1 Introduction

AISI 4140 steel is a low-cost low alloy (chromium–molybdenum) carbon steel. It is a general-purpose high-tensile steel with a wide field of applications, that mainly includes machinery parts and tools. However, AISI 4140 steel is vulnerable to corrosion and wear. Thus, stainless steels are preferable for applications in corrosive environments instead of the uncoated AISI 4140 steel, although they are much more expensive. AISI 4140 steel could become a much more competitive material after a surface hardening treatment and especially after a coating treatment such as nitriding, boriding, and borochromizing. Thus, the application field could be expanded to more demanding sectors such as the marine industry.

The conventional boride coatings provide satisfactory protection in plenty of technological applications. However, they show several disadvantages that affect their efficacy. Specifically, Kulka et al. (Ref 1) report high brittleness even in Fe2B monolayers, which is revealed under tensile or depressing stresses and abrupt loading and can lead to complete failure (delamination).

Regarding the resistance against electrochemical corrosion, Kayali et al. (Ref 2) showed that the corrosion current of boronized coatings is quite low in thin layers, which are formed under mild thermal processing conditions. However, the corrosion was increasing as the coating became thicker. At first sight, it was a peculiar behavior, but it is probably explained by the fact that in thicker layers it is more likely, even inevitable, for microcracks to appear. These locally limited failures, even if they are not detectable, can facilitate the diffusion of ions that cause corrosion into the substrate.

The cases mentioned above are not unique, but they are indicative of the major problems related to the limits of conventional boride coatings. For those important reasons, the boride coatings can be enriched with transition metals to form ternary borides in the general form of MxM′yBz. One of the most significant elements in forming ternary borides with iron is chromium.

These borides show excellent mechanical and anti-corrosion properties. According to Huang et al. (Ref 3) and Ma et al. (Ref 4), Cr-doped Fe2B shows enhanced hardness and reduced brittleness in comparison with the pure Fe2B phase. However, the results are controversial regarding the fracture toughness because Huang et al. (Ref 3) claimed that it increases in low Cr concentrations and decreases when the Cr concentration becomes higher. Ma et al. (Ref 4) though found that fracture toughness increases gradually in all Cr concentration ranges. Moreover, Üçisik et al. (Ref 5) claimed that Cr has a negative effect on fracture toughness. Similar results for hardness were also noted by Balandin (Ref 6).

The first attempt to create a mixed boride-chromium coating was made by Kolesnikov et al. (Ref 7). They noticed that some chromium borides were formed, and the mixed boride layer showed improved resistance against tensile stresses. Ten years later Grachev et al. (Ref 8) used a fluidized bed reactor to produce boride coatings with the presence of chromium on a variety of steels achieving high resistance against wear and corrosion. However, they did not mention the formation of any kind of chromium borides. Suwattananont and Petrova (Ref 9) studied the addition of VIB group metals (including chromium) in the boronizing process of low-carbon steel. They reported the formation of CrB and Cr2B as well as a significant increase in the coating thickness. They also noticed the containment arising from the undesirable deboronizing of the boronized surface. Moreover, Khafri and Abady (Ref 10), Samadi and Habibolahzade (Ref 11), and Bartkowska et al. (Ref 12) experimented with a borochromizing process consisting of two stages. The first stage was chromizing, which was succeeded by boronizing. The results were satisfactory. After all, the most recent and most comprehensive study on the microstructure of a mixed Fe-Cr-B coating was published by Zeng et al. (Ref 13). The most interesting part of their research was about the evolution of the microstructure of the coating during the thermochemical process. They concluded that in brief time processes the dominant phases in the coating layer were Fe2B and FeCrB. In moderate durations mostly FeCrB was produced and in long-time processes mainly binary chromium borides were formed.

Electrochemical Impedance Spectroscopy (EIS) is a powerful tool for investigating corrosion resistance because it can provide information about the qualitative characteristics of corrosion. The widely used potentiodynamic method instead gives mainly the magnitude of the corrosion effect on materials. However, only a relatively small number of experimental studies include EIS in their diagnostic methods for evaluating corrosion resistance of boride coatings on steels (Ref 14,15,16,17,18,19,20,21,22) and some on non-ferrous alloys (Ref 23,24,25,26,27,28), most of them performed by the same researchers during the decade of 2010–2020. It is notable that in none of these studies a borochromized coating was investigated.

