Electrochemical Corrosion Behaviour of WC-Co Cemented Carbide in Acidic and Alkaline Solutions for PVD Coating Removal

Abstract

This study investigates the corrosion behaviour of a WC–6Co cemented carbide (94 wt% WC, 6 wt% Co) in acidic (pH 2) and alkaline (pH 13) electrolytes used for industrial PVD coating removal. The removal of the coating was not investigated, since no coatings were applied or analysed in this study. The objective was exclusively to simulate the corrosion response of the exposed substrate after the coating had been removed during electrochemical stripping. Potentiodynamic polarisation measurements were performed from OCP −0.2 V to +3 V at a scan rate of 1 mV·s⁻¹, followed by surface characterisation using SEM/EDS and laser profilometry to identify corrosion mechanisms and quantify material degradation. In an acidic solution, corrosion was dominated by cobalt dissolution, followed by the formation of a W–O-rich corrosion-product layer, as indicated by increased tungsten and oxygen contents in SEM/EDS analyses. The layer became increasingly porous and mechanically unstable at higher potentials. Progressive thickening of the corrosion-product layer and subsequent breakdown resulted in significant material loss, including surface abrasion up to ~8 µm. In alkaline electrolytes, SEM/EDS analyses revealed a Co–O-rich surface layer, suggesting cobalt-containing hydroxide/oxide corrosion products. These results suggest that surface-layer formation on WC–Co does not necessarily provide reliable corrosion protection, as stability and morphology strongly depend on pH. These findings provide valuable guidance for the use of cemented carbides in electrochemical stripping processes for PVD coating removal.

1. Introduction

Cemented carbides are a class of composite materials consisting of hard carbide particles embedded in a metallic binder matrix. The most commonly used system combines tungsten carbide (WC) with cobalt (Co) as the binder phase, providing an exceptional balance of hardness, toughness, and wear resistance, making these materials indispensable for cutting tools and wear-resistant applications [1]. Their microstructure typically consists of WC grains ranging from 0.2 to 6 µm dispersed within the cobalt matrix, where higher cobalt content improves toughness at the expense of hardness [2].

Due to their biphasic structure, cemented carbides are susceptible to galvanic corrosion, in which either the Co binder or WC grains dissolve depending on the pH of the medium [3,4]. Thermodynamic stability regions of Co and W in aqueous systems can be interpreted using Pourbaix diagrams. For cobalt, acidic conditions promote dissolution, whereas alkaline conditions favour passivation through formation of ${Co\left(OH\right)}_2$. In contrast, tungsten tends to passivate in acidic environments via ${WO}_3$ formation but becomes unstable under alkaline conditions, where soluble tungstate ions (${{WO}_4}^{2-}$) are formed [5].

In acidic solutions, WC forms a passive ${WO}_3$ film, while cobalt readily dissolves, leading to degradation of the matrix and eventual loss of mechanical integrity [4,6]. This galvanic effect is consistent with the polarisation measurements reported by Sutthiruangwong et al. [4], which show lower corrosion potential and higher current density for cobalt compared to tungsten carbide.

The reaction equations for tungsten at acidic pH, as described by Pourbaix [5], are as follows:

$$W+2H_{2}O\to {WO}_2+{4H}^{+}+{4e}^{-}$$

$$2{WO}_2+H_{2}O\to W_2{O}_5+{2H}^{+}+{2e}^{-}$$

$$W_2{O}_5+H_{2}O\to {WO}_3+{2H}^{+}+{2e}^{-}$$

For tungsten carbide the simplified passivation reaction in acidic media is given as follows [6]:

$$WC+5H_{2}O\to {WO}_3+{CO}_2+{10H}^{+}+{10e}^{-}$$

Tungsten carbide forms a WO₃ layer on the surface of its grains. Galvanic coupling reactions between WC and Co result in cathodic protection of WC grains, while the Co binder matrix acts as the anode and undergoes dissolution, as shown in the equation below [2,5]:

$$Co\to {Co}^{2+}+{2e}^{-}$$

$${2H}^{+}+{2e}^{-}\to H_2$$

In the alkaline environment, tungsten is thermodynamically unstable and forms soluble tungstate species ${{WO}_4}^{2-}$, while Co becomes passivated through the formation of a ${Co\left(OH\right)}_2$ layer [5,6]:

