Inhibition efficiency and mechanism of nitrilo-tris(methylenephosphonato)zinc on mild steel corrosion in neutral fluoride-containing aqueous media

1 Introduction

Fluorine is widespread in nature with its content in the earth’s crust ranging from 0.025 to 0.10 wt% (Koga and Rose-Koga 2018), which makes it commonly used in various fields, including development and processing of apatite and phosphorite minerals, metallurgy, cement production, construction, medicine, and many others (Dai et al. 2021). Nevertheless, steel corrosion in neutral fluoride-containing aqueous media is now less studied compared to other halide-containing media, and efficient technologies for its inhibition have not yet been properly developed to date. So, mechanisms of corrosion inhibition of engineering materials in fluoride-containing media are extremely relevant to be studied.

The fluoride ion differs markedly from other halide ions in its effect on the structure of an aqueous corrosive medium. Among halide ions, F⁻ ion has the smallest ionic radius (Dai et al. 2021), the largest electronegativity (Shannon 1976), the lowest polarizability (Pauling 1932; Tantardini and Oganov 2021), and the highest chemical hardness (Parr and Pearson 1983; Pearson 1985). In aqueous solution, fluoride ions interact favorably with the surrounding water molecules to create a thick and structured hydration shell, contrarily to other halides (Dai et al. 2021; Trompette 2014; Trompette et al. 2010).

There still remains significant disagreement within the previously published data regarding the effect of fluoride ion on the steel corrosion. On the one hand, it is considered that a large hydration shell and high dehydration energy prevent fluoride ion from initiating pitting corrosion (Vásquez Moll et al. 1985), but the fluoride ion, on the other hand, has the lowest pitting potential among the halide ions (De Castro and Wilde 1979). However, the authors of (Ogura and Ohama 1981) has previously found a tendency of the pitting potential for iron to get more positive on going from SO₄²⁻ to Cl⁻ to Br⁻ to I⁻ to F⁻. It has been observed (Vásquez Moll et al. 1985) that fluoride ion is capable of pitting toward mild steel, but the pitting potential is more positive than in the presence of sulfate or chloride ions. At the same time, simultaneous presence of F⁻ and Cl⁻ ions in a corrosive medium has been noted to increase the corrosion rate for mild steel (Wang et al. 2020). It has been earlier shown (Singh et al. 2002) that the rate of corrosion of mild steel in alkaline fluoride-containing solutions is reduced at fluoride concentrations of 25 ppm and above.

The rate of steel corrosion may be reduced through the formation of passive layers on its surface. For the aforementioned reasons, the effect of fluoride ions on the formation of passive layers and their dynamic equilibrium with a corrosive medium is markedly different from that of other halide ions.

Within the framework of the adsorption theory (Kuznetsov et al. 2020), fluoride ions are expected to have lower specific adsorption because of their habit to retain the hydration shell, as compared to other halide ions (Trompette 2014; Trompette et al. 2010). Taking this expectation, Vásquez Moll et al. (1985) explain the lower corrosivity of fluoride ions and consider that inhibition of the pitting corrosion is initiated by competitive adsorption between F⁻ and either OH⁻ ions or H₂O molecules, followed by the reactions of oxide formation leading to repassivation of the metal surface. The destruction of the passive layer by fluoride ions is commonly indicated by the formation of soluble complexes with iron ions of complex composition (Khalil et al. 1985; Ogura and Ohama 1981; Strehblow et al. 1979), which is preceded by the adsorption of F⁻ ions.

Within the framework of the 3D thin film theory (Kuznetsov et al. 2020), the effect of fluoride ions on the steel passivation is considered in terms of structure and solubility of the resulting fluorine-containing phases. Both anhydrous iron(II) fluoride and iron fluoride hydrate are slightly soluble in water (Ferey et al. 1975; Perry 2011), while iron(III) fluoride, on the contrary, is highly soluble in water (Ferey et al. 1975). Iron(III) fluorides in aqueous media tend to replace fluoride ions with OH⁻ and O²⁻ ions resulting in the formation of complex hydroxo- and oxofluorides FeF_3−x(OH)_x·nH₂O and FeF_3−xO_x/2 (Burbano et al. 2018; Duttine et al. 2014; Lemoine et al. 2019b). Mixed Fe²⁺ and Fe³⁺ hydrate-fluoride complexes have been earlier noted (Lemoine et al. 2019a) to tend to precipitate as poorly soluble iron(III) hydroxofluorides under heating. Poorly soluble compounds in the passive film composition are apparently responsible for the increase in the corrosion resistance of steel in fluoride-containing media as compared to media containing other halide ions.

