Mo content-depended competition between Cr₂O₃ enrichment and selective dissolution of CoCrFeNiMo_x high entropy alloys

Abstract

The corrosion and passivation behavior of the as-cast CoCrFeNiMo_x (x = 0, 0.2, 0.4, 0.6, 0.8, and 1.0) high entropy alloys (HEAs) are investigated. With the addition of Mo, increasing Mo and Cr riched σ phase precipitates from the primary FCC phase. The corrosion resistance of the CoCrFeNiMo_x HEAs strongly depends on the phase structure and the passive film. Generally, Mo promotes the selective dissolution of the FCC phase and the enrichment of Cr₂O₃ in the passive film, the former deteriorates the corrosion resistance while the latter enhances it. Interestingly, the hypoeutectic Mo_0.6 exhibits a higher corrosion resistance than that of Mo_0.4, eutectic Mo_0.8, and hypereutectic Mo₁. Analysis of surface passive film and the composition of alloy phases implies that the Cr₂O₃ enrichment dominates in Mo_0.6 while selective dissolution governs in Mo_x (x = 0.4, 0.8, and 1.0). Mo content-depended competition between the Cr₂O₃ enrichment and selective dissolution is discussed.

Introduction

High entropy alloys (HEAs), which consist of multiple principal elements in equimolar or near-equimolar ratios have attracted extensive attention due to their breakthrough in the traditional alloy design concept and the performance particularity caused by the multi-principal disordered solid solution phase structure^1,2. In recent years, various HEAs that exhibit good hardness, high wear resistance, and good structural stability at high temperatures have developed^3,4,5. Among them, the CoCrFeNi alloy with a face-centered cubic (FCC) structure has received wide interest due to its good ductility and fracture toughness⁶. In addition to the mechanical properties, most HEAs generally have good corrosion resistance due to their easy formation of a single solid solution structure and contain various passivation alloy elements, such as Cr, Ti, and Mo, and have high application potential in the field of anti-corrosion engineering⁷.

Recently, the corrosion behavior of some kinds of CoCrFeNi-based HEAs alloy in aqueous solution has been reported. Ye et al.⁸ reported CrMnFeCoNi coating has better corrosion resistance than A36 steel substrate in H₂SO₄ solution due to its obvious passivation behavior. However, Luo et al.⁹ investigated the low content of Cr and the extensive formation of metal hydroxide in the passive film are responsible for the lower corrosion resistance of CrMnFeCoNi HEA than 304 L stainless steel in H₂SO₄ solution. In NaCl solution, the corrosion starts from Cr-depleted interdimeric, forming galvanic corrosion with Cr-rich dendrites and making CrMnFeCoNi HEA corrosion resistance degrade⁸. Moreover, Qiu et al.¹⁰ and Wang et al.¹¹ investigated the corrosion resistance of Al₂CrFeNiCoCuTi_x and CoCrFeNiTi_x HEAs, which showed the beneficial effect of Ti in forming the dense passive film rich in TiO₂ and Cr₂O₃. The influence of Al on the corrosion of HEAs has been widely reported. Kao et al.¹² revealed the corrosion resistance of Al_xCoCrFeNi (x = 0, 0.25, 0.50, 1.00) in H₂SO₄ solution. Although the addition of Al increased the thickness of the passive film, Al₂O₃ was easy to form a porous film on the metal surface. And the selective dissolution of Al and Ni has a negative effect on the corrosion resistance of the alloy^12,13,14. At the same time, some studies have shown that the addition of Al has no obvious negative effect on the corrosion resistance of the alloy, such as in CrFeMoVAl_x¹⁵ and FeCoNiCrCu_0.5Al_x¹⁶.

As to the effect of Mo on the corrosion behavior of HEAs, most of them focused on the galvanic corrosion of the alloys and the role of Mo on pitting resistance in NaCl solution. Generally, Mo addition improved the pitting resistance and self-repairing ability of passive film in NaCl solution, which due to the MoO₄²⁻ species could inhibit pitting and promote the formation of Cr oxide-based inner layer¹⁷. Niu et al.¹⁸ studied the corrosion behavior of the as-annealed CoCrFeNiMo_x HEAs (x = 0, 0.2, 0.5, 0.8, 1) in artificial seawater solution, the result showed that the passive film became denser, and the overall dissolution rate decreased by appropriate Mo addition, which could depend on the formation of Mo-oxide film, besides Cr₂O₃ film. However, although Mo-oxides can improve the self-repairing capabilities of the passive film in NaCl solution, the multiple role of Mo in the corrosion resistance of HEAs in acidic solution is rarely reported. For instance, Chou et al.¹⁹ evaluated the corrosion behaviors of Co_1.5CrFeNi_1.5Ti_0.5Mo_x HEAs in 0.5 M H₂SO₄ solution, which were fairly similar. This might be due to the highly acidic environment greatly facilitating the corrosion part, while leaving the passivation part a secondary process. Therefore, in order to further reveal the effect of Mo content on the passivation and dissolution behaviors of CoCrFeNiMo_x HEAs in acidic conditions, more detailed studied on the HEAs in weaker acidic environments are urgently needed.

