*3.1. Potentiodynamic Polarization*

Figure 1 presents Tafel polarization curves for the Cu5Zn5Al1Sn alloy exposed to 3.5 wt.% NaCl solution with and without the Cys inhibitor. The trend of polarization curves suggests that a low amount addition of cysteine (10−<sup>4</sup> M, 10−<sup>5</sup> M) in 3.5 wt.% NaCl solution moves the curves to the lower current region in the anodic potential route compared with the blank solution. Nevertheless, the higher concentration (10−<sup>3</sup> M, 10−<sup>2</sup> M) of the inhibitor in solution causes a shift of the curves to a more cathodic potential region and decreases the corrosion current meaningfully. Consequently, cysteine influences both the cathodic and anodic reactions. Hence, cysteine is shown to act as a mixed-type inhibitor to the Cu5Zn5Al1Sn alloy in 3.5 wt.% NaCl and to retard majorly cathodic corrosion reactions [22–25]. The corrosion current density (Icorr) and corrosion potential (Ecorr) were calculated by fitting the experimental results with Gamry Echem Analyst software 5.61 and listed in Table 1. The inhibition efficiency (μ%) was also obtained by Equation (1) using parameters listed in Table 1 [26].

$$
\mu\% = \frac{\text{Icorr} - \text{Icorr}(\text{inh})}{\text{Icorr}} \times 100\% \tag{1}
$$

**Figure 1.** Potentiodynamic polarization curves for the Cu5Zn5Al1Sn alloy after 60 min immersion in 3.5 wt.% NaCl in the absence (Blank) and presence of different concentrations of cysteine (10−<sup>5</sup> to <sup>10</sup>−2).


**Table 1.** Polarization parameters for the Cu5Zn5Al1Sn alloy measured in 3.5 wt.% NaCl solution in the absence and presence of di fferent concentrations of cysteine.

In this equation Icorr (inh) and Icorr are the current densities of the working electrode in the 3.5 wt.% NaCl solution with and without cysteine, respectively.

The Icorr is inversely related to the inhibitor concentration, i.e., an increase in inhibitor concentration results in a lower value of Icorr. This is probably due to the improved protection of the Cu5Zn5Al1Sn alloy by cysteine molecules. Therefore, the mitigation of the Cu5Zn5Al1Sn alloy deterioration in 3.5 wt.% NaCl solution may associated with the development of a protective layer from the adsorbed inhibitor species on the alloy surface [27]. The adsorption layer of cysteine molecules on the alloy surface can hinder the movement of corrosive species, resulting in a reduction in corrosion rate [2,23,24,28,29].

It is evident from Table 1 that both βc and βa are varied on the addition of the inhibitor. From this behavior, it is suggested that both anodic and cathodic reactions are retarded by the protective layer formed by inhibitor molecules and the alloy/electrolyte interface [3,23,29]. In addition, βc values are greater than βa values at all inhibitor concentrations and the variation in βc with the inhibitor concentration is greater than the variations in βa. Both the results give the same evidence that cysteine is more e fficient in inhibition of cathodic reactions than that of anodic reactions [3,30]. The e fficiency of cysteine increases with its concentration and reaches a maximum value of 97.2% at the concentration 10−<sup>2</sup> M.

#### *3.2. Electrochemical Impedance Spectroscopy*

Nyquist and Bode diagrams of the Cu5Zn5Al1Sn alloy in 3.5 wt.% NaCl solution with and without di fferent amounts of cysteine are illustrated in Figure 2a,b, correspondingly. From Figure 2a, the Nyquist diagram for Cu5Zn5Al1Sn alloy in 3.5 wt.% NaCl solution contains a capacitive loop at intermediate frequency and a straight line in the low frequency region. A straight line at low frequency area may be attributed to the presence of the Walberg constant as a result of the movement of soluble metal species from the alloy surface to the bulk solution [23]. The diameter of the capacitive loop rises in the presence of Cys and the straight line in the low frequency region is removed at higher Cys concentrations, which may be attributed to the inhibitive e ffect of Cys on the corrosion process [31]. The results of Bode plots (Figure 2b) show that the frequency range with maximum phase becomes larger with increasing Cys concentration. There is a shift in the phase maximum to a lower frequency region and the phase angle is raised to about 80◦, in the inhibitor concentration of 10−2 M. These results sugges<sup>t</sup> that the protective ability of the Cu5Zn5Al1Sn alloy is increased with inhibitor concentration, showing that inhibitor particles e ffectively adsorb on the alloy surface [32,33].

