On the Application of Scanning Electrochemical Probe Microscopies to Investigate Galvanic Corrosion Processes

Abstract

This study focuses on a group of scanning electrochemical probe microscopies used to reveal the early stages of galvanic coupling corrosion reactions, based on the use of microelectrochemical sensors for measuring local potentials and currents associated with chemical reactions occurring at anodic and cathodic sites, and their correlation with results obtained with conventional electrochemical techniques. Although galvanic corrosion between dissimilar metals is generally analyzed by assuming that the anodic and cathodic half-cell processes occur in different metals, the use of microelectrochemical techniques reveals that the corrosion process is actually more heterogeneous. Cathodic activity is present in both metals, but to very different degrees. Anodic activity is also localized, as the surface of the more reactive metal is not fully available to undergo anodic dissolution. Because galvanic corrosion processes are heterogeneously distributed over the surface of the coupled materials, even in model systems, the identification of cathodic sites and reactions is often insufficient when monitored by conventional electrochemical methods. These observations are particularly relevant when corrosion protection measures aim to minimize or eliminate the activity of cathodic reaction sites.

1. Introduction

Metals corrode under the effect of chemical attack, usually caused by water and ambient air. The corrosion process requires two redox half-reactions: an anodic and a cathodic half-reaction. During the oxidation (anodic) half-reaction, metals transition to their ionic state, creating a local anodic zone, and are released into the aqueous solution. This provides electrons, through the metal itself, to the local cathodic zone. During the reduction (cathodic) half-reaction, an oxidizing species decreases its oxidation state by consuming electrons from the anodic reaction. If the oxidizing agent is molecular oxygen, which is partially soluble in water, these half-reactions can be expressed as follows:

$$\mathrm{A}\mathrm{n}\mathrm{o}\mathrm{d}\mathrm{i}\mathrm{c} \mathrm{h}\mathrm{a}\mathrm{l}\mathrm{f}\text{-}\mathrm{c}\mathrm{e}\mathrm{l}\mathrm{l} \mathrm{r}\mathrm{e}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}: M\left(s\right)⟶M^{n+}\left(aq\right)+ n e^{-}$$

(1)

$$\mathrm{C}\mathrm{a}\mathrm{t}\mathrm{h}\mathrm{o}\mathrm{d}\mathrm{i}\mathrm{c} \mathrm{h}\mathrm{a}\mathrm{l}\mathrm{f}\text{-}\mathrm{c}\mathrm{e}\mathrm{l}\mathrm{l} \mathrm{r}\mathrm{e}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}: {\mathrm{O}}_2\left(aq\right)+2 {\mathrm{H}}_2\mathrm{O}\left(l\right)+4 e^{-}⟶4 \mathrm{O}{\mathrm{H}}^{-} \left(aq\right)$$

(2)

These processes can be modified by the action of pH, which intervenes directly in the cathodic half-reaction and can in turn promote complexation equilibria and the precipitation of hydroxides of the released metal cation.

Corrosion can also occur through the action of a bimetallic cell, a process called galvanic corrosion. Thus, if two electrodes composed of two different metals are electrically connected and the circuit is closed by an electrolytic medium, galvanic coupling occurs, constituting a galvanic cell. Through this connection, the more reactive metal tends to oxidize and releases a flow of electrons towards the cathode, which is the more noble metal of the galvanic pair. For galvanic corrosion to occur, an electrolytic connection is necessary between the contact areas of the two metals. This mechanism has both a detrimental effect, accelerating degradation, and a protective effect, inhibiting corrosion of the more noble metal by constantly supplying it with electrons. Furthermore, the degree of deterioration or attack depends largely on the ratio of the anode to cathode surface areas. If the attack is spread over a large anodic surface, its impact on a smaller cathodic surface is negligible. Conversely, if the anodic surface is small compared to that of the cathode, the anodic attack intensifies because a much larger cathodic surface must be protected compared to the anodic surface [1].

As a phenomenon based on redox half-reactions, corrosion phenomena can be analyzed using conventional electrochemical techniques, such as open-circuit potential (OCP) measurement, potentiodynamic polarization, and Tafel analysis [2,3,4,5]. Combining these methods provides information on the mechanism, the kinetics of the reaction under study, and the potential deposition of corrosion products. However, this information is always averaged over the entire metal surface. Since corrosion reactions begin at the micrometer scale, and therefore occur heterogeneously across the material surface, it is necessary to complement these measurements with localized electrochemical investigation techniques.