In the present work, we produced as-cast CoCrFeNiMo_x (x = 0, 0.2, 0.4, 0.6, 0.8, and 1.0) HEAs and compared their corrosion behavior in 5 mM H₂SO₄ solution. The microstructure characterizations were investigated by X-ray diffraction (XRD), scanning electron microscope (SEM), and electron backscatter diffraction (EBSD). And the Volta potential maps were acquired by scanning kelvin probe force microscopy (SKPFM). The electrochemical behavior was evaluated by potentiodynamic and potentiostatic polarization, Mott–Schottky, and electrochemical impedance spectroscopy (EIS). In addition, the passive film was analyzed with X-ray photoelectron spectroscopy (XPS) detailed depth profile. Finally, this paper concentrated on the effect of Mo-induced evolution in alloy structure and passive film composition, and their influences on the corrosion resistance in dilute H₂SO₄ solution. Especially, the competition between passivation and dissolution was further discussed.

Results and discussion

Phase composition and microstructure

Figure 1 shows the XRD patterns of as-cast CoCrFeNiMo_x HEAs. The CoCrFeNi matrix alloy displays FCC single-phase solid solution structure. With the addition of Mo, the diffraction peaks of the FCC phase shift toward the lower 2θ angle, due to the solute atom Mo can induce lattice expansion¹⁸. Significantly, for the Mo₀ and Mo_0.4 HEAs, there appears obvious [200] texture seen from the abnormal peak intensity ratios²⁰. When the molar ratio of Mo is over 0.4, the precipitated phase gradually appears to mitigate the lattice distortion. The secondary phase is identified as the σ-phase, and its structure is close to a tetragonal FeCrMo phase, according to JCPDS card number 09‐0050²¹. The total intensity of the σ-phase is significantly increasing with elevating Mo content. In the Mo₁, the σ-phase becomes the main phase structure beyond the FCC phase.

Fig. 1: XRD patterns of the CoCrFeNiMo_x HEAs.

XRD results of CoCrFeNiMo_x (x = 0, 0.2, 0.4, 0.6, 0.8, 1) HEAs, including FCC phase and σ-phase.

Figure 2 shows the SEM images of CoCrFeNiMo_x (x = 0, 0.2, 0.4, 0.6, 0.8, 1) HEAs. The EBSD phase maps of CoCrFeNiMo_x HEAs are present in Fig. 3, in which the blue and yellow stand for the FCC phase and σ-phase, respectively. CoCrFeNi HEA has a single-phase FCC structure (Figs. 2a and 3a), which is consistent with the XRD results. It can be seen from the SEM image (Fig. 2b) that the Mo_0.2 has a similar dendrite structure. And according to EDS results (Supplementary Fig. 1), the dendrite structure is rich in Mo, but XRD and EBSD results show both these two phases determined to be the FCC phase. This phenomenon has also been reported in previous literature. Both Shun et al.²¹ and Guo et al.²² investigated that in the CoCrFeNiMo_0.2~0.3, the interdendrite and dendrite are determined to be the FCC phase, and their major difference is the content of Mo. The second phase gradually precipitates as dendrite from the FCC matrix when the molar ratio of Mo increases to 0.4 (Figs. 2c and 3c). The EDS mappings are shown in Supplementary Fig. 1 and the chemical compositions for each phase are summarized in Table 1, in which the σ-phase is rich in Mo and Cr. According to the electronhole theory, as applied to superalloys¹⁸, Cr and Mo are two key elements in the formation of σ-phases, since they both have a larger electron hole number than most transition elements. For Mo_0.6 (Fig. 2d, g), the microstructure change to typically hypoeutectic: one of which is a bulk-like structure and another is a eutectic-like structure. The hypoeutectic structure also can be inferred from the EBSD results (Fig. 3d), in which the bulk is the FCC phase and the eutectic-like region is the eutectic structure composed of the FCC phase and σ-phase. The fully eutectic layered structure is obtained in Mo_0.8 (Fig. 3e). And the lamellar spacing of the Mo_0.8 alloy is about 320 nm (Fig. 2h). With the further increase of Mo content, the microstructure of the alloy changes from eutectic structure to hypereutectic structure in Mo₁ HEAs (Fig. 2f, i), and the FCC phase is no longer the main phase (Fig. 3f). According to SEM patterns and XRD analyses, it can be concluded that the σ-phase enriched Cr and Mo gradually precipitate from the CoCrFeNiMo_x HEAs with the increase of Mo.

Fig. 2: SEM images of CoCrFeNiMo_x HEAs.

SEM images of a Mo₀, b Mo_0.2, c Mo_0.4, d and g Mo_0.6, e and h Mo_0.8, f and i Mo₁.

Fig. 3: Phase maps of the EBSD of CoCrFeNiMo_x HEAs.

Phase maps of a Mo₀, b Mo_0.2, c Mo_0.4, d Mo_0.6, e Mo_0.8, f Mo₁, the blue in the phase map stands for the FCC phase, the yellow stands for the σ-phase, and the green stands for unrecognized area.

Table 1 Chemical compositions for each phase of CoCrFeNiMo_x (x = 0.4, 0.6, 0.8 1) HEAs (at. %).