In the present study, we aimed to investigate the influence of the short-term exposure in 10% vol. acid solutions of HCl and H2SO4 on the corrosion resistance of boronized and borochromized layers developed on AISI 4140 steel. The borochromizing technique we used, which is described in detail in our previously published article: (Ref 29), leads to the formation of a double boride layer consisting of an external (Fe,Cr)B layer grown on an already existing Fe2B coating. We took advantage of the powerful EIS technique to examine the mechanism of corrosion in the boronized and borochromized coatings and their progress through time of exposure to a corrosive environment. This way, we tried to expand the field of EIS applications to the corrosion evaluation of multicomponent coatings with boron and chromium.

2 Materials and Methods

Specimens of AISI 4140 steel were cut from a rod into cylindrical pieces of about 2 cm in diameter and 0.5 cm in length. Half of them underwent a single boronizing treatment at 950 °C for 8 h. The rest of them underwent the borochromizing treatment (boronizing at 950 °C for 8 h and chromizing at 1050 °C for 6 h). Both thermochemical treatments were conducted by using the pack cementation method. The chemical composition of a typical AISI 4140 steel is presented in Table 1.

Table 1 Chemical composition of AISI 4140 steel
Full size table

The boronizing agent was a powder mixture consisting of boron carbide (B4C, 230 mesh, Sigma-Aldrich), potassium tetrafluoroborate (KBF4, powder, Alpha Aesar, > 99.5%), and silicon carbide (SiC, 400 grit, Struers). Moreover, the chromizing powder consisted of ferrochrome (Fe, 325 mesh, Acros Organics, 99%, and Cr, powder, Acros Organics, > 99%), ammonium chloride (NH4Cl, powder, Merck, 99.8%), and aluminum oxide (Al2O3, powder, Panreac, > 99.5%). The quantities and the purpose of use of each component are referred to in Table 2.

Table 2 Specifications of the boronizing and chromizing powders
Full size table

A conventional CARBOLITE CWF 1300 furnace was facilitated for the boronizing and borochromizing treatments, and the experiments were performed under an argon atmosphere. Before the corrosion tests, the boronized or borochromized samples were examined via XRD and SEM/EDS. XRD measurements were conducted via a BRUKER D9 diffractometer, which is equipped with a Cu Ka (1.5406 Å) radiation system. For SEM/EDS analysis a JEOL JSM-6390LV scanning electron microscopy system was utilized, which is additionally equipped with an Oxford INCA EDS system and software. Afterward, the specimens were subjected to a corrosion test by immersion in an acid bath that contained 10% vol. HCl or H2SO4 for 0.5-4 h. After each 0.5 h step, the specimens were measured via EIS. EIS analysis was conducted in a frequency range of 5 × 10–2-105 Hz and 10 mV voltage amplitude. For the EIS experiments, a VoltaLab PGZ 402 potentiostat was used, which was controlled by a Voltamaster 4 software package. All EIS experiments were performed in an electrochemical cell consisting of three electrodes at room temperature (25 °C), using a 3.5% w/v NaCl aqueous solution. A saturated calomel electrode (SCE) and a platinum plate were used as reference and counter electrodes, respectively. The experimental installation of the electrochemical corrosion apparatus complied with ASTM G69-20 and ASTM G71-81 standards. To ensure a steady state, before each EIS experiment, the samples were immersed in the solution for at least 30 min at open circuit potential (OCP). The above-referred ASTM standards impose the 3.5% w/v saline solution to be the electrolytic medium for the conduction of the EIS experiments. Possible contamination of the saline or acidic solution by traces of each other was avoided by cleaning the samples with deionized water properly and drying them right after the removal from each solution.