Scanning electrochemical probe microscopy (SEPM) techniques allow for in situ monitoring of surface electrochemical activity in the early stages of corrosion and its evolution over time due to a powerful combination of electrochemical sensitivity and spatial resolution. Among them, the Scanning Vibrating Electrode Technique (SVET) [6,7,8] and Scanning Electrochemical Microscopy (SECM) [9,10,11] have been successfully introduced into the corrosion laboratory. The SVET technique uses a passive microelectrode acting as a pseudo-reference in the microelectrochemical cell [6]. By vibrating this sensor near the surface of the material to be characterized, the local potential gradients that form around the cathodic and anodic sites are measured with high precision. These gradients result from the ionic fluxes induced by corrosion reactions [6]. Conversely, in SECM, the amperometric probe is an active microelectrode that is scanned close to the surface of the material under study while electrochemically interrogating its local redox activity, either by collecting reactants or corrosion products, or by converting a redox mediator added to the electrolytic environment for this purpose [12,13]. Therefore, several operating modes are used to study corrosion reactions [14], including generation-collection [15], redox competition [16] and feedback [17]. In addition to amperometric operation, passive microelectrode tips can serve as probes for potentiometric operation in SECM [18], a method also known as Scanning Ion Electrode Technique (SIET) by the corrosion research community [19]. This technique has the advantage for not requiring the electrochemical conversion of a redox mediator for these measurements and exhibits high selectivity [20], since the potential at the microelectrode probe depends on the activity of an ion of interest which constitutes an ion-selective microelectrode (ISME). Thus, pH microsensors have been developed using dual metal/metal oxide systems based on antimony, iridium or tungsten microelectrodes [21], as well as many ISME’s designed mainly for corrosion applications, which have been used to map the concentration distribution of different ions such as Al³⁺ [22], Cu²⁺ [20], Mg²⁺ [23] and Zn²⁺ [20].

The study of galvanic corrosion by scanning electrochemical probe methods was initiated by studying the model Fe–Zn pair couple using SVET [24] and SECM in redox competition mode [25]. This work quickly expanded to the characterization of corrosion processes occurring in scratches [26,27] and cut edges of galvanized steels [28,29]. Different geometric configurations of galvanic couples are also frequently imaged by SVET and SECM to study the specific effects of corrosion inhibitors on dissimilar metals [30,31,32]. This application has been subsequently extended to the characterization of the localized galvanic coupling between the matrix and different secondary phases of the substrate in complex multiphase alloys [33,34,35] and metal-coating systems [36,37,38,39,40,41], or to the mapping of the heterogeneous distribution of corrosion sites due to thermally and mechanical differences generated along welds [42,43,44,45,46], or grain boundaries [47,48]. Another area of application concerns the experimental validation of numerical models developed to simulate galvanic corrosion processes [49,50] or to overcome the experimental limitations of microelectrochemical techniques due to the development of electric fields in the electrochemical cell [51]. Therefore, the microscopic characterization of galvanic coupling systems has many applications and requires better correlation of local data with global observations obtained using conventional electrochemical techniques [52].

This work uses a multi-scale approach to evaluate the possibilities and limitations of various electrochemical methodologies available for the detection and characterization of the initial stages of galvanic corrosion. The objective is to determine the extent and surface distribution of this type of corrosion by using and correlating conventional electrochemical measurements with surface analysis measurements by SVET and SECM, using microelectrodes as electrochemical sensors. It is particularly relevant to determine whether average surface area analysis is valid for describing galvanic corrosion processes, especially when described within the framework of mixed potential theory. For example, if the entire exposed surface of the most active metal is considered a massive anode, while the cathode is homogeneous and comprises only the most noble metal. Furthermore, the study aims to establish a comparison between experimental results and theoretical estimates by interpolation of observed trends. To this end, three metals widely used in technological society, although exhibiting different degrees of reactivity, namely Fe, Cu and Zn, were selected to produce several galvanic pairs through their possible combinations.

