Long-Term Corrosion Investigation of CoCrMoW Alloys under Simulated Physiological Conditions

1. Introduction

CoCr along with titanium alloys are the most frequently used materials in surgical implant applications, chiefly for hard tissue replacement [1,2]. The use of the CoCr based alloys as biomaterial, attracts a lot of attention because of their excellent properties such as high corrosion resistance, good biocompatibility and good mechanical properties [3, 4]. Among the properties required for using these materials for surgical applications, a very good corrosion resistance of the alloy is a mandatory one.

The corrosion resistance of these biomaterials is influenced by several factors like, the alloy itself (e.g. chemical composition, microstructure, surface state), the characteristics of the environment (pH, temperature), as well as the way of manufacturing these biomaterials [5,6]. For instance, by casting method, alloys with very good corrosion resistance were obtained compared to those ones where a manufacturing method was employed [7,8]. As regards the environment, for example, it was reported that the corrosion resistance of Co28Cr6Mo alloys is drastically reduced when these alloys are tested in NaCl compared to PBS [9] his behavior most probably is related to the fact that in such aggressive media the ion chlorides are attacking the protective film formed on the surface, leading to the pitting corrosion processes and finally to the film dissolution. Conversely, in PBS media, this type of alloy is less exposed to corrosion attack because during the immersion period, the phosphate ions adsorbed on the surface act as a efficient barrier between the alloy and environment [9,10].

As many studies revealed, the corrosion behavior of these types of alloys is drastically influenced by the chemical composition of the alloy as well. In this respect, a lot of research work is devoted [11,12]. For example, at Co29Cr alloy, a good resistance to corrosion attack was observed in SBF simulated biofluid [13] because a film, containing significant amount of Cr₂O₃ and small amounts of cobalt oxides, is formed [14]. It is well accepted that mainly the presence of Cr₂O₃ and/ or Co (OH)₃ species in the film plays a key role in providing good corrosion protection of these CoCr-based alloys. However, at long-time immersion in similar media (e.g. artificial saliva, Ringer solution and Hank solution), dissolution processes at interface alloy/biofluid are noticed, yielding to a notable instability of the film [12,14]. In order to answer to these issues, many approaches were tackled over the time and, recently, data attested that the addition of a small amount of molybdenum in the composition of the CoCr alloy is about to obtain films with conspicuous protective properties. As the electrochemical corrosion tests evidenced, the Mo itself stabilizes the passive film. These reported data opened new perspectives in finding other pathways for obtaining alloys with very good corrosion resistance. Other attempts were reported over the time, as manufacturing CoCr-based alloys with small amounts of W. For these CoCrW alloys, it was observed, that the addition W does not bring about a better stability of the film and rather a superior mechanical property of the alloy which does not necessarily improve corrosion resistance of the material [15].

As result, in the last few years, molybdenum (Mo) and tungsten (W) were added as alloying elements in CoCr-based alloys and promising outcomes were obtained. At these new alloys, the improved corrosion resistance emerges actually from the synergetic effect of Mo, W and Cr which promotes the formation of a compact passive film with good stability [11,16,17] which indeed acts like a barrier between the alloy surface and the biofluid, impeding hence the dissolution of the metallic elements [18]. In other words, Mo and W provide benefits towards passivation and protective role in the breakdown processes, yielding resistance to corrosion attack and hence to the dissolution of the film [17].

Concerning other aspects of using these CoCr based alloys for medical applications, their costs should be envisaged, and especially those ones related to chromium. In this respect, new gates were opened by means of manufacturing CoCr based alloys with lower content of Cr. However, by reducing the amount of Cr, the corrosion resistance of material is most probably diminished with undesirable consequences in their use for medical applications. Therefore, to reduce the content of Cr and at the same time to have CoCr-based alloys with good corrosion resistance, the addition of Mo and Was alloying elements is a valuable option. No literature data about the corrosion behavior of these new types of alloys, i.e. Co Cr MoW in biofluids were reported. Most frequently studies are devoted for corrosion behavior of CoCr, CoCrMo or CoCrMoW alloys with high amount of chromium (e.g 27-30 % Chromium) [11,19,20].

