2.1. Electrochemical impedance spectroscopy (EIS)

The electrodes were connected to the CH Instruments - Electrochemical Workstation by crocodile clips. The system was maintained for around 30 minutes to reach a steady state. To obtain the open circuit potential (OCP) values, the voltages between the working and reference electrodes were measured. 5 mV amplitude and 100 kHz to 0.2 Hz frequency range were set for the sinusoidal voltage. Then the impedances for this range of frequencies were measured by the potentiostat at the open circuit potential. The EIS data were obtained after 30 minutes of stabilization for each sample. These impedance data were fitted to the appropriate electrical circuit in the program EC-Lab-Analyst. The circuit used in this study is called the Randles circuit. By modelling with this circuit, the system characteristics like, R _p =polarization resistance (Ω), R _S = solution resistance (Ω), and Cp(eff)= electrochemical double-layer capacitance (μF) were obtained.

2.2. Potentiodynamic polarization technique

For potentiodynamic polarization analysis, the same experimental setup was used. The electrode potential was changed from - 1 to +1 V with a scan rate of 0.5 mVs^-1, and the corresponding corrosion current was measured. The formula employed for finding out the corrosion rate is,

where, Icorr corrosion current (A), K the constant which defines the corrosion rate units in mm year^-1, where K = 3272 mm (A cm year) ^-1 according to ASTM standard G 102), EW the equivalent weight (g/equivalent), ρ the density in (gcm^-3), and A the surface area in (cm²).

The microstructural images of polished surfaces of the alloys prior to and after corrosion were captured by OPTIKA, a typical optical microscope. In the case of the conventional optical microstructure of the aged alloys, the surfaces were metallographically etched by a typically suggested copper etchant, one of ammonium hydroxide + hydrogen peroxide (3%) in a 1:1 ratio. The SEM and EDX investigations were carried out using a JEOL scanning electron microscope with an attached X-ray analyzer.

3.1. Physical properties

The effects of Al and Zn content by 10 wt% each on the physical and mechanical properties of Cu, alloy experimental results are exposed in the following two charts. Figure 1 shows the physical properties of the alloys related to hardness and density at peak aged conditions. It is observable from the figure that, at peak aged conditions, both Cu-10Al and Cu-10Zn have higher hardness values than pure Cu. As solution strengthening occurs through the dissolved Zn and Al atoms in the alloys. As a result, α-brass and Al-bronze have higher hardness compared to Cu [¹⁵,¹⁶]. Again, Al-bronze has superior hardness to Zn-bearing α-brass. During age hardening, five stable intermetallics of Al and Cu are formed: Cu₉Al₄, Cu₃Al₂, Cu₄Al₃, CuAl and CuAl₂. Among them, two specials are Cu₉Al₄ and Cu₃Al₂, which cause higher hardness in the Cu-10Al alloy [¹⁷,¹⁸]. These stable intermetallic phases have been identified in Cu-Al alloys in several previous investigations [¹⁹-²¹]. When the concentration of Zn is below 35 wt.%, the phase of brass is α, which has a substitutional and disordered face-centred-cubic (fcc) structure. As aresult, Cu-10Zn shows a single fcc state of α-phase with no discernible age-hardening effects, which makes Cu-10Zn alloy softer than Cu-10Al [²², ²³]. Additionally, in Fig. 1 depicts the variation of density among the experimental alloys. Pure copper has the highest density, whereas Al-bronze is the lightest, as the densities of Cu, Zn, and Al follow this sequence: Cu>Zn>Al. It can be seen from Table I. that Al and Zn are present in the alloys by 10 wt.%.

Figure 1 Variation of hardness and density of the pure Cu, Al and Zn added Cu alloys. 