2. Materials and Methods

2.1. Materials

Samples of iron, copper, and zinc of varying dimensions were used. Conventional electrochemical and SVET measurements were performed on commercially available metal samples with a purity greater than 97%, machined to the desired dimensions. For conventional electrochemical and SVET measurements, 1 mm diameter Fe and Cu wires were used, while the Zn pieces were cut from a plate to a size of 1.5 mm × 1.0 mm (purchased from Iturrino, Madrid, Spain). On the other hand, the samples for SECM measurements were prepared using wires with a diameter of 125 µm and a purity greater than 99.9% supplied by Goodfellow (Cambridge, UK). Both SECM and SVET measurements were performed on surfaces embedded in the resin from the EpoFix kit (Struers, Ballerup, Denmark) fabricated using PVC molds. Connections were made at the rear of the mold using tin solder and adhesive (Total Tech; Ceys, L’Hospitalet de Llobregat, Spain). Before each measurement, all samples were prepared by grinding on silicon carbide abrasive papers ranging from 180 up to a 2400 grit size (Buehler, Lake Bluff, IL, USA). Then, the metal surfaces were rinsed with distilled water to remove abrasive residues and particles. In this way, samples containing Cu-Fe, Cu-Zn and Fe-Zn galvanic pairs were prepared. Immersing the metal samples in the resin requires the use of a mold that retains the polymer matrix until it solidifies. This mold consists of pre-cut PVC segments with carefully sanded cross-sections (see Figure 1A). One end of the cylinder is covered with adhesive tape and Parafilm (Sigma-Aldrich, St. Louis, MO, USA) to prevent resin leakage during pouring. To achieve this, it is first necessary to fix the metal wires to a support that ensures their immobilization during the addition of resin and drying (see Figure 1B). Molds were systematically prepared with two facing metals, thus producing the three galvanic pairs: Cu-Fe, Cu-Zn, and Fe-Zn. Several samples of each pair were included in the mold to mitigate any risk of failure of an electrical connection previously made by tin soldering during the processing and immobilization. To finalize the preparation, a homogeneous mixture of the resin and hardener, in the proportions indicated by the manufacturer (EpoFix kit; Struers, Ballerup, Denmark), is poured into the mold, taking care to avoid the formation of bubbles. Since this is a cold-curing resin, the sample is left to rest for 24 h to ensure complete hardening of the polymer matrix. Once the PVC mold is removed, the sample surface is ground using silicon carbide abrasive papers. An electrochemical cell is then fabricated on this surface to create an additional container for the electrolytic solution, also using PVC molds and commercial insulating silicone.

The solutions were prepared using pro-analysis quality reagents and ultrapure water, with a resistivity of 18.0 MΩ cm⁻¹. The corrosion test solution was prepared with a concentration of 5 mM NaCl (Panreac Química S.A., Barcelona, Spain), as the aim was to determine the initial activation and passivation stages, given that a higher concentration of chlorides only showed intense activation effects. This solution was naturally aerated and kept at ambient temperature (18 ± 2 °C). For initial surface localization during SECM experiments in negative feedback mode, a 0.5 mM ferrocene-methanol + 5 mM NaClO₄ solution (both from Sigma-Aldrich, St. Louis, MO, USA) was used, taking advantage of the low corrosive power of perchlorate ions. The hexachloroplatinic acid (H₂PtCl₆) used to prepare the platinum black deposit on the SVET probe was purchased from Sigma-Aldrich.

2.2. Methods

Open-circuit potential (OCP), potentiodynamic polarization, and zero-resistance ammetry (ZRA) measurements were performed using a DropSens multichannel potentiostat (Metrohm DropSens, Oviedo, Spain). A flat electrochemical cell with a 2.8 cm diameter opening at the base was used to expose the mold containing the galvanic couple, and using a saturated calomel electrode as a reference and an auxiliary platinum electrode. Prior to each experiment, the samples were left non-polarized to stabilize in the test solution for 30 min to reach their spontaneous open circuit potential (E_OCP) in the electrolyte. Potentiodynamic polarization measurements were initiated at E_OCP −0.20 V, while the switching potential was set at +1.00 V vs. SCE. A scan rate of 1 mV s⁻¹ was used.