Herewith, we studied the corrosion behavior of two types of alloys with different concentrations in Cr and with different alloying elements as W and Mo. These studies are actually aimed at opening new perspectives in finding strategies for obtaining alloys, with low content of Cr and with excellent corrosion behavior. The evidence of the role played by alloying elements in the corrosion behavior of these CoCr-based alloys is another goal of this research work. Thus, the corrosion behavior of two types of alloys after long term immersion period, i.e. 1000 hours, were studied in PBS solution by electrochemical methods (i.e. open circuit potential (OCP), potentiodynamic polarization (PP) and electrochemical impedance spectroscopy (EIS)) The first alloy contains a high amount of Cr (29 % Cr) and as alloying element W, labeled Co29Cr7W, whereas the second one contains a lower amount of Cr (21 %) and as alloying elements Mo and W ones, labeled Co21Cr8Mo7W. In order to better understand the corrosion behavior of these alloys, additional information will be gained mainly from X-ray photoelectron spectroscopy (XPS) and inductively coupled plasma mass spectroscopy (ICP) techniques. In our opinion, by studying the corrosion behavior of these alloys after long term immersion instead of short ones, very often reported in literature, added value is brought to the study because more accurate information about their corrosion behavior is gained.

2. Materials and Methods

Co21Cr8Mo7W and Co29Cr7W alloys were tested in the form of rods with 8 mm in diameter. The chemical composition of these alloys, in cast state, is presented in Table 1.

Before experiments, the samples were mechanically prepared by grinding with abrasive paper and suspensions of micrometric alumina powder, up to the metallographic quality, then degreased in ethanol in an ultrasonic bath and rinsed with distilled water.

The experiments were performed in a 0.1 M phosphate buffer saline solution (PBS) with the following composition: 8 gL^-1 NaCl, 0.2 gL^-1 KCl, 1.44 gL^-1Na₂HPO₄ and 0.25 gL^-1 KH₂PO₄. The solution was maintained at a pH= 7.4 and a temperature of 37ºC. All reagents were analytical grade (Sigma-Aldrich), and the solution was prepared using bidistilled water.

2.1. Electrochemical Measurements

The open-circuit potential was monitored during 1000 hours of immersion, at 37°C in natural aerated solutions, and the potential values are related to Ag/AgCl reference electrode. Three identical samples of 1 cm² were immersed in a 50 mL PBS solution in a Thermocal 20 incubator. At predetermined time intervals the stationary potentials were measured in open circuit (OCP). After different periods of immersion (168 and 1000 hours) the Tafel curves were recorded, in a potential range of 250 mV vs. Ag/AgCl around the corrosion potential, with a rate of 2.5 mV s^-1 to evaluate the corrosion rate of specimens. At the same time the EIS data acquired potentiostatic at rest potential, between 1MHz and 10^-2 Hz, with a perturbation of 10 mV. At the end of the immersion period, the anodic polarization curves were acquired between -300 mV and + 1000 mV with a potential sweep rate of 2.5 mVs^-1. All electrochemical measurements were performed using a PARSTAT 4000 potentiostat/galvanostat in a three-electrode electrochemical cell with a platinum sheet as counter electrode, Ag/AgCl as a reference electrode and the samples as working electrode. The acquired data were processed using the CorrView 3.3d and the Zview 2.70 dedicated software. To certify the reproducibility, three identical samples of each alloy (Co21Cr8Mo7W and Co29Cr7W) were tested.

2.2. Surface Characterization and Electrolyte Analysis

The surface chemical composition of these alloys was investigated by X-ray Photoelectron Spectroscopy (XPS) using an Escalab Xi+ system from Thermo Scientific provided with Al Ka gun. Acquisition steps of 1 eV and 0.1 eV were used for general spectra and high-resolution spectra, respectively. The deconvolution of high-resolution spectra was performed by Thermo Scientific Avantage Software. The C1s line at 284.4 eV was used as a reference to correct the binding energies for charge energy shift. A Shirley background was subtracted from the spectra. Least-square curve-fitting of the spectra was performed based on a mixture of Gaussian –Lorentzian functions.

The changes in hydrophobicity of the surface because of corrosion products depositions were highlighted by wetting analysis. The contact angle measurement was done with Drop Shape Analysis System apparatus, 133 DSA1 model (FM40 Easy Drop) from KRÜSS GmbH Germany. The samples were placed on a plane support and drops of deionized water with a volume of 3 µL each were put on the surface with a dispensing micro-syringe. The contact angle values were collected in static regime at room temperature, initially and at 30 seconds after the drop of water was placed on the sample surface. Static contact angle was recorded using Sessile Drop Fitting method for the angles in between 30–90^o and Circle Fitting method for those less than 30^o. Over 7 measurements were averaged for each sample.