The results of the EIS experiment were fitted with a simulated circuit (Figure 3) and the value of electrical elements that make up the equivalent circuit are presented in Table 2. The equivalent circuit parameters are solution resistance (Rs), film resistance (Rf), film capacitance (Qf), charge transfer resistance (Rct), double layer capacitance (Qdl), and Warburg resistance (W), which result from ionic di ffusion of the corrosion product to the electrode surface. The EIS fitting result of the Cu5Zn5Al1Sn alloy in blank 3.5%wt NaCl solution shows a Warburg resistance and a lower value of Rct, as shown in Figure 2a and Table 2, respectively. The Warburg impedance is included to account for the di ffusion of corrosion products to the bulk solution and/or dissolved oxygen to the electrode surface. With a lower Cys concentration, the Warburg impedance remains unchanged, and Rf and Qf are observed to address the film formed between the Cys molecules and alloying elements. The presence of W is attributed to the low inhibition efficiency of the lower concentration of Cys, which may be related to the thinner and unstable film formed which can be easily dissolved to the bulk solution. When more than 10−<sup>4</sup> M Cys is added to the solution, the Warburg impedances vanish and the Rct values become larger. This shows the presence of high resistance to transfer charge from bulk solution to specimen surface and/or and specimen surface to bulk solution; hence, the diffusion process is controlled. This may be attributed to the development of a thick and protective Cys/metal film on the alloy surface.

**Figure 2.** Nyquist (**a**) and Bode (**b**) plots for the Cu5Zn5Al1Sn alloy in 3.5 wt.% NaCl with and without different concentrations of cysteine (10−<sup>5</sup> to <sup>10</sup>−2) (scattered: experimental data; lines: fitting data).

**Figure 3.** Electrical equivalence circuit for Cu5Zn5Al1Sn alloy in 3.5 wt.% NaCl without R(Q(RW)) (**a**), with 10−<sup>5</sup> M Cys, R(Q(R(Q(RW)))) (**b**) and 10−4–10−<sup>2</sup> M Cys, R(Q(R(QR))) (**c**). Legend: solution resistance (Rs), film resistance (Rf), film capacitance (Qf), charge transfer resistance (Rct), double layer capacitance (Qdl), and Warburg resistance (W).

**.** 

**Table 2.** Impedance parameters for the Cu5Zn5Al1Sn alloy in 3.5 wt.% NaCl with and without different concentrations of cysteine obtained by using ZSimpWin 3.50.


Inhibition efficiency was obtained by using Equation (2) [34–36], the values of which are listed in Table 2. The inhibition efficiency of Cys for Cu5Zn5Al1Sn alloy corrosion in 3.5 wt.% solution is seen to increases with its concentration, and a maximum inhibition efficiency value of 96.44% is recorded for the case of the Cys concentration being 10−<sup>2</sup> M. The inhibition efficiencies obtained from polarization and EIS investigations are similar and follow the same trend, which implies that the explanation for Cu5Zn5Al1Sn alloy corrosion characteristics on the basis of Icorr and Rp could be well-thought-out as valid and reliable.

$$\mu\% = \frac{\text{Rp} (\text{inh}) - \text{Rp}}{\text{Rp} (\text{inh})} \times 100\% \tag{2}$$

where μ, Rp, and Rp (inh) are inhibition efficiency, and polarization resistance for uninhibited and inhibited surfaces, respectively.

