCrMo alloy microstructure is also single-phase and fine grain, but more equiaxed than CP-Ti and with substantial twinning. This is seen in the FESEM image of a chemically etched surface in Fig. 3a. The results of EDS spot analyses conducted at six different locations (see Fig. 3b and Table 2) confirm the alloy is compositionally homogeneous.

Fig. 3: FESEM micrographs of CoCrMo alloy locating EDS spots.
figure 3

a Fine grain twinned microstructure and b high-magnification micrograph.

Full size image
Table 2 EDS analyses of CoCrMo alloy (wt %).
Full size table

Electrochemical corrosion results

In standardized methods, two features of the CPP scans are used to quantify the response: (1) the breakdown or transpassive potential during the forward anodic scan and (2) the potential at which the hysteresis loop in the reverse cathodic polarization scan crosses the forward scan. In this investigation, the effects of surface stabilization by immersion prior to CPP scans and scan parameter values on values derived from scans using CP-Ti and CoCrMo materials were determined. The forward CPP scans are used to show the effects of immersion time on Ezcf, Izcf, and Eb values. The reverse scans are used to indicate the occurrence of hysteresis and measure Ep, Ezcr, and Izcr values.

Effect of immersion times at open circuit on resting potential (E r)

The results of OC immersion in Fig. 4a, b show that the potentials of CP-Ti and CoCrMo increase with immersion time until stable Er values are attained. Measured values are included in parentheses. The rapid increase in the potentials measured within the first eight hours can be attributed to the formation and initial growth of an oxide film on the electrode surfaces. The potentials approach stable values as the oxide layers stabilize. The stable values approached in measurements made with different electrodes of the same material differ most likely due to impurity concentrations in the CP-Ti and minor compositional differences in CoCrMo. In the measurements with CP-Ti electrodes, values of −0.52 and −0.37 VSCE attained after immersion times of 1 and 8 h indicate the initial growth of the oxide film, which stabilized after ~72 h at −0.25 VSCE. The resting potential remained essentially constant beyond 72 h, with a value of −0.27 VSCE measured after 168 h. This indicates that the oxide film formed on CP-Ti remained nearly stable beyond about 72 h.

Fig. 4: Potentials measured during immersion for 1 h, 8 h, 72 h, and 168 h at room temperature in PBS solution.
figure 4

The values of Er are given in parentheses. a CP-Ti and b CoCrMo alloy electrodes.

Full size image

Similar behavior occurred in the measurements with CoCrMo electrodes, where the potential increased rapidly during the first few hours to values of −0.33 and −0.31 VSCE after 1 and 8 h, respectively. The faster stabilization of Er suggests a more spontaneous formation and faster initial growth of the oxide film on CoCrMo than on CP-Ti. The Er values measured for CoCrMo were similar beyond 8 h, reaching values of −0.10 VSCE after 72 h and −0.08 VSCE after 168 h. The potential became unstable after ~55 h in the 72-hour immersion test, but returned to the same Er value near −0.10 VSCE. This probably indicates the occurrence of localized corrosion and repassivation of the electrode surface. The sudden increase to −0.10 VSCE after 144 h in the 168 h test may be a measurement artifact.

EIS analyses were performed to measure the surface electrical properties of the electrode surfaces generated after different OC immersion times. Figure 5a, b show Bode and Nyquist plots of CP-Ti and CrCoMo alloy materials, respectively. The Bode phase angle plots for CP-Ti in Fig. 5a (top) indicate highly capacitive surface films formed during the 72- and 168-hour immersions, based on phase angles being higher than −80° over a wide frequency range. The film formed within an immersion time of 1 h also resulted in phase angles higher than −80°, but over a narrower frequency range. The increase in frequency ranges with high phase angles (>−80°) suggests that the surface film becomes more stable after longer immersion times. The same frequency ranges have high phase angles in analyses of the oxide films developed during the 72- and 168-h OC immersions, which indicates the passivating film is stable.

Fig. 5: EIS analyses were performed to measure the electrical properties of the surfaces generated after different OC immersion times.
figure 5

a CP-Ti and b alloy CoCrMo after different OC immersion times in PBS solution at room temperature as Bode phase angle (top), Bode magnitude (middle), and Nyquist (bottom) plots.