$$W+{4H}_{2}O\to {{WO}_4}^{2-}+{8H}^{+}+{6e}^{-}$$

$$WC+{14OH}^{-}\to {{WO}_4}^{2-}+{{CO}_3}^{2-}+{7H}_{2}O+{10e}^{-}$$

$$Co+{2OH}^{-}\to {Co\left(OH\right)}_2+{2e}^{-}$$

The aim of this study is to investigate the electrochemical behaviour of WC–6Co cemented carbide in acidic (pH 2) and alkaline (pH 13) environments using potentiodynamic polarisation, SEM/EDS, and confocal laser microscopy. Particular attention is given to identifying the corrosion mechanisms and evaluating the protective effectiveness of passivation layers under different pH conditions. The electrolytes used in this study, characterized by their high electrical conductivity are proprietary solutions developed for the electrochemical removal of PVD coatings. Therefore, the objective is to assess their influence on the underlying cemented carbide substrate during decoating processes. The electrochemical corrosion behaviour of the uncoated substrate is used to evaluate the corrosive impact on the surface once the coating has been removed and the substrate remains exposed to the stripping bath.

2. Theory—Reconditioning in Surface Coating Technology

Physical Vapour Deposition (PVD) is a widely used technique for applying extremely hard and ultra-thin coatings, typically only a few micrometres thick. This technology enables the deposition of layers that significantly enhance the performance and lifespan of cutting tools, forming tools and precision components. Coating properties such as hardness, toughness and durability can be precisely tailored by selecting the appropriate metallic targets and reactive process gases [7].

To fully exploit these benefits throughout the tool’s service life, it is essential that coatings not only perform reliably but also can be removed without damaging the underlying substrate. This makes reconditioning a critical step, as it enables selective stripping of worn or defective layers followed by recoating. Reconditioning offers both economic and ecological benefits by extending tool lifetime, reducing material consumption, and minimising waste generation [8].

In industrial practice, stripping hard ceramic coatings without damaging the substrate is critical. Typical situations requiring controlled stripping include:

• Tool refurbishment: Expensive coated cutting tools require sharpening and recoating before reuse. Therefore, the old coating must be completely removed [8].
• Defective coatings: Layers exhibiting poor adhesion, delamination, or incorrect thickness must be stripped prior to recoating [8].

Coating removal (stripping) can be achieved through physical, chemical, electrochemical, or plasma-based methods [9]. Physical techniques such as sandblasting risk altering substrate geometry, whereas chemical and electrochemical approaches offer greater control and substrate preservation. Electrochemical stripping allows process optimization through the adjustment of electrolyte composition, temperature, current density, and pH, while additives can enhance selectivity and surface protection. A major challenge arises when the coating and substrate share similar chemical compositions, requiring precise control to avoid substrate damage and maintain dimensional integrity [7,8].

Electrochemical Stripping Processes

In electrochemical processes, the mechanism involves the oxidation of the coating elements induced by the application of electrical power. The workpieces are connected as the anode and immersed in an electrolyte that serves as the medium for ionic transfer. To minimize ohmic loss, the electrolyte must exhibit high electrical conductivity [10]. The cathode surrounds the anode, and its geometry is adapted to the shape and size of the part being stripped. Cathodes are typically made of highly durable materials, most commonly stainless steel. An appropriate cathode-to-anode surface area ratio is essential to avoid current-density limitations that would otherwise reduce the stripping rate [11]. Furthermore, electrolyte agitation or controlled movement of the workpiece is required to remove anodically dissolved ions from the surface, thereby preventing local concentration gradients and improving stripping uniformity [10,11]. The process can be controlled either potentiostatically, where the applied voltage is regulated and the resulting current is monitored, or galvanostatically, where the current is kept constant and the voltage varies accordingly.

In a potentiostatically controlled process, once the coating is fully removed, a pronounced change or drop in current is observed, serving as an indicator to stop the process, thereby preventing unnecessary substrate corrosion. A notable advantage of the electrochemical stripping process is the long lifetime of the electrolyte. Unlike in wet chemical etching, the electrolyte serves as an ion-transfer medium and is not consumed during stripping, making the process more cost-effective and sustainable. Additionally, the electrochemical decoating process can be monitored in real time by observing the current response. Furthermore, the stripping time is considerably shorter than in wet chemical processes.

3. Experimental

3.1. Materials

The specimens used in this study were uncoated cemented carbide samples composed of 94 wt% tungsten carbide (WC) and 6 wt% cobalt (Co), commonly referred to as WC–6Co (see Figure 1). No coated specimens were investigated, and no PVD coatings were deposited or analysed. All experiments were performed on the bare WC–6Co substrate.