Electrochemical measurements

Figure 4a, b displays the polarization curves of as-cast CoCrFeNiMo_x HEAs in 5 mM H₂SO₄ solution at room temperature. All the alloys exhibit a transition from active to passive behavior, with a wide passivation range and low current density, indicating that protective passive films can be formed on the surface of these alloys. Table 2 summarizes the corrosion current density (i_corr), corrosion potential (E_corr), and the passivation current density (i_pass) of all alloys. The i_pass is the current density at 0.4 V_SCE in the middle of the passivation range. With the Mo content increased from 0.2 to 0.4, E_corr negatively shifts and the i_corr gradually increases, which shows the deterioration of corrosion resistance. However, a low i_pass of 0.874 μA cm⁻² and a positive E_corr is observed in the Mo_0.6 showing the improvement of corrosion resistance. Then, the performance of Mo_0.8 and Mo₁ deteriorates again. In the passivation range, the level of i_pass reflects the dissolution of the oxide film. The smaller i_pass of the oxide film, the better the stability of the film¹⁹. The i_pass of Mo_0.4, Mo_0.8, and Mo₁ increases gradually, indicating the poor stability of passive films. The lower i_pass in Mo_0.6 and Mo_0.2 suggests the higher stability of oxidation film on the surface, which is also in accordance with the higher corrosion resistance. Moreover, the current density of the HEAs with Mo addition decreases in the middle of the passivation range, especially in Mo_0.6, which may be attributed to the formation of molybdenum hydroxide or molybdate on the surface. And its adsorption enhances the passivation effect²³. But the overall corrosion resistance of the alloy is still affected by microstructure, specifically the phase composition and distribution. Accordingly, the performance of Mo_0.8 and Mo₁ deteriorates.

Fig. 4: Potentiodynamic and potentiostatic polarization results.

a potentiodynamic polarization curves, b the enlarged region showing passive potential window, c potentiostatic polarization curves, d corresponding steady-state currents at 0.4 V_SCE for 10 h, the standard deviations (error bars) of the three tests for the steady-state currents.

Table 2 Key parameters from potentiodynamic polarization of CoCrFeNiMo_x HEAs in 5 mM H₂SO₄ solution.

Table 3 The fitted parameters for EIS of the as-cast CoCrFeNiMo_x HEAs in 5 mM H₂SO₄ solution.

The current density versus time plots at 0.4 V_SCE are also provided to have a better insight into the passivation process (Fig. 4c, d). At the initial stage, the current density after potentiostatic polarization (i_p) decreases sharply, attributed to metal surface dissolution and the formation of passive film^24,25. Then, the current remains relatively stable with extending polarization time, indicating that the formation and dissolution of the passive film have reached a balance after 10 h of potentiostatic polarization. For Mo_0.4, there are some sudden increases in local current density at the initial stage of potentiostatic polarization, which reflects the formation and annihilation of metastable pits on the surface. It indicates that local surface dissolution is obviously accelerated, the passivation ability of the surface is insufficient, and the protection of the generated oxide is insufficient. The higher i_p of alloys containing Mo than Mo₀ indicates a high dissolution rate on the surface of these alloys, which is associated with Cr, Mo-depletion regions caused by the second phase enriched in Mo and Cr²⁶. Especially, the i_p of Mo₁ increases after half an hour of polarization, which indicates that the surface passive film is dissolved, which is consistent with the fact that there is no stable passivation range in the dynamic polarization curve. The order of i_p from large to small is Mo₁, Mo_0.8, Mo_0.4, Mo_0.6, Mo_0.2, Mo₀, and the corrosion resistance increases in turn, which is consistent with the potentiodynamic polarization results. In view of the abnormal phenomenon that the corrosion resistance of Mo_0.6 is higher than Mo_0.4, the phase structure and passive film composition will be discussed later.

EIS measurements were conducted to further investigate the corrosion resistance of as-cast CoCrFeNiMo_x HEAs. The Nyquist plots and Bode plots of the original alloys in the H₂SO₄ solution are illustrated in Fig. 5a, b. Nyquist plots shown in Fig. 5a are characterized as unfinished semi-arcs relating to the difference in passive film properties²⁷. The corrosion resistance of alloys can be evaluated by the diameter of semi-arc in Nyquist plots. The larger diameter of the semi-arc, the better the corrosion resistance²⁸. The semi-arc diameter of Mo_0.6 and Mo₀ HEA is significantly larger than that of other alloys, indicating a higher corrosion resistance of the two alloys. In the Bode plot (Fig. 5b), the impedance moduli display a good linear relationship with a slope close to −1 in the frequency range of 10⁻² ~ 10 Hz. Meanwhile, the phase angles of all the specimens reached the maximum in this frequency region, indicating the pseudocapacitive nature of the passive films on the CoCrFeNiMo_x HEAs²⁹. Furthermore, the value of |Z | in the Bode plot at 0.01 Hz usually corresponds to the polarization resistance reflecting the corrosion resistance of the alloy³⁰. Compared with other CoCrFeNiMo_x HEAs, the Mo_0.6 and Mo₀ HEAs have a higher |Z | and larger phase angles at low frequency, showing better corrosion resistance.

Fig. 5: EIS measurements in 5 mM H₂SO₄ solution.

a Nyquist plots, b Bode plots, c, d the variation of fit resistance and capacitance, the standard deviations (error bars) of the three tests for Q₁, Q₂, R₁, R₂.

The electrical equivalent circuit used for fitting the experimental data has also been inserted in Fig. 5a. This model is often adopted for a non-uniform passive film, which consists of a porous outer layer and a compact inner layer³¹. The R_s represents the solution resistance, R₁//Q₁ represents the electrical resistance and the capacitance of the outer layer, and R₂//Q₂ represents the equivalent resistance and capacitance of the inner layer^31,32,33. Considering the rough electrode surface or uneven surface current, Q is used to signify the possibility of a non-ideal capacitance (CPE). It can be defined by the following Eq. (1)³⁴

$$Z_{{{{\mathrm{CPE}}}}} = \frac{1}{{Y_0}}({{{\mathrm{j\omega }}}})^{ - n}$$

(1)

Where Y₀ is the modulus, ω is the angular frequency, j² = −1 is the imaginary number and n is the CPE exponent. When n = 1, the CPE is considered an ideal capacitor, and non-ideal when 0.5 < n < 1.