3 Results and Discussion

In Fig. 1, metallographic images in a magnitude of 200× and 500× of boronized and borochromized specimens are depicted. In the SEM images of Fig. 2, one can observe that a coarse superficial porosity appeared on the boronized specimen. This porosity was created on the surface of the single boride coating because the boron atoms are diffused inward with a different diffusivity in comparison with the iron atoms that are diffused outward. This condition favors the accumulation of cavities due to their supersaturation and, subsequently, the formation of macroscopic porosity (Ref 30). This is known as the Kirkendall effect. As the Kirkendall effect occurs when two or more atom species diffuse in opposite directions, the development of porosity near the surface, where the diffusion is more intense, is inevitable. As also can be observed in Fig. 2, the superficial porosity is absent in the case of the borochromized specimen. Additionally, the coating–substrate interface has lost the tooth-like morphology, which appeared in the previous stage of boronizing. It is presumed that the elimination of porosity and the alteration of the coating–substrate interface are both due to the local boron diffusion phenomena occurring at the extremely high temperature of the chromizing stage (1050 °C). Moreover, in the borochromized specimen a compact 20-30 μm thick external layer can be distinguished mostly in the optical image (Fig. 1), but also in the SEM image (Fig. 2). This layer is consisted mostly of the mixed iron-chromium boride Fe1.1Cr0.9B0.9 (hereinafter this phase will be referred to as (Fe,Cr)B) near the surface and a single Fe2B layer below it, as the XRD measurements confirm (Fig. 3).

Fig. 1
figure 1

Optical images of the boronized and the borochromized specimens

Full size image
Fig. 2
figure 2

SEM images of the boronized and the borochromized specimens

Full size image
Fig. 3
figure 3

XRD pattern of the surface of a boronized (up) and a borochromized (down) specimen

Full size image

Although a few peaks of the boronized specimen (Fig. 3 up), such as those in 2θ = 45.1, 50.7, and 57.7 degrees correspond also to the FeB phase (021), (121), and (211) planes, respectively, the presence of the FeB phase in the boronized layer cannot be asserted, since the strongest peaks (111), (210) and (101) at 2θ = 41.3, 47.8, and 37.7 degrees, respectively (according to 00-900-8944 card from COD), do not appear in the XRD diagram.

The Nyquist plots of the boronized and borochromized specimens for all corrosion times are presented in the next diagrams (Fig. 4).

Fig. 4
figure 4

Nyquist plots after various corrosion durations

Full size image

The boronized specimens show an initial (before corrosion) Nyquist curve at about 103 \(\Omega\,{\mathrm{cm}}^{2}\) and the borochromized ones at about 3-4*103 \(\Omega \, {\mathrm{cm}}^{2}\). These values are in the same order of magnitude as those of other researchers, who have worked on the corrosion evaluation of boride coatings by EIS: (Ref 1,2,3,4). For comparison reasons, a representative impedance value of a non-borided carbon steel is about 160 \(\Omega \, {\mathrm{cm}}^{2}\) (Ref 1). Thus, the beneficial effect of the boriding and especially the borochromizing process in the corrosion resistance of steels is obvious.

The behavior of the borochromized specimens during corrosion was quite different in comparison with the boronized ones. Particularly, their impedance started with remarkably higher values than the boronized samples, but it dropped quickly after 1-1.5 h of immersion and eventually tended to reach the level of the impedance of the boronized specimens.

The shape of the Nyquist curves of the borochromized specimens (Fig. 4c and d) appears to be extended along the real part axis direction, diverging from the almost ideal semicircular form of the boronized specimens (Fig. 4a and b). The Bode diagrams of the boronized and the borochromized specimen after immersion in 4 h in HCl are presented in Fig. 5. The previously described behavior of the Nyquist curves of the borochromized specimens is clearer in these conditions and thus, they are selected to be shown in the Bode arrangement.

Fig. 5
figure 5

Bode diagrams of the boronized and the borochromized specimen after immersion in HCl for 4 h

Full size image

As shown in Fig. 5, in the Bode diagram of the borochromized specimen two separated gradients are distinguished, each one of them corresponding to a different time constant of the system. Each time constant corresponds to a different ohmic-capacitive loop in the equivalent electrical circuit. Based on this observation, the equivalent electrical circuit that we propose as appropriate to simulate the boronized and the mixed borochromized layer is presented in Fig. 6.

Fig. 6
figure 6

Equivalent electrical circuit for the boronized and the borochromized layer

Full size image

Other researchers also propose similar equivalent circuits to simulate boride coatings, e.g., in the following studies: (Ref 17, 18, 25).