The most advanced conceptualization of passivation processes (Popov 2005) is based on the ideas about ionic electrodiffusion in films having composition varying through the thickness. Within the framework of the electrodiffusion theory, the effect of fluoride ions on the formation and destruction of passive films is mainly predicated by their ionic radius r(F⁻) = 1.30 Å which is noticeably smaller compared to 1.38 and 1.35 Å for O²⁻ and OH⁻ ions, respectively, and significantly smaller than for other halide ions Cl⁻, Br⁻, and I⁻, which are, respectively, 1.81, 1.96, and 2.06 Å (Shannon 1976).

Nitrilotris-methylenephosphonic acid (NTP) complexes have been previously found to demonstrate high anticorrosion activity (Chausov et al. 2020a; Demadis et al. 2005; Kuznetsov 1996; Somov and Chausov 2014). As has been recently shown (Chausov et al. 2020a), Zn and Cd NTP complexes possessing different coordination structures have different anticorrosion activity, other things being equal. Among various Zn NTP complexes, tetrasodium nitrilotris(methylenephosphonato)zincate tridecahydrate Na₄[Zn{N(CH₂(PO₃)₃}]·13H₂O (ZnNTP) with a chelate structure exhibits the highest inhibition efficiency for steel corrosion (Chausov et al. 2020a).

Passivation of mild steel surface in neutral aqueous media containing Cl⁻ ions and the mechanism of inhibiting action of ZnNTP has been earlier studied in (Kazantseva et al. 2022). The structure of ZnNTP is shown in (Somov and Chausov 2014). Counter diffusion of ZnNTP ions moving from the solution to the metal surface and Fe²⁺ ions moving from the metal surface to the solution proceeds through the oxide passive layer as follows (Chausov et al. 2020a):

resulting in polymeric heterometallic complex [Fe_1/2Zn_1/2(H₂O)₃μ − H₄{N(CH₂PO₃)₃}]_n (FeZnNTP), which is practically insoluble in water and has the structure shown in (Chausov et al. 2020b) and zinc hydroxide.

This paper studies the passivation of mild steel E 235 in neutral aqueous media containing fluoride ions and proposes the corrosion inhibition mechanism of ZnNTP on this steel based on the results obtained.

2 Materials and methods

3 Results and discussion

Supplementary Table S2 provides data for the potential and anodic dissolution current density at the characteristic points of the measured polarization curves.

Figure 1 shows the behaviour of the most characteristic polarization curves measured under polarization of steel E 235 specimens in BBS (pH = 7.4).

Figure 1:

Anodic polarization curves measured under polarization of steel in BBS (pH = 7.4) without and with added F⁻ ions, with F⁻ ion concentration (mmol/L) being numerically shown on the respective curves. Dependence of the current density for anodic metal dissolution, i, on the anode potential, E.

Anodic polarization curves measured from E 235 steel in pure BBS (pH = 7.4) reveal its electrochemical corrosion behaviour characterized by the following parameters: the open-circuit potential, E_oc, is −0.73 V, the primary passivation potential, E_pp, is −0.33 V/SSCE, and the transpassivity potential, E_tp, is 1.14 V. Critical anodic current density, i_c, which corresponds to starting the active-to-passive transition, is 1.17 A/m², and the minimum current density in the passive region, i_p, is 0.46 A/m². The addition of F⁻ ions in amount of 0.025–0.56 mmol/L slightly changes the behaviour of polarization curves which demonstrate higher values of i_c and i_p. With increasing F⁻ concentration up to 0.98 mmol/L and higher, the current density in the passive region grows significantly. E_tp changes insignificantly with adding 0.025–1.4 mmol/L F⁻ ions. At 2.8 mmol/L F⁻ ions, the curves in the potential range of 0.6–1.0 V demonstrate a segment with a relatively constant current density, which is hardly considered as a passive one because of high i values. E_oc does not change significantly with increasing concentration of F⁻ ions in the corrosive medium.

Thus, in contrast to Cl⁻ ions, the addition of which significantly reduces the ability of steel to resist corrosion in a corrosive medium when adding as low as 0.14 mmol/L Cl⁻ ions (Kazantseva et al. 2022), F⁻ ions induce significant changes in the corrosion-electrochemical behaviour of E 235 steel only at high enough concentrations exceeding 0.98 mmol/L.

Active-to-passive transition ranging from (−0.4) to (−0.1) V involves the following reactions:

(E ∼ −0.7 V/SSCE [Pourbaix 1974]),

(E ∼ −0.64 V/SSCE [Pourbaix 1974]),

In the potential range of (−0.1)–0.2 V, the current density increases, which may be related to oxidation reactions with involvement of Fe²⁺:

(E ∼  0.1 V/SSCE [Pourbaix 1974]),

(E∼ 0.2 V/SSCE [Doe et al. 2008; Wang et al. 2019]).