3.2. Mechanical properties

Figure 2 shows the mechanical properties of the pure copper, Al and Zn added Cu alloys in terms of tensile strength and % elongation. It is noted from the figure that both the Al-bronze (Cu-10Al) and the α-brass (Cu-10Zn) show better tensile strength but reduced ductility than Cu. Previous studies suggest that adding some alloying elements to copper improves tensile strength [²⁴]. The results of the study support that zinc-added copper has more tensile strength and less % elongation than pure copper itself [²⁵]. The higher content of zinc in Cu-10Zn increases brittleness [²⁴]. The solid solution hardening of the α-phase of brass during the solidification process is responsible for this increased strength and reduced ductility. Again, the intermetallics formed during the solidification and ageing of Cu-10Al alloy, specially Cu₉Al₄ and Cu₃Al₂ intermetallics, increase the tensile strength and brittleness [¹⁷]. These precipitated particles hinder the dislocation movement, which leads to high tensile strength, but low ductility [¹⁸]. Furthermore, it is noticed from the figure that the Al added Cu has higher tensile strength and lower ductility than Cu-10Zn. The solid solution strengthening of Zn added Cu is not more prominent than Al added Cu alloy since Al-bronze consists of different intermetallics.

Figure 2 Variation of tensile strength and % elongation of the pure Cu, Al and Zn added Cu alloys. 

3.3. Impedance spectroscopy studies

The impedances were generated for variable frequencies between 100 kHz to 0.2 Hz by the potentiostat, CH Instruments - Electrochemical Workstation. The impedances were modeled with the electrical circuit in Fig. 3 to get the values of corresponding circuit components. The circuit components are solution resistance R _S , polarization resistance R _p , which is equivalent to charge transfer resistance Rct [²⁶], and effective double layer capacitance Cp(eff), the values of which are presented in Table II for the respective alloys. X² denotes the chi-squared value of goodness of fit. It has been previously found that the polarization resistance changes inversely with the corrosion rate of a metal, indicating the reactivity rate of the surface with the environment [²⁷,²⁸]. The R _p and Rct are equivalent as they both indicate corrosion resistance.

Figure 3 The equivalent electrical circuit for the impedance data. 

From Table II, the observed phenomenon is that the OCP shifts towards a negative direction as alloying elements are added into the Cu, but the open circuit potential value of Zn added Cu alloy falls below Al added Cu alloy. The more positive open circuit potential expresses better corrosion resistance of metal [²⁹]. Again, R _p is higher for pure copper rather than Al and Zn added alloys, and α-brass has the lowest value of R _p . This trend suggests that α-brass has the lowest corrosion resistance, while pure copper has the best corrosion protection among them. The addition of Al and Zn degrades the corrosion performance of Cu alloy as dealloying of Cu-Al and Cu-Zn occurs on the alloy surfaces due to the selective dissolution of Al and Zn. Moreover, the Cu-Zn solid solution is more prone to dealloying than Cu-Al solid solutions, which is why Zn-added α-brass shows the worst corrosion resistance [³⁰]. Normally Al-bronze and α-brass are involved to produce a compact film composed of Al₂O₃ and ZnO along with Cu₂O, where Al-related oxide is more stable than that of Zn [³⁰,³¹]. Pure copper shows the best performance as Cu is susceptible to forming cuprous oxide (Cu₂O) protective layer or to dope the film with the alloying elements’ ions such as iron, and each of the experimental alloys has a trace amount of iron in it [³²-³⁴]. Copper, when immersed in NaCl solution, forms Cu₂O by this reaction [³²,³⁵]:

But, when Cu bonds with Al and Zn to produce solid solution strengthening and age hardening particles, they do not let Cu to easily form oxide. Zn forms a zinc hydroxy chloride (Zn₅ (OH)₈ Cl₂) surface film by the following reaction [³⁶]:

Again, Assaf, et al. [³⁶], investigated that passive zinc oxide film is formed by the following reaction:

The aluminium oxide formation is shown in the equation below [³⁷]:

However, the values of R _S show a decreasing trend from Cu to Al and Zn added alloys in the table, and Cp(eff) shows the opposite trend. As Rs dominates the total path resistance, it follows a similar trend as R _p [³⁸].

Figure 4 displays the Nyquist diagram for the copper alloys. This diagram follows the capacitive-resistive semicircle model, which provides the real (Zr) and imaginary (Zi) parts of the impedances, respectively. The charge transfer process regulates the electrochemical corrosion of pure copper and copper alloys, as indicated by a full semicircular shape on the Nyquist diagram [³⁹].