Full size image

The CP-Ti Bode magnitude plots in Fig. 5a (middle) show that the oxide surface films become highly capacitive within a short immersion time, attaining impedances of ~350 kΩ cm2 at 10 mHz after 1-hour immersion and increasing to 700–800 kΩ cm2 after 72-h OC immersion. These results show continuous increases in phase angle and impedance with immersion time at OC that indicate surface passivation and stabilization continues through the longest test duration of 168 h. The Nyquist curves in Fig. 5a (bottom) show other effects of surface film stabilization. The Warburg tail observed in the EIS analysis of the surface formed after 1 h (indicated by the dashed line) reveals that diffusion occurs during the initial formation of the oxide film, but a Warburg tail was not detected after longer immersion times as the film stabilized.

The Bode and Nyquist plots for immersed CoCrMo alloy electrodes are shown in Fig. 5b. The Bode phase angle plots (top) indicate that highly capacitive surfaces developed during the 8-h and 168-h OC immersion times, both reaching phase angle values higher than −80° over a broad frequency range. The Bode phase angle plots for analyses of the 1-h and 72-h OC immersion electrode surfaces showed lower phase angles of −71° and −68° at 10 mHz, although both surfaces had values near −80° at higher frequencies. The phase angle values measured for the 72-h OC immersion sample are lower than those measured for samples with 8-h and 168-h immersion times. This is consistent with the instability that was detected in the open circuit potentials measured for the 72-h electrode prior to the EIS measurements (see Fig. 4b). The high phase angle values observed at high frequencies (>50 kHz) measured after the 72-h immersion is most likely due to inductance caused by the potentiostat and not related to the response from the surface.

The high impedance values in the Bode magnitude plots shown in Fig. 5b (middle) indicate highly capacitive surface films had formed after short OC immersion times. The impedances measured for all electrodes at 10 mHz range between ~370 and 450 kΩ cm2. The Nyquist plots show well-defined curves with no evidence of a Warburg tail, which is consistent with the rapid formation of a surface oxide film on CoCrMo. Note that the Nyquist curves indicate films formed during 1-h and 72-h OC immersions are slightly less capacitive than those formed during 8-h and 168-h immersion tests.

Comparing the Bode phase angle plots after 1-h and 8-h OC immersion times (see top plots in Fig. 5a, b) shows that the phase angles at 10 mHz are higher for CoCrMo alloy than for CP-Ti. This suggests that the oxide film on CoCrMo stabilized more rapidly. The phase angles are higher for the CP-Ti surface oxide than for the CoCrMo alloy surface oxide after immersion times longer than 8 h. The Bode magnitude plots show higher impedance values for the passive layer formed on CP-Ti compared to that formed on CoCrMo alloy for immersion times beyond 8 h. The Nyquist curves (see Fig. 5a, b bottom) show a steep climb in impedance (Zim) values for the surface formed after 8-h immersion time for both materials, which indicates the surfaces became more stable after immersion for 8 h. It is further observed that the Nyquist curve for CP-Ti after 168-h immersion has the steepest slope and a higher impedance than the film formed on the CoCrMo alloy. This is most likely because TiO2 film formed on the surface is more capacitive than the film formed on CoCrMo alloy.

Effect of open circuit immersion on CPP forward scans

Figure 6a, b show the forward scan segments of CPP with CP-Ti and Alloy CoCrMo after different OC immersion times, respectively. The CPP scans were started at 0.01 VSCE below the voltage measured at the end of the immersion time to minimize damage to the oxide films that had formed during OC immersion by imposing cathodic potentials at the beginning of the CPP scans. The CPP scans were run to a vertex voltage of 1.5 VSCE. The results of scans of freshly polished (bare) CP-Ti and CoCrMo electrode surfaces are included for comparison; scans with bare surfaces were initiated at −1.0 and −0.72 VSCE, respectively. Results for both alloys show that Ezcf values increase and Izcf values decrease with increasing OC immersion times. The surfaces generated at different immersion times result in oxide films of different stabilities that provide some protection, but evolve to fully passivating layers as the potential increases. Longer immersion times result in more robust oxide films, as indicated by higher Ezcf and lower Izcf. This results in a slower increase in the current as the surface is polarized than is observed for the bare surface (i.e., a finite slope), and higher potentials are required to reach the limiting current.