A representative sample was used as an internal standard for analytical investigations. The sample was a polished, indexable cutting insert with dimensions of 12 × 12 mm and a surface roughness of Ra = 0.025 µm. This material closely reflects the composition and microstructure of cemented carbide tools widely used in industrial applications. Its standardized nature ensures reproducibility and comparability in surface and corrosion studies.

3.2. Electrolytes

Two electrolytes relevant to electrochemical stripping were investigated: an acidic and an alkaline electrolyte, both representing electrolytes used in internal, standardized electrochemical stripping processes. The acidic electrolyte was an internally prepared nitrate-based aqueous electrolyte to provide ionic conductivity. The solution was acidified with acetic acid to adjust the pH to 2. The alkaline electrolyte was a commercially available hydroxide-based alkaline electrolyte containing minor amounts of phosphate species and a surfactant. Before use, it was diluted with deionized water (conductivity < 0.5 µS·cm⁻¹).

Owing to confidential internal process information, the exact composition, concentrations, and preparation details of both electrolytes cannot be disclosed. Therefore, only the general chemical character and the experimentally relevant properties are reported. The measured pH values used in this study were pH 2 for the acidic electrolyte and pH 13 for the alkaline electrolyte. The corresponding conductivities were 242 mS·cm⁻¹ and 38.5 mS·cm⁻¹, respectively. The reported pH values and conductivities refer to the electrolytes after equilibration to room temperature (20 °C). During electrochemical testing, the electrolyte temperature was monitored and remained at 20.0 ± 0.5 °C.

3.3. Electrochemical Testing

Electrochemical measurements were performed in a flat corrosion cell using a three-electrode configuration and an electrolyte volume of 100 mL. The cemented carbide sample served as the working electrode (WE), an Ag/AgCl (3 M KCl) electrode was used as the reference electrode (RE), and a stainless-steel electrode served as the counter electrode (CE), with an exposed surface area at least five times larger than that of the working electrode. All potentials are reported with respect to the Ag/AgCl (3 M KCl) reference electrode, which corresponds to +211 mV versus the standard hydrogen electrode (SHE) at 20 °C. The electrolyte temperature was monitored using a thermometer and remained at 20.0 ± 0.5 °C.

The exposed surface area of the sample was 0.63 cm². In the flat cell, the sample was positioned horizontally at the bottom of the cell. This orientation allowed gas bubbles to detach directly from the surface and disperse into the solution. Due to this position of the sample, stirring could not be applied. To closely replicate the conditions of a real industrial stripping process, the electrolyte was not purged with nitrogen. Although nitrogen purging is commonly used to reduce oxygen-induced corrosion, it does not reflect typical production environments.

As the carbide samples were already highly polished, no further surface polishing was performed prior to the measurements. To remove any potential residues, the surfaces were first rinsed with ethanol, followed by deionized water (conductivity < 0.5 µS·cm⁻¹), ensuring a clean surface condition for accurate electrochemical analysis.

Potentiodynamic Polarisation

Potentiodynamic polarisation was used to evaluate the electrochemical behaviour of WC–6Co in both acidic (pH 2) and alkaline (pH 13) environments. In contrast to linear polarisation (LPR), which is restricted to a narrow potential interval around the corrosion potential to assess uniform corrosion, potentiodynamic polarisation involves sweeping the potential over a wide anodic range in order to investigate electrochemical responses at higher potentials [12].

Before each potentiodynamic polarisation measurement, the electrode was allowed to stabilise under open-circuit conditions for 10 min. The open-circuit potential (OCP) was subsequently recorded for 60 s at a sampling interval of 0.1 s and used as the reference potential for defining the start of the polarisation scan. For OCP acquisition, a potential-drift acceptance criterion of dE/dt = 1 × 10⁻⁶ V·s⁻¹ was set in the measurement software.

For the WC–6Co measurements, the working electrode was polarised from OCP −0.2 V up to +3 V at a scan rate of 1 mV·s⁻¹. This potential range was selected based on the stability-window analysis performed using a Pt electrode to avoid excessive hydrogen evolution in the acidic electrolyte at more cathodic potentials and to focus on anodic phenomena such as passivation or pseudo-passivation, anodic metal dissolution, and material degradation. The scan rate of 1 mV·s⁻¹ was chosen to ensure quasi-equilibrium conditions while minimizing capacitive distortions.

For each electrolyte, three independent potentiodynamic polarisation measurements were performed under identical experimental conditions to assess reproducibility. The variability of key electrochemical parameters (E_corr, j_corr) between replicate runs remained below ±5%. In addition, the anodic branches exhibited consistent characteristic behaviour across all measurements, which occurred within reproducible potential ranges and current density levels. The polarisation curves presented are therefore representative of the repeated measurements, and the corresponding tested surfaces were selected for detailed post-test characterisation.