Figure 4 Nyquist diagram for the pure Cu, Al and Zn added Cu alloys. 

The Bode magnitude and phase plots are shown in Figs. 5a) and b), respectively. At the y-axis of the Bode magnitude plot, the modulus of impedance (|Z|/Ω) is presented, whereas the negative values of phase angles (-φ/º) are denoted at the y-axis of the Bode phase plot. On the other hand, the x-axes of both the Bode magnitude and phase plots denote the frequency (f/Hz). The impedance can be determined by the following equation:

Figure 5 Bode a) magnitude, and b) phase diagrams of the pure Cu, Al and Zn added Cu alloys. 

From the equation, it is known that at high frequency, the right term of the equation gets close to zero, and the impedance is close to the solution resistance. Figure 5a) shows a similar trend. The Bode magnitude plot for pure Cu is situated higher than Al-bronze, where the curve for α-brass lies lower than the alloy with a higher amount of Al, which can be explained by Eq. (6) and Table II. From the table, both R _S and R _p are higher in Al-bronze than α-brass, the pure Cu having the maximum value. For higher values of R _S and R _p , the impedance must be higher, according to Eq. (5). In Fig. 5b), no phase shift is there at lower and higher frequencies, indicating a resistive response. On the other hand, at mid-range frequencies, the phase shifts are closer to -90, which is a capacitive response.

3.4. Potentiodynamic polarization analysis

Figure 6 shows the potentiodynamic polarization curves obtained under scan rate of 0.5 mVs^-1 applying electrode potentials from -1 V to +1 V, for the commercially pure Cu, Al added Cu alloy as Al-bronze, and Zn added Cu alloy as α-brass. The experimental results of the potentiodynamic polarization analyses obtained through Eq. (1) are furnished in Table III. The corrosion current is expressed by Icorr, the corrosion potential or rest potential is expressed by Ecorr, and the corrosion rate is denoted in mm/year unit. Inspecting the Tafel plots, it is noticed that the slope of the curve for pure Cu is higher than the other two curves for the Zn and Al-added Cu alloys. The higher the slope, the slower the corrosion should take place, as the change in corrosion current is smaller [⁴⁰].

Figure 6 Tafel polarization curves for the pure Cu, Al and Zn added Cu alloys. 

The experimental data suggest that the corrosion current and corrosion rate are higher in Al, and Zn added alloys, where Zn added alloy showing the maximum corrosion. This can be interpreted as the pure copper exhibits the best corrosion resistance, whereas the α-brass depicting the worst resistance followed by Al-bronze, which behavior is like the EIS analysis.

3.5. Optical micrographic observation

Figure 7 presents the optical micrographs of Cu, Cu-10Al and Cu-10Zn alloys prior to and post corrosion process in 0.1 M NaCl solution. The surfaces of the alloys were highly polished with emery papers for getting smooth surfaces. There are slight scratches due to polishing. Although not very much difference can be observed from this type of unetched microstructures, but as there were different alloying elements present, a varied tone can be detected in the alloys. Pure copper shows orange-tinted red, Al-bronze exhibits dull gold tones, and α-brass shows reddish gold before corrosion in Figs. 7a1) to 7c1). Again, some differences are discernible among the micrographs of the metal surfaces after corrosion in Figs. 7a2) to 7c2) and no polished marks are visible. A uniformly distributed corroded surface is found in Fig. 7a2), as there is no intermetallic formed in pure copper. However, different intermetallics of copper aluminates associated with trace impurities cause localized pitting corrosion in Cu-10Al alloy, which can be observed in Fig. 7b2) [⁴¹]. Figure 7c2) shows an extensive amount of pitting corrosion caused by the Zn solid solution, as well as trace impurities into the matrix and the preferential dissolution in the case of Cu-10Zn alloy [⁴²]. As Cu and Zn have a large difference in equilibrium potential difference, selective dissolution of Zn occurs. It has been previously inspected that both physical and surface phases were remarkably changed due to anodic dissolution or free corrosion of brass [⁴³]. In the micrographs of corroded surfaces, pure copper is turned into yellow and green colour as typical of Cu₂O and CuO, which is visible in Fig. 7a2) [⁴⁴].