Fig. 6: The forward segments of CPP scans after different OC immersion times in PBS solution.
figure 6

a CP-Ti and b CoCrMo electrodes.

Full size image

Passivation of CP-Ti persists to the vertex potential, whereas breakdown of passivation on the CoCrMo alloy occurs at ~0.5 VSCE, followed by a pseudo-passive breakdown at ~0.8 VSCE. Breakdown occurs at slightly different potentials in scans with CoCrMo alloy electrodes that had been immersed for different durations. Table 3 summarizes the characteristic values derived from the anodic CPP scans shown in Fig. 6. The Izcf values were calculated using Tafel analysis for the scan with bare sample, and values for the OC immersed samples were calculated manually using the slope of the anodic legs and the corrosion potential because of the lack of a Tafel region. The anodic scans for CP-Ti show that Ezcf values become more cathodic after longer immersion times at OC through 72 h. The Ezcf and Izcf for CP-Ti are the same after immersion times of 72 and 168 h, which implies that the oxide film becomes stable within 72 h. The anodic scans to potentials above 1.1 VSCE (indicated by the dashed line) show the contribution of currents due to oxygen evolution. The anodic scans of CoCrMo alloy show Ezcf is more cathodic after longer immersion times through 72 h (see Fig. 6b). The Izcf after the 168-h immersion (3.5 nA cm−2) is only slightly lower than after the 72-h immersion (10 nA cm−2); this is consistent with the EIS results that show surface stability is reached after 72 h of OC immersion. The partial currents due to oxygen evolution at high potentials are insignificant relative to the corrosion currents of CoCrMo alloy.

Table 3 Characteristic values derived from the forward CPP scans after different immersion times at OC.
Full size table

The anodic scans of all CP-Ti electrodes show passivity from potentials above Ezcf up to the vertex potential of 1.5 VSCE. Minor disruptions in surface passivation occurred in the scan with the 72-h OCP electrode that resulted in current spikes. The extended passive range of CP-Ti can be attributed to the high ohmic resistance of the titanium oxide film that stabilized during the scan and remained protective up to the high vertex voltages, even in the presence of the chloride ions. It was also observed that longer OC immersion times resulted in more stable layers such that the steady passive current density of ~0.7 μA cm−2 was not attained until higher potentials were reached, as seen in Fig. 6a.

The anodic scans of CoCrMo alloy electrodes also show higher Ezcf potentials after longer OC immersions, as shown in Fig. 6b, but the breakdown potentials were all ~0.4–0.5 VSCE. The passive current densities measured in all scans are ~0.7 μA cm−2. The second pseudo-passive region observed between 0.65 VSCE and 0.78 VSCE was not affected by the immersion OC times. It is expected that Cr2O3 and MoO2 form at the lower passive potentials, and the pseudo-passive zone is likely due to the formation of MoO42− based oxides27,28,29.

The passive oxides formed during OC immersions provide effective corrosion resistance of both materials. This is observed in the slow increases in corrosion currents as the electrodes are polarized above Ezcf relative to currents measured in the scan with the bare surface. Surface films formed during longer OC immersions attenuated corrosion over a larger voltage range, but both materials had the same responses at potentials above ~0.3 VSCE, as observed in Fig. 6. However, the passive films formed on CP-Ti during the forward scan remained protective to near the vertex potential, whereas the passive films formed on CoCrMo degraded at a breakdown potential that depended on the immersion time. Breakdown occurred at ~0.4 VSCE in scans with the bare electrode and the electrodes immersed at OC for 1-h and 8-h OC, and ~0.5 VSCE in scans with the electrodes immersed at OC for 72 and 168 h.

Effects of vertex potential on CPP reverse scans

The CPP reverse scans were conducted after reaching the vertex potential down to 0.00 VSCE at the same scan rate used in the CPP forward scans. For clarity, the transition from net anodic to net cathodic currents during the reverse CPP scan is referred to as Ezcr. None of the CPP scans with CP-Ti in Fig. 7a show a positive hysteresis loop. This is probably because the vertex potential is within the passive region. All reverse scans show significantly lower current densities than were measured during the forward anodic scans at the same potential. The lower current density values and the shapes of the reverse curves indicate that the surfaces remain passivated during the reverse scan until Ezcr occurred. Measured values of Ezcr were not sensitive to the OC immersion times or the stabilities of the passive films formed during the OC immersions and anodic scans. This is expected because all scans were run to the same vertex potential, which would result in the formation of passive films with the same degree of stabilization.

