Microstructural Characteristics and Governing Mechanism of Anomalous Corrosion Behavior in a CoCrNiCu Medium-Entropy Alloy

Abstract

To clarify the anomalous corrosion behavior in Cu-containing CoCrNi-based medium-entropy alloys, in which an enhanced corrosion driving force is accompanied by a reduced overall corrosion rate, the phase constitution, microstructure, electrochemical behavior, post-corrosion morphology, and surface chemical states of CoCrNi, CoCrNiCu, and CoCrNiCuFe alloys were systematically compared. The results show that Cu addition induces pronounced phase separation in the CoCrNi matrix, leading to the formation of a Cu-depleted FCC1 phase, a continuous Cu-rich FCC2 intergranular network, and dispersed nanoscale Cu-rich precipitates, with an FCC2 area fraction of about 0.145. In 3.5 wt.% NaCl solution, CoCrNiCu exhibits a stronger thermodynamic tendency for corrosion, whereas its overall corrosion rate does not increase, but instead shows the lowest corrosion current density and higher impedance, indicating an anomalous electrochemical response. Post-corrosion SEM morphology, EDS elemental mapping, and XPS valence-state analyses further reveal that corrosion is mainly concentrated in the Cu-rich phases and their adjacent narrow regions, while the Cu-rich phases themselves remain relatively stable as non-sacrificial cathodes. Semi-quantitative thermodynamic and mass-transport calculations indicate that although Cu-induced phase separation enhances the micro-galvanic corrosion driving force, with an estimated interphase potential difference of about 0.337 V, the overall corrosion rate remains constrained by the oxygen diffusion supply during cathodic oxygen reduction on the Cu-rich regions. Therefore, the anomalous corrosion response of CoCrNiCu can be attributed to the synergistic effect of the enhanced micro-galvanic corrosion driving force caused by Cu-induced phase separation and the restricted cathodic oxygen supply.

1. Introduction

CoCrNi-based medium-entropy alloys have attracted wide attention because of their excellent strength-ductility synergy and good damage tolerance, and they have become one of the most representative FCC multi-principal-element alloy systems [1,2,3]. For long-term service structures used in marine engineering, chemical equipment, and salt-spray environments, not only good mechanical properties but also high corrosion resistance and stability are required [4,5]. As a typical multi-principal-element alloy system, CoCrNi-based alloys generally show a strong tendency to form simple solid solutions. However, their actual phase constitution is not determined solely by high configurational entropy, but is also jointly affected by mixing enthalpy, atomic size mismatch, and interelement interactions [6,7]. Therefore, during alloying regulation, this system may still evolve from a single-phase state to a multiphase state, accompanied by short-range ordering, elemental segregation, phase separation, and local compositional inhomogeneity, thereby affecting local electrochemical characteristics [8,9].

Previous studies have generally suggested that the formation and stability of a dense Cr-rich oxide passive film on the alloy surface are a key basis for the corrosion resistance of CoCrNi-based alloys [10,11]. More recent work has further shown that spontaneous passivation behavior and passive-film properties can change markedly with alloy composition and polarization conditions [12,13,14]. This indicates that corrosion behavior in Cu-containing systems cannot be judged solely from Cu content itself, but should be analyzed together with passive-film evolution and local microstructural features [15]. For systems with obvious phase separation and compositional heterogeneity, corrosion behavior is not governed only by film stability, but is also jointly influenced by local elemental distribution, interphase boundary states, and electrochemical differences between phases [16,17,18,19].

Cu is one of the common alloying elements used to regulate CoCrNi-based alloy systems. Appropriate Cu addition can modify the microstructure, surface state, and overall performance to a certain extent [20]. However, because Cu has relatively positive mixing enthalpies with Co, Cr, and Ni, its solid solubility is limited, and segregation, Cu-rich phase formation, or phase separation can readily occur [21,22], thereby significantly changing microstructural uniformity, phase characteristics, and local electrochemical behavior [23]. Existing studies have not reached a consistent understanding of the role of Cu in medium- and high-entropy alloys. Available evidence suggests that Cu segregation and Cu-rich phases can enlarge the interphase potential difference, intensify micro-galvanic effects, and alter passive-film composition and stability, thereby reducing corrosion resistance in chloride-containing media [15,18]. Another group reports that, in certain alloy systems, stabilization of Cu-rich phases, refinement of segregation, optimization of the solidification structure, or improvement of the surface film structure may enhance repassivation ability and polarization response, thus suppressing localized corrosion to some extent and yielding relatively more stable electrochemical behavior [24,25,26]. More importantly, studies from the past two years further indicate that the key factor controlling corrosion resistance is not simply whether Cu is present, but rather how the Cu-rich phase forms, how it is spatially distributed, and how it affects film stability and local potential differences [27,28,29,30].

This complexity is particularly pronounced in CoCrNiCu alloys with obvious phase separation, where localized corrosion behavior and macroscopic electrochemical response often do not show a simple one-to-one correspondence, and the underlying control mechanism still lacks a unified explanation [25]. Therefore, in this work, CoCrNi, CoCrNiCu, and CoCrNiCuFe alloys were selected as model materials to systematically compare differences in phase structure, microstructure, and corrosion behavior, with emphasis on the effect of Cu-induced microstructural evolution on corrosion response. By combining microstructural characterization, electrochemical testing, post-immersion corrosion morphology, XPS results, and thermodynamic and mass-transport analysis, the intrinsic relationship between localized corrosion behavior and macroscopic electrochemical response under Cu-induced phase separation is discussed, providing experimental and theoretical support for understanding the corrosion-control mechanism of Cu-containing CoCrNi-based medium-entropy alloys.

2. Materials and Methods

2.1. Alloy Preparation and Microstructural Characterization

CoCrNi, CoCrNiCu and CoCrNiCuFe alloys were prepared by vacuum arc melting. All elemental raw materials had purities higher than 99.9 wt.%, and the charge was weighed according to the nominal equiatomic ratios shown in

Table 1. Nominal charge composition and mass fractions of the alloys.