The resulting polarisation curves were analysed using Tafel extrapolation to estimate electrochemical parameters such as the corrosion potential (E_corr), corrosion current density (j_corr) and anodic and cathodic Tafel slopes (βa and βc). These parameters were obtained by fitting the polarisation data within the full potentiodynamic scan, specifically in a narrow potential range of approximately ±20 mV around E_corr, ensuring that the analysis was restricted to the near-equilibrium region of the curve. While these parameters provide a useful comparative description of the electrochemical behaviour close to the corrosion potential, they do not capture the strongly potential-dependent processes that dominate at the higher anodic potentials relevant to electrochemical stripping.

Potentiodynamic polarisation measurements were performed using a Metrohm (Filderstadt, Germany) Autolab PGSTAT302N potentiostat operated via Metrohm NOVA 2.1.6 software.

3.4. Surface Characterisation

Surface characterisation was carried out both before and after potentiodynamic polarisation testing to evaluate morphological changes and corrosion-induced alterations on the WC–6Co samples. Laser profilometry (Keyence, Frankfurt am Main, Germany), scanning electron microscopy (SEM; Zeiss, Oberkochen, Germany), and energy-dispersive X-ray spectroscopy (EDS; Zeiss, Oberkochen, Germany) were employed to assess material loss, surface topography, and elemental composition.

Surface topography and profile measurements were conducted using a Keyence (Frankfurt am Main, Germany) VK-X160K laser scanning microscope at 500× magnification (NA = 0.8, working distance = 0.54 mm). Measurements focused on profiling the interface between untreated substrate and electrochemically treated surface to identify features such as oxide film formation or surface abrasion due to uniform corrosion. These measurements also allowed the detection of localized corrosion features, such as pitting. Data analysis was performed using Keyence MultiFileAnalyser software (version 1.3.1.120).

High-resolution imaging and elemental analysis were performed with a Zeiss (Oberkochen, Germany) Sigma 300 SEM equipped with a Schottky field emission gun (Zeiss, Oberkochen, Germany) and a Smart EDS (EDAX) (Zeiss, Oberkochen, Germany) detector. SEM measurements were conducted at 15 kV acceleration voltage and a working distance of 10 mm using magnifications of 100× and 5000×. EDS measurements were conducted at 1000× and elemental data were reported in weight percent (wt%) using standardless quantification.

Cross-sectional analysis presented in this work was conducted on an actual industrial tool after real electrochemical stripping in the acidic electrolyte (pH 2) investigated in this study. This measurement was performed using a Thermo Fisher SCIENTIFIC (Dreieich, Germany) Scios 2 HiVac Dual-Beam FIB-SEM, which combines a focused ion beam with high-resolution field-emission SEM imaging. Imaging was conducted at 20,000× magnification and at an accelerating voltage of 5 kV.

4. Results

4.1. Electrochemical Investigation of Alkaline and Acidic Electrolytes Using a Pt Working Electrode

To determine the electrochemical stability windows of both electrolytes and to identify potential ranges to be avoided during polarisation measurements on WC–6Co, potentiodynamic polarisation was first performed using an inert platinum working electrode (Pt). The measurements were conducted from −1 V to +3 V at a scan rate of 1 mV·s⁻¹ under stirred conditions in a 1 L corrosion cell at 20.0 ± 0.5 °C. Figure 2 shows the resulting polarisation curves recorded in the alkaline (pH 13) and acidic (pH 2) electrolytes.

In the cathodic region (E < E_corr), the acidic electrolyte exhibits significantly higher current densities than the alkaline electrolyte. This behaviour is characteristic of intensified hydrogen evolution, which is strongly favoured at low pH. The high cathodic currents therefore indicate accelerated proton reduction (2H⁺ + 2e⁻ → H₂). Platinum is an excellent catalyst for hydrogen evolution, which further enhances cathodic activity in acidic media. In contrast, the alkaline electrolyte shows much lower cathodic currents, consistent with the slower hydrogen evolution kinetics associated with water reduction (2H₂O + 2e⁻ → H₂ + 2OH⁻), a well-established effect reflecting a decrease of approximately two orders of magnitude in HER rates on Pt in alkaline compared to acidic conditions [13].