The R1 element corresponds to the resistance of the electrolyte, it usually has small values, and it is not important for understanding the electrochemical phenomena that are related to the coating layer. The R3 element describes the charge transfer resistance of the superficial porosity network, and it is an important factor in the corrosion resistance of the overall system. Also, the capacitive load which is accumulated on the coating/substrate interface is represented by the imperfect capacitor Q2. The selection of an imperfect capacitor instead of an ideal one is explained by the fact that the interfaces do not function as two completely separated plates. On the contrary, charged particles can pass from one side to the other giving a partially ohmic character to the interfaces’ system. The imperfect capacitor is described by the constant phase element’s (CPE) admittance formula: Y = Q(jω)α, where Q and α are frequency-independent parameters. A pure resistor is characterized by α = 0, a perfect capacitor by α = 1, and imperfect capacitors lie between 0 and 1. The units of parameter Q are complicated and non-standard. Specifically, the unit is Siemens*cm−2*(second)α, so that the admittance Y is given always in Siemens*cm−2, but the parameter α can vary between 0 and 1. The external loop R2–Q1, around the R3–Q2 loop, represents the capacitive and ohmic behavior of the coating itself.

After the execution of fitting calculations (the fitting curves are provided as supplemental material), using an appropriate software package, we concluded with some interesting observations. These observations refer to the evolution of the values of the ohmic elements in the two types of coating during corrosion (Fig. 7), as well as the parameters Q and α of the imperfect capacitors (Fig. 8).

Fig. 7
figure 7

Evolution of the ohmic elements during the corrosion

Full size image
Fig. 8
figure 8

Evolution of the imperfect capacity parameters Q and α during the corrosion

Full size image

As it is shown in Fig. 7a, the resistance of the solution (R1) had low values in all cases, as expected, and remained stable during the corrosion process, except for the case of corrosion of the boronized specimen in H2SO4, where a slight increase in the R1 resistance can be observed. This increase could be caused by a respective decrease in the ionic concentration in the solution or it could be only a side effect of the fitting calculations process. Either way, this is an insignificant variation that is not directly correlated with the condition of the coating itself.

From Fig. 7b is also extracted that a negligible value of no more than 6 \(\Omega \, {\mathrm{cm}}^{2}\) corresponds to the resistor R3 in the boronized specimens. This leads to the conclusion that the existence of the internal R3–Q2 loop is unnecessary in the equivalent circuit of the simple boride coating. Thus, the R3–Q2 loop must be removed, so that the circuit will be simplified to a single R–Q loop (the R2–Q1 loop), which now represents the coating itself without the contribution of porosity. According to this modification, the coarse and open porosity of the boronized specimens does not play a significant role in the total electrochemical behavior of the Fe2B coating. This could be explained by presuming that the pores are large and open enough so that they are not capable of accumulating charges and creating an additional capacitive double layer inside them.

The evolution of the R3 resistance of the borochromized specimens during the corrosion process is presented in Fig. 7b. It seems that R3 values were not significantly altered by the corrosion since they varied approximately between 2500 and 5500 \(\Omega \, {\mathrm{cm}}^{2}\) without any clear trend in increase or decrease during the corrosion process. Maybe only a slight decrease in the case of HCl would be observed. The overall observations in Fig. 7b are the following: (1) the hardly detectable via microscopy porous part of the mixed (Fe,Cr)B coating appears to contribute significantly to the total impedance, and (2) the coarse and open porosity of the iron boride layer, which is clearly visible through the optical and SEM images (Fig. 1 and 2, respectively), is not dense enough to accumulate and hold capacitive charges and thus it does not contribute to the total impedance at all.

The most important result regarding the resistance comes from Fig. 7c, where the resistance of the coating in the borochromized and boronized specimens is presented. In both HCl and H2SO4 cases, the resistance of the coating (R2) in the borochromized specimens has an initial value of 7200-7850 \(\Omega \, {\mathrm{cm}}^{2}\), and subsequently, it was significantly decreased during the first 1.5 h of corrosion and stabilized at a very low value (1100-1250 \(\Omega \, {\mathrm{cm}}^{2}\) for HCl and 0-150 \(\Omega \, {\mathrm{cm}}^{2}\) for H2SO4) for the rest of the duration of the corrosion process. As shown in Fig. 1, the compactness of the internal layer of the borochromized specimen (50-150 μm beneath the surface) was significantly decreased due to the inevitable local boron diffusion during the chromizing process. Thus, the only compact layer of the borochromized coating was the external (Fe,Cr)B layer. Once this layer was penetrated by the corrosive ions, the corrosion rapidly proceeded to the base metal through the non-compact Fe2B sublayer. This could explain the above-described elimination of the coating resistance of the borochromized specimen.