Alloy	Co (wt.%)	Cr (wt.%)	Ni (wt.%)	Cu (wt.%)	Fe (wt.%)
CoCrNi	34.7	30.7	34.6	–	–
CoCrNiCu	25.3	22.3	25.2	27.2	–
CoCrNiCuFe	20.4	18.0	20.3	22.0	19.3

. Melting was carried out under an Ar protective atmosphere. To improve compositional homogeneity, each alloy ingot was flipped and remelted at least five times, and was then suction-cast into a bar-shaped ingot using a water-cooled copper mold. The ingots were cut by electrical discharge wire cutting into specimens of approximately 10 mm × 10 mm × 3 mm. Before testing, the exposed surfaces of the specimens were successively ground with 400# to 2000# SiC papers, mirror-polished using a 0.25 μm diamond polishing agent, ultrasonically cleaned in absolute ethanol for 10 min, and then dried for subsequent use.

Phase analysis was carried out using a Rigaku SmartLab X-ray diffractometer (Rigaku Corporation, Tokyo, Japan) equipped with a D/teX Ultra 250 detector (Rigaku Corporation, Tokyo, Japan) and Cu K$\alpha$ radiation at an operating voltage of 40 kV and a current of 40 mA. The scanning ranged from 20° to 100° with a step size of 0.02°. Microstructural observation and compositional analysis were performed using a Zeiss GeminiSEM 360 field-emission scanning electron microscope (Carl Zeiss AG, Oberkochen, Germany) equipped with Oxford Instruments EDS (Oxford Instruments plc, Oxford, UK). At least five points were measured in each phase region and averaged to obtain the phase composition and its fluctuation range. The grain size was determined by Oxford C-nano EBSD (Oxford Instruments plc, Oxford, UK) using the equivalent circle diameter method. Transmission electron microscopy (Thermo Fisher Scientific, Waltham, MA, USA) was used to characterize the fine structure of the alloys and intragranular precipitates.

2.2. Electrochemical Testing and Surface Analysis

Electrochemical experiments were conducted in naturally aerated 3.5 wt.% NaCl solution at 25 °C under stagnant conditions. A conventional three-electrode system was employed, in which the specimen with an exposed area of 1 cm² served as the working electrode, a platinum plate served as the counter electrode, and a saturated calomel electrode served as the reference electrode. A schematic illustration of the electrochemical cell of three-electrodes is shown in Figure 1. After immersion, the samples were stabilized for 1 h to obtain a relatively steady open-circuit potential. Potentiodynamic polarization tests were performed at a scan rate of 1 mV s⁻¹. The corrosion potential $E_{\mathrm{corr}}$, corrosion current density $i_{\mathrm{corr}}$, and Tafel slopes were obtained by fitting the polarization curves, and the polarization resistance $R_{\mathrm{p}}$ was calculated using the Stern–Geary relationship [31,32].

Electrochemical impedance spectroscopy was measured at the open-circuit potential over a frequency range from 100 kHz to 10 mHz with an AC perturbation amplitude of 10 mV. The impedance data were fitted using ZView software (version 3.0, Scribner Associates, Southern Pines, NC, USA) together with equivalent circuit models [33]. To analyze the post-corrosion surface chemical state, the samples were immersed in 3.5 wt.% NaCl solution for 72 h, then removed, rinsed with deionized water and dried. The surface morphology and local elemental distribution after immersion were examined by SEM and EDS, while the chemical-state of the surface was characterized by X-ray photoelectron spectroscopy (XPS) using a Thermo Fisher Scientific K$\alpha$ X system with a monochromatic Al K$\alpha$ X-ray source.

2.3. Thermodynamic and Mass-Transport Calculations

To clarify the synergistic control of interphase micro-galvanic coupling and cathodic oxygen diffusion during the corrosion of the CoCrNiCu alloy, theoretical calculations of the interphase potential difference and the oxygen diffusion-limited current density were carried out on the basis of thermodynamic and mass transport theory. The calculation temperature was set at $T=298.15$ K. The 3.5 wt.% NaCl solution was approximated as a 0.6 mol kg⁻¹ NaCl background electrolyte, and all potentials were referenced to SHE. It should be clarified that the analysis in this section is not based on software-based thermodynamic simulations, but on analytical and semi-quantitative evaluations using the Pitzer model, the Nernst equation, Henry’s law, the Setschenow correction, and the Stokes–Einstein equation.

2.3.1. Calculation of Interphase Potential Difference

The ionic activities in the NaCl background electrolyte were corrected using the Pitzer model. The equilibrium potential of each element was calculated using the Nernst equation, and the phase potential was obtained as the atomic-fraction-weighted sum of the elemental potentials within each phase. The Pitzer model parameters were taken from the 25 °C NaCl single-electrolyte system [34,35].

The ionic strength of the solution was calculated as follows:

$$I={\displaystyle \frac{1}{2}}\sum m_i{z}_i^2$$

(1)

where $m_i$ is the molality of ion i, and $z_i$ is its charge number.

For an electrolyte ${\mathrm{M}}_{{\nu }_M}{\mathrm{X}}_{{\nu }_X}$, the mean activity coefficient is expressed as:

$$ln{\gamma }_{\pm }=\left|z_M{z}_X\right|f^{\gamma }+m{\displaystyle \frac{2v_M{v}_X}{\nu }}{B}_{MX}^{\gamma }+m^2{\displaystyle \frac{2{\left(v_M{v}_X\right)}^{\frac{3}{2}}}{\nu }}{C}_{MX}^{\gamma }$$

(2)

where m is the molality of electrolyte ${\mathrm{M}}_{{\nu }_M}{\mathrm{X}}_{{\nu }_X}$, $z_M$ and $z_X$ are the charge numbers of cation M and anion X, respectively, ${\nu }_M$ and ${\nu }_X$ are the stoichiometric coefficients of the cation and anion, and $\nu ={\nu }_M+{\nu }_X$. The terms in the equation are defined as follows:

$$f^{\gamma }=-A_{\phi }\left[{\displaystyle \frac{\sqrt{I}}{1+b\sqrt{I}}}+{\displaystyle \frac{2}{b}}ln\left(1+b\sqrt{I}\right)\right]$$

(3)

$$B_{MX}^{\gamma }=2{\beta }_{MX}^{\left(0\right)}+{\displaystyle \frac{2{\beta }_{MX}^{\left(1\right)}}{{\alpha }^{2}I}}\left[1-\left(1+\alpha \sqrt{I}-{\displaystyle \frac{1}{2}}{\alpha }^{2}I\right)e^{-\alpha \sqrt{I}}\right]$$