In the anodic region, the acidic electrolyte displays a relatively broad plateau range between approximately 0.3 and 1.2 V with current densities around 0.02 mA·cm⁻². Beyond approximately 2.1 V, a steep rise in current density occurs, corresponding to oxygen evolution and the onset of electrolyte decomposition. The alkaline electrolyte exhibits a narrower plateau region between ~0.05 and 0.6 V at around 0.01 mA·cm⁻². At higher anodic potentials, the anodic current in the alkaline electrolyte remains higher than in the acidic electrolyte up to approximately 2.1 V. Above this potential, however, the acidic electrolyte exhibits a steeper current increase and eventually reaches higher current densities than the alkaline electrolyte. This behaviour is consistent with stronger oxygen evolution and earlier electrolyte decomposition in the acidic medium, whereas the alkaline electrolyte shows a more gradual increase in anodic current. This enhanced stability of the alkaline electrolyte may be associated with the presence of surfactants and phosphate-based inhibitors, which can adsorb onto the platinum surface and supress charge transfer, thereby mitigating oxygen evolution.

Based on the stability-window analysis obtained from the Pt electrode, the potential range for the WC–6Co measurements was adjusted to start at OCP −0.2 V in order to avoid enhanced hydrogen evolution in the acidic electrolyte at more cathodic potentials and to prevent unintended surface modifications. In addition, starting the scan at OCP −0.2 V ensured that the potential interval required for Tafel analysis around E_corr was included within the full potentiodynamic polarisation curve. The upper limit was set to +3 V, as the focus of these measurements was to examine anodic processes relevant to electrochemical stripping such as anodic active metal dissolution, passivation or pseudo-passivation. In this context, precise corrosion-rate determination was of secondary importance, since electrochemical stripping involves the application of an external potential rather than free-corrosion conditions.

4.2. Corrosion Behaviour of WC–6Co in Alkaline and Acidic Electrolytes

The corrosion behaviour of WC–6Co cemented carbide in alkaline (pH 13) and acidic (pH 2) electrolytes used for electrochemical stripping was investigated using potentiodynamic polarisation. The potential was scanned from OCP −0.2 V to +3 V at 1 mV·s⁻¹, a range selected based on the stability-window analysis obtained from the Pt electrode. All measurements were performed in a flat corrosion cell containing 100 mL of the electrolyte under unstirred conditions at 20 °C. Following the electrochemical tests, surface analyses were carried out to compare corrosion mechanisms in two electrolytes. The resulting polarisation curves are shown in Figure 3 as (a) logarithmic and (b) linear forms to facilitate comparison of cathodic and anodic reaction regions.

To complement the qualitative evaluation of the polarisation curves, Tafel-derived electrochemical parameters from the potential region around E_corr within the full potentiodynamic scan are summarized in

Table 1. Tafel analysis values of WC–6Co in alkaline and acidic electrolytes.

Electrolyte	OCP [V]	Ecorr [V]	jcorr [mA·cm−2]	βa [V·dec−1]	βc [V·dec−1]
Alkaline (pH 13)	−0.209	−0.210	1.48 $\times$ 10−3	0.135	0.319
Acidic (pH 2)	−0.158	−0.150	3.03 $\times$ 10−3	0.013	1.378

. These parameters provide a useful comparative indication of the relative corrosion tendency under conditions represented by the Tafel fit but are less suitable for describing the corrosion mechanisms dominating at the higher anodic potentials relevant to electrochemical stripping, where metal dissolution, pseudo-passivation, oxide growth, and oxide-film breakdown govern material removal.

The corrosion current density (j_corr) is higher in the acidic electrolyte than in the alkaline electrolyte (3.03 × 10⁻³ vs. 1.48 × 10⁻³ mA·cm⁻²), indicating higher corrosion tendency under the conditions represented by the Tafel analysis parameters.

In the acidic electrolyte, anodic metal dissolution begins immediately above E_corr (−0.15 V). The anodic current then rises rapidly and reaches clearly higher values than in the alkaline electrolyte in the initial active dissolution range. At intermediate potentials, approximately 0.65 to 0.85 V vs. Ag/AgCl (3 M KCl), the current density becomes temporarily attenuated. However, the measured current densities in this interval (~0.6–0.7 mA·cm⁻²) remain substantially higher than typical passive current densities (~10 µA·cm⁻² [14]), indicating that a fully developed passive state is not achieved. This behaviour is therefore described as pseudo-passivation rather than true passivation and is associated with the formation of a porous W-O-rich surface layer commonly associated with ${WO}_3$-rich corrosion products on the tungsten-rich surface. As the anodic potential increases further, the current density rises again, indicating renewed anodic dissolution at higher potentials rather than a transpassive transition. According to Song [15], a transpassive region cannot be assigned here because no true passive state is established beforehand. The high-potential behaviour is therefore interpreted as continued oxidation and growth of a porous W-O-rich corrosion-product layer, followed by mechanical instability and local film breakdown. In the linear representation (Figure 3b), the acidic electrolyte reaches a maximum current density of approximately 100 mA·cm⁻² at around 1.9 V. At approximately 2.8 V, a pronounced current change is observed, consistent with partial breakdown and local detachment of the thick, porous surface layer. This is reflected in a sharp current spike from ~25 mA·cm⁻² to ~95 mA·cm⁻², indicating the exposure of the fresh substrate material. The subsequent decrease in current density is consistent with the formation of a new W–O-rich surface layer. This interpretation is consistent with the place-exchange and field-assisted ion-transport mechanism described by Sato [16], which explains oxide growth through coupled migration of oxygen ions and metal ions during anodic polarisation.