The values of the Q and α parameters of the coating region of the boronized and the borochromized specimens are presented in Fig. 8(a) and (b). The boronized layers appear with a more stable capacitive character (Fig. 8b) and can hold larger capacitive loads (Fig. 8a) than the borochromized ones. Moreover, the capacitive character of the borochromized coatings tends to turn into ohmic during the corrosion with H2SO4. This results from the fact that the α1 parameter decreased from 0.8 to 0.5 (Fig. 8b). This transition from capacitive to ohmic behavior could also be correlated with the decrease in the ohmic resistance of those coatings (R2). This possible correlation is based on the fact that the decrease in the capacitive character of the coating leads to a higher current passing through the coating, especially in low frequencies, and corroding it.

SEM images of the surface of the boronized and borochromized specimens before corrosion are presented in Fig. 9.

Fig. 9
figure 9

SEM micrographs of the surface of the boronized (up) and borochromized (down) coatings before the corrosion test

Full size image

The roughness of the boronized coating’s surface is intense and some microcracks are present, too. Contrariwise, the surface of the borochromized coating is smoother and no microcracks were detected. This finely structured and complicated surface morphology can explain the initially high resistance of the porosity of the borochromized specimens.

Moreover, a concentration profile of chromium across the coating layer of the borochromized specimen, obtained by EDS analysis, is presented in Fig. 10.

Fig. 10
figure 10

Concentration gradient of chromium (SEM/EDS) across the borochromized coating before corrosion

Full size image

As is depicted in Fig. 10, the diffusion depth of chromium is ~ 50 μm from the surface, which means that below this depth only unaffected Fe2B from the boronizing stage can exist. However, this underlying Fe2B layer could not be detected by XRD (Fig. 3, down) because the penetration depth (skin depth) of x-rays at Cu Ka wavelength (1.5406 Å) in a ferrous substrate is approximately 63 μm. This result can be obtained by the fact that the attenuation coefficient of x-ray radiation at 8.0478 keV (which corresponds to a wavelength of 1.5406 Å) is μ = 2036 cm−1 (Ref 31). By using Lambert–Beer’s law of absorption (I = I0 exp(− μ x)), it can be obtained that at 63 μm below the surface the intensity of radiation is diminished to the 1/e fraction (~ 37%) of its initial value. For the above calculation, the bulk density of the Fe2B phase (7.43 g/cm3) and an angle of incidence for the x-rays equal to the one that corresponds to the main peak of the pattern of Fe2B (2θ = 45° or θ = 22.5°) were considered (see Fig. 3, up). According to this, even the main peak of the underlying Fe2B phase was unlike to be recorded.

A semi-quantitative chemical analysis of the surface of the borochromized specimen before corrosion is presented in Fig. 11.

Fig. 11
figure 11

Chemical analysis of the surface of the borochromized specimen before corrosion

Full size image

The results of the chemical analysis showed that the initial superficial concentration of chromium was 70 ± 4 at. % (normalized to Fe and Cr only).

The chemical analysis of the surface of the borochromized coatings was repeated after the corrosion tests. The EDS results are presented in Fig. 12.

Fig. 12
figure 12

Chemical analysis of the surface of the borochromized specimen after corrosion in HCl and H2SO4

Full size image

Notably, the atomic percentage of chromium on the surface of the borochromized specimens was diminished after the corrosion tests in HCl or H2SO4. In particular, the concentration of chromium after the corrosion test with HCl was 55.8 ± 0.1 at.% and with H2SO4 was 57.0 ± 0.1 at.%. These values are approximately 20% lower than the concentration of chromium on the unaffected coating. This fact indicates a dechromizing procedure occurred on the surface of the borochromized coatings and it can be correlated with the decrease in their resistance. The possibility of a part of Cr being consumed for the formation of Cr2O3 on the surface can be excluded because it would still be detectable by EDS. Moreover, a Cr2O3 film on the surface would cause passivation effects and eventually an increase in the total impedance of the coating, and not a decrease in it. However, the decrease in the superficial chromium concentration, in both HCl and H2SO4 cases, could be sufficiently explained by the possibility of the formation of water-soluble chromium(III) aquo complex ions \({\left[\mathrm{Cr}{\left({\mathrm{H}}_{2}\mathrm{O}\right)}_{6}\right]}^{3+}\).