SECM measurements were carried out with a Sensolytics instrument (Bochum, Germany). The sensing probe consisted of a Pt wire of 25 μm diameter encased in glass, and the microelectrochemical cell contained an Ag/AgCl/KCl(sat.) as reference and a platinum wire as auxiliary. The distance between the sensing probe (tip) and the sample was fixed at 25 µm by recording approach curves on the resin at a position equidistant from the two metals in the galvanic couple in a solution of 0.5 mM NaCl + 5 mM NaClO₄ with the tip polarized at +0.50 V relative to Ag/AgCl/KCl(sat.). SECM maps of galvanic corrosion processes were made in a 5 mM NaCl solution using molecular oxygen as the redox mediator, while the potential applied to the microelectrode tip was −0.70 V vs. Ag/AgCl/KCl(sat.).

SVET equipment was fabricated by Applicable Electronics (New Haven, CT, USA). The sensing probe was a Pt-Ir needle of 10 µm diameter, platinized at the tip by electroreduction in a hexachloroplatinic acid solution to produce a spherical platinum black deposit [6]. The probe vibration was set at 185 Hz in the X direction and 78 Hz in the Z direction. Measurements were conducted in the XY plane at a constant distance of 100 µm in the Z axis and monitored using the video microcamera placed vertically over the sample.

All electrochemical tests were performed with a newly ground sample and carried out in three replicates to ensure reproducibility of the measurements.

3. Results and Discussion

3.1. Conventional Electrochemical Characterization of Galvanic Pairs

To facilitate the interpretation of the galvanic couplings, potentiodynamic polarization curves were measured for each metal individually upon their exposure to a naturally aerated 5 mM NaCl solution, and they are shown in Figure 2. Next, the electrochemical parameters obtained by a Tafel analysis are given in

Table 1. Electrochemical parameters obtained during the Tafel analysis of the polarization curves in Figure 2, which were obtained for each metal immersed in an aerated 5 mM NaCl solution.

Metal	Ecor (V vs. SCE)	jcor (A cm−2)	βc (mV dec−1)	βa (mV dec−1)
Cu	−0.151	7.86 × 10−8	−110	92.6
Fe	−0.512	1.16 × 10−6	−122	107
Zn	−0.995	1.56 × 10−6	−168	96.2

. Copper exhibited the most positive corrosion potential (−0.151 V vs. SCE) and the lowest current density, indicating lower electrochemical activity and greater nobility for this metal. The opposite behavior was observed for zinc, which displayed the highest current density and the most negative corrosion potential (−0.995 V vs. SCE), making it the most reactive metal in this medium. Iron (Fe) exhibited intermediate E_cor and j_cor values between those metals. Although the anodic Tafel slopes, β_a, were similar for the three metals, Zn exhibited a steeper cathodic slope, indicating lower efficiency in activating the cathodic process. Finally, comparison of the anodic branches, both during sweeps to more positive potentials and after potential reversal at +1.00 V vs. SCE, reveals a distinctive behavior for copper. Specifically, only copper reforms a more stable passive layer after depassivation than when the bare metal surface is exposed to the test environment, as evidenced by a more positive repassivation potential than the E_cor value. Conversely, more negative repassivation potentials are observed for iron and zinc.

Expected values for the mixed potential and the corrosion current density of reactions that occur in the galvanic couple can be determined using the mixed potential approximation. That is, by adequately combining the Tafel curves for each metal [53]. By taking the Cu-Fe couple as an example, the method consists of extrapolating the anodic branch of the individual copper polarization curve to a theoretical mixed potential, obtained by the intersection of the anodic branch of iron and the cathodic branch of copper. We can then determine the current density at which the anodic process would theoretically occur in copper at that mixed potential by substituting it into the equation of the line corresponding to the fitted anodic Tafel line for this metal. This current density would correspond to that generated by the oxidation reaction of copper. Although very weak, this value allows us to deduce that two half-reactions predicted by the galvanic series occur preferentially during oxidation, but that other secondary reactions, such as the oxidation of copper in this specific case, are not completely inhibited. Figure 3 shows the graphical procedure used to estimate the electrochemical parameters corresponding to the three galvanic pairs studied, while

Table 2. Comparison of the electrochemical parameters obtained by the graphical extrapolation method (Figure 3) with the values from the Tafel analysis of the polarization curves, obtained for the three galvanic couples immersed in an aerated 5 mM NaCl solution (cf. Figure 2).