This segment of the passive region, which exhibits increasing current density caused by the oxidation of Fe²⁺ ions to Fe³⁺, may be termed the first segment of the passive region. The segment of the passive region ranging from 0.1–0.2 V up to E_tp may be, respectively, termed the second segment of the passive region.

OH⁻ ions needed for reactions (3²), (4), and (4¹) are formed in the oxygen reduction reaction.

(E ∼  0.2 V/SSCE [Bard et al. 1985]).

Figure 2 shows SEM micrographs of the surface of E 235 steel coupons polarized in BBS (pH = 7.4) containing 1.4 mmol/L F⁻ ions at potentials E = 0.3 (the first segment of the passive region) and E = 0.7 V (the second segment of the passive region). The surface of the coupon polarized at E = 0.3 V (Figure 2a) is covered with porous iron oxides and hydroxides visible as islands which are conjugated with increasing E and form a highly porous island film covering the specimen surface (Figure 2b). The composition of different surface areas of coupons is given in Supplement S3.

Figure 2:

SEM micrographs of the steel surface polarized in BBS (pH = 7.4) containing 1.4 mmol/L F⁻ ions at potentials E = 0.3 (a) and E = 0.7 V (b), taken at 7000x magnification.

Figure 3 shows anodic polarization curves measured from steel E 235 in BBS (pH = 7.4) with added 0.28 mmol/L F⁻ ions and ZnNTP inhibitor of different concentrations. The presence of the inhibitor in the corrosive medium reduces the values of critical current density and current density in the passive region, with i_c reducing to minimum at 0.5 g/L ZnNTP and gradually growing with further increasing ZnNTP concentration (Figure 3b). The value of E_ос, on the contrary, is maximal at 0.5 g/L ZnNTP. In the passive region, the higher C_inh in a corrosive medium, the lower value of i_p. The E_tp value tends to gradually reduce with increasing ZnNTP concentration.

Figure 3:

Anodic polarization curves measured from steel in BBS (pH = 7.4) with added 0.28 mmol/L F⁻ ions and various amounts of ZnNTP inhibitor (a). ZnNTP concentration (g/L) is numerically indicated against the respective curves. Dependence of the current density for anodic dissolution of metal, i, on the potential, E. The inset gives segments of the polarization curves shown as current density plotted on log-scale in the vicinity of the open circuit potential E_oc. Dependences of main parameters indicating the efficiency of metal protection, including critical current density of anodic metal dissolution, i_c, (curve 1), transpassivity potential, E_tp, (curve 2), and open circuit potential, E_oc, (curve 3) on the ZnNTP inhibitor concentration, C_inh (b).

The anodic polarization curves measured from steel E 235 in BBS (pH = 7.4) with added 1.4 mmol/L of F⁻ ions and different amounts of ZnNTP inhibitor are given in Figure 4. E_oc reaches its maximum at 0.2 g/L ZnNTP, but higher concentrations of the inhibitor reduce its value. The minimal i_c is observed at 0.5–1.0 g/L ZnNTP. A higher ZnNTP concentration induces a significant increase in i_c, but its value is not as higher as those observed in the inhibitor-free medium. The current density in the passive region decreases with increasing ZnNTP concentration. E_tp slightly drops with increasing C_inh.

Figure 4:

Anodic polarization curves measured from steel in BBS (pH = 7.4) with added 1.4 mmol/L F⁻ ions and different amounts of ZnNTP inhibitor (a). ZnNTP concentration (g/L) is numerically indicated against the respective curves. Dependence of the current density for anodic dissolution of metal, i, on the potential, E. The inset gives segments of the polarization curves shown as current density plotted on log-scale in the vicinity of the open circuit potential E_oc. Dependences of main parameters indicating the efficiency of metal protection, including critical current density of anodic metal dissolution, i_c, (curve 1), transpassivity potential, E_tp, (curve 2), and open circuit potential, E_oc, (curve 3) on the ZnNTP inhibitor concentration, C_inh (b).

The anodic polarization curves measured from steel E 235 in BBS (pH = 7.4) with added 5.6 mmol/L F⁻ ions and ZnNTP inhibitor are shown in Figure 5. Addition of 0.05–1.0 g/L ZnNTP leads to decreasing anodic dissolution current density in the active-to-passive transition region. With increasing ZnNTP concentration in the corrosive medium, i_c reduces down to its minimum at C_inh of 1 g/L and then drastically increases, and its value exceeding that observed in the inhibitor-free medium, which is apparently caused by the following processes:

Figure 5:

Anodic polarization curves measured from steel in BBS (pH = 7.4) with added 5.6 mmol/L F⁻ ions and different amounts of ZnNTP inhibitor (a). ZnNTP concentration (g/L) is numerically indicated against the respective curves. Dependence of the current density for anodic dissolution of metal, i, on the potential, E. The inset gives segments of the polarization curves shown as current density plotted on log-scale in the vicinity of the open circuit potential E_oc. Dependences of main parameters indicating the efficiency of metal protection, including critical current density of anodic metal dissolution, i_c, (curve 1), transpassivity potential, E_tp, (curve 2), and open circuit potential, E_oc, (curve 3) on the ZnNTP inhibitor concentration, C_inh (b).