Equivalent electrical circuits were employed to analyze the EIS data, as shown in Fig. 5c, d and Table 3. For all HEAs, R₂ is significantly higher than R₁, and Q₁ is higher than Q₂, which is close to 1, showing a denser inner film and loose outer film. The low R₁ can be ascribed to the low stability of some metal oxide species such as Fe-oxide which has poor stability and might selective dissolution in acidic solutions with low pH³². But the presence of stable elements such as Cr- and Mo-based ions will maintain surface adsorption^30,35, resulting in the formation of an external porous layer. The impedance R₂ of the inner passive film can be used as an effective index to evaluate the corrosion resistance of the alloy. As shown in (Fig. 5d, R₂ decreases with the increase of Mo content but rises at x = 0.6, indicating a better corrosion resistance of Mo_0.6 than that of Mo_0.4, Mo_0.8, and Mo₁. The increase in Q₂ over Mo contents shows that the addition of Mo leads to great changes in the thickness and composition of the passive film^36,37. The Q₂ of Mo_0.8 and Mo₁ HEAs is larger (Fig. 5d), which may be due to the increase of capacitance caused by the decrease of electron dissipation layer thickness and the increase of carrier number^9,32, resulting in the deterioration of corrosion resistance of these two alloys.

The semiconducting properties of the passive films formed on as-cast CoCrFeNiMo_x HEAs were characterized by the Mott–Schottky analysis. According to the Mott–Schottky plots, the potential-capacitance relationship in the space charge layer under the depletion condition can be expressed through Eq. (2):

$$\frac{1}{{C^2}}{{{\mathrm{ = \pm }}}}\frac{2}{{e\varepsilon \varepsilon _0NA^2}}\left( {E - E_{{{{\mathrm{FB}}}}} - \frac{{{{{\mathrm{kT}}}}}}{e}} \right)$$

(2)

where e is the charge of a single electron (1.6 × 10⁻¹⁹ C), k is the Boltzmann constant (1.38 × 10⁻²³ J K⁻¹), ε and ε₀ refer to the dielectric constant of the semiconductor (15.6), and the permittivity of vacuum (8.85 × 10⁻¹⁴ F m⁻¹), respectively. The E_FB is the potential of flat-band (V_SCE), which corresponds to the voltage intercept of the linear segment on the Mott–Schottky plots. T is the absolute temperature (K) and N represents the carrier density (cm⁻³) including the donor concentration (N_D) and acceptor concentration (N_A).

Figure 6a shows the Mott–Schottky plots of the passive film on the CoCrFeNiMo_x HEAs after the potentiostatic polarization in 5 mM H₂SO₄ solution at 0.4 V_SCE for 10 h. All plots show two linear regions and a narrow flat region (–0.2~ 0.4 V_SCE), which contains flat-band potential E_FB. A similar phenomenon has been reported for the passive film formed on stainless steel³⁸. At a potential lower than E_FB, the negative slope represents the p-type semiconductor corresponding to the outer layer of the passive film. While the positive slope at a potential higher than E_FB indicates the n-type semiconductor, which corresponds to the inner layer of the passive film³⁹. The donor (N_D) and acceptor densities (N_A) of the passive films formed on CoCrFeNiMo_x HEAs were calculated and shown in Fig. 6b. All carrier densities are in the range of 10²¹ ~ 10²² cm³, which presents a similar passive film character to that of the stainless steels^40,41,42. Particularly, the N_A of all CoCrFeNiMo_x HEAs is an order of magnitude larger than N_D, indicating the dense inner layer (n-type) and the loosened outer layer (p-type). In addition, with the addition of Mo, the carrier densities increase gradually, indicating the deterioration of the corrosion resistance of the alloy. The carrier concentration of Mo₀, Mo_0.2, and Mo_0.6 are low, showing better corrosion resistance. The carrier densities of Mo_0.8 and Mo₁ HEAs are higher than that of other alloys, suggesting more defects in the passive film. and higher electrochemical activity at the solution-metal interface, which is consistent with the previous electrochemical test results.

Fig. 6: Mott–Schottky results of the passive films.

a Mott–Schottky plots of the passive film after the potentiostatic polarization at 0.4 V_SCE for 10 h, b The donor (N_D) and acceptor densities (N_A) of the n-type and p-type semiconductors.