Particularly, in a Cl-rich environment, the spontaneously formed chromium(III) hexahydrate cations \({\left[\mathrm{Cr}{\left({\mathrm{H}}_{2}\mathrm{O}\right)}_{6}\right]}^{3+}\) are in equilibrium with partially Cl-substituted complexes as follows (Ref 32):

$${\left[\mathrm{Cr}{\left({\mathrm{H}}_{2}\mathrm{O}\right)}_{6}\right]}^{3+}+{\mathrm{Cl}}^{-}\rightleftharpoons {\left[\mathrm{Cr}{\left({\mathrm{H}}_{2}\mathrm{O}\right)}_{5}\mathrm{Cl}\right]}^{2+}+{\mathrm{H}}_{2}\mathrm{O}$$
(1a)
$${\left[\mathrm{Cr}{\left({\mathrm{H}}_{2}\mathrm{O}\right)}_{5}\mathrm{Cl}\right]}^{2+}+{\mathrm{Cl}}^{-}\rightleftharpoons {\left[\mathrm{Cr}{\left({\mathrm{H}}_{2}\mathrm{O}\right)}_{4}{\mathrm{Cl}}_{2}\right]}^{+}+{\mathrm{H}}_{2}\mathrm{O}$$
(1b)
$${\left[\mathrm{Cr}{\left({\mathrm{H}}_{2}\mathrm{O}\right)}_{4}{\mathrm{Cl}}_{2}\right]}^{+}+{\mathrm{Cl}}^{-}\rightleftharpoons \left[\mathrm{Cr}{\left({\mathrm{H}}_{2}\mathrm{O}\right)}_{3}{\mathrm{Cl}}_{3}\right]+{\mathrm{H}}_{2}\mathrm{O}$$
(1c)

The higher the Cl− concentration is the more favorable the formation of the Cl-substituted complexes is. The neutral complex entity \(\left[\mathrm{Cr}{\left({\mathrm{H}}_{2}\mathrm{O}\right)}_{3}{\mathrm{Cl}}_{3}\right]\) (1c) can be formed in very high Cl− concentrations such as these in our experiments. All those chromium aqua complexes are quite soluble in water and can theoretically support the chromium leakage from the coating surface.

Also, an elevated concentration of chlorine is detected in Fig. 11, due to the adsorption of Cl− ions during corrosion with HCl. As it is presented in Fig. 12, a local chemical analysis of various spots on the surface of this specimen revealed that most of the adsorbed chlorine ions are located inside cavities created during corrosion. These cavities are distinguished as dark regions in the SEM images of Fig. 13 (spectrum 2 and spectrum 3). The Cl− ions accumulated in these cavities and promoted corrosion.

Fig. 13
figure 13

Spot chemical analysis on the surface of the borochromized specimen after corrosion in HCl

Full size image

The elevated capacitance of the double layer (Q2) in the case of the borochromized coating immersed into HCl can be attributed to the adsorption of the Cl− ions on the coating as shown in Fig. 13.

According to Márquez-Herrera et al. (Ref 33), the main electrochemical reactions correlated with corrosion (anodic reactions, oxidation) are:

$$\mathrm{Fe}\stackrel{{\mathrm{H}}^{+}}{\to }{\mathrm{Fe}}^{2+}+2 {\mathrm{e}}^{-},\mathrm{ Cr}\stackrel{{\mathrm{H}}^{+}}{\to }{\mathrm{Cr}}^{2+}+2 {\mathrm{e}}^{-},\mathrm{ B}\stackrel{{\mathrm{H}}^{+}}{\to }{\mathrm{B}}^{3+}+3 {\mathrm{e}}^{-}$$
(2)

Simultaneously, the cathodic reaction (reduction) occurs when oxygen produces OH− as follows:

$${\mathrm{O}}_{2}+2 {\mathrm{H}}_{2}\mathrm{O}+4 {\mathrm{e}}^{-}\to 4 {\mathrm{OH}}^{-}$$
(3)

