2016 Volume 57 Issue 12 Pages 2033-2040
Air-formed surface oxide films on four types of Co-Cr- alloys were characterized using X-ray photoelectron spectroscopy (XPS) and five types of Co-Cr alloys were anodically polarized, to identify the effects of the addition of N, Mo, and W to Co-Cr alloys containing high Cr on the surface composition and corrosion resistance. Co-20Cr-15W-10Ni (ASTM F90), Co-30Cr-6Mo, Co-33Cr-5Mo-0.3N, and Co-33Cr-9W-0.3N were employed for XPS and the above four alloys and another Co-30Cr-6Mo (ASTM F75) were employed for anodic polarization. The surface oxide film on the Co-Cr alloys consisted of oxide species of Co, Cr, Mo, W and/or Ni contains a large amount of OH− with a thickness of 2.6–3.2 nm. Cations existed in the oxide as Co2+, Cr3+, Mo4+, Mo5+, Mo6+, W6+ and Ni2+. Cr and Mo are enriched and Co and Ni are depleted in the surface oxide film. W was enriched in the case of Co-20Cr-15W-10Ni but depleted in the case of Co-33Cr-9W-0.3N. On the other hand, Cr, Mo, W and Ni were enriched and Co was depleted in the substrate alloy just under the surface oxide film in the polished alloy. During rapid formation of the surface oxide film, Cr was preferentially oxidized and the oxidation of Co and Ni delayed, according to the oxidation and reduction potentials of these elements. The Co-Cr alloys essentially have high localized corrosion resistance that is not easily affected by a small change of composition. Co-33Cr-5Mo-0.3N shows higher corrosion resistance compare than conventional Co-Cr alloys.
Co-Cr alloys show excellent mechanical properties such as strength and toughness, castability, corrosion resistance, and wear resistance. Their corrosion resistance is better than that of stainless steel, and their wear resistance is better than those of stainless steel and Ti alloys. Because of the above properties, Co-Cr alloys are extensively used for medical devices despite their plasticity and workability are lower than those of stainless steel and Ti alloys. In dentistry and orthopedics, Co-Cr-Mo alloys are used for dental denture base and the head and stem of artificial joints.
However, Co-Cr alloys show low ductility at room temperature owing to the formation of carbides, σ phase, and/or martensitically transformed ε phase having hexagonal-close packed (hcp) structure. In order to improve the ductility, the addition of N with high Cr content was found to be effective for Co-Cr-Mo alloys. The solubility of N in Co-Cr-Mo alloys increases with increasing Cr content from 29 to 34 mass%.1) N contributes to the suppression of the martensite and σ phase formation. The high Cr and N content in Co-Cr-Mo alloy results in a significant improvement in mechanical properties such as yield strength, tensile strength, and elongation, and the improvement cannot be achieved only with high Cr in the alloy. The yield strength of N-containing 33Cr satisfied the type 5 criteria in ISO 22674 and the elongation was even higher that in the type 5 criteria.2) In addition, metal elution from N-containing 33Cr was lower than that of Co-29Cr-6Mo alloy.3) Thus, the N-containing 33Cr alloys are promising for use in dentures with adjustable clasps through one piece casting.
On the other hand, the corrosion resistance and tissue compatibility of Co-Cr alloys is maintained by the surface oxide film as a passive layer and their-resistance in Hanks' solution and cell culture medium is sufficiently high compared to other metals.3) The surface oxide film on Co-Cr alloys has been characterized in previous studies. In Co-Cr-Mo alloys, the oxide film consists of oxides of Co and Cr without Mo,4) in contrast, when the alloy polished mechanically in de-ionized water, the oxide film consists of oxide species of Co, Cr and Mo with a thickness of approximately 2.5 nm.5) This surface film contains a large amount of OH−, i.e., the hydrated or oxyhydroxidized oxide and can also traced Cr0 and Mo0 distributed at the inner layer of the film. The corrosion was found to proceed by selective dissolution of the Co in neutral or acidic solutions.6) During immersion of Co-Cr-Mo alloys in Hanks' solution, culture medium and cell culture, Co dissolves from the film, and the film composition changes to mostly comprise chromium oxide with a small amount of molybdenum oxide with formation of calcium phosphate on or in the film.5) After the alloy is anodically polarized in Hanks' solution in a passive range, the surface oxide film consists predominantly of Cr2O3 and Cr(OH)3 with a thickness of 3.1 nm.7) In another study, a Cr-enriched and Co-depleted passive layer formed on the alloy.8)
In this study, air-formed surface oxide films on four types of Co-Cr alloys containing large amounts of Cr and N were characterized using X-ray photoelectron spectroscopy (XPS) and five types of Co-Cr alloys were anodically polarized, to identify the effects of the addition of N, Mo, and W to Co-Cr alloys containing high Cr on the surface composition and corrosion resistance. The information will enhance the understanding of the corrosion resistance and tissue compatibility of the alloy.