The selective dissolution of alloys

To confirm the electrochemical potential difference between FCC and σ-phase, Volta potential maps were acquired by SKPFM for the alloys, as indicated in Fig. 7. The bright area demonstrates more positive potential than the dark area in the SKPFM images. The precipitated σ-phase shows a larger Volta potential value (~20 mV) than the FCC phase (~ −40 mV), as shown in Fig. 7d. Undoubtedly, the Volta potential difference leads to micro-galvanic activity between the σ-phase and FCC phase, at which the latter acts as the micro-anode to erode preferentially whilst σ-phase acts as the micro-cathode^43,44. However, the Volta potential difference between these two phases depends on the Mo content. With the increase of Mo, the Volta potential difference increases gradually, which means that the risen driving force for galvanic coupling between those regions⁴⁵. For Mo_0.4, there is no obvious Volta potential difference between the two phases. And Volta potential difference over the σ-phase and FCC phase is low for Mo_0.6 and Mo_0.8 alloys. But in the following observation of corrosion morphology, these three alloys still show weak galvanic corrosion. For the Mo₁, the Volta potential difference is about 60 mV at the interface between the σ-phase and FCC phase, which may be attributed to the quite different proportion of the two phases in the hypereutectic structure. The increased Volta potential difference indicates the enhanced micro-galvanic corrosion in Mo₁ alloy⁴³. The aggravated selective corrosion makes Mo₁ show higher i_p and poor corrosion resistance in electrochemical tests. Nevertheless, the 60 mV Volta potential difference is relatively low and only shows the relatively weak micro-galvanic corrosion in CoCrFeNiMo_x HEAs⁴³. Meanwhile, the Volta potential difference will affect the formation of the passive films on the surface of the alloys, the corresponding discussion comes later.

Fig. 7: SKPFM results of CoCrFeNiMo_x HEAs.

Surface voltage potential maps of (a) Mo_0.4, (b) Mo_0.6, (c) Mo_0.8, (d) Mo₁, (e) corresponding line-profile analysis of relative Volta potential in map (d).

Microstructural characterization also was conducted to support the selective corrosion of the FCC matrix (Fig. 8). The pitting occurs on the Mo_0.4 alloy surface (Fig. 8a). With the content of Mo in the alloy more than 0.4, the main corrosion way of the alloy changes into gradually enhanced galvanic corrosion. The secondary electron images show that there is an obvious contrast difference between the two phases, indicating that the FCC phase with darker color tends to dissolve. For the Mo_0.6, the corrosion type appears as galvanic corrosion at the interfaces between FCC matrix and σ-phase, as shown in Fig. 8b. In Mo_0.8 eutectic alloy, the increase of phase interface provides more preferential corrosion areas (Fig. 8c), which leads to the deterioration of corrosion resistance. And with increasing Mo content, the Volta potential difference increases, especially for Mo₁. As a result, the corrosion at the two-phase interface is particularly aggravated in Fig. 8d. In summary, when σ-phase appears, the FCC matrix is selectively dissolved. And with the increase of Mo content, the galvanic corrosion of the two-phase interface gradually intensified.

Fig. 8: SEM images after the potentiostatic polarization.

SEM images of (a) Mo_0.4, (b) Mo_0.6, (c) Mo_0.8, (d) Mo₁ after the potentiostatic polarization at 0.4 V_SCE for 10 h.

The composition and structure of passive films

Figure 9 shows the depth profiles of the Cr, Co, Ni, O, Fe, and Mo elements in the passive films of as-cast CoCrFeNiMo_x HEAs after the potentiostatic polarization at 0.4 V_SCE for 10 h in 5 mM H₂SO₄ solution. The amount of O and Cr is higher than other elements during the initial stage of sputtering. With the increase of sputtering time, the amount of O and Cr decreases gradually, indicating that Cr-enrich passive film is formed on the surface of each alloy. The inner layer and outer layer of the passive film are divided by the highest point of Cr content. When the differential of Cr content versus time is always greater than −3, it means that the Cr content tends to be stable and the sputtering reaches the substrate. The atomic ratio of Fe, Co, and Ni increases with the sputtering, suggesting that the three elements are selectively dissolved and have limited contribution to the passive film^46,47. The selective dissolution of these elements leads to the formation of a Cr-enrich passive film, which is similar to the passivation of stainless steel⁹.

Fig. 9: XPS depth profiles of the elements in the passive films.

XPS depth profiles of the elements in the passive film after the potentiostatic polarization at 0.4 V_SCE for 10 h in 5 mM H₂SO₄ solution, a Mo₀, b Mo_0.2, c Mo_0.4, d Mo_0.6, e Mo_0.8, and f Mo₁, as well as the comparison in element g O, h Cr, and i Mo distribution.

Figure 9g compares the O content of each HEA. The content of O increases with the addition of Mo, indicating an increased amount of metal oxides and hydroxides in the passive film. The surface O content of Mo₁ is higher than other alloys. Combined with the high carrier densities, the surface electrochemical activity of Mo₁ is higher, which may lead to the accumulation of a larger number of loose corrosion products⁴⁸. As shown in Fig. 9h, the atomic ratio of Cr in Mo₀ gradually decreases with the increase of sputtering time. In the alloys containing Mo, the peak value of the atomic ratio of Cr appears in the sub-surface layer of the passive film. The more Mo content is, the more obvious this phenomenon is. This is due to the selective dissolution of the primary phase caused by the precipitation of the second phase, which may cause the loss of surface Cr and nonuniform distribution of Cr in the passive film¹⁸. In particular, in the passive film formed on Mo_0.8 and Mo₁, Cr has a sudden change in the depth direction. The decrease of Cr content in the outer passive film and the nonuniform distribution of Cr will have a negative impact on the corrosion resistance. At the same time, the more selective dissolution of Ni, Cr, and Fe due to the enhanced selection corrosion, may lead to the reduction of the densification of the passive films. In particular, the content of Cr in the surface layer of Mo_0.6 is stable, indicating that the distribution of Cr in the passive film is uniform, therefore, the passive film has good stability. The content of Mo increases slightly with the increase of sputtering time as shown in Fig. 9j, and the upward trend is more obvious in Mo₁ HEAs than in other alloys. Furthermore, the changing trend of Mo content shows that Mo tends to dissolve rather than passivate in the alloy passive film with a molar ratio greater than 0.6.