Fig. 7: Cyclic potentiodynamic polarization scans conducted with CP-Ti electrodes.
figure 7

a After different OC immersion times and b bare electrodes scanned to different vertex potentials (gold arrows indicate CPP scan direction).

Full size image

The CPP scans of freshly polished and cleaned surfaces of CP-Ti electrodes (bare electrodes) performed up to vertex potentials of 0.5, 1.0, and 1.5 VSCE are shown in Fig. 7b. Minor differences in the Ezcf values measured in the three forward scans could be due to small differences in the surface finish after metallographic preparation. However, the Ezcr values on the reverse scans are significantly different and correlate with the vertex potentials. The highest Ezcr value was measured in the CPP scan run to the highest vertex potential (1.5 VSCE), and the lowest Ezcr occurs in the CPP scan run to the lowest vertex potential (0.5 VSCE). The Izcr values were similar in all scans. No evidence of pitting corrosion was detected during the microscopic examinations of the CP-Ti samples after the CPP scans.

Figure 8a shows CPP scans with CoCrMo electrodes subjected to different OC immersion times and with a polished electrode (bare surface) that was conducted to a vertex potential of ~1.6 VSCE. As was discussed in the context of the anodic scans, the immersion times at OC affected the measured values of Ezcf and the shapes of the anodic legs, but did not alter the breakdown potential. The stabilizing effects of the passive films formed during immersion at OC are apparent at voltages above Ezcf, but do not persist at voltages above the breakdown potential (Eb) in the anodic scans (see Table 3).

Fig. 8: Cyclic potentiodynamic scans conducted with CoCrMo alloy electrodes.
figure 8

a Scans up to a vertex voltage of approximately 1.6 VSCE for surfaces generated after different immersion times and b scans with polished surfaces to three vertex potentials (gold arrows indicate CPP scan direction).

Full size image

None of the reverse scans have a positive hysteresis loop. The reverse scans with the polished electrode (bare surface) and the 72-h and 168-h immersion electrodes follow the same paths as the forward scans at voltages higher than the passive region. (The scan with the bare surface electrode overlays the 168-h scan at voltages below ~0.6 VSCE.) The reverse scans with the 1-h and 8-h immersion electrodes are stabilized relative to the forward scans and have Ezcr values of 0.70 and 0.82 VSCE.

The difference in the responses during the reverse CPP scans with CoCrMo electrodes immersed for different OC hold times is attributed to differences in the surface morphologies occurring at the vertex potential. Chromium and molybdenum were most likely leached to different extents during the OC immersions as passive layers formed such that the surface compositions at the vertex potential were different. Unlike CP-Ti, the CoCrMo alloy surfaces were not stabilized approaching the vertex in the forward scan and corroded actively. The different currents attained at the vertex potential may reflect different concentrations of Co, Cr, and Mo at the corroded surfaces. The reverse scans indicate the propensities of different surface compositions to passivate. For example, scans with the 1-h and 8-h samples that did not form stable layers during immersion appear to passivate at ~1.0 VSCE, but passivation of the 72-h and 168-h electrodes that had formed stable passive layers during immersion, and the bare surface electrode does not occur until ~0.5 VSCE in reverse CPP scans (Fig. 8a). This indicates the response in the reverse scan is characteristic of the surface existing at the vertex potential and is not characteristic of the initial CoCrMo alloy surface.

Figure 8b shows CPP scans of polished CoCrMo electrodes that were conducted to three different vertex voltages that are all higher than the breakdown potential of the pseudo-passive region. Significantly different maximum currents were attained at the different vertex potentials, as indicated by arrows in Fig. 8b. The reverse scans followed the paths of the forward scans and indicated the propensity of the damaged surface to repassivate. These results show there is no correlation between the Ezcr values and either Ei, Ezcf, Eb, Ev or the damage that occurred during polarization in the transpassive region of the CPP forward scan. The variance seen in the Ezcr values is possibly due to the differences in the surface microstructures of the electrode surfaces at the vertex potential, including the relative stability of residual surface oxides and the composition of the underlying alloy. Both are affected by the vertex potential attained.