Galvanic Pair	Calculation Using the Mixed Potential Approximation to the Polarization Curves of Individual Metals
	Ecor (V vs. SCE)	jcor (A cm−2)
Cu-Fe	−0.388	1.23 × 10−5
Cu-Zn	−0.804	4.47 × 10−5
Fe-Zn	−0.807	4.24 × 10−5

allows these estimates to be compared with the experimental values determined from the polarization curves given in Figure 2.

In the Fe-Zn galvanic couple, the extrapolated corrosion current density for iron is considerably lower than that of zinc. This indicates that, in this system, the most significant electronic contribution comes from zinc, which acts as the anode. Furthermore, the current density of iron is five orders of magnitude lower when coupled to zinc. These results confirm that zinc acts as a sacrificial anode, protecting the iron, whose corrosion rate is significantly reduced in its presence. In the Cu-Zn couple, copper exhibits the lowest corrosion density, on the order of femtoamperes per square centimeter, meaning that, for this couple, zinc is very effective as a sacrificial anode due to its active metal nature, unlike copper, which is a rather noble metal.

The characterization of the electrochemical behavior of the galvanic pairs formed by the association of Cu, Fe, and Zn was experimentally performed by measuring and analyzing the potentiodynamic polarization curves obtained upon their exposure to a naturally aerated 5 mM NaCl solution, and they are shown in Figure 4. It is found that the Fe-Zn couple (blue) exhibits the most negative corrosion potential, in contrast to the Cu-Fe couple (black), which has the most positive corrosion potential. This result is consistent with the thermodynamic stability order of the three metals in the galvanic series. Unlike the Cu-Fe couple, the curves measured for Fe-Zn and Cu-Zn couples overlap considerably. The repassivation potential of the Fe-Zn couple is very close to the open-circuit potential of the Cu-Zn couple. Furthermore, the repassivation potential of the Cu-Zn couple is more positive relative to its own open-circuit potential than that of the Fe-Zn couple, under the same analysis. This suggests that the Cu-Zn galvanic couple (red) requires more positive potentials to repassivate. In other words, the Fe-Zn couple (blue) progresses more rapidly toward passive film formation.

Tafel analysis yielded the electrochemical parameters presented in

Table 3. Electrochemical parameters obtained during the Tafel analysis of the potentiodynamic polarization curves in Figure 4, which were obtained for the galvanic pairs immersed in an aerated 5 mM NaCl solution. The currents measured using a ZRA unit inserted in the electrical connection between the two metals of the galvanic couple, at the rear of the electrochemical cell, are also shown. These values, from Figure 5, correspond to those recorded 45 min after the electrical connection between the two metals was established.

Galvanic Pair	Ecor (V vs. SCE)	jcor (A cm−2)	βc (mV dec−1)	βa (mV dec−1)	jZRA (A cm−2)
Cu-Fe	−0.430	4.48 × 10−6	−175	158	4.06 × 10−5
Cu-Zn	−0.933	1.12 × 10−6	−202	158	3.86 × 10−5
Fe-Zn	−0.997	7.90 × 10−6	−204	193	2.84 × 10−5

. These results indicate that the corrosion current density (j_cor) is of the same order of magnitude for all three galvanic pairs. Furthermore, the Cu-Fe pair exhibits the most positive corrosion potential, significantly higher than that of the most active pair (Fe-Zn). It is also noteworthy that, while a corrosion potential intermediate between those of the individual metals was expected, the presence of Zn shifts the potential of the pair towards much more negative values, close to those of Zn alone. This suggests that the nature of Zn and its behavior during anodic potential sweeping—with the formation of corrosion products—are the processes that mainly determine the electrochemical behavior of the galvanic couples.

A distinctive behavior is observed regarding the Tafel slopes. Although the cathodic slopes are quite similar for all the galvanic pairs considered, a particular trend in the anodic slopes is observed for the Cu-Fe and Cu-Zn pairs, with practically identical values, whereas the Fe-Zn pair exhibits different behavior. This behavior also indicates that, although copper acts as the cathode when galvanically coupled to two different metals, the nature of these metals influences the kinetics of the cathodic reaction. The same is true for zinc, which, although it acts as the anode in the Zn-Cu and Zn-Fe pairs, has its anodic slope influenced by the nature of the cathode.