The inhibition efficiency results obtained from potentiodynamic (PD) and EIS experiments were used to observe the adsorption behavior of cysteine on the Cu5Z5Al1Sn alloy surface. Therefore, different adsorption isotherms were tested to determine the best suitable model to explain surface phenomena [23,37,38].

$$\frac{\text{C}}{\text{\(\frac{\text{C}}{\text{\(\text{Sads}}}\text{\(\text{\(\text{Sads}}}\text{\)}\text{\(\text{\(\text{Sads}}}\text{\)}\text{\(\text{\(\text{Sads}}}\text{\)}\text{\(\text{\(\text{Sads}}}\text{\)}\text{\(\text{\(\text{Sads}}}\text{\)}\text{\)}}\text{\)}}$$

$$\Theta = \frac{\text{RT}}{\text{b}} \text{lnKads} + \frac{\text{RT}}{\text{b}} \text{lnC} \tag{4}$$

In this equation, C, θ, Kads, and b are inhibitor concentration, the inhibitor's surface coverage, the equilibrium constant related to the interfacial molecular adsorption-desorption phenomena, and a constant relevant to the properties of both adsorbent and adsorbate in the Temkin model, respectively.

Figure 4 shows plots of Langmuir and Temkin adsorption isotherms determined by the respective equations listed in Equations (3) and (4), respectively. As observed in Figure 4a, a slope of 0.99 with R<sup>2</sup> = 0.99 derived from EIS and PD results which is close to unity can be observed from a C/θ versus C linear plot for the case of the Langmuir adsorption isotherm. On the other hand, a regression of 0.89 and 0.83 acquired from EIS and PDP, respectively, was obtained from the plot of θ versus ln C based on the Temkin adsorption isotherm (Figure 4b). By comparing the two isotherms, the Langmuir adsorption isotherm, which was seen to have less fitting error, was chosen as the more suitable model to describe the surface inhibition mechanism of cysteine for Cu5Z5Al1Sn alloy corrosion.

**Figure 4.** Langmuir (**a**) and Tamkin (**b**) adsorption plots of Cu5Z5Al1Sn alloy in 3.5 wt.% NaCl in the presence of the addition of different concentrations of cysteine obtained by EIS and PD experiments.

## *3.3. SEM Analysis*

Further analysis was performed to examine the surface morphology and elemental characterizations and their effects on corrosion behavior. Figure 5 illustrates the SEM images of the Cu5Zn5Al1Sn alloy exposed to 3.5 wt.% NaCl solution without and with different amounts of Cys. However, the uninhibited copper alloy surface (Figure 5b) is highly corroded and becomes rough as a result of the aggressive attack the from corroding solution. A very different surface morphology is observed in the presence

of 10−<sup>4</sup> M Cys as, shown in Figure 5c. The inhibitor molecules, which partly cover the surface, are seen to be formed. By contrast, Figure 5d does not show any corrosion attack and has almost the same morphology as the unexposed surface, suggesting that the addition of 10−<sup>2</sup> M Cys leads to a more inhibited corrosion, the surface of which is almost the same as an unexposed polished one (Figure 5a). Thus, it is obvious that corrosive attack is considerably restricted by 10−<sup>2</sup> M Cys. It is assumed that inhibitor molecules adsorb on the alloy surface and a smoother surface forms compared to the surface treated with the blank 3.5 wt.% NaCl solution.

**Figure 5.** SEM images of Cu-5Zn-5Al-1Sn alloy before (**a**) and after immersion in 3.5 wt.% NaCl, without (**b**) and with 10−<sup>4</sup> M of cysteine (**c**) and 10−<sup>2</sup> M cysteine (**d**), respectively.