The chemical analysis of electrolyte after 168 hours of immersion was performed using an inductively coupled plasma optical emission spectrometer with axial and radial viewing plasma configuration (ICP-OES, Optima 2100 DV Perkin Elmer), operating at a 40 MHz free-running ratio-frequency. For each extraction environment, two blank samples are measured as a reference.

3. Results and Discussions

3.1. Electrochemical Evaluation of CoCrMoW Alloy

The electrochemical characterization of Co21Cr8Mo7W and Co29Cr8Mo7W samples was carried out in order to analyze the corrosion behavior of these alloys in PBS.

3.1.1. OCP evolution

Figure 1 presents the overtime monitoring of the OCP for both alloys immersed in PBS solution during 1000 hours at 37°C. The evolution of the OCP with time is almost the same for both samples. During the first 24 hours, the initial potentials increase significantly with an approximative rate of 7.7 mV h^-1(up to +50 mV for Co21Cr8Mo7W and -50 mV for Co29Cr7W), indicating the spontaneous formation of a protective layer on surface materials in the biofluid. At both samples, between 24 and 230 hours of immersion, the OCP values are changing with time, suggesting that some processes of adsorption or/and dissolution occur at the interface, leading inevitably to both surface composition changes and apparently a slight instability of the protective layer. According to literature data [21, 22] among the possible processes of adsorption/dissolution that could occur at the interface, the adsorption of the phosphate ions on the alloy surface (mainly as chromium and cobalt phosphate complexes (e.g. Co(H₂PO₄)₂) followed by their partial dissolution might not be disregarded. As other reports suggest [9,21,23], by the formation of these chromium and cobalt phosphate species on the surface, the corrosion attack might be mitigated. Indeed, in our case this phenomenon is observed. Thus, as Figure 1 illustrates, it was noticed that the OCP values remain approximately constant between 230 and 1000 hours, revealing that during this period of immersion, the layer formed on both samples becomes a stable one with protective properties against the corrosion processes. However, the OCP values of Co21Cr8Mo7W alloy are situated at more positive potentials compared to those ones of Co29Cr7W alloy, i.e. a shift of 100 mV, regardless the immersion period (Figure 1). One may infer that the protective layer formed on the surface of Co21Cr8Mo7W hampers better the further dissolution of the alloys and increases the corrosion resistance of the material [13].

3.1.2. Anodic Polarization

In order to find out more information about corrosion behavior of both samples, after 1000 hours of immersion in PBS, the Tafel plots and potentiodynamic polarization curves were recorded and they are exhibited in Figure 2. From polarization curves, at both samples, a current density (j) less than 10 µA cm^-2 was observed in a potential range of about 630 mV (Figure 2). These results attest that both samples have a comparable passive domain in which the film formed on surface protects the surface against corrosion attack. These findings appear to attest that both samples can be used successfully for biomaterial applications, because the potential domain, in which the dissolution of oxide layer is obstructed, is actually large. For other CoCrMo alloys, a similar domain passivation was evidenced [9,10,21]. Literature reports attested that the resistance to passive dissolution of CoCr based alloys is mainly due to the formation of chromium oxide and cobalt and chromium phosphate complexes [10,20,22] on the surface. Given that the chemical composition of our samples is not too different from that of these CoCrMo alloys already studied, one may presume that a similar phenomenon takes place in our case. After 0.6 V vs. Ag/AgCl, a transpassive domain is observed, regardless of the type of specimen, revealing that at surface of both samples, a dissolution of the film starts to occur. One may presume that the attack of the surface after 0.6 V, in this neutral environment, is most probably mainly due to both the dissolution of chromium oxide and phosphate chromium ion complexes formed on the surface and water oxidation [10]. The existence of a transpassive domain after 0.6 V was also reported for other cast cobalt chromium alloys [9,24]. However, as Figure 2 illustrates, in the transpassive domain, a slight increase of the current density (j) is noticed for Co21Cr8Mo7W samples, i.e. from 10 µA cm^-2 to 43 µA cm^-2, whereas, for Co29Cr7W samples, a sudden increase of the current density, i.e. from 10 µA cm^-2 to 550 µA cm^-2 is observed. This behavior points out clearly that the layer formed on Co21Cr8Mo7W surface has better protective properties against the corrosion attack and is more stable. At first sight, this demeanor appears to be surprising, because the Co29Cr7W alloy, which has a higher amount of chromium compared to Co21Cr8Mo7W (Table 1) has a slightly lower corrosion resistance. In most cases, for different cast CoCrMo alloys [11,25], the presence of a high amount of chromium, mainly as Cr₂O₃, provides inhibiting properties of the surface to corrosion processes. Based on these observations, we assume that, in our case, not necessary the presence of a higher amount of Cr in the alloy brings about a better stability of the film and rather the synergetic effect of Cr, Mo and W present in the alloy composition prevents the corrosion attack most probably due to the formation of a very stable and compact layer on their surface. Since at other CoCrMo alloys, it was observed that the presence of Mo facilitates the formation of a compact and stable chromium oxide film, our findings appear not to be quite unexpected [16,17].