(4)

$$C_{MX}^{\gamma }={\displaystyle \frac{3}{2}}{C}_{MX}^{\phi }$$

(5)

For aqueous solution at 25 °C, the Debye–Hückel constant was taken as $A_{\phi }=0.3915{\mathrm{kg}}^{1/2}{\mathrm{mol}}^{-1/2}$, with $b=1.2{\mathrm{kg}}^{1/2}{\mathrm{mol}}^{-1/2}$ and $\alpha =2.0{\mathrm{kg}}^{1/2}{\mathrm{mol}}^{-1/2}$. ${\beta }_{MX}^{\left(0\right)}$, ${\beta }_{MX}^{\left(1\right)}$, and $C_{MX}^{\phi }$ are the ion interaction parameters in the Pitzer model. For the background electrolyte of NaCl of 0.6 mol kg⁻¹, the parameters were taken as ${\beta }^{\left(0\right)}=0.0765$, ${\beta }^{\left(1\right)}=0.2664$, and $C^{\phi }=0.00127$.

After applying the activity correction under the background electrolyte condition, the equilibrium potential of each element was estimated using the Nernst equation:

$$E_i=E_i^{\Theta }+{\displaystyle \frac{RT}{n_{i}F}}ln\left({\displaystyle \frac{a_{M^{n_i+}}}{a_M}}\right)$$

(6)

where $E_i^{\Theta }$ is the potential of the standard electrode relative to the standard hydrogen electrode, R is the gas constant, T is the thermodynamic temperature, $n_i$ is the number of transferred electrons and F is the Faraday constant. For solid metallic components, $a_M$ was approximated as 1, and the activity of metal ions in solution is expressed as:

$$a_{M^{n_i+}}={\gamma }_{M^{n_i+}}{m}_{M^{n_i+}}$$

(7)

In the equilibrium calculation, $m_{M^{n_i+}}={10}^{-6}\mathrm{mol}{\mathrm{kg}}^{-1}$ was used.

The phase potential of each alloy phase was approximated as the atomic-fraction-weighted average of the elemental potentials:

$$E_{\mathrm{phase}}=\sum _i\left(x_i·E_i\right)$$

(8)

The interphase potential difference was defined as:

$$\Delta E=E_{\mathrm{FCC}2}-E_{\mathrm{FCC}1}$$

(9)

where $x_i$ is the atomic fraction of element i in the corresponding phase. $E_{\mathrm{phase}}$ and $\Delta E$ were used to characterize the relative electrochemical potential difference between the FCC1 and FCC2 phases.

2.3.2. Calculation of Oxygen Diffusion-Limited Current

In air-saturated 3.5 wt.% NaCl solution at 25 °C, the cathodic oxygen reduction reaction:

$$O_2+2H_{2}O+4e^{-}\to 4O{H}^{-}$$

has a theoretical limiting diffusion current density of:

$$i_{L,\mathrm{theory}}={\displaystyle \frac{nF{D}_{O_2}{C}_{O_2}^{\ast }}{{\delta }_{\mathrm{diff}}}}$$

(10)

where the number of electrons transferred was taken as $n=4$, and the partial pressure of oxygen was taken as $P_{O_2}=0.21$ atm. The thickness of the oxygen diffusion layer ${\delta }_{\mathrm{diff}}$ was assigned an empirical value of 0.05 cm according to the relevant corrosion mass transport models [36].

The oxygen diffusion coefficient $D_{O_2}$ was approximately calculated using the Stokes–Einstein equation [37,38]:

$$D_{O_2}={\displaystyle \frac{k_{B}T}{6\pi \eta r_{O_2}}}$$

(11)

The equilibrium oxygen solubility in air-saturated pure water, $C_{\mathrm{water}}$, was calculated using Henry’s law:

$$C_{\mathrm{water}}=K_H{P}_{O_2}$$

(12)

The equilibrium oxygen solubility in the solution, $C_{O_2}^{\ast }$, was calculated after correction using the Setschenow equation [39,40,41]:

$$log\left({\displaystyle \frac{C_{\mathrm{water}}}{C_{O_2}^{\ast }}}\right)=k_{s}I$$

(13)

In the above equations, F is the Faraday constant, T is the absolute temperature, $\eta$ is the viscosity of the solution, $k_B$ is the Boltzmann constant, $r_{O_2}$ is the hydrated radius of the oxygen molecule, $k_s$ is the salting-out constant, $K_H$ is Henry’s constant and I is the ionic strength of the solution. It should be noted that the weighted elemental phase-potential calculation adopted in this work is only intended for a semi-quantitative comparison of the relative electrochemical tendency between FCC1 and FCC2, rather than a rigorous prediction of the true phase potentials in a multi-principal-element alloy.

3. Results and Discussion

3.1. Phase Constitution and Microstructural Features

The XRD results in Figure 2 show that the CoCrNi alloy mainly exhibits the characteristics of a single FCC phase, while the CoCrNiCu and CoCrNiCuFe alloys display peak broadening or partial peak splitting, indicating that the addition of Cu and Fe drives the system from a relatively homogeneous single-phase structure toward a more complex dual-phase microstructure. For the CoCrNiCu alloy, the diffraction peaks can be assigned to two FCC phases with lattice parameters of approximately 3.632 Å and 3.578 Å, which are hereafter referred to as FCC1 and FCC2, respectively. The addition of Cu induces pronounced phase separation in the CoCrNi matrix, which is consistent with the commonly observed formation of Cu-rich phases in Cu-containing high-entropy and medium-entropy alloys [18].

The SEM morphology and EDS mapping results in Figure 3 show that the CoCrNiCu alloy exhibits a dual-phase microstructure composed of a matrix phase and an intergranular phase. The results of the EDS point analysis listed in

Table 2. Chemical composition of two phases in CoCrNiCu alloy determined by EDS point analysis.