#### *3.4. XPS and Auger Results*

In order to further understand the surface composition of the uninhibited and inhibited copper alloy, high-resolution XPS Cu2p and Auger CuLM2 spectra of the Cu5Zn5Al1Sn alloy surface without and with the addition of various concentrations of cysteine were recorded and are shown in Figure 6. The Cu 2p profiles of Cu(0) and Cu(I) are similar, and as a result it is difficult to use the XPS Cu 2p spectrum alone to differentiate them, whereas the binding energy of the Auger peak of Cu(I) is about 2 eV higher than the Auger peak of Cu(0), meaning the CuLM2 spectrum is necessary to further investigate the valence state of copper. In all cases, there are two deconvoluted peaks of Cu2p 3/2 at different binding energies (Figure 6a). The interpretation of the peaks is different as the reaction taking place on the alloy surface is not same in the blank and Cys-inhibited solutions. In the blank solution, the corrosion products may exist as the form of oxides by reactions with the dissolved oxygen. While in the presence of inhibitors, the inhibitor particles have the opportunity to participate in surface reaction. In blank 3.5 wt.% NaCl solution, the deconvoluted Cu2p peaks at 932.99 and 934.83 eV (Figure 6a) may be assigned to Cu/Cu+ and Cu2+, respectively. In addition to the XPS result, the Auger spectrum obtained in the blank solution shows an Auger peak at 570.11 eV, which is a typical characteristic of cuprous ion (Cu+), with a band broadening to the lower binding energy. The band broadening is an indication of the presence of metallic copper, which may be attributed to the uncovered alloy surface. Generally, the XPS and Auger results of the blank solution show the presence of Cu, Cu<sup>+</sup>, and Cu2+ species on the alloy/solution interface, which may be attributed to the uncovered alloy surface. With regard to the Cu2p 3/2 obtained after immersion in solution with 10−<sup>4</sup> M Cys, the deconvoluted peaks are located at 932.53 and 934.41 eV, respectively. Similarly, when the surface is inhibited with 10−<sup>2</sup> M Cys, the Cu2p 3/2 peak is deconvoluted into peaks located at 932.62 and 934.10 eV. The peaks may be assigned to Cu/Cu+ and Cu2+ ions, respectively, in both cases [39–42]. In the solution containing different concentrations of cysteine (10−4and 10−<sup>2</sup> M), the Auger peaks obtained at 571.05 and 571.02 eV may both be assigned to Cu<sup>+</sup>. The binding energy of the Cu+ peak on the Auger spectra increases in the presence of Cys, which may be associated with the formation of a film containing organic inhibitor molecules typically centered at higher binding energies than the binding energy for the oxide species [43–47]. Both in the absence and presence of Cys, the Cu2p XPS spectra demonstrate the Cu (II) satellite peaks attributed to the presence of copper (II). Furthermore, there is no Auger peak related to Cu2+ other than a band broadening in the blank solution, which may be ascribed to the lower content of the Cu2+ species. The uninhibited alloy surface and that of the inhibited one with lower Cys concentration (10−<sup>4</sup> M) have a higher peak intensity, which could be attributed to the poor coverage area of corrosion products and the metal/inhibitor film, respectively [48].

**Figure 6.** High-resolution XPS Cu2p 3/2 spectra (**a**) and Auger spectra (**b**) of the Cu5Zn5Al1Sn alloy surface without and with addition of inhibitors in 3.5 wt.% NaCl. Legend: BE, binding energy.

High resolution XPS Zn2p spectra (Figure 7a) show two peaks for the alloy surface, both without and with different concentrations of Cys, which may be assigned to Zn2p 3/2 and Zn2p 1/2. The Zn2p 3/2 peak at 1021.75 and 1021.76 eV in the blank and 10−<sup>4</sup> Cys inhibited surfaces may be assigned to Zn2+, which is associated with the presence of corrosion product [49–51]. However, the peak observed on the surface inhibited with 10−<sup>2</sup> M Cys is almost negligible due to low Zn2<sup>+</sup> species on the alloy surface. This phenomenon indicates that the alloy surface is not fully protected by inhibitor molecules in the presence of lower cysteine concentration. Obviously, the dezincification rate of the Cu5Zn5Sn1Al alloy is strongly bounded by the surface film/inhibitor film of the alloy surface. A more dense/integrated surface film leads to a lower dezincification rate. It can also be deduced that the formed inhibited film may rise from other alloying metals of the Cu5Zn5Sn1Al alloy, which will probably be a copper-cysteine film.