In the case of HCl as the corrosive substance, the Cl− ions can contribute to the oxidation, according to the following reactions:

$$\mathrm{Fe}+2 {\mathrm{Cl}}^{-}+4 {\mathrm{H}}_{2}\mathrm{O}\to {\mathrm{FeCl}}_{2}.4{\mathrm{H}}_{2}\mathrm{O}+2 {\mathrm{e}}^{-}$$
(4a)
$${\mathrm{FeCl}}_{2}.4{\mathrm{H}}_{2}\mathrm{O}+2 {\mathrm{OH}}^{-}\to \mathrm{Fe}{\left(\mathrm{OH}\right)}_{2}+2 {\mathrm{Cl}}^{-}+4 {\mathrm{H}}_{2}\mathrm{O}$$
(4b)
$$4\mathrm{ Fe}{\left(\mathrm{OH}\right)}_{2}+2 {\mathrm{H}}_{2}\mathrm{O}+{\mathrm{O}}_{2}\to 4\mathrm{ Fe}{\left(\mathrm{OH}\right)}_{3}$$
(4c)

It is important to emphasize that the Cl− ions play a catalytic role in the anodic reaction sequence (4a–c) because, while they are being consumed in the reaction (4a), they are recovered in the reaction (4b). Moreover, Fe2+ in the form of Fe(OH)2 can be further oxidized to Fe3+ in the presence of oxygen (4c) and form Fe(OH)3.

As proposed by Rosas-Beccera et al. (Ref 25) and Mejía-Caballero et al. (Ref 34), when H2SO4 is used as the corrosive agent, the SO42− ions can produce another anodic reaction, which includes the boron contained in the boride compounds of the coating layer:

$$24 {\mathrm{H}}^{+}+8\mathrm{ B}+3 {{\mathrm{SO}}_{4}}^{2-}\to 8 {\mathrm{B}}^{3+}+3 {\mathrm{S}}^{2-}+12 {\mathrm{H}}_{2}\mathrm{O}$$
(5)

This additional corrosion mechanism (5) in the case of H2SO4 as a corrosive agent could be responsible for the extreme decrease in the resistance of the coating of the borochromized specimen corroded with H2SO4 (Fig. 7c).

The cathodic reactions during the pre-corrosion stage of immersion in acid solutions could be the following:

$$2 {\mathrm{H}}^{+}+2 {\mathrm{e}}^{-}\to {\mathrm{H}}_{2}\uparrow$$
(6a)
$${\mathrm{O}}_{2}+4 {\mathrm{H}}^{+}+4 {\mathrm{e}}^{-}\to 2 {\mathrm{H}}_{2}\mathrm{O}$$
(6b)

without (6a) and with (6b) the contribution of the atmospheric oxygen, respectively. The role of the reactions (6a–b) is to neutralize the H+ cations produced by the ionization process of the HCl and H2SO4 acids and absorb the electrons produced by the anodic reactions.

Additionally, some local failures of the boronized and borochromized coatings that are detected via SEM are depicted in Fig. 14. Most of those failures are microcracks, shallow cavities, and limited delamination.

Fig. 14
figure 14

Post-corrosion local failures on (a) boronized in HCl, (b) boronized in H2SO4, (c) borochromized in HCl, and (d) borochromized in H2SO4

Full size image

4 Conclusion

An 80-100 μm thick tooth-like single-phase Fe2B coating was formed as a result of a boronizing process at 950 °C for 8 h. A subsequent chromizing process at 1050 °C for 6 h modified the Fe2B layer to a 160 μm thick mixed Fe-Cr boride coating, which includes a compact external layer consisting mainly of Fe1.1Cr0.9B0.9 and a disturbed tooth-shaped Fe2B zone beneath.

The borochromized layers showed much higher electrical impedance values before the corrosion test than the boronized ones. This is attributed to the bulk part of the coating, the contribution of which was eliminated after 1.5 h of corrosion, in the case of the borochromized coatings. The contribution of the porous region of the coating was not significantly changed during the corrosion in all types of coating and corrosive agents.

The equivalent electrical circuit that was adopted in the simulation was an R–Q loop for the coating, which includes another R–Q loop for the porosity. This circuit simulated satisfactorily the behavior of the borochromized specimens, but, in the case of the boronized ones, the internal loop, corresponding to the porosity, showed negligible impedance values. For this reason, the equivalent circuit of the boronized coatings could be replaced by a single R–Q loop, the purpose of which is to simulate the bulk region of the coating only.

SEM images and EDS analysis before and after the corrosion process showed that the surface of all coatings suffered damage, which included chromium loss and corrosive ion adsorption for the borochromized specimens and microcracks and light delamination for the boronized ones.