Additionally, the same electrochemical cells containing the galvanic pairs could be monitored by introducing a zero-resistance ammeter (ZRA) in the electrical connection clamp used at the rear of the sample to establish the galvanic coupling condition. In this case, a mixed potential was spontaneously developed by the galvanic pair immersed in the test electrolyte by establishing the electrical connection, and the currents measured with the ZRA were recorded as a function of time since the connection was established. The recorded currents are shown in Figure 5. It is readily observable that the galvanic pairs formed with Zn are more reactive than the Cu-Fe couple. In all cases, almost stationary values were attained solely for times exceeding 2000 s. For comparison, Table 3 presents the near-stationary values recorded 45 min after the electrical connection between the two metals was established. These j_ZRA values are approximately ten times higher than the j_cor values determined by Tafel analysis of the polarization curves of the given galvanic couple. Since no stationary conditions are reached during the potential scan, the data obtained by these two electrochemical techniques cannot be directly correlated. In contrast, the j_ZRA values in Table 3 are close to the estimates obtained using mixed theory (cf. Table 2), although the latter were determined by extrapolating polarization curves. This demonstrates that the applicability of the mixed potential method is limited to a qualitative description of the electrochemical behavior of the simulated galvanic couple, as it does not take in account kinetic effects such as the mass transfer control of O₂ in the cathodic reaction or the eventual formation of the passive film over the active metal in the real system.

3.2. SVET Characterization of Galvanic Pairs

The three galvanic pairs were analyzed by SVET on millimeter-sized surfaces. The maps cover the entire galvanic pair, physically separated by an insulating resin but electrically interconnected at the rear of the mold. Figure 6, Figure 7 and Figure 8 present the results for the Cu-Fe, Fe-Zn, and Cu-Zn systems, respectively. In the maps labelled B, the sensing probe vibrated and collected ionic current data vertically (axis perpendicular to the scanning plane), showing the distribution of the collected ionic current j along this Z-axis. Similarly, the maps labelled C were obtained with the sensing probe vibrating in a plane parallel to the sample surface, and the corresponding maps show the distribution of the collected ionic current j along the Y-axis.

Based on the obtained maps, the initial hypotheses are generally confirmed by the results. The nobler metal exhibits cathodic reactivity (blue areas in Figure 6 and Figure 8 for copper, and in Figure 7 for iron), while the more active metal in each pair exhibits anodic reactivity. Furthermore, in all three cases, the cathodic zone on the nobler metal is relatively homogeneous, unlike the anodic zone, which shows greater heterogeneity, even localized pitting in the case of the Cu-Zn pair, which proved to be the most active galvanic couple. Indeed, according to the data in

Table 4. Maximum and minimum vertical current density values recorded on the SVET maps for the three galvanic pair systems. Positive values (j_max) correspond to the anodic process in the less noble metal, and negative values (j_min) to the cathodic reaction in the noble metal.

Galvanic Pair	jmax (μA cm−2)	jmin (μA cm−2)
Cu-Fe	46.4	−28.9
Cu-Zn	350	−265
Fe-Zn	238	−106

, the Cu-Zn galvanic pair exhibits the highest maximum and minimum current densities. In fact, although the color scale is similar in Figure 7 and Figure 8, the local reactivity on Zn reaches nearly 350 µA cm⁻². This result can be compared to that of the other pairs in Table 4. Furthermore, in this Cu-Zn couple, it is interesting to note the presence of cathodic regions around the anodically active zones on the more reactive metal, contributing to the heterogeneous distribution imaged in Figure 8B. This localized corrosion behavior is characteristic of zinc and appears to be due to its high reactivity, strongly amplified by the action of copper. Thus, the large number of electrons released at the anodic sites react to some small extent in the vicinity of the anodes [54]. It should be noted that signs of cathodic activity are also observed in the active metal of the Fe-Zn pair (Figure 7B), although this activity is close to the resolution limit of the technique. In contrast, no cathodic activity is observed in the active metal of the Cu-Fe couple on the corresponding SVET map (Figure 6B). In fact, this pair exhibits consistently lower current densities than the other two galvanic systems, by an order of magnitude. These results could be expected from the different reactivities in the galvanic series, as copper combined with iron is the galvanic couple with the smallest potential difference. However, it should be noted that the spatial resolution of SVET does not allow us to determine whether the current density over a very small area results from purely anodic (or cathodic) reactivity, or whether both processes are involved, since microcathodes on the more reactive metal are orders of magnitude less reactive than the anodic sites (see Figure 8B).