Four types of Co-Cr alloys containing high Cr were employed for XPS in this study: Co-20Cr-15W-10Ni (ASTM F90, L605), Co-30Cr-6Mo_1 (Cobaltan®, Shofu, Kyoto, Japan), Co-33Cr-5Mo-0.3N (Cobarion®, Mo), and Co-33Cr-9W-0.3N (Cobarion®, W). On the other hand, Co-30Cr-6Mo_2 (ASTM F75, Vitallium), in the addition to the above four alloys, was employed for anodic polarization. The characterization of Co-30Cr-6Mo_2 using XPS was not performed in this study, because the composition of Co-30Cr-6Mo_2 was almost the same as that of Co-30Cr-6Mo_1 and F799-95 whose composition is similar to that of F75 has been already performed elsewhere.5) Co-20Cr-10W-5Ni, Co-30Cr-6Mo_1 and Co-30Cr-6Mo_2 alloys were purchased. Co-33Cr-5Mo-0.3N and Co-33Cr-9W-0.3N alloys were melted (Eiwa Co. Ltd., Kamaishi, Iwate, Japan). Compositions of the Co-Cr alloys are summarized in Table 1.
| Alloy | Cr | Mo | W | Ni | Fe | Si | Mn | N | C | Co |
|---|---|---|---|---|---|---|---|---|---|---|
| Co-20Cr-15W-10Ni | 20.18 | — | 15.11 | 9,93 | 2.01 | 0.01 | 1.51 | 0.03 | 0.07 | Bal. |
| Co-30Cr-6Mo_1 | 29.83 | 6.04 | — | 0.02 | 0.02 | 0.38 | <0.01 | <0.25 | <0.35 | Bal. |
| Co-33Cr-5Mo-0.3N | 32.3 | 4.89 | — | 0.01 | — | 0.60 | 0.59 | 0.39 | 0.074 | Bal. |
| Co-33Cr-9W-0.3N | 32.6 | — | 9.02 | 0.01 | — | 0.57 | 0.58 | 0.38 | 0.055 | Bal. |
| Co-30Cr-6Mo_2 | 29.86 | 5.83 | 0.11 | 0.03 | 0.03 | 0.63 | <0.02 | <0.25 | <0.35 | Bal. |
To conduct XPS and anodic polarization, respectively, under the same condition, each alloy ingot was cut into small pieces, re-melted and cast in a sand mold consisting mainly of silica using a centrifugal casting machine to obtain the same size of specimens. The mold was kept at room temperature before casting. The mold was cooled down in air at room temperature after casting. The alloy was supplied as a rod with a diameter of 11 mm. Disks that were 2 mm thick were cut from the rod. The disks were metallographically polished and finished by polishing with #1000 SiC polishing paper, followed by ultrasonic rinsing in acetone and ethanol for 900 s each and drying in a stream of nitrogen (>99.9%). The alloy was stored in a desiccator for 7 d to stabilize the surface oxide film until analysis by XPS. On the other hand, the disk specimens were also used for anodic polarization. The Co-Cr alloy disks were metallographically polished and finished by polishing with #800 SiC polishing paper, followed by ultrasonic rinsing in acetone and ethanol for 900 s each and drying in a stream of nitrogen (>99.9%). The anodic polarization measurement had been started immediately after the specimen preparation.