Further analyses of O, Cr, and Mo spectra were performed and provided in Figs. 10–12 to determine the depth distribution of the species with different oxidation states. As shown in Fig. 10, the O 1 s spectra were characterized by the presence of three components O²⁻ (530.7 eV), OH⁻ (532.5 eV), and H₂O (532.8 eV). The Mo_0.2 (Fig. 10a), Mo_0.4 (Fig. 10b), Mo_0.6 (Fig. 10c), and Mo₁ (Fig. 10d) are selected for analysis. The Mo₀ and Mo_0.8 are shown in Supplementary Fig. 3. The intensity of O1s spectra decreases gradually with the increase of sputtering time. And except for Mo₁, the peak of O²⁻ is the main peak in the O 1s spectra of other alloys. It shows that the passive films with alloy oxides as the main component are formed on the surface of the alloys (Supplementary Fig. 2a). The spectral intensity of O²⁻ in Mo_0.2 and Mo_0.6 decreases slowly in the depth direction of the passive films (Fig. 10a,c). When the sputtering time is 30 s, there is still an obvious oxide peak. The oxide content of Mo_0.6 is always higher than Mo_0.4, inferring that more metal oxides in the passive film lead to its corrosion resistance better than Mo_0.4. In addition, Mo₁ shows a different spectrum (Fig. 10b). When the sputtering time is 6 s, the hydroxide is the main specie, which is more than that of other alloys.

Fig. 10: The detailed XPS spectra of O 1 s of the passive films.

a Mo_0.2, b Mo_0.4, c Mo_0.6, and d Mo₁ HEAs.

Fig. 11: The detailed XPS spectra of Cr 2p_3/2 of the passive films.

a Mo_0.2, b Mo_0.4, c Mo_0.6, and d Mo_0.8 HEAs.

Fig. 12: The detailed XPS spectra of Mo 3d of the passive films.

a Mo_0.2, b Mo_0.4, c Mo_0.6, and d Mo_0.8 HEAs.

According to the fitting results, Fig. 13a displays the ratio of O²⁻/OH⁻ with increasing depth. Mo_0.6 has a stable and highest ratio of O²⁻/OH⁻, indicating that its passive film is relatively uniform and dense. In addition, Mo₁ provides a higher ratio of O²⁻/OH⁻ after 4 cycles. However, the ratio of O²⁻/OH⁻ changes sharply, and the ratio of the surface layer film is lower than other alloys, suggesting that the outer layer of the passive film is mainly composed of loose metal hydroxide, which cannot resist corrosive ions, resulting in poor corrosion resistance of Mo₁.

Fig. 13: Special species distribution summarized from XPS depth profiles.

a XPS depth profiles of the ratio of O²⁻/OH⁻, b the ratio of Cr_ox/Cr_hy, c the atomic percentage of Mo species, d the atomic percentage of Cr₂O₃ for the passive films and reciprocal of passivation current density.

The Cr 2p_3/2 spectra were separated into Cr(OH)₃ (578.7 eV), Cr₂O₃ (576.5 eV), and Cr⁰ (574.4 eV) (Fig. 11). the Mo_0.2 (Fig. 11a), Mo_0.4 (Fig. 11b), Mo_0.6 (Fig. 11c), and Mo_0.8 (Fig. 11d) are selected. The Mo0 and Mo1 are shown in Supplementary Fig. 3. Cr element exhibits the special property to combine a relatively high heat of adsorption of oxygen (or OH⁻) with a relatively low Cr-Cr bond strength, which is a crucial factor for the nucleation of the oxide⁴⁹. In addition, compared to Cr(OH)₃, Cr₂O₃ has a more thermally stable anhydrous structure and lower point defect density⁵⁰, so it plays an important role in corrosion resistance. With the increase of sputtering time, the main peak of Cr gradually moves from Cr(OH)₃ and Cr₂O₃ with binding higher energy to Cr metal with lower binding energy (Supplementary Fig. 2b). In the passive films, Cr₂O₃ is the main species of Cr. When the time increases to 30 s and 42 s, the percentage of Cr₂O₃ in the spectrum of Mo_0.2 and Mo_0.6 is always higher than that in Mo_0.4 and Mo₁ indicating that Cr₂O₃ is deeply and evenly distributed in the passive film, which makes the passive films of Mo_0.2 and Mo_0.6 have good stability.

According to the XPS analysis results, the ratio of Cr₂O₃ and Cr(OH)₃ (Cr_ox/Cr_hy) for the passive films formed on CoCrFeNiMo_x is shown in Fig. 13b. Mo_0.6, Mo_0.2, and Mo₀ have a high ratio of Cr_ox/Cr_hy. And the passive films contain more Cr₂O₃, which can well protect the alloys and have good corrosion resistance. The ratio of Cr_ox/Cr_hy for Mo_0.4, Mo_0.8, and Mo₁ decrease in turn, which corresponds to the gradually decreased corrosion resistance and is consistent with the results of the electrochemical test. It can be concluded that the content of Cr₂O₃ in the passive film is the decisive factor affecting the corrosion resistance.