Furthermore, data relating to the horizontal component of the current density were recorded, measured approximately by placing the SVET probe on the resin between the two metals in their central zone. These data are presented in

Table 5. Values of the horizontal current density (j_hor) values recorded in the central area of the SVET maps, designated by C in Figure 6, Figure 7 and Figure 8, for the three galvanic systems. For comparison, the currents determined for each pair using conventional electrochemical techniques in Table 2 and Table 3 are also shown.

Galvanic Pair	jhor (μA cm−2)	1 jcorr (μA cm−2)	2 jcorr (μA cm−2)	jZRA (μA cm−2)
Cu-Fe	3.2 × 10−6	1.23 × 10−5	4.48 × 10−6	4.06 × 10−6
Cu-Zn	7.3 × 10−6	4.47 × 10−5	1.12 × 10−6	2.84 × 10−5
Fe-Zn	5.4 × 10−6	4.24 × 10−5	7.90 × 10−6	3.86 × 10−5

¹ From the application of the mixed potential theory to individual polarization plots. ² From the Tafel analysis of the polarization curves measured to each galvanic couple.

for each pair, along with the corrosion current data measured by conventional electrochemistry, already presented in Table 2 and Table 3. It should be noted that this current value is not a quantitative measure of the actual current density flowing between the two electrodes, because its measurable effect decreases with distance (the current lines leaving the surface are not parallel at all distances, but gradually diverge from each other as the distance from the source increases). In fact, the measurement is not taken at zero height, and the metals are not perfectly aligned with the horizontal axis defined by the probe vibration. However, since it represents the “most direct path” of the current, its value is considered representative of the reactivity of the galvanic couple. Its amplitude, on the order of µA cm⁻², is consistent with the results of conventional electrochemistry. This horizontal current represents 5 to 10% of the expected value and is significantly higher than the current measured during the polarization of the couple and with the zero-resistance ammeter (ZRA) placed in series in the electrical connection at the rear of the sample.

3.3. SECM Characterization of Galvanic Pairs

SECM measurements revealed the cathodic oxygen consumption activity of each galvanic couple on small-area samples (125 µm in diameter). The results are presented in Figure 9, Figure 10 and Figure 11 as three-dimensional and two-dimensional (top view) maps of the Cu-Fe, Fe-Zn, and Cu-Zn galvanic couples, respectively. Using the redox competition mode with molecular oxygen as a mediator, the variations in the current measured at the tip during sample scanning indicate of variable oxygen availability in the electrolyte near the surface. This availability depends on its consumption during the cathodic half-cell reaction of a corroding substrate. Although the measurements are performed in close proximity to the surface, they are not close enough to account for surface roughness or tilt of the substrate. Therefore, the maps in Figure 9, Figure 10 and Figure 11 represent only the spatial distributions of electrochemical reactivity in the different galvanic couples studied.

Although the Cu-Fe galvanic pair is the less reactive according to the measurements above, it must be noticed that oxygen electroreduction can occur in both metals, as shown in Figure 9 (with Cu in the upper left corner of the diagrams). Yet, most of the cathodic current is produced in the more noble metal, as expected. The low amplitude of the cathodic signal produced in the active metal is below the detection threshold of SVET measurements, which were recorded with a greater distance between the tip and the sample. Therefore, for comparison purposes, only the signals recorded during oxygen electroreduction in the nobler metal of the galvanic couple were taken into account. The peaks observed on the maps in Figure 9, Figure 10 and Figure 11 correspond to the least negative measured current values, namely −1.22, −0.320, and −0.977 nA, respectively. This indicates a lower dissolved oxygen concentration near the electrode than the values obtained outside this area, which are approximately −2.8, −2.7, and −2.0 nA, respectively. Although the small tilt of the samples and variations in measurement height (which either facilitate or hinder oxygen diffusion toward the tip) make a direct comparison of quantitative values difficult, it is observed that the metal acting as the cathode consumes oxygen present in the solution, thus reducing the amount available for the reaction in the probe. Consequently, the current measured by the probe decreases when the tip scans over a cathodic site.