In contrast, the alkaline electrolyte does not exhibit a pronounced pseudo-passive region. Anodic metal dissolution begins at approximately −0.21 V and is followed by a steady increase in current density, indicating continued active anodic dissolution without the establishment of a comparably protective surface layer. Based on the electrochemical response, SEM/EDS observations, and Pourbaix considerations [5], tungsten dissolution in alkaline solution is interpreted to proceed via soluble tungstate species (${{WO}_4}^{2-}$), while the Co−O-rich surface layer is attributed to cobalt hydroxide-containing corrosion products, likely including ${Co\left(OH\right)}_2$. Although this surface layer may mitigate cobalt dissolution locally, it does not stabilise the overall surface, and WC grain leaching remains the dominant degradation mode.

A further distinction between the two electrolytes is reflected in the anodic current response relative to the commonly referenced value of 5 mA·cm⁻² from ASTM G61 for iron- nickel-, or cobalt-based alloys [17]. In the present work, this value is used only as a comparative benchmark for pronounced anodic activity and associated material degradation, not as a formal corrosion criterion for WC–6Co. On this basis, the acidic electrolyte reaches this benchmark at lower potentials than the alkaline electrolyte, indicating an earlier onset of strong anodic degradation under the applied polarisation conditions.

These differences are consistent with the microstructural composition of the WC–6Co substrate. Owing to the high WC content (94 wt%) and the relatively low cobalt fraction (6 wt%), the electrochemical response is governed predominantly by the WC phase. In the acidic electrolyte, tungsten oxidation is associated with pseudo-passivation through the formation of a ${WO}_3$-rich surface layer, whereas in alkaline media tungsten dissolution is interpreted to proceed via soluble ${{WO}_4}^{2-}$ species [5]. Although cobalt may form a hydroxide-containing surface layer in alkaline conditions, its low fraction in the substrate is insufficient to stabilise the overall surface. As a result, the acidic electrolyte exhibits pseudo-passivation followed by renewed anodic dissolution and film breakdown at higher potentials, whereas the alkaline electrolyte shows continued active dissolution without the formation of a comparably protective surface layer.

4.3. Surface Characterisation of WC–6Co After Potentiodynamic Polarisation in the Acidic Electrolyte

Figure 4 shows the WC–6Co sample after potentiodynamic polarisation testing in the acidic electrolyte (pH 2), revealing a partially detached corrosion-product layer on the surface. A thick yellowish-grey primary corrosion-product layer is loosely attached to the surface, while a second underlying layer is visible beneath it. SEM micrograph at 100× magnification (Figure 5c) reveals the underlying surface after removal of the fractured upper layer, showing a porous surface morphology resembling dried, cracked soil. The corresponding EDS spectrum (Figure 5d) indicates elevated tungsten and oxygen contents, while cobalt was not detected under the applied measurement conditions. For comparison, SEM and EDS analyses of the untreated surface prior to testing are shown in Figure 5a,b.