The Sn3d peak may be deconvoluted into two different peaks: Sn5/2 and Sn3/2. The peaks centered at 486.65 eV, 486.77 eV, and 486.58 eV correspond to the Sn5/2 peaks, which were obtained in blank solution, with 10−<sup>4</sup> M Cys and with 10−<sup>2</sup> M Cys, respectively (Figure 7b). In all cases, the peak values correspond to ionic tin (Sn<sup>4</sup>+) [39,52–56], which may indicate the formation of the compound of Sn on the alloy surface, meaning, therefore, the contribution of Sn2+ to the surface film should not be ignored. In addition to the main Sn3d peaks, satellite peaks at higher binding energies of 499.05 and 499.08 eV appear for the blank and the surface-inhibited surface with lower Cys concentration (10−<sup>4</sup> M), respectively, which may be attributed to the presence of more tin ion contents on the alloy surface [57]. In the meantime, the atomic ratio shown in Table 3 reveals that the inhibitor free alloy surface contains a greater amount of tin. On the other hand, the alloy surface inhibited with 10−<sup>2</sup> M Cys contains an almost insignificant amount of SnO2, which may result in the development of the inhibitor film and prevent the formation of the corresponding metal oxide.

**Figure 7.** High-resolution XPS of Zn2p (**a**) and Sn3d (**b**) spectra of Cu5Zn5Al1Sn alloy surface without and with addition of inhibitors in 3.5 wt.% NaCl solution.

**Table 3.** Composition (atomic %) of the surface of Cu5Zn5Al1Sn alloy samples before and after the addition of inhibitors in a 3.5 wt.% NaCl solution.


The deconvoluted C1s spectrum illustrated in Figure 8a for the golden alloy in the absence of cysteine may be attributed to three peaks which indicate three chemical forms of the C element present on the alloy surface. The three peaks described above are located at 284.75, 285.73, and 288.59 eV. The component at 284.6 eV is assigned to the non-oxidized carbon containing (C–C) composition. Moreover, the features at BE values of 285.73 eV and 288.59 eV can be assigned to the groups containing a carbon–oxygen bond, i.e., a C–O group like ether or hydroxyl, and adventitious carbon which is usually found on the metal surfaces and results from adsorbed oxidized carbonaceous species from the atmosphere, respectively [44,45,58]. For the case involving the presence of 10−<sup>4</sup> M cysteine, the C1s spectrum (Figure 7a) also shows three peaks at 288.54, 286.38, and 284.88 eV, and the three deconvoluted peaks for the case of 10−<sup>2</sup> M Cys are located at 288.68, 286.21, and 284.66 eV, which match with the COO<sup>−</sup>, C−NH2, and C− SH groups, respectively [39,40]. The increment of the atomic ratio of carbon (Table 3) of the 10−<sup>2</sup> M Cys inhibited surface can be ascribed to the existence of the high amount of cysteine on the alloy surface.

**Figure 8.** High-resolution XPS C1s (**a**), O1s (**b**), S2p, and N1s spectra of the Cu5Zn5Al1Sn alloy surface before and after addition of inhibitors in 3.5 wt.% NaCl solution (**<sup>c</sup>**,**d**).