Phase	Co (at.%)	Cr (at.%)	Ni (at.%)	Cu (at.%)
Matrix (FCC1)	33.2 ± 1.5	31.8 ± 1.3	26.4 ± 1.1	8.41 ± 0.65
Intergranular (FCC2)	11.9 ± 1.2	11.9 ± 1.1	22.7 ± 1.8	53.54 ± 2.1

indicate that Co and Cr are mainly enriched in the matrix phase, whereas Cu is preferentially segregated to the intergranular phase, forming a relatively continuous Cu-rich network, and Ni is distributed relatively uniformly between the two phases. Combined with XRD phase analysis, the matrix phase can be identified as the Cu-depleted FCC1 phase, while the intergranular phase can be identified as the Cu-rich FCC2 phase [9]. This indicates that Cu addition transforms the originally more homogeneous single-phase CoCrNi alloy into a phase-separated structure with pronounced multiscale compositional heterogeneity. The formation of this intergranular Cu-rich phase is closely related to the limited solid solubility and positive mixing enthalpy between Cu and the other principal elements. During solidification and subsequent cooling, Cu atoms are difficult to remain stably dissolved in the growing FCC matrix and are therefore gradually rejected and preferentially enriched in grain-boundary and interdendritic regions, eventually forming Cu-rich intergranular phases. Meanwhile, some Cu is retained inside grains as fine precipitates, reflecting a multiscale segregation feature involving both intergranular enrichment and intragranular precipitation. Fe addition further modifies the Cu-induced phase separation behavior by reducing the compositional difference and lattice-parameter mismatch between the Cu-lean FCC1 matrix and the Cu-rich FCC2 phase, making the grain-boundary Cu-rich network more discontinuous and semi-isolated.

Image statistics show that the Cu-rich FCC2 phase has an area fraction of about 0.145 and forms a relatively continuous network along grain boundaries. In contrast, the Cu-depleted FCC1 matrix occupies most of the microstructure. The TEM results in Figure 4 further show that, in addition to the continuously distributed Cu-rich phase along grain boundaries, uniformly dispersed Cu-rich nanoscale precipitates are also present inside the grains, which is consistent with the multiscale segregation features reported for Cu-containing CoCrNi-type alloys [19]. The EBSD statistical results in Figure 5 show that the average grain size of the CoCrNiCu alloy is about 32.7 μm. Taken together, the XRD, SEM, and TEM results indicate that the addition of Cu transforms the CoCrNi alloy from a relatively homogeneous FCC system into a multiscale heterogeneous microstructure with pronounced phase separation.

3.2. Electrochemical Corrosion Behavior and Surface Response

As shown in Figure 6a, the three alloys exhibit markedly different potentiodynamic polarization behaviors in 3.5 wt.% NaCl solution. CoCrNi and CoCrNiCuFe show similar anodic polarization features, in which the current density first increases with increasing potential, then decreases markedly, and changes again in the higher-potential region, overall reflecting the formation of a passive film, local passivity breakdown, and a subsequent repassivation tendency. In contrast, the polarization curve of CoCrNiCu is shifted toward lower potentials, and during anodic polarization it does not exhibit electrochemical features similar to those of the other two alloys, but instead maintains a relatively low current density over a wide potential range. Tafel fitting further shows that CoCrNiCu has the most negative $E_{\mathrm{corr}}$, the lowest $i_{\mathrm{corr}}$, and the highest $R_{\mathrm{p}}$, as summarized in

Table 3. Electrochemical parameters obtained from potentiodynamic polarization tests in 3.5 wt.% NaCl solution.

Alloy	${\mathit{E}}_{\mathbf{corr}}$ (V vs. SCE)	${\mathit{i}}_{\mathbf{corr}}$ (A/cm2)	${\mathit{\beta }}_{\mathbf{a}}$ (mV/dec)	${\mathit{\beta }}_{\mathbf{c}}$ (mV/dec)	${\mathit{R}}_{\mathbf{p}}$ ($\mathbf{\Omega }·{\mathbf{cm}}^2$)
CoCrNi	$-0.297\pm 0.015$	$(5.68\pm 0.23)\times {10}^{-6}$	$68.3\pm 3.2$	$121.5\pm 5.8$	$(1.50\pm 0.08)\times {10}^6$
CoCrNiCu	$-0.508\pm 0.012$	$(2.73\pm 0.11)\times {10}^{-6}$	$52.7\pm 2.5$	$89.4\pm 4.3$	$(2.62\pm 0.12)\times {10}^6$
CoCrNiCuFe	$-0.348\pm 0.018$	$(2.92\pm 0.14)\times {10}^{-6}$	$61.2\pm 3.0$	$103.7\pm 5.1$	$(2.27\pm 0.11)\times {10}^6$

. It should be noted that the more negative corrosion potential here reflects a shift in the mixed electrochemical state of the system, rather than directly indicating the overall corrosion rate. These results indicate that the corrosion potential, corrosion current density, and polarization resistance of CoCrNiCu do not follow the monotonic relationship commonly observed in conventional corrosion systems, suggesting that Cu addition does not simply improve or degrade the corrosion resistance of CoCrNi-based alloys, but more likely changes the rate-controlling step of the corrosion process [27,28,29].

As shown in Figure 6b–d, the EIS spectra of the three alloys exhibit pronounced frequency-dependent responses. In the Nyquist plots shown in Figure 6b, all three alloys show impedance responses dominated by capacitive arcs, among which the CoCrNiCu alloy displays the largest arc radius. In addition, CoCrNiCu exhibits an obvious tail extension and non-ideal dispersive feature in the low-frequency region. The Bode plots in Figure 6c further show that the impedance modulus $\left|Z\right|$ of all three alloys increases with decreasing frequency. Among them, the CoCrNiCu alloy maintains the highest values throughout the medium- and low-frequency range, with a low-frequency $log\left|Z\right|$ of 4.7, significantly higher than 3.9 for CoCrNiCuFe and 3.6 for CoCrNi. In the phase-angle plots shown in Figure 6d, all three alloys display one dominant peak, and the CoCrNiCu alloy shows the highest peak value of about 79°, higher than 72° for CoCrNi and 68° for CoCrNiCuFe. In the high-frequency region, the impedance modulus values of the three alloys are relatively close, indicating that the solution resistance under the test conditions is similar.

Notably, the low-frequency upturn and non-ideal dispersive response of the CoCrNiCu alloy suggest that its impedance response contains a diffusion-related contribution in addition to the interfacial charge-transfer process. Previous studies have shown that under conditions of finite diffusion length, interfacial non-uniformity, and pronounced CPE behavior, diffusion impedance does not necessarily appear as a typical 45° Warburg line, but more often manifests itself as low-frequency dispersion or upturn following a depressed capacitive arc [42,43,44]. For metallic corrosion systems, especially when charge transfer is coupled with oxygen mass transport under thin electrolyte layers or restricted oxygen supply, the Nyquist response often still remains a single arc, with deformation appearing only in the low-frequency region [45,46].