3.4. Final Remarks

The results agree with the prediction of the galvanic series, showing that zinc is the most electrochemically active metal, followed by iron, then copper. This sequence explains the ability of sacrificial anodes to prevent the degradation of a more noble metal once the galvanic couple is established, leading to faster degradation of that anode. The current and degradation rate of sacrificial anodes tend to decrease as the open-circuit potential difference between the metals involved increases, because anode dissolution is accelerated under these conditions.

The experimental results appear to agree with a more accurate theoretical interpolation when passive layer formation and oxygen mass transfer control phenomena are taken into account, as these phenomena are limiting factors for the establishment of anodic and cathodic processes, and for the corrosion current, respectively.

Unlike conventional electrochemical methods, scanning electrochemical microscopies allow the detection of local current densities at the metal surface. In this way, the use of SECM in the redox competition mode, by measuring the electroreduction of soluble molecular oxygen in the test electrolyte, confirms that the cathodic activity of each couple is preferentially localized in the nobler metal, although a smaller cathodic activity can be detected on the less noble metal near the anodic sites.

On the other hand, both the anodic and cathodic activities occurring on the more and less reactive metals of the galvanic pair, can be simultaneously imaged with SVET, and should be considered in numerical simulations of galvanic corrosion. Among the model galvanic couples considered in this work, the localized nature of pitting corrosion in the zinc is highlighted for the Cu-Zn couple, a characteristic of its electrochemical behavior, due to its coupling with a much nobler metal such as copper.

4. Conclusions

This study used a multi-scale approach to evaluate galvanic corrosion between copper, iron, and zinc in model two-metal configurations without physical contact immersed in the same electrolyte. The experimental setup employed electrochemical techniques, including conventional methods providing global information on the process, and electrochemically sensitive microscopies capable of detecting surface phenomena and providing localized information. These latter techniques include scanning electrochemical microscopy (SECM) and scanning vibrating electrode microscopy (SVET). The experimental findings led to the following conclusions:

• The individual study of each metal in a galvanic coupling system using potentiodynamic polarization allows us to understand its specific electrochemical behavior, thus facilitating the identification of the differences and effects that may occur when it is subjected to galvanic coupling conditions by application of the mixed potential theory. This leads to a deeper understanding of the electrochemical mechanisms involved in the initial phases of galvanic corrosion.
• The application of the mixed potential theory makes it possible to make predictions about the behavior of galvanic corrosion for individual metals or metallic phases when the corrosion process has kinetics controlled by polarization activation; that is, in the absence of passivity and precipitation of corrosion products. The main limitation of this method relies on the assumption that the anodic and cathodic processes are completely independent between the two metals, with the anodic corrosion reaction occurring exclusively on the less noble metal and the corresponding cathodic half-reaction only on the more noble metal. This cathodic half-reaction is considered identical to that which occurs in the corrosion process specific to each metal within the aggressive environment under consideration.
• Conventional electrochemical techniques allow adequate quantification of the phenomenon in terms of electrochemical parameters obtained by Tafel analysis of potentiodynamic polarization curves and by direct amperometric measurement of the current flowing in the electrical circuit linking the two metals.
• The use of scanning microelectrochemical techniques to measure local potentials and ion densities has shown that the cathodic process also occurs in localized areas of the more reactive metal, albeit to a lesser extent than in the more noble metal. Therefore, conventional electrochemical measurements, which require separating the anodic and cathodic half-reactions at different electrodes, do not fully capture the intensity of corrosion processes occurring in a real galvanized assembly, even when it appears homogeneous to the naked eye. This calls into question mechanistic analyses that fail to consider localized and in situ information.
• Scanning microelectrochemical techniques should be used to delve deeper into both reaction mechanisms and protection procedures in any real-world corrosion process, particularly in alloys, joints and welds, going beyond the simple model pairs considered in this work, even when their corrosion distribution appear more homogeneous. This conclusion is not limited to the use of SVET and SECM, as other local microelectrochemical techniques, such as SIET or potentiometric SECM, can also contribute to the multiscale characterization of galvanic corrosion.