2.2 XPSXPS was performed with an electron spectrometer (JPS-9010MC, JEOL, Tokyo Japan). All binding energies given in this paper are relative to the Fermi level, and all spectra were excited with the Al Kα line (1486.6 eV). The spectrometer was calibrated against Au 4f7/2 (binding energy, 84.07 eV) and Au 4f5/2 (87.74 eV) of pure gold and the Cu 2p3/2 (932.53 eV), Cu 2p1/2 (952.35 eV), and Cu Auger L3M4,5M4,5 line (kinetic energy, 918.65 eV) of pure copper. The energy values were based on published data.9) The binding energies were calibrated using the energy of the C 1s peak of contaminant carbon, i.e., 285.0 eV. The take-off angle of the photoelectron was 90° to the surface of alloy. To estimate the photoelectron peak intensities, the background was subtracted from the measured spectrum according to Shirley's method.10)
The composition and thickness of the surface oxide and the composition of the substrate were simultaneously calculated according to a published method.11,12) The quantitative determination was based on the assumption of a three-layer model consisting of an outermost contaminant hydrocarbon layer with uniform thickness, followed by a surface oxide film with uniform thickness and the underlying alloy surface with spectroscopically infinite thickness on XPS.
Empirical data13–16) and theoretically calculated data17) on the relative photoionization cross sections were used for the quantification. The relative photoionization cross sections used in this study are summarized in Table 2, where σij/σO1s represents the relative photoionization cross section of a level j electron of an element i to that of O 1s electrons. The reproducibility of the results was confirmed several times under the same conditions. The relative concentrations of elements in the surface oxide layer were calculated, with the exception of carbon as a contaminant.
| Photoionization cross-section | ||||
|---|---|---|---|---|
| Level | Co 2p3/2 (Oxide) |
Ni 2p3/2 (Metal) |
Ni 2p3/2 (Oxide) |
Cr 2p3/2 (Oxide) |
| σij | 3.9 | 2.3 | 7.47 | 1.7 |
| Reference | 13 | 14 | 15 | 16 |
| Level | Mo 3d5/2 | W 4f7/2 | ||
| σij | 3.5 | 5.8 | ||
| Reference | 16 | 17 | ||
The anodic polarization was performed with a potentiostat (HABF-501A, Hokuto Denko, Tokyo, Japan) with a function generator (HB-111, Hokuto Denko, Tokyo, Japan). The saturated calomel electrode (SCE) and Pt electrode were used as reference and counter electrodes, respectively. The specimen was fixed in a polytetrafluoroethylene holder with an o-ring. The exposed area contacting the electrolyte was 0.278 or 0.353 cm2. After immersing the specimens in 0.9%NaCl solution at 310 K, the open circuit potential (OCP) was measured for 10 min. Thereafter, the linearly-increasing anodic potential was applied at a constant sweep rate of 1 mV s−1 from the initial potential which was −50 mV lower than the final value of OCP. The measurement was stopped when the applied potential reached to the final value of 1.5 VSCE. The reproducibility of the results was confirmed by conducting polarizations three times under the same conditions.
Survey XPS spectra form four alloys are shown in Fig. 1. Carbon was detected on all alloys. It was concluded that none of the alloy contained carbonate because no peak was detected at the characteristic C 1s peak of carbonate in the energy region of 289–290 eV.18) Therefore, carbon detected from the alloy can be ascribed to the so-called contaminant carbon. Apart from carbon, Co, Cr, Mo, W, Ni, and O were detected on the polished Co-Cr alloys using XPS, according to their nominal compositions. The other elements contained in the alloys as impurities were not detected in the alloy. N was detected under the contaminant level, so N was not used for quantification. The XPS spectra of the binding energy regions of Co 2p3/2, Cr 2p, Mo 3d, W 4f, and Ni 2p3/2 electrons obtained from the alloy are shown in Figs. 2–5. After subtracting the background by Shirley's method,10) each spectrum was decomposed into spectra according to the binding energy data.19,20) Co 2p, Cr 2p, Mo 3d, W 4f, and Ni 2p peaks showed both oxide and metal states in all alloys, respectively. The Co 2p peaks were decomposed into peaks originating from Co0 (metal state), Co2+ and satellite (Fig. 2). The Cr 2p peaks were originated from Cr0 (metal state) and Cr3+ (Fig. 3). The Mo 3d peaks were decomposed into peaks originating from Mo0 (metal state), Mo4+, Mo5+, and Mo6+ (Fig. 4). The W 4f peaks were decomposed into W0 (metal state) and W6+ (Fig. 5). The Ni 2p peak was decomposed into peaks originating from Ni0 (metal state), Ni2+ and satellite (Fig. 6). The surface oxide film of the alloys was very thin because the metallic-state elements in the substrate were detected through the film. According to the binding energy, Ni2+ exists as Ni(OH)2, but the chemical species of Co2+, Cr3+, Mo4+, Mo5+, Mo6+, and W6+ are unknown because no binding energy data exist to determine the species. In other words, we could not identify the following states: CoO or Co(OH)2, Cr2O3 or Cr(OH)3, MoO2 or Mo(OH)4, Mo2O5 or Mo(OH)5, MoO3 or Mo(OH)6, and WO3 or W(OH)6, respectively.