To further examine whether this low-frequency response contains a diffusion-related contribution, equivalent-circuit fitting was first carried out using the simplified $R\left(QR\right)$ model shown in the inset of Figure 6c, which was applied uniformly to the three alloys for direct comparison. Under this simplified description, the apparent $R_{\mathrm{ct}}$ of CoCrNiCu is 209.4 k$\Omega ·{\mathrm{cm}}^2$, about 55 times that of the CoCrNi alloy, and the detailed fitting parameters are listed in

Table 4. Fitting parameters from EIS data using the $R\left(QR\right)$ equivalent circuit model.

Alloy	${\mathit{R}}_{\mathbf{s}}$ ($\mathbf{\Omega }·{\mathbf{cm}}^2$)	Q (S·sn/cm2)	n	${\mathit{R}}_{\mathbf{ct}}$ (k$\mathbf{\Omega }·{\mathbf{cm}}^2$)
CoCrNi	$7.61\pm 0.35$	$(3.88\pm 0.19)\times {10}^{-5}$	$0.809\pm 0.015$	$3.79\pm 0.18$
CoCrNiCu	$10.04\pm 0.48$	$(3.05\pm 0.15)\times {10}^{-5}$	$0.893\pm 0.012$	$209.4\pm 10.2$
CoCrNiCuFe	$8.12\pm 0.39$	$(7.65\pm 0.37)\times {10}^{-5}$	$0.827\pm 0.014$	$9.78\pm 0.47$

. Meanwhile, the CoCrNiCu alloy has a relatively lower Q value but an n value as high as 0.893, indicating that its dominant impedance response is closer to a depressed but relatively complete capacitive arc. Combined with the polarization results, CoCrNiCu shows higher impedance parameters and a lower corrosion current density on the one hand, but the most negative corrosion potential on the other, indicating that it has the strongest corrosion tendency while the overall interfacial reaction process is not accelerated accordingly and is instead more strongly constrained. Therefore, the fitted $R_{\mathrm{ct}}$ in the simplified model is more appropriately regarded as an apparent interfacial resistance parameter, rather than a pure charge-transfer resistance.

For CoCrNiCu, the diffusion-containing Randles-type $R\left(Q\right(RW\left)\right)$ model shown in the inset of Figure 6d was further introduced for comparison. The fitting results, summarized in

Table 5. Comparison of equivalent-circuit fitting parameters for CoCrNiCu using the simplified $R\left(QR\right)$ model and the diffusion-containing $R\left(Q\right(RW\left)\right)$ model.

Model	${\mathit{R}}_{\mathbf{s}}$ ($\mathbf{\Omega }$)	CPE-T	CPE-P	${\mathit{R}}_{\mathbf{ct}}$ or ${\mathit{R}}_2$ (k$\mathbf{\Omega }$)	${\mathit{W}}_{\mathit{R}}$ (k$\mathbf{\Omega }$)	${\mathit{W}}_{\mathit{T}}$ (s)	${\mathit{W}}_{\mathit{P}}$
$R\left(QR\right)$	10.04	$3.05\times {10}^{-5}$	0.893	209.4	—	—	—
$R\left(Q\right(RW\left)\right)$	10.07	$3.027\times {10}^{-5}$	0.894	188.45	139.76	11.89	2.389

, show that the high-frequency interfacial parameters remain essentially unchanged between the two models. Specifically, $R_s$ changes only from 10.04 to $10.07\Omega$, the CPE-T value from $3.05\times {10}^{-5}$ to $3.027\times {10}^{-5}$, and the CPE-P value from 0.893 to 0.894. In contrast, when the diffusion branch is included, the previously fitted high-low-frequency impedance can be further resolved into a charge-transfer resistance of $R_2=188.45$ k$\Omega$ together with an additional diffusion-related contribution, with $W_R=139.76$ k$\Omega$ and $W_T=11.89$ s. These results indicate that the large low-frequency impedance response of CoCrNiCu cannot be assigned to pure charge transfer alone, but is more consistently interpreted as a coupled response involving both charge transfer and diffusion-related mass transport.

Considering that the electrochemical tests were carried out in naturally aerated stagnant NaCl solution, the cathodic oxygen reduction process is inevitably influenced by dissolved oxygen supply and diffusion rate. Therefore, the impedance response of the CoCrNiCu alloy more likely reflects the cooperative action of interfacial charge transfer and oxygen mass transport, rather than a single pure charge-transfer process.

3.3. Post-Corrosion Morphology and High-Resolution XPS Analysis

As shown in Figure 7, the SEM morphologies of the three alloys after electrochemical testing differ significantly. The surface of the CoCrNi alloy remains relatively smooth and compact overall, with only a few small pits and fine scratches and no obvious accumulation of corrosion products. Higher-magnification observations reveal uniformly distributed nanoscale corrosion pores, indicating a relatively uniform corrosion feature. In contrast, the addition of Cu markedly changes the corrosion distribution mode of the CoCrNi-based alloy, causing the CoCrNiCu alloy to exhibit pronounced localized corrosion. Corrosion is mainly concentrated in the regions adjacent to the Cu-rich intergranular phase, especially in the transition zone between this phase and the matrix, whereas the Cu-rich phase itself remains relatively stable during corrosion.

The CoCrNiCuFe alloy also exhibits nonuniform corrosion characteristics, but the addition of Fe changes the morphology and distribution of corrosion products on the CoCrNiCu alloy surface. The previously dense and localized accumulation of corrosion products is weakened, resulting in a product structure composed of both particles and flakes, with local regions prone to spallation. This change may also be related to the participation of Fe-containing oxide/hydroxide species in the corrosion-product evolution, which tend to form relatively loose lamellar products with poorer adhesion. This is unfavorable for the formation of effective surface protection, and therefore its overall corrosion resistance is slightly lower than that of the CoCrNiCu alloy.