Surface profile measurements obtained using laser microscopy revealed a highly rough surface with an average profile depth of approximately 3.4 µm. At higher magnification, individual depressions of up to 6 µm in depth were observed. These features indicate the highly porous and non-uniform nature of the corrosion-product layer, consistent with the cracked and irregular morphology observed in the SEM micrograph in Figure 5c. To estimate the possible thickness of the W–O-rich corrosion-product layer, an additional surface profile measurement was carried out on a WC–6Co sample polarised at a higher scan rate of 5 mV·s⁻¹. In contrast to the polarisation curve recorded at 1 mV·s⁻¹, the curve obtained at 5 mV·s⁻¹ did not exhibit a pronounced anodic current peak at 2.8 V, indicating that the initially formed layer had not yet detached. This behaviour can be attributed to the significantly shorter exposure time at higher scan rates, which delayed film fracture. For comparison, the corresponding polarisation curve at 5 mV·s⁻¹ is presented in its linear representation as an overlay with the 1 mV·s⁻¹ measurement in Figure 6a. The higher current densities observed at 5 mV·s⁻¹ are attributed to the increased scan rate, which leads to a deviation from quasi-equilibrium conditions and therefore higher transient current responses. It should be noted that all other electrochemical measurements were performed at 1 mV·s⁻¹ to ensure comparability, and the higher scan rate applied here serves solely to illustrate the behaviour of the corrosion-product layer under reduced exposure time. The corresponding surface profile indicates an apparent maximum thickness of approximately 44 µm (see Figure 6c), suggesting substantial anodic accumulation of a porous W–O-rich corrosion-product layer under these conditions. The SEM micrograph presented in Figure 6b further confirms the brittle and porous nature of the layer, revealing extensive cracking and locally detached fragments. These features are consistent with the electrochemical response shown in Figure 3, where high anodic currents persist despite continuous layer formation, indicating that the increasingly porous and ion-permeable W–O-rich layer formed under prolonged anodic polarisation is mechanically unstable and prone to cracking.

Since the corrosion-product layer must be removed prior to recoating, geometric change to the substrate must be considered. To assess the depth of uniform corrosion after removal of the layer, the sample was subjected to ultrasonic treatment in ethanol. Subsequent surface profile measurements revealed a uniform material loss with an average surface abrasion of approximately 8 µm (see Figure 6d).

The FIB (Focused Ion Beam) cross-section shown in Figure 7b, obtained from a cemented carbide cutting tool after an industrial electrochemical stripping process in an acidic electrolyte (pH 2), supports the laboratory findings. However, the industrial stripping process and the laboratory electrochemical measurements were performed under different conditions and should not be interpreted as directly equivalent. The industrial FIB/SEM observations are presented only as qualitative support for the corrosion mechanisms identified under controlled laboratory conditions. The cross-section reveals loose WC grains and loss of the cobalt binder matrix in the near-surface region. To further corroborate this behaviour, an additional SEM top-view micrograph of the same tool was analysed (Figure 7a). The surface image clearly shows exposed WC grains and the pronounced depletion of the cobalt matrix.

4.4. Surface Characterisation of WC–6Co After Potentiodynamic Polarisation in Alkaline Electrolyte

Surface profile measurements performed after the potentiodynamic polarisation in the alkaline electrolyte (pH 13) revealed neither the formation of a thick surface layer nor any evidence of film breakdown. The average material removal from the surface was approximately 0.3 µm. Local profilometric analysis of smaller areas revealed depressions up to 3 µm in depth. A representative local profile is shown in Figure 8.

However, pronounced leaching of WC grains resulted in significant surface roughening, which, in industrial applications, typically necessitates subsequent micro-blasting to remove the remaining loose, soft cobalt-rich matrix. This treatment does not increase the actual material removal depth but eliminates the mechanically weakened cobalt matrix until the original carbide structure is reached.

As shown in Figure 9, the exposed surface exhibited a faint pinkish coloration, which is consistent with cobalt hydroxide-containing corrosion products and suggest cobalt passivation under alkaline conditions. This observation is consistent with the Pourbaix predictions for the Co–H₂O system [5], where cobalt is thermodynamically stabilised as ${Co\left(OH\right)}_2$ in high-pH environments.

SEM micrograph and EDS spectrum analyses (Figure 10) further show leached WC grains and a Co–O-rich matrix (highlighted in red), consistent with the expected corrosion behaviour under alkaline conditions and with the predictions from the Pourbaix diagrams [5].

It should be noted that the parameters of the potentiodynamic polarisation process differ significantly from those used during industrial electrochemical stripping, where much higher voltages are applied for only a few seconds or minutes. In contrast, during the laboratory polarisation measurement, the samples were exposed to lower potentials but for significantly longer duration, which resulted in more pronounced surface damage. Nonetheless, these measurements provide valuable insights into the corrosion mechanisms occurring during the stripping process.