The reacted surfaces of all CrCoMo alloy electrodes showed similar levels of localized corrosion regardless of the immersion time prior to the CPP scans. Photomicrographs of a freshly polished surface and a corroded surface are shown in Fig. 9. The FESEM image in Fig. 9a is an etched surface that shows significant twining. The FESEM image shown in Fig. 9b is the surface of the electrode that had been immersed in PBS solution for 72 h and then subjected to a CPP scan to a vertex potential of 1.6 VSCE (red curve in Fig. 8a). The corroded surface is covered with a brittle oxide layer that has a blistered appearance. The oxide layer probably formed during OC immersion and was damaged when scanned above the breakdown potential.

Fig. 9: FESEM photomicrographs of CoCrMo alloy electrode.
figure 9

a Polished and etched surface and b corroded surface after CPP scan with 72-h OC immersion electrode.

Full size image

Comparing CPP scans with cathodic polarization scans

Cathodic polarization scans were conducted to compare the responses of freshly polished CP-Ti and CoCrMo alloy surfaces with those of surfaces subjected to the anodic polarization leg of the CPP scan. Cathodic scans of freshly polished CP-Ti and CoCrMo electrodes initiated at 1.6 VSCE (green) are compared with CPP scans of polished CP-Ti and CoCrMo electrodes made to vertex potentials of 1.5 or 1.6 VSCE (black) in Fig. 10a, b. The Ezcr value for the cathodic scan with CP-Ti is lower than the value in the CPP scan, whereas the Ezcr value for the cathodic scan with CoCrMo is higher than the value in the CPP scan. The initial current in the cathodic scan with CP-Ti at 1.6 VSCE decreases quickly as passivation occurs and continues to decrease at a rate similar to that measured in the CPP. In contrast, the initial current in the cathodic scan with CoCrMo decreases at a low rate similar to that measured at the vertex potential in the CPP scan, and the two scans follow the same path until ~0.9 VSCE, at which point the cathodic polarization scan exhibits Ezcr and the CPP scan shows continued active corrosion. The Ezcr value measured during the cathodic polarization scan with CoCrMo differs from the value measured in the CPP scan because the surface morphologies are different at the vertex potential. The surface of the electrode at the vertex potential of the CPP scan has been altered by the passivation and degradation processes that occurred during the forward (anodic) part of the CPP scan. This indicates the CoCrMo surface used in the cathodic polarization scan had a sufficiently high Mo content to passivate, but the leached surface of the CoCrMo electrode that existed at the vertex of the CPP scan did not. However, sufficient chromium remained to repassivate the CoCrMo surface at a lower voltage in the CPP scans (~0.4 VSCE). The shape of the cathodic polarization curve indicates passivation by chromium occurs, but the current increases at voltages below ~0.1 VSCE.

Fig. 10: Overlayed the CPP and cathodic polarization scans were conducted to compare the responses of freshly polished surfaces with those of surfaces subjected to the anodic polarization leg of the CPP scan.
figure 10

a CP-Ti and b CoCrMo. SEM photomicrographs of CoCrMo surface after cathodic polarization scan. c Low and d moderate magnifications.

Full size image

Figure 10c, d is SEM images of the CoCrMo surface after the cathodic polarization scans that show bright oxide particles embedded in a uniform brittle oxide layer. Oxides were most likely generated at voltages higher than Ezcr (>0.65 VSCE) and persisted during the scan. Table 4 shows the results of EDS analyses of the bright particles and the brittle oxide layer after the cathodic scan. Measurements show that the bright particles (Spot 1) are primarily cobalt oxide, and the dark gray oxide layer (Spot 2) is composed of both chromium- and cobalt-based oxides. Bright particles were not observed on the surface of the oxide layer after the CPP scan with CoCrMo alloy (see Fig. 9b), but probably formed beneath the previously formed oxide layers to create the blistered appearance. Although further tests and analyses are required to elucidate the mechanism, the results demonstrate the vertex potential affects the behavior of the return scan. This is likely due to the effects of the oxides formed on the surface during the scan to the vertex potential. The effects of the vertex potential on the surface microstructure can be determined by characterizing electrode surfaces that have been scanned up to different vertex potentials.