Because the sample surfaces after electrochemical testing are usually already in a severely over-corroded state, the final morphology cannot directly reflect the specific roles of different microstructural regions during corrosion. To further observe the corrosion behavior of different microstructural regions, the samples were immersed in 3.5 wt.% NaCl solution for 72 h and then examined by SEM. As shown in Figure 8, the surface of the CoCrNiCu alloy can be clearly divided into the intergranular Cu-rich phase region, the intragranular matrix region, and the narrow intergranular transition zone. At the same time, EDS elemental mapping shows that the O signal is relatively low in the intergranular phase region but increases markedly beginning from the narrow transition zone, indicating that the matrix region and the narrow transition zone are more prone to forming and retaining corrosion oxides.

As shown in Figure 9, the high-resolution XPS spectra obtained after 72 h of immersion show that the Cu 2p spectrum on the surface of the CoCrNiCu alloy still retains a clear characteristic peak of metallic Cu, together with a small Cu₂O peak, indicating that only limited oxidation occurs on the surface of the Cu-rich phase during corrosion and that this phase remains chemically stable overall [18]. By contrast, Cr, Co, and Ni all exhibit oxide and hydroxide signals to varying degrees, while the O 1s spectrum contains characteristic peaks of oxides, hydroxides, and adsorbed water, indicating that the matrix region undergoes more pronounced anodic dissolution during corrosion, accompanied by the deposition and coverage of corrosion products [47,48,49,50]. These results further indicate that Cu addition changes the distribution of passive-film chemistry across different microstructural regions: the Cu-rich regions remain dominated by metallic Cu with only slight Cu₂O formation, whereas the Cu-depleted matrix is more readily covered by oxide and hydroxide species, reflecting clear differences in film composition and protectiveness between different phases.

The above experiments and observations indicate that corrosion in the CoCrNiCu alloy does not proceed uniformly over the entire surface, but is concentrated in specific microstructural regions and their neighboring areas. Meanwhile, the Cu-rich intergranular phase remains relatively intact after corrosion, indicating that different microstructural regions play different electrochemical roles during the corrosion process.

3.4. Corrosion Mechanism Governed by Micro-Galvanic Coupling and Cathodic Oxygen-Supply Limitation

As shown by the polarization and impedance results in Section 3.2, the CoCrNiCu alloy exhibits an electrochemical response distinctly different from that of conventional corrosion systems, namely, a negative shift in corrosion potential $E_{\mathrm{corr}}$, no corresponding increase in corrosion current density $i_{\mathrm{corr}}$, but instead higher impedance response and polarization resistance. Meanwhile, the non-ideal dispersive response observed in the low-frequency region of EIS indicates that the corrosion process of this system is not controlled solely by a single interfacial charge-transfer step, but more likely involves the participation of mass transport. This indicates that the addition of Cu does not simply improve or deteriorate the corrosion resistance of the alloy but rather changes the control step of the corrosion process. Related studies in the past two years have further shown that the corrosion response of Cu-containing alloys is increasingly understood to be jointly determined by the distribution of Cu-rich phases, the difference between phases’ potential, and the mode of microstructural regulation [27,28,29,30]. However, polarization and impedance parameters alone cannot directly determine whether this anomalous response mainly arises from changes in passive-film behavior, enhanced interphase electrochemical heterogeneity, or mass-transport limitation of the cathodic process. Therefore, it is necessary to combine the microstructural characteristics, post-corrosion morphology, and surface chemical state of the CoCrNiCu alloy with quantitative analysis of the interphase potential difference and the theoretical oxygen diffusion-limited current density to further clarify its corrosion-control mechanism.

To facilitate a discussion of the mechanism controlling the anomalous electrochemical response of the CoCrNiCu alloy,

Table 6. Thermodynamic and electrochemical parameters of CoCrNiCu alloy.

Parameter	Symbol	Value	Unit	Source/Meaning
FCC1 phase potential	$E_{\mathrm{FCC}1}$	$-0.524$	V vs. SHE	$E_i=E_i^{\Theta }+{\displaystyle \frac{RT}{n_{i}F}}ln\left({\displaystyle \frac{a_{M^{n_i+}}}{a_M}}\right)$
FCC2 phase potential	$E_{\mathrm{FCC}2}$	$-0.187$	V vs. SHE	$E_i=E_i^{\Theta }+{\displaystyle \frac{RT}{n_{i}F}}ln\left({\displaystyle \frac{a_{M^{n_i+}}}{a_M}}\right)$
Interphase potential diff.	$\Delta E$	$0.337$	V vs. SHE	$\Delta E=E_{\mathrm{FCC}1}-E_{\mathrm{FCC}2}$
O2 diffusion coeff.	$D_{{\mathrm{O}}_2}$	$1.37\times {10}^{-5}$	cm2 s−1	Stokes-Einstein eq.
O2 solubility	$C_{{\mathrm{O}}_2}^{\ast }$	$2.34\times {10}^{-5}$	mol cm−3	Henry’s law
Theoretical limiting current density	$i_{L,\mathrm{theory}}$	$(2.5\pm 1.2)\times {10}^{-5}$	A cm−2	$i_L={\displaystyle \frac{nF{D}_{{\mathrm{O}}_2}{C}_{{\mathrm{O}}_2}^{\ast }}{{\delta }_{\mathrm{diff}}}}$
Corrosion groove width	w	$0.8\pm 0.1$	μm	SEM image analysis
FCC2 phase area fraction	$f_{\mathrm{FCC}2}$	$0.145$	—	SEM image analysis
Average grain size	L	$32.7\pm 1.5$	μm	EBSD
Effective anode area fraction	$f_{\mathrm{eff}}$	$0.051$	—	$f_{\mathrm{eff}}={\displaystyle \frac{4}{\sqrt{3}}}{\displaystyle \frac{w}{L}}-{\displaystyle \frac{\pi }{2}}{\left({\displaystyle \frac{w}{L}}\right)}^2$

summarizes the key parameters calculated according to the methods described in Section 2.3, including those related to the interphase micro-galvanic driving force, the local anodic area effect, and the limitation of cathodic oxygen diffusion.