Figure 7 Optical micrographic images of the surfaces of pure Cu, Al and Zn added Cu alloys before and after corrosion in 0.1 M NaCl solution. 

There is a film of Al₂O₃ on the surface of Cu-10Al in Fig. 7b2), which is white in general. The ZnO together with CuO shows the white and green tone on the surface of Cu-10Zn alloy in Fig. 7c2) [⁴⁵]. The micrographs also display the highest damaged surface of the Zn added alloys and followed by Al-added alloys and then pure Cu. These results also comply with the experimental results of the corrosion rate of the experimental alloys.

The conventional optical microstructures of the etched pure Cu, Al and Zn added Cu alloys are shown in Fig. 8, after cold rolling and peak aged conditions. Due to rolling by 40% deformation, elongated and broken grains are formed in the direction of rolling. The grains become elongated in the direction of the applied stress, irreversibly changing the microstructure during cold working. The distribution of the size of the Cu grains is not uniform, and there are some smaller grains visible in the middle of the coarse grains, where several twins also exist within the grains [Fig.8a)] [⁴⁶]. Different phases such as α-phase, several κ-phases, and retained β' phase are there in Al-bronze. The specific phases also refine the grain structure [⁴⁷]. Cu-Zn alloy shows a single-phase fcc condition because the level of Zn content is below 35 wt.%. The single-phase brass is formed when Zn is melted into Cu and develops a uniform crystal structure, revealed by the Cu-10Zn alloy’s resultant microstructures [Fig. 8c)]. Similar observations are also made by the earlier investigators [⁴⁸].

Figure 8 Optical microstructures of the etched samples of pure Cu, Al and Zn added Cu alloys. 

3.6. SEM and EDX observation

The surface micrographs of the corroded surfaces of the pure Cu, Al-bronze and α-brass using SEM are presented in Fig. 9.

Figure 9 SEM images of a) pure Cu, b) Al added Cu, c) Zn added Cu alloys after corrosion in 0.1 M sodium chloride medium. 

The uneven surface topography is visible due to the formation and demolishment of oxide layers for each sample. It is conspicuous that the dissolution of pure copper is homogeneous, which is depicted in Fig. 9a), whereas there is selective dissolution in the case of Al-added Cu and Zn-added Cu in Figs. 9b) and c). Furthermore, surface defects such as microcracks and pinholes are visible, which are, to a greater extent, on the surfaces of Al and Zn-added Cu alloys. The pinholes are the effects of pitting corrosion which seems excessively on the surface of Cu-10Zn alloy, caused by the Cu-Zn solid solution matrix and the preferential dissolution as discussed previously. One significant observation is that Zn bearing alloy surface shows thick oxide layers with internal massive damage. The pitting potentials in Cu-Al and Cu-Zn alloys are higher than in pure copper, which may be due to the presence of the mixed passive films Cu₂O and Al₂O₃ in Cu-Al alloy and Cu₂O and ZnO in Cu-Zn alloy [³⁰, ³¹]. The dissolution rate of oxide film ZnO is higher and not too stable than that of Al₂O₃.

According to the quantitative EDX analysis as tabulated in Table IV, the alloys show following chemical compositions in wt.%:

The compositions can be interpreted as the wt.% of Al is reduced from 9.46% to 5.45% in Al-bronze after corrosion. On the other hand, the wt.% of Zn is reduced during corrosion from 10.3% to 1.61%, depicting that dissolution of Zn bearing particles in α-brass is more prominent than Al bearing particles in Al-bronze. During the corrosion process, the element Zn and Al unalloyed form the α-brass and Al-bronze, which dissolved easily. As a result, such lower levels of Zn remain in the alloy. The percentages of oxygen in every alloy indicate the layers of oxides formed.