Survey XPS spectra of the Co-Cr Alloys.
The XPS spectra of the binding energy regions of Co 2p3/2 electrons obtained from the Co-Cr alloys.
The XPS spectra of the binding energy regions of Cr 2p electrons obtained from the Co-Cr alloys.
The XPS spectra of the binding energy regions of Mo 3d electrons obtained from the Co-Cr alloys.
The XPS spectra of the binding energy regions of W 4f electrons obtained from the Co-Cr alloys.
The XPS spectra of the binding energy regions of Ni 2p3/2 electrons obtained from the Co-Cr alloys.
The XPS spectrum of the O 1s electron binding energy region consists of at least three peaks originating from O2−, i.e., hydroxide or hydroxyl groups, OH−, and hydrate and/or adsorbed water,15) as shown in Fig. 7. The proportions of oxygen species and the [OH−]/[O2−] ratio are 2.6–3.5. The surface oxide after polishing contained a large amount of OH−, and the [OH−]/[O2−] ratio is almost the same as previously determined for a Co-Cr-Mo alloy (2.9).5)
The XPS spectra of the binding energy regions of O 1s electrons obtained from the Co-Cr alloys.
Table 3 reveals the compositions of the surface oxide film on the alloys. The compositions shown in Table 3 were calculated assuming that the detected carbon originated from the so-called contaminant carbon. The surface oxide film consisted of Co2+, Cr3+, Mo4+, Mo5+, Mo6+, W6+, and Ni2+, as well as oxygen species such as O2−, OH−, and H2O. Co and Cr were contained in almost the same concentration like a previous study.5) Assuming that the surface oxide film consists of CoO, Cr2O3, MoO3, WO3, and Ni(OH)2, the oxygen concentration was calculated as 43.7, 42.7, 45.3, and 45.6 mol% in Co-20Cr-15W-10Ni, Co-30Cr-6Mo_1, Co-33Cr-5Mo-0.3N, and Co-33Cr-9W-0.3N, respectively, which is smaller than the oxygen concentration (67.9, 69.0, 66.7, and 67.3 mol%, respectively). The extra oxygen originated from OH− and H2O. Co(OH)2 and Cr(OH)3 possibly exist in the oxide film, but molybdenum hydroxide is hardly formed because more Mo exists in the oxide.5) Regardless of the composition, the oxide film is not simply an anhydrous oxidic substance: it is an oxyhydroxide.11,14)
| Alloy | Relative concentrations (mol%) | Thickness of the film (nm) | |||||
|---|---|---|---|---|---|---|---|
| Co | Cr | Mo | W | Ni | O | ||
| Co-20Cr-15W-10Ni | 14.4 | 15.0 | — | 1.4 | 1.3 | 67.9 | 3.2 |
| Co-30Cr-6Mo_1 | 12.4 | 17.0 | 1.6 | — | — | 69.0 | 3.0 |
| Co-33Cr-5Mo-0.3N | 13.5 | 18.4 | 1.4 | — | — | 66.7 | 2.6 |
| Co-33Cr-9W-0.3N | 17.7 | 17.4 | — | 0.6 | — | 67.3 | 2.8 |
Cation fractions in the surface oxide film and the nominal composition of the alloy in atomic percentages are summarized in Table 4. From this estimation, the atomic bulk composition was converted from the values of Table 1. Comparing both fractions, it is clear that Cr and Mo are enriched and Co and Ni are depleted in the surface oxide film. W was enriched in the case of Co-20Cr-15W-10Ni but depleted in the case of Co-33Cr-9W-0.3N. Enrichment of Cr in the surface oxide is explained by its lower potential compared with Co and Ni. In addition, Co and Ni preferentially were dissolved during polishing process. Mo may be remained in the surface oxide film and eventually resultant cation farction increased. The preferential oxidation and dissolution of an element is governed by the relative relation among component elements in an alloy. Therefore, the difference in behavior of W between Co-20Cr-15W-10Ni and Co-33Cr-9W-0.3N may be caused by the existence of Ni.