In order to have a better insight about the corrosion behavior of these alloys in perspective of using them in medical fields, the ion release investigations were performed at both samples after 168 hours of immersion according to ISO 10271:2001 [26]. The results, shown in the Table 3, reveals a significant quantity of chromium ion released at Co29Cr7W, i.e. 73.7 µg cm^-2 compared to Co21Cr8Mo7W where a negligible amount of chromium ions is released, i.e. 0.15 µg cm^-2.

These results are of great importance in terms of medical applications of these types of alloys, because they clearly attest that only the Co21Cr8Mo7W is suited for surgical implants, (i.e 0.15 µg cm^-2 concentration of chromium ion released) because the concentration of chromium ion released is much lower than that worldwide accepted. (i.e. maximum 0.5 µg cm^-2 [27]).

Moreover, as a higher amount of chromium ions, mainly associated with the dissolution of chromium oxide or chromium hydroxide from the passive film, is observed to be released at Co29Cr7W compared to Co21Cr8Mo7W, the less stability of the film at Co29Cr7W is hence expected to be noticed.

This corrosion behavior of our CoCrMoW alloys deserves to be further in detail investigated to clearly certify that Co21Cr8Mo7W is more appropriate for biomaterial applications and to elucidate the main factors responsible for such behavior.

Better protective properties of the film formed on Co21Cr8Mo7W against corrosion processes is also clearly evidenced from the corrosion parameters (Table 2), estimated from Tafel plots (inset Figure 2). Therefore, the corrosion potentials (E_cor), determined after two different period of immersion, are situated at more positive potentials (Table 3) at Co21Cr8Mo7W, i.e. -0.061 V and -0.059 V at 168 and 1000 hours of immersion for Co21Cr8Mo7W and -0.104 V and -0.114 V for Co29Cr7W at same periods of immersion. Furthermore, the corrosion current density (j_cor), appraised, after the same periods of immersion, is lower (Table 3) at Co21Cr8Mo7W, i.e. 0.038 µA cm^-2 and 0.049 µA cm^-2 at Co21Cr8Mo7W and 0.427 µA cm^-2 and 0.180 µA cm^-2 at Co29Cr7W. As expected, a similar trend was observed for the evolution of corrosion rate, measured at the same periods of immersion, where slower corrosion rates were observed at Co21Cr8Mo7W, i.e. 1.09 µm y^-1 and 2.81 µm y^-1. From these results, it is also interesting to notice that the corrosion parameters of Co21Cr8Mo7W, i.e. E_cor, i_cor and corrosion rates, are not notably changing with immersion time, revealing that the protective film formed on CoCrMoW alloy remains stable. One may conclude that the Co21Cr8Mo7W has better corrosion resistance and is more appropriate for surgical applications.

Table 2. Corrosion parameters of Co21Cr8Mo7W and Co29Cr7W alloys. after immersion in 0.1M PBS at 37°C.

Sample	icor. (µA cm-2)		Ecor vs. Ag/AgCl (mV)		Corrosion rate (µm y-1)
	168 h	1000 h	168 h	1000 h	168 h	1000 h
Co21Cr8Mo7W	0.038	0.049	-61	-59	1.09	2.81
Co29Cr7W	0.427	0.180	-104	-114	12.24	6.57

Table 2. Corrosion parameters of Co21Cr8Mo7W and Co29Cr7W alloys. after immersion in 0.1M PBS at 37°C.