These processes result in binding of Fe²⁺ into stable complexes, in the bulk of the electrolyte, not in the near-electrode layer. The current density in the potential range (−0.3)–1.2 V turns out to be significantly lower in the presence of the inhibitor than without it. The value of E_oc is maximal at 0.2 g/L ZnNTP and then drops with increasing inhibitor concentration in the solution.

In media containing both F⁻ ions and ZnNTP inhibitor, the same processes (2–5) occur, in general, as without the inhibitor. The decrease in i_c and i_p and the increase in E_tp in the presence of ZnNTP are due to partial transformation of the oxide–hydroxide layer into the FeZnNTP heterometallic polynuclear complex following the reaction (1). Since reaction (1) involves Fe²⁺ ions, the inhibitor exhibits its best protective properties in aerated media with oxygen as a natural oxidant.

Figure 6 shows SEM micrographs of the surface of E 235 steel coupons polarized in BBS (pH = 7.4) containing 1.4 mmol/L F⁻ ions and 5 g/L ZnNTP at potentials E = −0.05 (the first segment of the passive region) and E = 0.65 V (the second segment of the passive region). The data of elemental energy-dispersive analysis, the surfaces of the specimens are covered with a continuous film with iron oxides and hydroxides as the main components (Supplementary Table S3).

Figure 6:

SEM micrographs of the steel surface polarized in BBS (pH = 7.4) containing 1.4 mmol/L F⁻ ions and 5 g/L ZnNTP at potentials E = −0.05 (a) and E = 0.65 V (b) taken at 7000x magnification.

Supplementary Table S4 provides peak positions and integrated intensities for main peaks in the XPS spectra measured with layer-by-layer etching of passive films formed on steel E 235 in BBS (pH = 7.4). Figure 7 shows P2p XPS depth-profiling spectra for passive films formed on steel E 235 in BBS (pH = 7.4) with added 1.4 mmol/L F⁻ ions and 5 g/L ZnNTP inhibitor at potentials of −0.05 V (a) and 0.65 V (b) which correspond to the first and second segments of the passivity region, respectively. The reference XPS spectra measured from the initial ZnNTP inhibitor and FeZnNTP heterometallic complex taken from the previously published data (Chausov et al. 2020a) are also shown for comparison.

Figure 7:

P2p XPS spectra measured with layer-by-layer etching of passive films formed on the steel surface in BBS (pH = 7.4) containing 1.4 mmol/L F⁻ ions and 5 g/L ZnNTP inhibitor at potentials E = −0.05 (a) and E = 0.65 V (b) as photoelectron flux intensity depended on the binding energy. Numerical designations of the curves correspond to the etching depth of the passive film in nm. The reference spectra for the initial ZnNTP inhibitor and FeZnNTP heterometallic complex taken from (Chausov et al. 2020a) are also shown for comparison.

For the layers related to the interface between the passive film and inhibited corrosive medium, which were formed at both potentials of −0.05 V and 0.65 V, the measured P2p XPS spectra match that of the initial ZnNTP inhibitor measured as a reference. The only peak observed at a binding energy of 133.3–133.5 eV indicates the chemical equivalence of all phosphorus atoms in the PO₃ groups of the ZnNTP inhibitor, and each group being bound with a metal atom serving as an electron donor (Somov and Chausov 2014). Such a strong resemblance of the P2p XPS spectra measured from the surface passive layer to that of the reference ZnNTP inhibitor indicates its adsorption on this surface. As ZnNTP diffuses deep into the passive film, spectral features appear at 134.4–134.6 eV and 135.2–135.3 eV in the P2p XPS spectrum, with their integral intensity being related to each other as 2:1. These spectral contributions match those characteristic of the reference FeZnNTP complex and correspond to the states of phosphorus atoms in the chemically non-equivalent PO₃ groups in the FeZnNTP complex (Chausov et al. 2020a). Two PO₃ groups participate in the coordination of the metal atom serving as an electron donor and the respective P2p electrons have a lower binding energy of 134.4–134.6 eV, while the third PO₃ group is not coordinated by the metal atom and the respective binding energy is 135.2–135.3 eV. The appeared spectral contributions characteristic of the FeZnNTP complex indicate the reaction (1) between the initial ZnNTP inhibitor which diffuses from the inhibited corrosion medium deep into the passive film and Fe²⁺ ions which move in the opposite direction, from the passive film/metal surface interface to the passive film/corrosive environment interface. For the films formed at both −0.05 V and 0.65 V, the contribution at 133.3–133.5 eV assigned to the initial ZnNTP inhibitor is seen to be weaker due to its conversion in reaction (1). The diffusion of the ZnNTP inhibitor through the passive film is obviously promoted by decreasing its concentration in course of reaction (1). At the same time, the spectral features at 134.4–134.6 eV and 135.2–135.3 eV attributed to the reaction product (1) get stronger. At the potential of −0.05 V (Figure 7a), ZnNTP inhibitor is almost completely conversed into FeZnNTP complex, and the residual concentration of ZnNTP inhibitor at a depth of 15–30 nm being insignificant as compared to that of FeZnNTP. On the contrary, at the potential of 0.65 V (Figure 7b), the conversion of ZnNTP inhibitor into FeZnNTP complex turns out to be quite slow, and the intensity of the corresponding P2p spectral features changes less markedly compared to those observed for the films formed at −0.05 V for the same analyzed depths. For the potential of 0.65 V, the reaction (1) slows down sharply at a depth of 15–20 nm and is not completed.