| Cation fraction (mol%) | Co | Cr | Mo | W | Ni |
| Co-20Cr-15W-10Ni | 44.9 | 46.9 | — | 4.2 | 4.0 |
| Co-30Cr-6Mo_1 | 40.0 | 54.9 | 5.1 | — | — |
| Co-33Cr-5Mo-0.3N | 40.6 | 55.1 | 4.3 | — | — |
| Co-33Cr-9W-0.3N | 45.0 | 53.1 | — | 1.9 | — |
| Nominal (mol%) | Co | Cr | Mo | W | Ni |
| Co-20Cr-15W-10Ni | 67.8 | 23.6 | — | 3.3 | 5.2 |
| Co-30Cr-6Mo_1 | 64.0 | 32.4 | 3.6 | — | — |
| Co-33Cr-5Mo-0.3N | 60.5 | 36.5 | 3.0 | — | — |
| Co-33Cr-9W-0.3N | 59.0 | 38.1 | — | 2.9 | — |
The compositions of the substrate alloys estimated from the XPS data and the nominal composition are summarized in Table 5. The substrate metal elements were detected through the surface oxide film by XPS. Therefore, the detected depth was small, and only the substrate just under the surface oxide film was detected. Cr, Mo, W and Ni were enriched and Co was depleted in the substrate alloy just under the surface oxide film in the polished alloy. Cr and Mo were also enriched as the same in the surface oxide. In general, enriched elements usually depleted in the substrate alloy just under the surface oxide after stabilizing the oxidation. However, this is not appeared in this study, probably because the polished surface is still unstable and the oxidation and dissolution were not completed. On the contrary, Ni concentrated layer was formed just under the surface oxide film because the oxidation of Ni was delayed and other elements were preferentially oxidized. This phenomenon is usually found in Ti-Ni alloys.21)
| Cation fraction (mol%) | Co | Cr | Mo | W | Ni |
| Co-20Cr-15W-10Ni | 49.2 | 31.1 | — | 4.6 | 15.1 |
| Co-30Cr-6Mo_1 | 56.5 | 39.4 | 4.1 | — | — |
| Co-33Cr-5Mo-0.3N | 55.0 | 41.3 | 3.7 | — | — |
| Co-33Cr-9W-0.3N | 41.0 | 43.7 | — | 5.3 | — |
| Nominal (mol%) | Co | Cr | Mo | W | Ni |
| Co-20Cr-15W-10Ni | 67.8 | 23.6 | — | 3.3 | 5.2 |
| Co-30Cr-6Mo_1 | 64.0 | 32.4 | 3.6 | — | — |
| Co-33Cr-5Mo-0.3N | 60.5 | 36.5 | 3.0 | — | — |
| Co-33Cr-9W-0.3N | 59.0 | 38.1 | — | 2.9 | — |
The surface oxide films on the Co-Cr alloys are schematically illustrated in Fig. 8. Cr is enriched and Co, W, and Ni are depleted in the oxide film and Ni, Cr, Mo, W are concentrated just under the oxide film, especially in Ni. During rapid formation of the surface oxide film, Cr was preferentially oxidized and the oxidation of Co, Mo, W and Ni delayed, according to the oxidation and reduction potentials of these elements. Therefore, the above enrichment and depletion are observed. In particular, the enrichment of Ni in the substrate alloy just under the surface oxide is also observed in Ti-Ni alloy.21) Cr exists as Cr3+ in the oxide and the oxide contains a large amount of OH−. During polishing, preferentially Co and Ni are dissolved, and the surface oxide changes to largely comprise chromium oxide (Cr3+) with small amounts of cobalt, molybdenum, and tungsten oxides. Simultaneously, Cr and Ni increased and Co decreased in the substrate alloy just under the surface oxide. Therefore, Co decreased both in the surface oxide and in the substrate alloy just under the surface oxide. In the previous study, a large amount of OH− was detected.5,22) This property may induce a high reactivity with molecules containing positive charge. More poly(ethylene glycol) terminated with amines are immobilized on a F75 Co-Cr alloy than 316L type stainless steel and grade 2 Ti, because of large OH− concentration on the surface.23)
Schematic illustration of the surface of Co-Cr alloys.