Table 4 EDS point analyses of the CoCrMo surface after cathodic polarization scan in atom %.
Full size table

Evaluating the use of CPP scans to indicate localized corrosion susceptibility

The observation of positive hysteresis loops in CPP scans is often considered to indicate the pitting susceptibility of a material. No hysteresis loops occurred in the CPP scans of CP-Ti because the highest vertex potential used (1.5 VSCE) was within the passive region. All reverse scans seen in Fig. 7a followed the same path and reached the same zero-current potential (Ezcr). This was anticipated because the polarization of all electrodes in the forward scans most likely resulted in the formation of passivating TiO2 films having similar thicknesses and compositions. The shapes of the return CPP scans in runs to different vertex potentials (see Fig. 9b) were similar and showed a consistent stabilizing effect of the passivated surface. The passive layers present on each CP-Ti electrode at the vertex potential stabilized the surfaces during the reverse scans until the decreasing anodic currents became equal to the cathodic currents (Izcr). The value of Ezcr at which Izcr was attained in CPP scans run to different vertex potentials depended on the scan time required for the anodic current to decrease to Izcr, not on the vertex potential. The anodic currents measured in the reverse scans were controlled by properties of the passivated electrode surface, but were not directly related to the corrosion susceptibility of the material.

Significant hysteresis loops were not observed in the CPP scans with CoCrMo alloy electrodes. This is probably because the passivating films generated during the OC immersion and forward scans were damaged at voltages above the breakdown potential, and the surfaces were actively corroding at the vertex potential. The vertex potentials were all higher than the breakdown potential of the pseudo-passive region (at ~0.8 VSCE). Differences in the current behaviors observed in the reverse scan and the measured Ezcr values are consistent with different extents of damage to the passive layers and the leaching of the underlying metal occurring on the different electrodes. This can be confirmed by additional surface and cross-sectional analyses of reacted electrodes. The responses are not characteristic of the initial CoCrMo alloy composition. Some electrode surfaces restabilized more rapidly and at higher Ezcr values than others, but the Izcr values were similar for all electrodes.

Corrosion assessment based on the responses from cyclic potentiodynamic polarization scans

Results of CPP scans with CP-Ti and Alloy CoCrMo show that immersion at OC prior to the scan and the vertex potential Ev can affect the responses and key values that are commonly extracted to assess the corrosion susceptibility of the electrode material. The state of the electrode surface can change due to the formation of passive layers or leaching that occurs during immersion prior to and during the forward (anodic) CPP scan. This can affect the current response and the extracted Ezcf and Izcf values commonly used to characterize the alloy corrosion behavior. Similarly, the state of the surface at the vertex potential can affect the response during the reverse (cathodic) scan and the Ezcr value. For example, the reverse scan can be affected by the presence of stable or destabilized surface oxides and the leaching of the underlying metal surface. In a CPP scan to a vertex potential that does not exceed the breakdown potential of the passivating oxide, the surface will remain stable during the reverse scan, and the current will not show a hysteresis loop; this was the case in CPP scans with CP-Ti. In contrast, in a CPP scan to a vertex potential higher than the breakdown potential of the passivating oxide, such as the CPP scans with CoCrMo, the surface will continue to corrode actively during the reverse scan until the oxide film restabilizes at a lower voltage. The presence of damaged passive films and the composition of the underlying alloy surface are expected to affect the characteristics of the reverse scan and Ezcr and Izcr values. The occurrence of a hysteresis loop and values extracted from the reverse scan (including Ep) are characteristics of the surface condition at the vertex potential and the scan rate rather than the alloy itself; neither indicates the propensity of the alloy for pitting corrosion.

The formation of passivating layers and the occurrence of active corrosion prior to and during CPP scans should be taken into account when conducting and assessing CPP scans to provide confidence in the physical significance of extracted values and insights regarding corrosion behavior. In many cases, the responses represent the behavior of a partially stabilized or partially destabilized surface and are not characteristic of the material being analyzed. The use of standardized measurement conditions (such as the vertex potential) may not provide physically meaningful values for some materials or represent the corrosion performance in a particular service environment. Changes in the electrode surface that occur prior to and during the measurement should be taken into account when selecting the measurement conditions and interpreting the response and compositional or environmental dependencies.