According to the calculations in Section 2.3, the approximate phase potentials of FCC1 and FCC2 are −0.524 V and −0.187 V, respectively, corresponding to an interphase potential difference of about 0.337 V versus SHE. This indicates that the Cu-rich FCC2 phase has a higher potential and is therefore more likely to act as the cathode, whereas the Cu-depleted FCC1 phase has a lower potential and is more likely to undergo preferential anodic dissolution. Although this phase potential is only an approximate estimate based on composition-weighted elemental contributions and does not explicitly consider the complex elemental interactions, local chemical environment, or passive-film/interfacial kinetics in multicomponent solid solutions, it is nevertheless consistent with the phase-partitioning features revealed by EDS, the selective dissolution morphology after corrosion, and the general laws of micro-galvanic corrosion in multiphase alloys, and can therefore serve as an important auxiliary basis for identifying cathodic FCC2 and anodic FCC1 [51]. In other words, the Cu-induced microstructural heterogeneity likely establishes a relatively stable anodic-cathodic partition and thereby provides the thermodynamic driving force for localized corrosion; however, this cathodic/anodic assignment is still based mainly on the above mutually consistent indirect evidence and remains to be directly verified by localized electrochemical techniques such as SKPFM or SVET. Recent studies have likewise pointed out that the geometric distribution, continuity, and stability of Cu-rich phases can significantly reshape local potential differences and corrosion response, so the specific form of microstructural heterogeneity is more critical than the mere presence of Cu [29].

The immersion-corrosion morphology in Figure 8 further shows that preferential dissolution does not spread uniformly across the entire Cu-depleted matrix, but is mainly concentrated in the narrow transition zone adjacent to the Cu-rich phase, which can be approximately regarded as the dominant anodic region. To quantitatively characterize the area fraction of the local preferential dissolution region and evaluate the degree of current concentration within this local area, the effective anodic area fraction can be expressed, based on a hexagonal grain geometry approximation, as:

$$f_{\mathrm{e}\mathrm{f}\mathrm{f}}={\displaystyle \frac{4}{\sqrt{3}}}{\displaystyle \frac{w}{L}}-{\displaystyle \frac{\pi }{2}}{\left({\displaystyle \frac{w}{L}}\right)}^2$$

(14)

where L is the grain size and w is the width of the narrow preferential-dissolution band. Substitution gives a dominant anodic area fraction of $f_{\mathrm{eff}}\approx 5.51\%$. The current density in the dominant anodic region can then be written as:

$$i_{a,\mathrm{l}\mathrm{o}\mathrm{c}\mathrm{a}\mathrm{l}}={\displaystyle \frac{i_{\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{r}}}{f_{\mathrm{e}\mathrm{f}\mathrm{f}}}}$$

(15)

The calculation shows that the current density in the dominant anodic region is about 18.1 times the macroscopic average corrosion current density. This indicates that the main consequence of micro-galvanic action is not a marked increase in the total corrosion current, but rather the concentration of a limited corrosion current in the regions adjacent to phase boundaries, thereby leading to the coexistence of low macroscopic $i_{\mathrm{corr}}$ and obvious localized corrosion [52]. This interpretation is supported jointly by the immersion-corrosion morphology, EDS mapping, and electrochemical results: Figure 8 shows that preferential dissolution is mainly confined to the narrow transition zone near the Cu-rich phase, whereas the polarization and EIS results indicate that the overall corrosion rate does not increase in parallel with the local corrosion driving force.

However, this increase in local current density does not break the balance between anodic dissolution and cathodic reduction. It should be emphasized that a higher local anodic current density does not mean that the total corrosion current of the system can increase without limit; instead, it must be balanced by the reduction current that can be sustained on the Cu-rich cathodic phase. When the effective cathodic area is limited, the total corrosion current is further constrained by the available cathodic current [53,54]. The current-conservation relationship can be expressed as:

$$i_{a,local}·f_{\mathrm{e}\mathrm{f}\mathrm{f}}=i_{c,FCC2}·f_{FCC2}$$

(16)

If the macroscopic corrosion current is distributed over the entire area of the Cu-rich cathodic phase, the average cathodic current density on FCC2 can be approximated as

$$i_{c,\mathrm{F}\mathrm{C}\mathrm{C}2}={\displaystyle \frac{i_{\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{r}}}{f_{\mathrm{F}\mathrm{C}\mathrm{C}2}}}$$

(17)

Substituting $i_{\mathrm{corr}}=2.73\times {10}^{-6}\mathrm{A}·{\mathrm{cm}}^{-2}$ and $f_{\mathrm{FCC}2}=0.145$ gives $i_{\mathrm{c},\mathrm{FCC}2}\approx 1.88\times {10}^{-5}\mathrm{A}·{\mathrm{cm}}^{-2}$. This result indicates that, under the condition of a limited area of the Cu-rich cathodic phase, the cathodic current density per unit area required to sustain corrosion is already at a relatively high level. In other words, to balance the anodic dissolution current, the Cu-rich cathodic phase with limited area must support a stronger oxygen reduction reaction. Because the electrochemical tests were conducted in naturally aerated stagnant NaCl solution, the cathodic process is sensitive to dissolved oxygen supply, and thus whether the total corrosion current can further increase is more readily constrained by the oxygen supply available for cathodic oxygen reduction.

The EIS results further show that the CoCrNiCu alloy exhibits a stronger interfacial impedance response under open-circuit conditions. Its larger Nyquist arc and higher apparent $R_{\mathrm{ct}}$ indicate that the overall electrochemical reaction process is more strongly constrained. For this type of corrosion system, in which mass transport may participate, the apparent $R_{\mathrm{ct}}$ should not simply be equated with pure charge transfer resistance, but rather should be understood as the combined response of coupled interfacial charge transfer and diffusion processes [46]. Combined with the mass-transport calculations in Section 2.3, the theoretical limiting diffusion current density $i_{L,\mathrm{theory}}\approx 2.5\times {10}^{-5}\mathrm{A}{\mathrm{cm}}^{-2}$ is of the same order of magnitude as the average cathodic current density on the FCC2 phase, $i_{c,\mathrm{FCC}2}\approx 1.88\times {10}^{-5}\mathrm{A}{\mathrm{cm}}^{-2}$, indicating that the oxygen reduction process on the Cu-rich cathodic phase has approached the oxygen-supply-limited regime. Meanwhile, the non-ideal dispersive response observed in the low-frequency region of EIS further supports this interpretation, indicating that charge transfer and oxygen mass transport jointly participate in the overall corrosion reaction of the CoCrNiCu alloy [55]. Therefore, although the interphase potential difference can continuously provide the driving force for micro-galvanic corrosion, the overall corrosion rate of the system cannot increase in parallel with the strengthening local corrosion driving force. In other words, in this system, micro-galvanic action mainly controls the spatial distribution of localized corrosion, whereas the overall corrosion rate is further constrained by the oxygen availability at the cathodic sites.