Figure 8 shows the change in the atomic percentage of phosphorus in the composition of the initial inhibitor ZnNTP x(P_ZnNTP) and the reaction product (1) x(P_FeZnNTP) calculated based on the procedure described in Supplementary Table S5 as depended on the etching depth of the passive film.

Figure 8:

The atomic percentage of phosphorus in the initial ZnNTP inhibitor (curves 1, 3) and in the FeZnNTP heterometallic complex (curves 2, 4) as depended on the etching depth δ of the passive film obtained from the XPS analysis with layer-by-layer etching of passive films formed on steel in BBS (pH = 7.4) with added 1.4 mmol/L F⁻ ions and 5 g/L ZnNTP inhibitor at potentials E of −0.05 V (curves 1, 2) and 0.65 V (curves 3, 4).

For the passive film formed at −0.05 V, the behaviour of x(P_ZnNTP) (curve 1) and x(P_FeZnNTP) (curve 2) depended on the etching depth, δ, are seen to be close to exponential. This indicates that the order of reaction (1) with respect to the initial ZnNTP inhibitor is close to 1, which agrees with the reaction occurring under conditions of relatively constant concentrations of other reagents. For the passive film formed at 0.65 V, on the contrary, these dependences (curves 3 and 4) end at the etching depth of 10 nm because of probably the lack of Fe²⁺ ions needed for reaction (1), which are spent at 0.65 V in competing reactions (3¹), (3²), and (4).

The depth-profiling Fe2p XPS spectra measured from passive films formed in BBS (pH = 7.4) with added 1.4 mmol/L F⁻ ions and 5 g/L ZnNTP inhibitor at potentials of −0.05 V and 0.65 V are shown in Figure 9a and b, respectively.

Figure 9:

Fe2p_3/2 XPS spectra measured with layer-by-layer etching of passive films formed on the steel surface in BBS (pH = 7.4) with added 1.4 mmol/L F⁻ ions and 5 g/L ZnNTP inhibitor at potentials E = −0.05 (a) and E = 0.65 V (b) as photoelectron flux intensity depended on the binding energy. Numerical designations of the curves correspond to the etching depth of the passive film. The reference spectra for the initial ZnNTP inhibitor and FeZnNTP heterometallic complex taken from (Chausov et al. 2020a) are also shown for comparison.

Analysis of the measured Fe2p_3/2 XPS spectra shows that the films formed at −0.05 V and 0.65 V contain oxidized forms of iron predominating in the surface layers of the passive film of 1–5 nm, primarily Fe²⁺, with the maximum being observed at 709.3–710.5 eV. The spectra measured from 10 nm layer and deeper in the passive film formed at −0.05 V (Figure 9a) are dominated by metallic iron, with a maximum being observed at 706.7–707.2 eV. The oxidized forms are visible in the respective spectra as low-intensity maxima at 709.3–710.5 eV and 710.8–711.7 eV assigned to Fe²⁺ and Fe³⁺, respectively. On the contrary, the Fe2p_3/2 XPS spectra for the same etching depth of the passive film formed at 0.65 V (Figure 9b) contain a significant contribution of the oxidized form of Fe³⁺ in addition to metallic iron. This indicates reactions Fe(OH)₂ + OH⁻ = Fe(OH)₃ + e^– and Fe(OH)₂ + OH⁻ = FeOOH + H₂O + e^– occurring with increasing polarization potential from −0.05 to 0.65 V. The observed low percentage of Fe³⁺ ions in the surface layers (1–5 nm) of the passive film is apparently caused by the depassivation processes, Fe(OH)₃ + 3F⁻ = FeF₃ + 3OH⁻ and FeOOH + H₂O + 3F⁻ = FeF₃ + 3OH⁻, which, however, have no significant effect on the deeper layers of the passive film (deeper than 15 nm).