Methods

Metallographic characterization

The sample surfaces were polished to a 1-µm surface finish. The electrodes fabricated from as-received CP-Ti and CoCrMo alloy were characterized by using a FESEM (beam voltage 20 keV, working distance between 11–15 mm) with associated EDS. Both backscattered electron and secondary electron imaging of the polished and corroded electrode surfaces were used to highlight the compositional and structural features of the surfaces. The microstructures in several regions were documented before each electrochemical test for later comparisons with those same regions of corroded surfaces to identify active phases and regions. This included compositions and elemental distributions within and between grains.

Electrochemical tests

The test system is illustrated in Fig. 11 and includes a standard three-electrode electrochemical cell system with a graphite counter electrode and a saturated calomel (SCE) reference electrode in which an electrode fabricated using CP-Ti or CoCrMo alloy served as the working electrode. A piece of metal was embedded in epoxy resin and polished to produce an electrode having an exposed area of ~1 cm2. All electrochemical tests were conducted at room temperature in PBS solution having the chemical composition given in Table 5 and a pH of 7.4. The chloride concentration in the PBS is ~5460 ppm, which represents aggressive human body fluids. The working electrodes were electrochemically cleaned by immersing in PBS solution prior to OCP measurements and applying a potential of −0.80 VSCE for ~5 min. A freshly polished and cleaned electrode was used for each series of measurements.

Fig. 11
figure 11

Schematic of the electrochemical test system.

Full size image
Table 5 Chemical composition of the PBS solution.
Full size table

Terminology given in ASTM F212930,31,32 is used for characteristic values that were extracted from the CPP scans. These include the breakdown potential (Eb), the vertex potential (Ev), and the protection potential (Ep). In addition, the zero-current potential measured in the forward scan is referred to as Ezcf and the potential at which zero net current is measured in the reverse scan is referred to as Ezcr, with corresponding estimated corrosion currents Izcf and Izcr. The initial CPP potential is Ei and the final scan potential Ef. These are identified using a generic CPP scan in Fig. 12. These values are characteristics of the CPP scan, but may not represent intrinsic property values of the material being evaluated. In addition, the resting potential (Er) is defined as the stable voltage measured under open circuit conditions (commonly referred to as open circuit potential). The value of Er may differ from the value of Ezcf that is measured during the CPP forward scan.

Fig. 12
figure 12

Generic cyclic potentiodynamic polarization curves identifying corrosion terminology.

Full size image

Electrodes were immersed an open circuit to measure changes in resting potential (Er) in separate tests for durations of 1, 8, 72, and 168 h in PBS solution at room temperature. EIS analyses were performed after each test to characterize the changes in surface properties that occurred during immersion. The EIS analyses were conducted between 105 Hz to 10–3 Hz with a 10 mV perturbation potential. Following the EIS analysis, a CPP scan was conducted at 0.167 mV/s starting from 10 mV below the open circuit potential measured at the end of the immersion duration to a peak vertex potential (Ev) of 1.5 VSCE. All reverse cycles were stopped at 0 VSCE. For clarity, the EIS and CPP scans are identified using the open circuit immersion durations of 1, 8, 72, and 168 h. The actual immersion times include the time required to complete EIS analyses, which are ~30 min.

A second series of CPP scans were conducted with polished electrodes of each material that were not subjected to significant OC immersion durations prior to analyses but were electrochemically cleaned at −0.80 VSCE in PBS solution immediately prior to the CPP scan. This series was performed to assess the effect of the vertex potential on the response during the reverse scan. Scans with CP-Ti were to vertex potentials of 0.5, 1.0, and 1.5 VSCE, and scans with Alloy CoCrMo were to vertex potentials of 1.0, 1.2, and 1.6 VSCE. Finally, cathodic polarization scans with freshly polished CP-Ti and CoCrMo alloy electrodes were performed starting at a vertex potential of 1.6 VSCE to assess the effects of changes in the surfaces occurring during the anodic scans of the CPP measurements on characteristics of the cathodic return scan.