The post-corrosion morphology and surface chemical state support the above interpretation. SEM results show that preferential dissolution is mainly concentrated in the regions adjacent to the Cu-rich phase. XPS further shows that the Cu-related regions are still dominated by metallic Cu and slightly oxidized Cu₂O, indicating that the Cu-rich phase retains high stability during corrosion and is therefore more likely to exist as the cathodic phase [17]. TEM further reveals that, in addition to the Cu-rich phase continuously distributed along grain boundaries, dispersed Cu-rich nanoscale precipitates are also present within the grains. The continuously distributed Cu-rich phase along grain boundaries is more likely to be the main carrier of micro-galvanic coupling and cathodic oxygen reduction, whereas the intragranular dispersed Cu-rich nanoscale precipitates may exert an auxiliary effect on the overall corrosion behavior by regulating the distribution of local cathodic sites [56], thereby contributing to local micro-galvanic effects. At the same time, these nanoscale Cu-rich precipitates may also influence the local composition and stability of the passive film, although their individual contribution to the corrosion behavior cannot yet be fully isolated based on the present results.

Overall, the corrosion process of the CoCrNiCu alloy can be summarized into several steps, as shown in Figure 10: The addition of Cu induces phase separation in the CoCrNi matrix, establishing a pronounced potential difference between Cu-depleted FCC1 and Cu-rich FCC2, thereby forming stable micro-galvanic coupling; meanwhile, anodic dissolution is concentrated within limited regions adjacent to phase boundaries, leading to amplification of the local current density and producing a distinct intergranular corrosion morphology; at the same time, the oxygen reduction process on the Cu-rich cathodic phase exhibits response characteristics consistent with cathodic oxygen-supply limitation, so that the overall corrosion rate is constrained by the oxygen availability at the cathodic sites [45]. Therefore, the corrosion behavior of the CoCrNiCu alloy is essentially not the result of Cu simply improving or degrading corrosion resistance. Rather, Cu reconstructs the microstructural heterogeneity and the spatial distribution of anodic and cathodic sites, so that localized corrosion distribution and the overall corrosion rate are governed by different controlling factors, namely micro-galvanic driving force for the former and cathodic oxygen-supply limitation for the latter [53,54,55]. This interpretive framework is also consistent with the current focus of related studies, namely, understanding the anomalous corrosion response and intrinsic mechanism of Cu-containing medium- and high-entropy alloys by considering Cu-content regulation, Cu-rich phase stability [57,58], and passive-film evolution together [59]. It should be noted, however, that the role of oxygen transport in the present work is still inferred mainly from indirect evidence, including electrochemical response, corrosion morphology, XPS results, and semi-quantitative mass-transport analysis, and will require direct verification by dedicated experiments under controlled oxygen concentration or hydrodynamic conditions in future work.

If the overall corrosion rate of the CoCrNiCu alloy is considered to be mainly constrained by the effective area of the Cu-rich cathodic phase and the diffusion-limited oxygen reduction process, then the macroscopic corrosion current can be approximately expressed as:

$$i_{\mathrm{c}\mathrm{o}\mathrm{r}\mathrm{r}}\approx f_{\mathrm{F}\mathrm{C}\mathrm{C}2}·i_{L,\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{o}\mathrm{r}\mathrm{y}}=f_{\mathrm{F}\mathrm{C}\mathrm{C}2}·{\displaystyle \frac{nF{D}_{O_2}{C}_{O_2}^{\ast }}{{\delta }_{\mathrm{d}\mathrm{i}\mathrm{f}\mathrm{f}}}}$$

(18)

where $f_{\mathrm{FCC}2}$ is the area fraction of the Cu-rich cathodic phase and $i_{\mathrm{L},\mathrm{theory}}$ is the limiting diffusion current density of the cathodic oxygen reduction reaction. This relationship indicates that, unlike the traditional corrosion-resistance mode mainly relying on passive-film barrier protection, the overall corrosion rate of this system no longer depends solely on the anodic dissolution capability, but is mainly governed jointly by the effective cathodic area and the upper limit of oxygen diffusion mass transport, reflecting the reconstructive effect of microstructural heterogeneity on the kinetic controlling step of corrosion.

4. Conclusions

The effect of Cu-induced phase separation on the corrosion behavior of CoCrNi-based medium-entropy alloys was investigated. The main conclusions are as follows:

(1) Cu addition to the CoCrNi-based alloy induced pronounced phase separation, leading to a dual-phase microstructure composed of a Cu-depleted FCC1 matrix and a Cu-rich FCC2 intergranular phase. The Cu-rich phase was distributed continuously along grain boundaries, and dispersed Cu-rich nanoscale precipitates were also present within the grains, providing the microstructural basis for electrochemical heterogeneity and micro-galvanic coupling.

(2) In 3.5 wt.% NaCl solution, the CoCrNiCu alloy exhibited a more negative $E_{\mathrm{corr}}$, a lower $i_{\mathrm{corr}}$, a higher $R_p$, and a pronounced high-impedance EIS response. These results indicate an anomalous electrochemical response, in which the corrosion driving force was enhanced whereas the overall corrosion rate remained restricted.

(3) Phase-potential calculations, post-corrosion surface observations, and XPS results showed that a pronounced interphase potential difference of about 0.337 V existed between the Cu-depleted FCC1 phase and the Cu-rich FCC2 phase. The Cu-rich FCC2 phase tended to act as the cathode, whereas the Cu-depleted FCC1 phase and the regions adjacent to the phase boundaries served as the main anodic dissolution sites, resulting in localized corrosion.

(4) The corrosion behavior of the CoCrNiCu alloy was governed by the combined effects of phase-separation-induced micro-galvanic coupling and cathodic oxygen-supply limitation. Equivalent-circuit comparison further indicated that its low-frequency response cannot be attributed to pure charge transfer alone, but is better interpreted as a coupled charge-transfer and oxygen mass-transport process. Cu addition modified the microstructural heterogeneity and the spatial distribution of anodic and cathodic sites, causing localized corrosion distribution and the overall corrosion rate to be governed by different controlling factors. This provides a new perspective for understanding the anomalous corrosion mechanism of CoCrNiCu-based high-entropy alloys.