Figure 10 shows the curves for the atomic percentage of the main elements in the composition of passive films formed in BBS (pH = 7.4) with added 1.4 mmol/L F⁻ ions without an inhibitor (a) and with the additionally introduced 5 g/L ZnNTP inhibitor (b) as depended on the polarization potential. In a corrosive medium not containing an inhibitor (Figure 10a), iron and oxygen are the main components of the passive film, with the fluorine fraction being not higher than 5 at.% over all the potential range. In the region of active dissolution at the potential Е ranging from −0.75 V to −0.35 V, the concentration of iron and oxygen increases in the passive film, with the fluorine concentration growing slightly, which indicates the beginning of the formation of a passive oxide-hydroxide layer. In the first segment of the passivity region at E = 0.3 V, the fractions of iron and oxygen drop to their minima, with the fluorine concentration reaching its maximum, which is apparently associated with the passivation of the surface as a result of reaction (3¹). An increase in iron and oxygen fractions with simultaneous decrease in the fluorine fraction observed with further increasing the polarization potential up to 0.7 V (the second segment of the passivity region) may apparently be explained by reaction (5), which leads to the dissolution of fluorides in the passive layer followed by the transfer of the dissolution products into corrosive medium, which is also confirmed by the microanalysis data (Supplementary Table S3). On the surface polarized at 1.45 V (transpassivity region), iron and oxygen fractions slightly decrease apparently due to the destruction of the oxide-hydroxide constituent of the film.

Figure 10:

Elemental composition of passive films formed on the steel surface in BBS (pH = 7.4) with added 1.4 mmol/L F⁻ ions without ZnNTP inhibitor (а) and with added 5 g/L ZnNTP inhibitor (b) at different potentials. The concentration of main elements c as depended on the polarization potential E.

The behaviour of the iron and fluorine concentrations in the passive films formed in BBS (pH = 7.4) with added 1.4 mmol/L F⁻ ions and 5 g/L ZnNTP resembles in general that observed in the inhibitor-free medium (Figure 10b). A significant amount of oxygen (∼85 at.%) observed on the surface at Е = −0.77 V is probably due to the adsorption of the ZnNTP inhibitor. In the transpassivity region (E = 1.35 V), oxides and hydroxides are destructed, which leads to decreasing the oxygen content. Insignificant change in the concentration of fluorine in the surface layer, also confirmed by the microanalysis data (Supplementary Table S3), arises from decreasing amount of poorly soluble FeF₂ because of occurrence of competing processes, primarily (1). The concentration of N, P, and Zn being constituents of the inhibitor, grows with increasing potential, which points to increasing the content of the FeZnNTP heterometallic complex which is remained in the passive film even in the transpassivity region (Е = 1.35 V), in contrast to oxides and hydroxides. As seen, the iron fraction in the surface layer formed at the transpassivity potential in the presence of the inhibitor does not only decrease, but even slightly increases, apparently due to low iron ion permeability of the FeZnNTP-doped layer.

As presented in (Kazantseva et al. 2022), surface layer formed at E = 0.5 V in pure BBS (pH = 7.4) and in BBS with the addition of 5 g/L of ZnNTP inhibitor are characterized by a uniform change in the content of iron, oxygen and components of the corrosive medium. The thickness of the oxide-hydroxide constituent matching the molar O-to-Fe ratio of 1:1 is estimated to be about 5 nm in pure BBS and about 15 nm in BBS with added ZnNTP.

F⁻ ions added in amount of 1.4 mmol/L to BBS slightly change the behaviour of elemental composition profiles for the surface layers formed in the passive region (Figure 11). For the coupon polarized at E = 0.3 V corresponding to the first section of the passivity region (Figure 11a), the iron concentration gradually increases from ∼10 at.% in the surface layer to ∼20 at.% at a depth of 15 nm and then drastically grows up to 45 at.% at a depth of 30 nm. The oxygen concentration decreases from ∼65 at.% in the surface layer to 60 at.% at a depth of 5 nm and keeps almost the same to a depth of 15 nm, with further decreasing to ∼35 at.% at a depth of 30 nm. The thickness of the passive film formed at 0.3 V in the presence of 1.4 mmol/L F⁻ ions is estimated to be ∼27 nm. B and Na elements adsorbed from the corrosive medium are evenly distributed through the film thickness. The fluorine concentration increases from ∼3 at.% in the surface layer to 7 at.% at a depth of 30 nm.

Figure 11:

Elemental composition profiles for the surface layers formed on steel in BBS (pH = 7.4) containing 1.4 mmol/L F⁻ ions at potentials 0.3 V (a) and 0.7 V (b). Concentration of the main elements, c, depending on the etching depth, δ.