In the previous study, surface oxide of Co-29Cr-6Mo (F799-95) was about 2.5-nm thick and contained a large amount of OH−. Co is depleted and Cr is enriched in the surface oxide film and Co is enriched and Cr depleted in the substrate alloy just under the surface oxide. This also supports the above results in this research.
3.5 Anodic polarizationAnodic polarization curves of the alloys are shown in Fig. 9. In all alloys, no pitting was observed. Increase of current density over 0.6 VSCE due to transpassive dissolution and oxygen evolution was observed. The shapes the curves are almost identical. However, current densities of Co-33Cr-5Mo-0.3N and Co-33Cr-9W-0.3N near corrosion potential were smaller than those of Co-20Cr-10W-5Ni and Co-29Cr-6Mo. Current density near corrosion potential was decreased with Cr content. On the other hand, the passive currents of Co-33Cr-5Mo-0.3N and Co-33Cr-9W-0.3N were smaller than those of Co-20Cr-15W-10Ni and Co-30Cr-6Mo_1. Passive current densities of Co-Cr alloys at 0.4 VSCE are shown in Fig. 10. The value was slightly smaller in Co-33Cr-5Mo-0.3N, indicating that the passive film is more protective in the alloy. However, the cause of the decrease in passive current is not clear. In a previous study that Co-Cr-Mo-N alloys are immersed in lactic acid, N does not influence the corrosion resistance.3) In any case, according to the above results, Co-Cr alloys essentially have high localized corrosion resistance that is not easily affected by a small change of composition. This is also supported by the XPS results. Co-33Cr-5Mo-0.3N shows higher corrosion resistance compare than conventional Co-Cr alloys. In the high Cr alloy, the contents of Mo and N may increase the corrosion resistance.
Anodic polarization curves of Co-Cr alloys.
Passive current densities of Co-Cr alloys at 0.4 VSCE.
Second phases such as σ and ε in Co-Cr alloys influence mechanical property. On the other hand, the surface oxide films or passive films on Co-Cr alloys are strongly protective. Therefore, the second phases hardly influence composition of surface oxide and corrosion resistance in biological environment. Also in this study, the corrosion resistance was the almost same among the Co-Cr alloys employed in this study.
The surface oxide film on a mechanically polished Co-Cr alloys consist of oxide species of Co, Cr, Mo, W and Ni and contains a large amount of OH− with a thickness of 2.6–3.2 nm. More OH− exists the surface most layer on the surface oxide film. Cations exist in the oxide as Co2+, Cr3+, Mo4+, Mo5+, Mo6+, W6+ and Ni2+. Cr and Mo are enriched and Co and Ni are depleted in the surface oxide film. W was enriched in the case of Co-20Cr-15W-10Ni but depleted in the case of Co-33Cr-9W-0.3N. On the other hand, Cr, Mo, W and Ni were enriched and Co was depleted in the substrate alloy just under the surface oxide film in the polished alloy. Cr was preferentially oxidized and the oxidation of Co and Ni delayed, according to the oxidation and reduction potentials of these elements. In all alloys, no pitting was observed. Co-Cr alloys essentially have high local corrosion resistance that is not easily affected by a small change of composition.
This work was supported by a Grant-in-Aid for revitalization promotion program (A-STEP) from Japan Science and Technology Agency of Japan.