The deconvoluted O1s spectra of the Cu5Zn5Al1Sn alloy in the 3.5 wt.% NaCl solution without and with the Cys inhibitor are shown in Figure 8b. There are three distinct peaks for the blank alloy surface in 3.5 wt.% NaCl solution, with the first peak located at 530.74 eV being attributed to <sup>O</sup>2<sup>−</sup>, which could be related to oxygen atoms bound to the constituent metal oxides [48,59,60]. Since the major corrosion product on the outermost part of the alloy is derived from copper, the dominant metal oxide relies on copper [8], whereas the contribution of other constituent metals (Zn, Sn, and Al) cannot be ignored. The presence of Zn and Sn oxides are confirmed by their respective XPS spectrum. On the other hand, it is impossible to obtain peak analysis for Al (neither Al2p nor Al2s can be used) because of the peak overlapping with Cu3p and Cu3s [8]. The second peak observed at 531.57 eV may be ascribed to OH<sup>−</sup>, which can be associated with the occurrence of hydrous copper, zinc, tin, and aluminum oxides.

The third peak at 532.33 eV may be assigned to atmospheric oxygen, which is similar to the carbon contamination [42]. With the addition of 10−<sup>4</sup> M of Cys, the O1s peak is deconvoluted into three peaks located at 530.74, 531.48, and 532.09 eV, which could be assigned to metal oxides, hydrous metal oxides, and C=O or –CON arising from cysteine molecules, respectively. The presence of metal oxide can be ascribed to partial coverage of the metal surface by the inhibitor film [42,61,62]. The O1s spectrum of the sample containing 10−<sup>2</sup> M Cys may be deconvoluted into three peaks at 531.12, 531.87, and 532.79 eV, which can be attributed to C-O from carboxyl and C=O or -CON, respectively [39]. In contrast to the uninhibited and the 10−<sup>4</sup> M Cys inhibited alloy surface, the alloy inhibited with the high Cys (10−<sup>2</sup> M) concentration shows no peak associated with the presence of metal oxide. It is

indicated that the alloy surface is fully covered by the inhibitor film after the addition of 10−<sup>2</sup> M cysteine [63].

Figure 8c shows the deconvoluted S2p spectrum of Cu5Zn5Al1Sn alloy in the absence and presence of cysteine. There is no peak found on the S2p spectrum of the blank Cu5Zn5Al1Sn alloy in 3.5 wt.% NaCl attributed to the nonappearance of any sulfur species. In the presence of 10−<sup>4</sup> M Cys, two peaks located at 162.95 and 167.95 eV can be recognized, and for the case of 10−<sup>2</sup> M Cys, the three deconvoluted peaks are located at 162.45, 163.42, and 168.1 eV. The lower binding energies (162.95 and 162.45) may be assigned to the metal–sulfur interaction and the upper ones (167.95 and 168.10) are ascribed to the -SH group of the cysteine molecule [42,61,63,64]. The deconvoluted peaks at 162.45 and 163.42 eV obtained from the 10−<sup>2</sup> M Cys inhibited surface result from band broadening and because of the presence of more sulfur content they are considered to be peaks derived from the same component [43,65]. The intensity of S2p peaks and percentage ratio of S (Table 3) increases with cysteine concentration, suggesting that the efficiency of cysteine is concentration-dependent and more Cys molecules are adsorbed on the Cu5Zn5Al1Sn alloy surface at the higher inhibitor concentration.

The N1s spectrum for the uninhibited/inhibited surface has only one deconvoluted peak, as shown in Figure 8d, for all cases. The peak at 400.02 eV obtained from the blank solution can be assigned to N raised from the atmosphere during sample preparation. The peaks at 399.65 and 399.37 eV are obtained from 10−<sup>4</sup> M and 10−<sup>2</sup> M, respectively, can be assigned to the secondary nitrogen (–NH) attributed to the presence of organic matrix belonging to cysteine [39]. The peak beyond 400 eV in the N1s spectra of the inhibited surfaces is absent, which shows that there is no -N<sup>+</sup>H− group [66]. Therefore, the above phenomenon indicates that the adsorption site of cysteine on the surface of the studied alloy does not rely on the N group. An atomic ratio of nitrogen (Table 3) in a surface inhibited with 10−<sup>2</sup> M Cys confirms that the greater amount of inhibitor species are adsorbed on the alloy surface, and hence a better protection effect is displayed.