The Mo 3d spectra were separated into Mo⁶⁺ (235.5 and 232.2 eV), Mo⁴⁺ (232.5 and 229.0 eV), and Mo⁰ (231.1 and 228.0 eV) (Fig. 12). The behavior of Mo 3d spectra are similar to the Cr 2p_3/2 spectra, with the increase of sputtering time, the metal peak increases, and the peaks of metal oxides decrease (Supplementary Fig. 2d). The content of high valence Mo⁶⁺ oxides gradually decreases with the increase of Mo and disappears in Mo_0.2 after a sputtering time of 30 s. With the increase of Mo content, the higher valence oxide of Mo appears in deeper passive films, and the proportion gradually increases in Mo_0.6 (Fig. 12c1), Mo_0.8 (Fig. 12d1), and Mo₁ (Supplementary Fig. 3e1). Figure 13c shows the atomic percentage of Mo in different valence states. With the increase of sputtering time, the content of Mo metal increases, and the Mo oxides including Mo⁶⁺ and Mo⁴⁺ contents decrease gradually. Moreover, with the molar ratio of Mo increasing from 0.4 to 0.6, the content of the oxides Mo⁶⁺ and Mo⁴⁺ increases gradually. With the increase of Mo content to more than 0.6, the percentage of Mo oxides remains stable, indicating that the oxides of Mo don’t increase with the increase of Mo content in the alloy, and there is an upper limit for the oxide species of Mo in the passive film. This may be attributed to a large amount of energy released upon adsorption of oxygen not easily disrupting a large number of Mo-Mo bonds⁴⁹. Because the transition from adsorbed overlayer to oxide nucleation requires the disruption of metal-metal bonds⁴⁹. Previous studies have shown that Mo oxides have a positive effect on the passive film^18,51, but the positive effect is limited due to the limitation of Mo oxides contents.

The role of Mo in microstructure and passive films

In general, the addition of Mo has a negative impact on the corrosion resistance of the alloy, with a relatively small negative impact on Mo_0.2 and Mo_0.6. Especially, the corrosion resistance of Mo₀ is better than other alloys (Figs. 4 and 5), but its Cr₂O₃ content is not higher than that of other alloys (Fig. 13d). On the one hand, Mo₀ is a single-phase structure with uniform composition, which is easy to form a uniform and dense passive film (Fig. 14a). On the other hand, the fact that Cr₂O₃ content in other alloys does not decrease significantly may be attributed to Mo promoting the enrichment of Cr₂O₃ in the passive film, and the selective dissolution of other elements promoted by phase selective corrosion. However, The Cr₂O₃ passive film formed by the massive dissolution of other elements is looser and less protective than the passive film formed by the diffusion and oxidation of Cr elements to the outer layer (Fig. 14). Therefore, the addition of Mo does not obviously change the percentage of Cr₂O₃ in the passive film, but the corrosion current density increases.

Fig. 14: The schematic diagram of galvanic corrosion and passivation.

a Mo₀, b Mo_0.4, c Mo_0.6, and d Mo_0.8 HEAs.

For Mo_0.2, the addition of a small amount of Mo contributes to the formation of Mo oxides in the passive film, and Cr₂O₃ is uniformly distributed in the passive film. There is no precipitated σ-phase in Mo_0.2 (Figs. 2 and 3), only elemental segregation of Mo, and thus very weak selective corrosion in the alloy structure. And Mo promotes the deprotonation of Cr hydroxide⁵², resulting in a higher content of Cr₂O₃ in Mo_0.2 than that in Mo₀. In contrast, the element segregation caused by the addition of Mo will also slightly increase the current density in corrosion. The corrosion resistance of Mo_0.4 deteriorates (Fig. 14b) due to the noticeable decrease of Cr₂O₃ content (Fig. 13d), even though a few amounts of Mo oxides are formed. More importantly, due to the precipitation of a small amount of discontinuous second phase rich in Mo and Cr, selective corrosion dissolution is conducted in the primary phase^18,51. Therefore, the limited Mo and Cr oxides and selective corrosion make the worse corrosion resistance of Mo_0.4 than Mo_0.2 and Mo_0.6.

As the atomic ratio of Mo increases to 0.6 (Fig. 14c), the corrosion resistance of the alloy improves, which is better than Mo_0.4. Ameer et al. proposed that the oxides of Mo can eliminate the defects on the surface of the passive film, thereby improving the pitting resistance of the alloy⁵³. The content of Mo oxides is more than Mo_0.4, which plays a synergistic effect in the passive film. In addition, Mo can promote the enrichment of Cr₂O₃ in the passive film, making the passive film contain a large amount of Cr₂O₃. The precipitation of a more connected second phase may alleviate the galvanic corrosion caused by pitting precipitation⁵¹. Under this composition, the increase of the Cr₂O₃ percentage caused by the addition of Mo reached the maximum value, and the protective effect of Cr₂O₃ dominates, rather than the corrosion resistance reduction caused by the selective dissolution. Hence, Mo has a little negative impact on the corrosion resistance of Mo_0.6.