Sample	icor. (µA cm-2)		Ecor vs. Ag/AgCl (mV)		Corrosion rate (µm y-1)
	168 h	1000 h	168 h	1000 h	168 h	1000 h
Co21Cr8Mo7W	0.038	0.049	-61	-59	1.09	2.81
Co29Cr7W	0.427	0.180	-104	-114	12.24	6.57

Table 3. Ions released of Co21Cr8Mo7W and Co29Cr7W alloys after 168 hours immersion in 0.1M PBS at 37°C.

Sample	Metal element	Quantity of ions released
		(mg L-1)	(µg cm-2)
Co21Cr8Mo7W	Co	0.679 mg/L	7.1
	Cr	< 0.02 mg/L	0.15
	Mo W	21.8 µg/L undetectable	0.22
Co29Cr7W	Co	0.335 mg/L	3.33
	Cr	7.40 mg/L	73.7
	W	undetectable

Table 3. Ions released of Co21Cr8Mo7W and Co29Cr7W alloys after 168 hours immersion in 0.1M PBS at 37°C.

Sample	Metal element	Quantity of ions released
		(mg L-1)	(µg cm-2)
Co21Cr8Mo7W	Co	0.679 mg/L	7.1
	Cr	< 0.02 mg/L	0.15
	Mo W	21.8 µg/L undetectable	0.22
Co29Cr7W	Co	0.335 mg/L	3.33
	Cr	7.40 mg/L	73.7
	W	undetectable

In terms of medical applications, another property of these materials which should not be disregarded is the hydrophobic/hydrophilic character effected by the protective film formed on the surface of the samples. Therefore, static contact angle measurements were carried out before and after 1000 hours of immersion at both specimens. The results revealed that at both types of samples, the character of the samples surface is changed from a hydrophobic one before the immersion (i.e. 99.33 ± 0.5 dgr for Co21Cr8Mo7W and 98.93 ± 0.5 dgr for Co29Cr7W) to a hydrophilic one after the immersion (i.e. 69.87 ± 0.5 dgr for Co21Cr8Mo7W and 67.79 ± 0.5 dgr for Co29Cr7W), suggesting that actually the protective film formed on both types of alloys has a hydrophilic character. Therefore, the evidence of the formation of a stable layer on these alloys, with hydrophilic character which both inhibits corrosive attack and provides good affinity for the adhesion of the cells, might open new reliable perspectives in obtaining CoCrMoW alloys with conspicuous properties for medical implants.

3.2. Electrochemical Impedance Spectroscopy

Additional information about the corrosion behavior of both types of samples were obtained from the EIS investigations, carried out potentiostatically at OCP, after 1000 hours of immersion in PBS solution.

As Figure 3a illustrates, within the frequency range from 0.01 to 10⁵ Hz, the impedance Z is much higher at Co21Cr8Mo7W compared to Co29Cr7W, suggesting that the oxide film formed on Co21Cr8Mo7W exhibits better protective properties against the corrosion processes. These findings are in good agreement with previous ones from anodic polarization investigations which attested a similar corrosion behavior at both samples.

In order to gain further information about the corrosion behavior of these samples, by considering an equivalent circuit (EEC) with two time constant in series (Figure 3b, inset), the impedance spectra of both specimens (Figure 3) were fitted. By using this EEC, the experimental data gave the best fit. The first time constant, constituted from CPE_in/R_in, is attributed to the oxide layer formed, during the immersion, on the surface and the second time constant, constituted from CPE_out/R_out, accounts for the dissolution processes at interface alloy/electrolyte and the inhomogeneity of the passive film. Due to these physical complex phenomena and heterogeneity of the surface, the constant phase elements (CPE_out and CPE_in) instead of capacitances were more adequate for fitting the EIS spectra [28], and the chi -squared values (X²) are lower than 1.3E-3 for both samples. This EEC was previously successfully used for describing corrosion behavior of other CoCrMo alloys in different media [9,10,12]. The values of the equivalent circuit parameters of both alloys in the PBS solution are shown in Table 4.