In the passive film formed at 0.7 V (second segment of the passivity region) in BBS with added 1.4 mmol/L F⁻ ions (Figure 11b), iron and oxygen elemental concentrations change similarly to those observed in the passive film formed at 0.3 V. The iron concentration gradually increases from ∼20 at.% in the surface layer up to ∼25 at.% at a depth of 5–15 nm, with further drastic increase up to ∼40 at.% at a depth of 30 nm. The oxygen concentration is 60–65 at.% at a depth of 15 nm and decreases to ∼40 at.% at a depth of 30 nm. Thus, the oxide/hydroxide constituent of the passive film is estimated to be ∼30 nm thick, which is slightly thicker than that estimated for the film formed at lower potential of 0.3 V. The concentration of B and Na adsorbed from the corrosive medium is not higher than 10 at.% and remains approximately the same through all the passive film thickness, whereas the fluorine concentration, on the contrary, increases in the deeper layers.

Thus, the elemental composition and molar O-to-Fe ratio of the films formed within the passivity region in BBS containing 1.4 mmol/L F⁻ ions change unevenly in-depth. The as-formed films are significantly thicker compared to those grown in both pure BBS and BBS containing ZnNTP. In addition, the fluorine concentration is considerably higher in the deeper layers of the film.

Figure 12a shows elemental composition profiles for surface layers formed in BBS containing 1.4 mmol/L F⁻ ions and 5 g/L ZnNTP inhibitor at the passivation potential of 0.05 V matching the first segment of the passivity region. The iron concentration gradually increases from ∼10 at.% in the surface layer up to ∼40 at.% at a depth of 30 nm. The oxygen concentration also changes gradually, decreasing from ∼55 at.% in the surface layer down to ∼40 at.% at a depth of 30 nm. The thickness of the oxide-hydroxide constituent of the film is estimated to be about 30 nm. Concentrations of Na, P, N, and Zn adsorbed from the corrosion medium and inhibitor somewhat decrease in-depth the film, while the fluorine content noticeably increases from ∼2 at.% in the surface layer up to 8 at.% at a depth of 30 nm.

Figure 12:

Elemental composition profiles for the surface layers formed on steel in BBS (рН = 7.4) with added 1.4 mmol/L F⁻ ions and 5 g/L ZnNTP under polarization at potentials of −0.05 V (а) and 0.65 V (b). Concentration of the main elements, c, depending on the etching depth, δ.

In the passive film formed in BBS containing 1.4 mmol/L F⁻ ions and 5 g/L ZnNTP at 0.65 V matching the second segment of the passivity region (Figure 12b), the iron concentration evenly increases from ∼10 at.% in the surface layer up to ∼40 at.% at a depth of 20 nm, while the oxygen concentration decreases from ∼60 at.% in the surface layer up to ∼40 at.% at a depth of 20 nm. The thickness of the oxide-hydroxide constituent of the passive film formed under these conditions is estimated to be about 20 nm. The concentration of Na decreases from ∼16 at.% in the surface layer to ∼8 at.% at a depth of 10–20 nm. P, N, Zn, and F are distributed relatively evenly through all the film thickness, with their concentration being not higher than 5 at.%.

In the films formed in the environment containing both fluoride ions and inhibitor, the concentration of the main elements (Fe and O) evenly changes in-depth. Na, P, N, and Zn adsorbed from the corrosive environment are evenly distributed through the film thickness. In the film formed at −0.05 V, the fluorine concentration turns out to be maximal in the deeper layers, while its concentration in the film formed at 0.65 V is approximately the same through all the film thickness. It can be assumed that with increasing the polarization potential, the passive film gets thinner due to the occurrence of reactions (1) and (5).

Thus, fluoride ions demonstrate their ambivalent effect on the corrosion-electrochemical behaviour of E 235 steel. On the one hand, they contribute to its passivation by forming poorly soluble iron(II) fluoride FeF₂ in the surface layer composition. On the other hand, they contribute to the destruction of this film by conversing poorly soluble compounds into soluble ones with Fe³⁺ ions and their further transfer into the corrosive medium.

The observed corrosion-electrochemical behaviour of steel E 235 in BBS (pH = 7.4) containing both fluoride ions and ZnNTP inhibitor suggests that in the potential range of −0.1 to 0.2 V these two additives work in the same direction. The higher E_tp value, the stronger corrosion resistance. For media containing low (0.28 mmol/L) and moderate (1.4 mmol/L) amount of F⁻ ions, the optimal concentration of ZnNTP inhibitor is 0.5–1.0 g/L. Other amounts of F⁻ ions have not been found to provide proper corrosion protection and facilitate the decrease in i_c because of apparently simultaneous occurrence of two competing reactions (1) and (3¹). In the passive region, the inhibitor added reduces the dissolution current density, i_p. E_tp slightly decreases with increasing ZnNTP concentration in the corrosive medium, which indicates a slight deterioration of the passive film resistance. To suppress corrosion processes at higher concentration of F⁻ ions in the corrosive medium more inhibitor is required to be added. At 5.6 mmol/L F⁻ ions in the corrosive medium, ZnNTP exhibits its protective properties being added in amount of 5 g/L or more.