When the content of Mo is greater than 0.8, the corrosion resistance of the alloy decreases, and the corrosion resistance of Mo₁ deteriorates sharply. Mo_0.8 is a typical lamellar eutectic structure (Fig. 14d). The lamellar structure will produce more phase boundaries, providing more preferential dissolution boundaries for corrosion, which will lead to the increase of corrosion current density (Fig. 8). From the composition of the passive film, with the further increase of the content of Mo to 0.8, the auxiliary Mo oxide did not increase compared with Mo_0.6 (Fig. 13c), while the content of Cr₂O₃ in the passive film decreased (Fig. 13d), leading to the decline of the quality of the passive film. The decrease in the percentage of Cr₂O₃ is mainly attributed to the high content of the decreasing diffusion of Cr to the outer layer of the passive film, making the outer layer of the passive film lose its protection, leading to the invasion of corrosive ions, which further worsens the protection of the passive film. The dissolution of Ni and other elements in Mo_0.8 due to the selective corrosion of the phase also reduces the density of the passive film. Mo₁ shows the high carrier density in the Mott–Schottky test and the high content of metal hydroxide in the passive film components in XPS analysis, indicating that there are a lot of defects in the passive film and the accumulation of corrosion products. The lower ratio of Cr and Mo of Mo₁ than that of other alloys leads to poor development of passive film and has poor stability and protection⁵⁴. In addition, the proportion of two phases in hypereutectic structure is quite different, forming a large cathode and a small anode, which will cause serious corrosion of the FCC matrix. And the more different Volta potential than that of other alloys enhances the micro-galvanic corrosion.

In summary, the addition of Mo deteriorates the corrosion resistance of the CoCrFeNiMo_x HEAs owing to the selective dissolution of the FCC matrix over the σ-phase. The preferential corrosion of phase interfaces amplifies in eutectic (Mo_0.8). And the increased Volta potential difference accelerate the micro-galvanic corrosion in hypereutectic structures (Mo₁). Mo promotes the enrichment of Cr₂O₃ in the passive film which enhances the corrosion resistance of the CoCrFeNiMo_x HEAs. But the limited Mo and Cr oxides and selective corrosion reduce the corrosion resistance of Mo_0.4. And the poor passive films resulting from the aggravated phase selective corrosion offer limited protection for Mo_0.8 and Mo₁, resulting in poor corrosion resistance. In Mo_0.6, the percentage of Cr₂O₃ in the passive film increases to the maximum value due to the addition of Mo. The beneficial effect of Cr₂O₃ enrichment exceeds the negative impact of Mo-induced selective corrosion. As a result, Mo_0.6 exhibits a lower corrosion rate than that of Mo_0.4, Mo_0.8, and Mo₁.

Methods

Materials preparation

The CoCrFeNiMo_x (x = 0, 0.2, 0.4, 0.6, 0.8, and 1.0) HEAs were cast via arc melting of mixture of the pure metal elements (purity > 99.99 %) in a 0.01 atm vacuum. The ingots were cast and re-melted at least 5 times to facilitate composition homogenization. Samples with dimensions of 10 × 10 × 4 mm were encapsulated by epoxy resin. Then, the only exposed surface of 1 cm² was ground sequentially from 600 to 2000 grit SiC paper. The EBSD samples were ground sequentially from 600 to 4000 grit SiC paper and were polished with silica polishing solution for 30 min. Finally, the specimens were cleaned with deionized water, alcohol and then dried in the cool air.

Electrochemical measurements

Corrosion behavior was carried out by an electrochemical workstation (Princeton PARSTAT 4000) using a conventional three-electrode electrochemical cell system, which consisted of a platinum (Pt) counter electrode, the sample as the working electrode, and a saturated calomel electrode (SCE) as the reference electrode. All the electrochemical measurements were performed at 25 °C in 5 mM H₂SO₄ solution. Before testing, the open circuit potential (OCP) was measured for 1 h in the test solution to ensure a stable state. The potentiodynamic polarization (PD) was performed with a scan rate of 0.167 mV s⁻¹ in a potential window of −0.2 V_OCP ~ 1.2 V_SCE. And the potentiostatic polarization (PS) was conducted with an applied voltage of 0.4 V_SCE for 10 h. The Mott–Schottky analyses were performed with a scan rate of 40 mV s⁻¹ and a fixed frequency of 1 kHz in a potential window of −0.8 V_SCE ~ 0.8 V_SCE. The EIS experiment was recorded over a frequency range from 10⁵ to 10⁻² Hz with a potential perturbation of 10 mV and fitted by ZsimpWin software. Each test was repeated at least three times to ensure reproducibility.

Surface composition and morphology analysis

The phase composition of as-cast CoCrFeNiMo_x HEAs was identified by X-ray diffractometer (XRD, Bruker D8 ADVANCE) within 2θ of 20° ~80° at a scanning rate of 2° min⁻¹. A scanning electron microscope (SEM, FEI Helios G4 CX) was applied to characterize the microstructure of the alloys. The electron backscatter diffraction (EBSD, FEI Helios G4 CX) patterns were acquired under the following conditions: an accelerating voltage of 20 kV, and a sample tilt of 70°. And the Volta potential maps were acquired by asscanning kelvin probe force microscopy (SKPFM, Bruker Dimension Icon). The chemical characteristics of the passive films formed at 0.4 V for 10 h were obtained using the X-ray photoelectron spectroscopy (XPS, PHI 5000 VersaProbe III) equipped with Al Kα Xray radiation operated at 50 W and 15 kV. The analysis area was set by an X-ray beam size of 200 μm in diameter. The surface was sputtered by Ar⁺ ion bombardment every 6 s for 12cycles followed by 12 s for 11cycles with an ion beam of 2 keV and a 3 × 3 mm² area to obtain depth profiles. The analyzer pass energy was 280 eV for high-resolution spectra of Cr 2p_3/2, Fe2p_3/2, Mo 3d, and O1s. All XPS peaks were calibrated to the standard carbon C1s peak (284.8 eV) and analyzed using the Multipak software.