From these parameters, the polarization resistance (R_p) of both specimens, calculated as the sum of R_out and R_in, was estimated and they are 8.6 ×10⁶ Ω cm² and 2.3 ×10⁶ Ω cm² for Co21Cr8Mo7W and Co29Cr7W respectively. As the R_p values of both samples are of the order of Megaohms, it appears that the film formed on both specimens have good corrosion resistance and, from this perspective, they can be further used for surgical applications. However, the R_p of Co21Cr8Mo7W alloy is cca. four times higher than that one of Co29Cr7W alloy, suggesting that a more stable oxide film with better resistance corrosion is formed on Co21Cr8Mo7W. These results are in good agreement with electrochemical corrosion investigations which attested better corrosion performance at Co21Cr8Mo7W alloy and a lower rate of passive layer dissolution (Table 2). These findings are supported by the results obtained for the inner resistances, R_in, of both samples which seems to attest a similar trend evolution (Table 4). The EIS results point out clearly that Co21Cr8Mo7W alloy has better surface protection.

3.3. X-ray Photoelectron Spectroscopy Investigations

The surface chemical composition of Co21Cr8Mo7W and Co29Cr7W alloys was investigated by XPS after their immersion in PBS solution at 37ᵒC for 1000 hours. The XPS analysis attested mainly the presence of cobalt, chrome, tungsten, oxygen, phosphorus on the surface of both specimens. Besides, on the Co21Cr8Mo7W surface, molybdenum is evidenced. These results are in line with those from chemical composition (Table 1) which revealed the presence of the cobalt, chromium, tungsten elements in the bulk of both types of alloys and molybdenum only in the Co21Cr8Mo7W bulk alloy.

Based on literature reports [9,13] and our findings which revealed the presence of a certain amount of phosphorus (i.e. cca. 18 %) on the surface of both specimens, one may assume that during immersion, a non -negligible amount of cobalt and chromium phosphate complexes (e.g. Co(H₂PO₄)₂) are formed on the surface.The presence of these adsorbed complexes on the protective film formed on alloys, as electrochemical results suggest, is expected to mitigate the corrosion attack, and hence to have a certain contribution in ensuring a large passivation domain of both alloys, i.e. about 630mV.

In order to go deeper in understanding the corrosion behavior of these alloys, the Cr 2p, W 4f, Mo 3d, and O 1s deconvoluted high resolution spectra were recorded and the deconvoluted high resolution spectra of both alloys are shown in Figure 4 and Figure 5.

The Cr-2p_3/2 spectra of both types of alloys were fitted with 5 peaks (Figure 4a and Figure 5a). These peaks are attributed to chromium metallic species (Cr⁰) (BE 547.5 ± 0.2 eV, [13,18]) Cr (III) oxide most probably present as Cr₂O₃ species (BE = 575.7 ± 0.2 eV, [13, 16]), Cr(OH)₃ species (BE 576.9 ± 0.2 eV, [13,16]), chromium mixed oxide species (BE 578.2 ± 0.2 eV, [29]) and Cr (VI) oxide, like CrO₃ species (BE 579.5 ± 0.2 eV, [23]). The corresponding relative fractions of the chemical species are shown in inset of figures. From deconvoluted Cr-2p_3/2 spectra one may assert that a significant amount of Cr₂O₃ and Cr(OH)₃ species are formed on the surface of both specimens. It is well known that the presence of these constituents in the passive film ensures good protective properties of the film against corrosion attack [9,13,16]. Thus, based on these findings, one may presume that, good protective properties of the film formed on the surface of both alloys is due to the presence of these non-negligible amounts of chromium oxide and chromium hydroxide species in the film. However, a higher concentration of Cr₂O₃ and Cr(OH)₃ species is present in the passive film formed on the Co29Cr7W surface (i.e. Cr₂O₃ 20.29 % and Cr(OH)₃ 30.90 %) compared to Co21Cr8Mo7W (i.e. Cr₂O₃ 14.91 % and Cr(OH)₃ 28.52 %). These results along with those from electrochemical investigations appear to reveal that the presence of a higher amount of Cr₂O₃ and chromium hydroxide species in the film not necessary brings about better protective properties of the film against corrosion processes. Therefore, in order to better explain the raison behind the corrosion behavior of these alloys, further XPS analysis was necessary to be achieved.