5. Discussion

Electrochemical investigations of WC–6Co in acidic (pH 2) and alkaline (pH 13) electrolytes revealed distinct corrosion mechanisms that are consistent with thermodynamic predictions from Pourbaix diagrams [5]. In the acidic electrolyte, the electrochemical response, SEM/EDS analyses, and surface morphology indicate preferential dissolution of the cobalt binder together with the formation of a W–O-rich surface layer. Based on the Pourbaix predictions for tungsten in acidic media, this layer is attributed to tungsten oxide corrosion products, likely including ${WO}_3$-rich species. This interpretation is consistent with previous work by Sutthiruangwong et al. [4], who reported that cobalt exhibits a lower corrosion potential and higher corrosion current density than WC, while WC shows a more noble corrosion potential and the occurrence of pseudo-passive plateaus. The difference in electrochemical nobility establishes a galvanic couple in which WC acts as the cathodic phase and cobalt as the anodic phase, thereby promoting selective cobalt dissolution in acidic environments.

In the alkaline electrolyte, the surface appearance, electrochemical response and SEM/EDS results indicate a different corrosion mechanism. Tungsten is thermodynamically unstable under alkaline conditions and is therefore interpreted to dissolve via soluble tungstate species (${{WO}_4}^{2-}$), whereas the observed Co–O-rich surface layer is attributed to cobalt hydroxide-containing corrosion products, likely including ${Co\left(OH\right)}_2$, in agreement with the alkaline Pourbaix stability regions [5]. Although this hydroxide-containing layer may locally reduce cobalt dissolution, the low cobalt content of the WC–6Co substrate is insufficient to stabilise the overall surface. As a result, degradation remains dominated by tungsten dissolution and progressive WC grain leaching.

A further distinction between the two electrolytes is reflected in their anodic current response. Using 5 mA·cm⁻² [17] only as a comparative benchmark for pronounced anodic activity, the acidic electrolyte reaches this value at lower potentials than the alkaline electrolyte, indicating an earlier onset of strong anodic dissolution under the applied polarisation conditions. At the same time, the polarisation behaviour suggests that surface-layer formation on WC–Co does not necessarily imply reliable corrosion protection. In the acidic electrolyte, the W–O-rich corrosion-product layer causes temporary current attenuation. However, the corresponding current densities remain far above those expected for a true passive state. The behaviour is therefore more appropriately described as pseudo-passivation rather than as true passivation.

At higher anodic potentials in the acidic electrolyte, the current increases again, indicating renewed anodic dissolution rather than a transpassive transition. A transpassive region cannot be clearly assigned in the present system because no well-defined passive state is established beforehand. The high-potential response is instead interpreted as continued oxidation and growth of a porous W–O-rich corrosion-product layer, followed by mechanical instability, local breakdown, and partial detachment. This interpretation is consistent with the place-exchange and field-assisted ion-transport mechanism proposed by Sato [16], which describes oxide growth through coupled migration of oxygen ions and metal ions during anodic polarisation. Once the W-O-rich corrosion-product layer fractures or detaches, fresh substrate is exposed and further material loss occurs.

Overall, the corrosion behaviour of WC–6Co is governed not only by electrochemical conditions, but also by the evolving morphology and mechanical stability of the surface layers formed during anodic polarisation. The combined evidence from electrochemical measurements, SEM/EDS analyses, profilometry, and FIB cross-sections supports the conclusion that acidic and alkaline electrolytes result in fundamentally different corrosion behaviours in WC–Co hard metals [4,8].

6. Conclusions

The corrosion behaviour of WC–6Co under electrochemical stripping conditions differs markedly between acidic and alkaline electrolytes. In the acidic electrolyte (pH 2), corrosion is dominated by preferential cobalt dissolution together with the formation of a W–O-rich corrosion-product layer attributed to ${WO}_3$-rich species. This layer causes only temporary current attenuation and is therefore more appropriately described as pseudo-passive rather than truly passive. At higher anodic potentials, the layer becomes increasingly porous and mechanically unstable, resulting in local breakdown, renewed anodic dissolution, and significant material loss.

In the alkaline electrolyte (pH 13), tungsten is thermodynamically unstable and corrosion is interpreted to proceed predominantly via the formation of soluble tungstate species (${{WO}_4}^{2-}$). At the same time, a Co–O-rich surface layer attributed to cobalt hydroxide-containing corrosion products such as ${Co\left(OH\right)}_2$ forms on the binder phase. However, this layer is insufficient to stabilise the overall WC–6Co surface, and degradation remains dominated by tungsten dissolution and WC grain leaching. These findings are consistent with the thermodynamic stability regions predicted by Pourbaix diagrams [5] and indicate that surface-layer formation on cemented carbides does not inherently guarantee corrosion protection.

Overall, the results show that both electrolytes induce distinct corrosion mechanisms and that the formation of surface layers alone does not ensure protection in WC–Co cemented carbides. Understanding these mechanisms is essential for optimising electrochemical stripping and recoating strategies to ensure surface integrity and extend the service life of WC–Co tools.