XPS data analysis has shown that small ionic radius of F⁻ ions facilitates their penetration into the deeper layers of the as-formed films. The films formed in BBS (рН = 7.4) in the presence of F⁻ ions are mainly comprised of oxides and hydroxides of iron(II) and (III) and poorly soluble FeF₂, and the concentration of fluorides being higher in the deeper layers of the film. It is worth to note that these conditions facilitate the formation of the thicker films with the nominal thickness of the oxide-hydroxide constituent of ∼30 nm, which remain over the entire passive region. Nevertheless, the potentiodynamic data, in particular, the observed i_p values, clearly point to the iron ion penetrability of such films.

The formation of passive film on steel E 235 in BBS containing F⁻ ions and ZnNTP inhibitor is schematically shown in Figure 13, along with the proposed structure of the as-formed film. For the films polarized in the range of potentials below 0.1–0.2 V, the XPS data analysis gives typical iron oxidation state to be 2+, with the oxide-hydroxide constituent of the film consisting of mainly iron(II) oxide and hydroxide, FeO and Fe(OH)₂. Under counter diffusion of Fe²⁺ and ZnNTP ions, reaction (1) proceeds leading to the formation of FeZnNTP heterometallic polynuclear complex in the film. In parallel, reaction (3¹) proceeds with sparingly soluble FeF₂ as a product. The as-formed surface layer of about 30 nm thick is continuous and low permeable by iron ions. In the potential range E above 0.1–0.2 V, the surface layer composition changes due to the oxidation of Fe²⁺ to Fe³⁺, and the respective oxide-hydroxide constituent containing significant amount of Fe³⁺ oxides and hydroxides. A lack of Fe²⁺ ions caused by the occurrence of competing reactions (1) and (4) as well as an excess of F⁻ ions results in the transformation of poorly soluble FeF₂ into highly soluble FeF₃ following the reaction (5), which then leads to thinning the film. In addition, the lack of Fe²⁺ ions due to the occurrence of parallel reactions (3¹) and (4) results in reducing the degree of conversion of ZnNTP inhibitor to FeZnNTP complex in this potential region. The oxide-hydroxide layer thickness for the film formed in this potential region is estimated to be about 20 nm, but the film still remains continuous and low permeable by iron ions and corrosive environment.

Figure 13:

Schematic of the formation of passive film on mild steel in BBS containing F⁻ ions and ZnNTP inhibitor, along with the proposed structure of the film. Numbers in “asterisks” correspond to reaction equations described in the text.

4 Conclusions

The corrosion-electrochemical behaviour of mild steel E 235 was studied in BBS (pH = 7.4) containing F⁻ ions with and without added ZnNTP inhibitor.

In the inhibitor-free medium, the corrosion-electrochemical behaviour of steel changes significantly at concentrations of F⁻ ions of 0.98 mmol/L and higher. The effect of F⁻ ions on the passive film formation is ambivalent, as their interaction with Fe²⁺ ions, on the one hand, produces poorly soluble FeF₂ and thus facilitates the passivation of steel. At the same time, their interaction with Fe³⁺ ions works for the destruction of the passive film via forming soluble compounds.

In the potential range E below 0.1–0.2 V/SSCE, F⁻ ions and ZnNTP inhibitor added simultaneously into the corrosive medium work in the same direction and contribute to the surface passivation due to concurrent formation of poorly soluble FeF₂ and FeZnNTP, which make passive layer more continuous. The depth-profiling XPS analysis has shown that the resulting passive film is comprised of mainly oxides and hydroxides of iron(II) and poorly soluble FeF₂ and FeZnNTP products.

In the potential range E above 0.1–0.2 V/SSCE, F⁻ ions and ZnNTP inhibitor work in contra-directional ways. While F⁻ ions are involved in the formation of soluble FeF₃ and thus destruct the passive film, the inhibitor promotes the formation of FeZnNTP and thus makes the passive layer more continuous.

The FeZnNTP complex has been shown to be the most stable constituent of the passive film, which partially remains even in the transpassivity region.

5 Recommendations

ZnNTP could serve as an effective inhibitor for the corrosion of carbon steel in neutral aqueous media containing a moderate amount of F⁻ ions. When the concentration of fluoride ions in the corrosive medium does not exceed 1.4 mmol/L, the optimal concentration of ZnNTP inhibitor is 0.5–1.0 g/L. For media containing F⁻ ions in amount of 5.6 mmol/L or higher, the recommended amount of ZnNTP inhibitor is at least 5 g/L.