The W-4f spectra of Co21Cr8Mo7W were deconvoluted by assuming one doublet associated to WO₃ species ( the W 4f_7/2 of this doublet is at BE 35.9 eV ± 0.2 eV, [30,31]) (Figure 4b) and that one of Co29Cr7W (Figure 5b) was deconvoluted by assuming three doublets attributed to W⁰ species (the W 4f_7/2 of the first doublet is at BE 31.4 ± 0.2 eV, [30,31]), W⁵⁺ species (the W 4f_7/2 of the second doublet is at BE 34.9 ± 0.2 eV, [32]) and WO₃ species (the W 4f_7/2 of the third doublet is at BE 36 ± 0.2 eV [30,31]). For each doublet we considered a spin orbit splitting of 2.1 eV and a W 4f_7/2 / W 4f_5/2 ratio of 1.33 [32]. The corresponding relative fractions of the chemical species are shown in insets of the corresponding figures. These results clearly point out that the tungsten constituent in the protective film formed on Co21Cr8Mo7W surface is only present as tungsten (VI) oxide species, whereas on Co29Cr7W surface, is also present as W metallic and W⁵⁺ species.

As regards the Mo 3d spectrum of Co21Cr8Mo7W, this spectrum was deconvoluted by considering one double associated to MoO₃ species (the Mo 3d_5/2 of this doublet is at 232.7 ± 0.2 eV, [22,29]) (Figure 4c). For this doublet we considered a spin orbit splitting of 3.13 eV and a Mo 3d_5/2 / Mo 3d_3/2 ratio of 0.67 [29]. It is obvious from these results that molybdenum constituent in the film is present only as MoO₃.

In summary, from the above XPS investigations and literature data [11,18,20], one may conjecture that in fact good corrosion resistance observed at Co29Cr7W alloys originate from the presence of both chromium species (i.e. Cr₂O₃ and Cr(OH)₃) in a significant amount and tungsten oxides species in a much lower amount in the passive film. Conversely at Co21Cr8Mo7W, the XPS results correlated with reported data [16,17, 19], revealed that good corrosion behavior of this alloy is mainly due to the synergetic effect of Cr, W and Mo present in the film as Cr₂O₃, Cr(OH)_3, WO₃ and MoO₃. Previous data from literature revealed that better corrosion resistance of the film formed on CoCr-based alloys surface might be gained when Mo and W are incorporated as alloying elements. The MoO_x and WO₃ species present in the passive film seem to ensure a good compactness of the film [33]. Other reports [16] demonstrated that an efficient barrier against the diffusion of species through the film formed on the surface of these types of alloys might be obtained when both the MoO₂ /similar products and the WO₃ are present in the protective oxide film. It was also evidenced that these oxide species were contributing to the sluggishness of the selective dissolution process of the metals beneath them.

As a conclusion, based on XPS results and electrochemical outcomes, one may emphasize that better corrosion performance observed at Co21Cr8Mo7W alloy is strongly related to the synergetic effect of Cr, W and Mo and cannot be hence disregarded.

These observations are conspicuous ones because they point out that indeed the presence of Cr in a higher amount in the alloy (i.e. 29 % for Co29Cr7W and 21 % for Co21Cr8Mo7W ) not necessary hampers very well the corrosion processes (i.e. r_cor is 2.81 µm y^-1 for Co29Cr7W and 1.09 µm y^-1 for Co21Cr8Mo7W ) and actually the synergetic effect of Cr, Mo and W constituents present in the passive film as oxides is responsible for excellent protective properties of the film against corrosion attack. In other words, one may conjecture that, in our case, at Co21Cr8Mo7W, the presence of W as only hexavalent tungsten oxide along with the presence of molybdenum as MoO₃ in the film brings about improved protectiveness properties of the passive film, impeding hence better the dissolution processes in the film. Besides, Cwalina K.L. et all. [17] demonstrated that the presence of WO₃ species in the passive film extend the passive range because the stability of the film is enhanced by means of interaction of W with water, which leads to the formation of insoluble WO₃ phase. Thus, we consider that the extended passive range and lowered passive current density observed at Co21Cr8Mo7W alloy (i.e. ΔE_pass = 0.63 V; i_pass = 0.4 µA cm^-2) compared to Co29Cr7W alloy (ΔE_pass = 0.49 V; i_pass = 1.2 µA cm^-2) is also due to the presence of a higher amount of WO₃ species in the passive film formed on Co21Cr8Mo7W surface (Figure 4b and Figure 5b).

The O-1s spectra of both types of alloys, deconvoluted with three peaks, associated to O^metal-oxide species (BE 530.4 ± 0.2 eV, [9, 13]), hydroxide or hydroxyl groups, OH^- (BE 531.7 ± 0.2 eV, [34]) and to chemisorbed water (BE = 532.8 ± 0.3 eV, [34] and/or metal-PO₄ [23], (Figure 4d and Figure 5c) support our above XPS results which attested manly the presence of oxygen bonded to the metal and the formation of Cr(OH)₃. The corresponding relative fractions of the chemical species are shown in insets of Figure 4d and Figure 5c.

Moreover, as O-1s spectra point out, a non-negligible amount of hydroxyde or hydroxyl groups is present for example as Cr(OH)₃ on the surface of both specimens, (i.e. around 46 % OH^- species no matter the type of sample, Figure 4d and Figure 5c). The presence of a considerable amount of these types of species on the surface of both alloys is of great importance because they could significantly contribute to the hydrophilic character of the surface, evidenced from static contact angle measurements. The hydrophilic character of the surface, in terms of surface affinity for the adhesion of the cells, is indispensable for medical applications.

One may conclude that in terms of medical applications, this strategy in adding as alloying elements W and Mo in the CoCr alloy composition is of great perspective in obtaining CoCr -based alloys with lower content of chromium than usually used for such applications.

5. Conclusions

The corrosion behavior of two cast CoCr based alloys, one with lower content of chromium (Co21Cr8Mo7W) and another one with higher content of chromium (Co29Cr7W), was systematically investigated in 0.1M PBS, by electrochemical methods, for long period of immersion.

After 1000 hours of immersion, the anodic polarization curves clearly outlined that both studied alloys present good corrosion resistance. Despite the fact that both alloys have similar corrosion behavior, from the estimation of the corrosion parameters, one may resume that Co21Cr8Mo7W alloy has better corrosion protection (i.e. i_cor = 49 nA cm^-2, E_cor = -59 mV and R_cor = 2.81 µm y^-1) compared to Co29Cr7W alloy (i.e. i_cor = 180 nA cm^-2, E_cor = -114 mV and R_cor = 6.57 µm y^-1). This behavior attests that on Co21Cr8Mo7W surface a more stable and compact passive film is formed. These outcomes are in line with EIS results which evidenced a higher polarization resistance (R_p) at Co21Cr8Mo7W alloy (i.e. 8.6 M Ω cm²) compared to Co29Cr7W alloy where a R_p of 3 M Ω cm² was determined.

The good corrosion resistance observed at Co29Cr7W, as XPS investigations point, is mainly due to the presence of chromium species (i.e. Cr₂O₃, Cr(OH)₃ ) and tungsten oxide species in the passive film. At Co21Cr8Mo7W, the XPS results suggest that the excellent corrosion behavior of this alloy mainly originates from the synergetic effect of Cr, W and Mo present in the film as Cr₂O₃, Cr(OH)_3, WO₃ and MoO₃. Moreover, the present investigations revealed that the presence of a higher amount of Cr₂O₃ and chromium hydroxide species in the film does not necessarily bring about better corrosion performance of these CoCr based alloys. In fact, the remarkable corrosion resistance observed at these CoCr based alloys with a lower content of Cr, namely Co21Cr8Mo7W, emerges, as stated previous, from the strategy of adding different alloying elements (i.e. Mo and W in the CoCr alloy composition.

Additionally, as the ion release measurements suggest, at Co21Cr8Mo7W a low concentration of chromium ion release (i.e. 0.15 µg cm^-2), very much beneath the value accepted medical, was evidenced whereas at Co29Cr7W, a high concentration of chromium ion release (i.e. 73.7 µg cm^-2), very much over value accepted medical, is observed.

These results are promising ones in terms of medical applications, because the Co21Cr8Mo7W alloy, herein studied, could be regarded as a valid alternative when aiming to develop new CoCr-based alloys for such applications. By using such type of alloy for surgical applications, with lower amount of Cr, new surgical devices with low costs and safely in terms of chromium ion release could be obtained.